starlingx/kernel
Code Issues Proposed changes

ice: Additional patches to support the customer use cases

Browse Source 
Intel listed total 28 commits that need us to back port. There are
9 commits that are already included in our code base. The commit
"ice: Add support for E825-C TS PLL handling" will not be back
ported since we're not dealing with E825 for 24.09. So we need
back port 18 commits. These commits were introduced in linux-6.9.y
and linux-6.10.y.

To back port these 18 commits successfully, we totally back ported
37 upstream commits.

1) The patches 1-15 are cherry picked to fix the conflicts for patch
   16 ("ice: introduce PTP state machine") and patch 36 "ice:
   Introduce ice_ptp_hw struct". Also will be helpful for the
   subsequent commits back porting.

2) The patches 24-27 are cherry picked to fix the conflicts for patch
   28 ("ice: Fix debugfs with devlink reload")

3) The minor adjust was done for the patches 17, 21, 23 and 33 to
   fit with the context change.

Verification:
- installs from iso succeed on servers with ice(Intel Ethernet
  Controller E810-XXVDA4T Westport Channel) and i40e hw(Intel
  Ethernet Controller X710) for rt and std.
- interfaces are up and pass packets for rt and std.
- create vfs, ensure that they are picked up by the new iavf
  driver and that the interface can come up and pass packets
  on rt and std system.
- Check dmesg to see DDP package is loaded successfully and
  the version is 1.3.36.0 for rt and std.

Story: 2011056
Task: 50950

Change-Id: I9aef0378ea01451684341093a167eaead3edc458
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
This commit is contained in:
Jiping Ma 2024-09-02 06:42:24 +00:00
parent a146061fae
commit 5be9ba03ba
78 changed files with 21614 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

684
kernel-rt/debian/patches/ice-VDF/0001-ice-Auxbus-devices-driver-for-E822-TS.patch Normal file
View File

@ -0,0 +1,684 @@
From ba0d88d4ff54805aac7aec77cc5b05d0df9114da Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:34 +0200
Subject: [PATCH 01/36] ice: Auxbus devices & driver for E822 TS
There is a problem in HW in E822-based devices leading to race
condition.
It might happen that, in order:
- PF0 (which owns the PHC) requests few timestamps,
- PF1 requests a timestamp,
- interrupt is being triggered and both PF0 and PF1 threads are woken
up,
- PF0 got one timestamp, still waiting for others so not going to sleep,
- PF1 gets it's timestamp, process it and go to sleep,
- PF1 requests a timestamp again,
- just before PF0 goes to sleep timestamp of PF1 appear,
- PF0 finishes all it's timestamps and go to sleep (PF1 also sleeping).
That leaves PF1 timestamp memory not read, which lead to blocking the
next interrupt from arriving.
Fix it by adding auxiliary devices and only one driver to handle all the
timestamps for all PF's by PHC owner. In the past each PF requested it's
own timestamps and process it from the start till the end which causes
problem described above. Currently each PF requests the timestamps as
before, but the actual reading of the completed timestamps is being done
by the PTP auxiliary driver, which is registered by the PF which owns PHC.
Additionally, the newly introduced auxiliary driver/devices for PTP clock
owner will be used for other features in all products (including E810).
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit d938a8cca88a5f02f523f95fe3d2d1214f4b4a8d)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 12 +
.../net/ethernet/intel/ice/ice_hw_autogen.h | 1 +
drivers/net/ethernet/intel/ice/ice_main.c | 11 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 393 +++++++++++++++++-
drivers/net/ethernet/intel/ice/ice_ptp.h | 30 ++
5 files changed, 430 insertions(+), 17 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index b9cd0113b859..0a3d76d184ba 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -671,6 +671,18 @@ static inline bool ice_vector_ch_enabled(struct ice_q_vector *qv)
return !!qv->ch; /* Enable it to run with TC */
}
+/**
+ * ice_ptp_pf_handles_tx_interrupt - Check if PF handles Tx interrupt
+ * @pf: Board private structure
+ *
+ * Return true if this PF should respond to the Tx timestamp interrupt
+ * indication in the miscellaneous OICR interrupt handler.
+ */
+static inline bool ice_ptp_pf_handles_tx_interrupt(struct ice_pf *pf)
+{
+ return pf->ptp.tx_interrupt_mode != ICE_PTP_TX_INTERRUPT_NONE;
+}
+
/**
* ice_irq_dynamic_ena - Enable default interrupt generation settings
* @hw: pointer to HW struct
diff --git a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
index 531cc2194741..6756f3d51d14 100644
--- a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
+++ b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
@@ -231,6 +231,7 @@
#define PFINT_SB_CTL 0x0016B600
#define PFINT_SB_CTL_MSIX_INDX_M ICE_M(0x7FF, 0)
#define PFINT_SB_CTL_CAUSE_ENA_M BIT(30)
+#define PFINT_TSYN_MSK 0x0016C980
#define QINT_RQCTL(_QRX) (0x00150000 + ((_QRX) * 4))
#define QINT_RQCTL_MSIX_INDX_S 0
#define QINT_RQCTL_MSIX_INDX_M ICE_M(0x7FF, 0)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 8a6acb5a722e..39cb6ee52abe 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3190,7 +3190,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (!hw->reset_ongoing)
+ if (!hw->reset_ongoing && ice_ptp_pf_handles_tx_interrupt(pf))
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
}
@@ -7444,8 +7444,13 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
}
/* configure PTP timestamping after VSI rebuild */
- if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_cfg_timestamp(pf, false);
+ if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) {
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
+ ice_ptp_cfg_timestamp(pf, false);
+ else if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL)
+ /* for E82x PHC owner always need to have interrupts */
+ ice_ptp_cfg_timestamp(pf, true);
+ }
err = ice_vsi_rebuild_by_type(pf, ICE_VSI_SWITCHDEV_CTRL);
if (err) {
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 3648d3cccacc..e3012608c9dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -255,6 +255,24 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
return ice_ptp_set_sma_e810t(info, pin, func);
}
+/**
+ * ice_ptp_configure_tx_tstamp - Enable or disable Tx timestamp interrupt
+ * @pf: The PF pointer to search in
+ * @on: bool value for whether timestamp interrupt is enabled or disabled
+ */
+static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
+{
+ u32 val;
+
+ /* Configure the Tx timestamp interrupt */
+ val = rd32(&pf->hw, PFINT_OICR_ENA);
+ if (on)
+ val |= PFINT_OICR_TSYN_TX_M;
+ else
+ val &= ~PFINT_OICR_TSYN_TX_M;
+ wr32(&pf->hw, PFINT_OICR_ENA, val);
+}
+
/**
* ice_set_tx_tstamp - Enable or disable Tx timestamping
* @pf: The PF pointer to search in
@@ -263,7 +281,6 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
struct ice_vsi *vsi;
- u32 val;
u16 i;
vsi = ice_get_main_vsi(pf);
@@ -277,13 +294,8 @@ static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
vsi->tx_rings[i]->ptp_tx = on;
}
- /* Configure the Tx timestamp interrupt */
- val = rd32(&pf->hw, PFINT_OICR_ENA);
- if (on)
- val |= PFINT_OICR_TSYN_TX_M;
- else
- val &= ~PFINT_OICR_TSYN_TX_M;
- wr32(&pf->hw, PFINT_OICR_ENA, val);
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
+ ice_ptp_configure_tx_tstamp(pf, on);
pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
}
@@ -674,9 +686,6 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
int err;
u8 idx;
- if (!tx->init)
- return;
-
ptp_port = container_of(tx, struct ice_ptp_port, tx);
pf = ptp_port_to_pf(ptp_port);
hw = &pf->hw;
@@ -774,6 +783,39 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
}
}
+/**
+ * ice_ptp_tx_tstamp_owner - Process Tx timestamps for all ports on the device
+ * @pf: Board private structure
+ */
+static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
+{
+ struct ice_ptp_port *port;
+ unsigned int i;
+
+ mutex_lock(&pf->ptp.ports_owner.lock);
+ list_for_each_entry(port, &pf->ptp.ports_owner.ports, list_member) {
+ struct ice_ptp_tx *tx = &port->tx;
+
+ if (!tx || !tx->init)
+ continue;
+
+ ice_ptp_process_tx_tstamp(tx);
+ }
+ mutex_unlock(&pf->ptp.ports_owner.lock);
+
+ for (i = 0; i < ICE_MAX_QUAD; i++) {
+ u64 tstamp_ready;
+ int err;
+
+ /* Read the Tx ready status first */
+ err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
+ if (err || tstamp_ready)
+ return ICE_TX_TSTAMP_WORK_PENDING;
+ }
+
+ return ICE_TX_TSTAMP_WORK_DONE;
+}
+
/**
* ice_ptp_tx_tstamp - Process Tx timestamps for this function.
* @tx: Tx tracking structure to initialize
@@ -2448,7 +2490,21 @@ s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
*/
enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf)
{
- return ice_ptp_tx_tstamp(&pf->ptp.port.tx);
+ switch (pf->ptp.tx_interrupt_mode) {
+ case ICE_PTP_TX_INTERRUPT_NONE:
+ /* This device has the clock owner handle timestamps for it */
+ return ICE_TX_TSTAMP_WORK_DONE;
+ case ICE_PTP_TX_INTERRUPT_SELF:
+ /* This device handles its own timestamps */
+ return ice_ptp_tx_tstamp(&pf->ptp.port.tx);
+ case ICE_PTP_TX_INTERRUPT_ALL:
+ /* This device handles timestamps for all ports */
+ return ice_ptp_tx_tstamp_owner(pf);
+ default:
+ WARN_ONCE(1, "Unexpected Tx timestamp interrupt mode %u\n",
+ pf->ptp.tx_interrupt_mode);
+ return ICE_TX_TSTAMP_WORK_DONE;
+ }
}
static void ice_ptp_periodic_work(struct kthread_work *work)
@@ -2557,6 +2613,187 @@ void ice_ptp_reset(struct ice_pf *pf)
dev_err(ice_pf_to_dev(pf), "PTP reset failed %d\n", err);
}
+/**
+ * ice_ptp_aux_dev_to_aux_pf - Get auxiliary PF handle for the auxiliary device
+ * @aux_dev: auxiliary device to get the auxiliary PF for
+ */
+static struct ice_pf *
+ice_ptp_aux_dev_to_aux_pf(struct auxiliary_device *aux_dev)
+{
+ struct ice_ptp_port *aux_port;
+ struct ice_ptp *aux_ptp;
+
+ aux_port = container_of(aux_dev, struct ice_ptp_port, aux_dev);
+ aux_ptp = container_of(aux_port, struct ice_ptp, port);
+
+ return container_of(aux_ptp, struct ice_pf, ptp);
+}
+
+/**
+ * ice_ptp_aux_dev_to_owner_pf - Get PF handle for the auxiliary device
+ * @aux_dev: auxiliary device to get the PF for
+ */
+static struct ice_pf *
+ice_ptp_aux_dev_to_owner_pf(struct auxiliary_device *aux_dev)
+{
+ struct ice_ptp_port_owner *ports_owner;
+ struct auxiliary_driver *aux_drv;
+ struct ice_ptp *owner_ptp;
+
+ if (!aux_dev->dev.driver)
+ return NULL;
+
+ aux_drv = to_auxiliary_drv(aux_dev->dev.driver);
+ ports_owner = container_of(aux_drv, struct ice_ptp_port_owner,
+ aux_driver);
+ owner_ptp = container_of(ports_owner, struct ice_ptp, ports_owner);
+ return container_of(owner_ptp, struct ice_pf, ptp);
+}
+
+/**
+ * ice_ptp_auxbus_probe - Probe auxiliary devices
+ * @aux_dev: PF's auxiliary device
+ * @id: Auxiliary device ID
+ */
+static int ice_ptp_auxbus_probe(struct auxiliary_device *aux_dev,
+ const struct auxiliary_device_id *id)
+{
+ struct ice_pf *owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ struct ice_pf *aux_pf = ice_ptp_aux_dev_to_aux_pf(aux_dev);
+
+ if (WARN_ON(!owner_pf))
+ return -ENODEV;
+
+ INIT_LIST_HEAD(&aux_pf->ptp.port.list_member);
+ mutex_lock(&owner_pf->ptp.ports_owner.lock);
+ list_add(&aux_pf->ptp.port.list_member,
+ &owner_pf->ptp.ports_owner.ports);
+ mutex_unlock(&owner_pf->ptp.ports_owner.lock);
+
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_remove - Remove auxiliary devices from the bus
+ * @aux_dev: PF's auxiliary device
+ */
+static void ice_ptp_auxbus_remove(struct auxiliary_device *aux_dev)
+{
+ struct ice_pf *owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ struct ice_pf *aux_pf = ice_ptp_aux_dev_to_aux_pf(aux_dev);
+
+ mutex_lock(&owner_pf->ptp.ports_owner.lock);
+ list_del(&aux_pf->ptp.port.list_member);
+ mutex_unlock(&owner_pf->ptp.ports_owner.lock);
+}
+
+/**
+ * ice_ptp_auxbus_shutdown
+ * @aux_dev: PF's auxiliary device
+ */
+static void ice_ptp_auxbus_shutdown(struct auxiliary_device *aux_dev)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+}
+
+/**
+ * ice_ptp_auxbus_suspend
+ * @aux_dev: PF's auxiliary device
+ * @state: power management state indicator
+ */
+static int
+ice_ptp_auxbus_suspend(struct auxiliary_device *aux_dev, pm_message_t state)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_resume
+ * @aux_dev: PF's auxiliary device
+ */
+static int ice_ptp_auxbus_resume(struct auxiliary_device *aux_dev)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_create_id_table - Create auxiliary device ID table
+ * @pf: Board private structure
+ * @name: auxiliary bus driver name
+ */
+static struct auxiliary_device_id *
+ice_ptp_auxbus_create_id_table(struct ice_pf *pf, const char *name)
+{
+ struct auxiliary_device_id *ids;
+
+ /* Second id left empty to terminate the array */
+ ids = devm_kcalloc(ice_pf_to_dev(pf), 2,
+ sizeof(struct auxiliary_device_id), GFP_KERNEL);
+ if (!ids)
+ return NULL;
+
+ snprintf(ids[0].name, sizeof(ids[0].name), "ice.%s", name);
+
+ return ids;
+}
+
+/**
+ * ice_ptp_register_auxbus_driver - Register PTP auxiliary bus driver
+ * @pf: Board private structure
+ */
+static int ice_ptp_register_auxbus_driver(struct ice_pf *pf)
+{
+ struct auxiliary_driver *aux_driver;
+ struct ice_ptp *ptp;
+ struct device *dev;
+ char *name;
+ int err;
+
+ ptp = &pf->ptp;
+ dev = ice_pf_to_dev(pf);
+ aux_driver = &ptp->ports_owner.aux_driver;
+ INIT_LIST_HEAD(&ptp->ports_owner.ports);
+ mutex_init(&ptp->ports_owner.lock);
+ name = devm_kasprintf(dev, GFP_KERNEL, "ptp_aux_dev_%u_%u_clk%u",
+ pf->pdev->bus->number, PCI_SLOT(pf->pdev->devfn),
+ ice_get_ptp_src_clock_index(&pf->hw));
+
+ aux_driver->name = name;
+ aux_driver->shutdown = ice_ptp_auxbus_shutdown;
+ aux_driver->suspend = ice_ptp_auxbus_suspend;
+ aux_driver->remove = ice_ptp_auxbus_remove;
+ aux_driver->resume = ice_ptp_auxbus_resume;
+ aux_driver->probe = ice_ptp_auxbus_probe;
+ aux_driver->id_table = ice_ptp_auxbus_create_id_table(pf, name);
+ if (!aux_driver->id_table)
+ return -ENOMEM;
+
+ err = auxiliary_driver_register(aux_driver);
+ if (err) {
+ devm_kfree(dev, aux_driver->id_table);
+ dev_err(dev, "Failed registering aux_driver, name <%s>\n",
+ name);
+ }
+
+ return err;
+}
+
+/**
+ * ice_ptp_unregister_auxbus_driver - Unregister PTP auxiliary bus driver
+ * @pf: Board private structure
+ */
+static void ice_ptp_unregister_auxbus_driver(struct ice_pf *pf)
+{
+ struct auxiliary_driver *aux_driver = &pf->ptp.ports_owner.aux_driver;
+
+ auxiliary_driver_unregister(aux_driver);
+ devm_kfree(ice_pf_to_dev(pf), aux_driver->id_table);
+
+ mutex_destroy(&pf->ptp.ports_owner.lock);
+}
+
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
@@ -2635,7 +2872,15 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
- if (!ice_is_e810(hw)) {
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL) {
+ /* The clock owner for this device type handles the timestamp
+ * interrupt for all ports.
+ */
+ ice_ptp_configure_tx_tstamp(pf, true);
+
+ /* React on all quads interrupts for E82x */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
+
/* Enable quad interrupts */
err = ice_ptp_tx_ena_intr(pf, true, itr);
if (err)
@@ -2650,8 +2895,16 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Store the PTP clock index for other PFs */
ice_set_ptp_clock_index(pf);
- return 0;
+ err = ice_ptp_register_auxbus_driver(pf);
+ if (err) {
+ dev_err(ice_pf_to_dev(pf), "Failed to register PTP auxbus driver");
+ goto err_aux;
+ }
+ return 0;
+err_aux:
+ ice_clear_ptp_clock_index(pf);
+ ptp_clock_unregister(pf->ptp.clock);
err_clk:
pf->ptp.clock = NULL;
err_exit:
@@ -2701,6 +2954,13 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E822:
+ /* Non-owner PFs don't react to any interrupts on E82x,
+ * neither on own quad nor on others
+ */
+ if (!ice_ptp_pf_handles_tx_interrupt(pf)) {
+ ice_ptp_configure_tx_tstamp(pf, false);
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
+ }
kthread_init_delayed_work(&ptp_port->ov_work,
ice_ptp_wait_for_offsets);
@@ -2711,6 +2971,101 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
}
}
+/**
+ * ice_ptp_release_auxbus_device
+ * @dev: device that utilizes the auxbus
+ */
+static void ice_ptp_release_auxbus_device(struct device *dev)
+{
+ /* Doing nothing here, but handle to auxbux device must be satisfied */
+}
+
+/**
+ * ice_ptp_create_auxbus_device - Create PTP auxiliary bus device
+ * @pf: Board private structure
+ */
+static int ice_ptp_create_auxbus_device(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev;
+ struct ice_ptp *ptp;
+ struct device *dev;
+ char *name;
+ int err;
+ u32 id;
+
+ ptp = &pf->ptp;
+ id = ptp->port.port_num;
+ dev = ice_pf_to_dev(pf);
+
+ aux_dev = &ptp->port.aux_dev;
+
+ name = devm_kasprintf(dev, GFP_KERNEL, "ptp_aux_dev_%u_%u_clk%u",
+ pf->pdev->bus->number, PCI_SLOT(pf->pdev->devfn),
+ ice_get_ptp_src_clock_index(&pf->hw));
+
+ aux_dev->name = name;
+ aux_dev->id = id;
+ aux_dev->dev.release = ice_ptp_release_auxbus_device;
+ aux_dev->dev.parent = dev;
+
+ err = auxiliary_device_init(aux_dev);
+ if (err)
+ goto aux_err;
+
+ err = auxiliary_device_add(aux_dev);
+ if (err) {
+ auxiliary_device_uninit(aux_dev);
+ goto aux_err;
+ }
+
+ return 0;
+aux_err:
+ dev_err(dev, "Failed to create PTP auxiliary bus device <%s>\n", name);
+ devm_kfree(dev, name);
+ return err;
+}
+
+/**
+ * ice_ptp_remove_auxbus_device - Remove PTP auxiliary bus device
+ * @pf: Board private structure
+ */
+static void ice_ptp_remove_auxbus_device(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev = &pf->ptp.port.aux_dev;
+
+ auxiliary_device_delete(aux_dev);
+ auxiliary_device_uninit(aux_dev);
+
+ memset(aux_dev, 0, sizeof(*aux_dev));
+}
+
+/**
+ * ice_ptp_init_tx_interrupt_mode - Initialize device Tx interrupt mode
+ * @pf: Board private structure
+ *
+ * Initialize the Tx timestamp interrupt mode for this device. For most device
+ * types, each PF processes the interrupt and manages its own timestamps. For
+ * E822-based devices, only the clock owner processes the timestamps. Other
+ * PFs disable the interrupt and do not process their own timestamps.
+ */
+static void ice_ptp_init_tx_interrupt_mode(struct ice_pf *pf)
+{
+ switch (pf->hw.phy_model) {
+ case ICE_PHY_E822:
+ /* E822 based PHY has the clock owner process the interrupt
+ * for all ports.
+ */
+ if (ice_pf_src_tmr_owned(pf))
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_ALL;
+ else
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_NONE;
+ break;
+ default:
+ /* other PHY types handle their own Tx interrupt */
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_SELF;
+ }
+}
+
/**
* ice_ptp_init - Initialize PTP hardware clock support
* @pf: Board private structure
@@ -2731,6 +3086,8 @@ void ice_ptp_init(struct ice_pf *pf)
ice_ptp_init_phy_model(hw);
+ ice_ptp_init_tx_interrupt_mode(pf);
+
/* If this function owns the clock hardware, it must allocate and
* configure the PTP clock device to represent it.
*/
@@ -2753,6 +3110,10 @@ void ice_ptp_init(struct ice_pf *pf)
if (err)
goto err;
+ err = ice_ptp_create_auxbus_device(pf);
+ if (err)
+ goto err;
+
dev_info(ice_pf_to_dev(pf), "PTP init successful\n");
return;
@@ -2781,6 +3142,8 @@ void ice_ptp_release(struct ice_pf *pf)
/* Disable timestamping for both Tx and Rx */
ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_remove_auxbus_device(pf);
+
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
clear_bit(ICE_FLAG_PTP, pf->flags);
@@ -2804,5 +3167,7 @@ void ice_ptp_release(struct ice_pf *pf)
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
+ ice_ptp_unregister_auxbus_driver(pf);
+
dev_info(ice_pf_to_dev(pf), "Removed PTP clock\n");
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 995a57019ba7..d94c22329df0 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -157,7 +157,9 @@ struct ice_ptp_tx {
* ready for PTP functionality. It is used to track the port initialization
* and determine when the port's PHY offset is valid.
*
+ * @list_member: list member structure of auxiliary device
* @tx: Tx timestamp tracking for this port
+ * @aux_dev: auxiliary device associated with this port
* @ov_work: delayed work task for tracking when PHY offset is valid
* @ps_lock: mutex used to protect the overall PTP PHY start procedure
* @link_up: indicates whether the link is up
@@ -165,7 +167,9 @@ struct ice_ptp_tx {
* @port_num: the port number this structure represents
*/
struct ice_ptp_port {
+ struct list_head list_member;
struct ice_ptp_tx tx;
+ struct auxiliary_device aux_dev;
struct kthread_delayed_work ov_work;
struct mutex ps_lock; /* protects overall PTP PHY start procedure */
bool link_up;
@@ -173,11 +177,35 @@ struct ice_ptp_port {
u8 port_num;
};
+enum ice_ptp_tx_interrupt {
+ ICE_PTP_TX_INTERRUPT_NONE = 0,
+ ICE_PTP_TX_INTERRUPT_SELF,
+ ICE_PTP_TX_INTERRUPT_ALL,
+};
+
+/**
+ * struct ice_ptp_port_owner - data used to handle the PTP clock owner info
+ *
+ * This structure contains data necessary for the PTP clock owner to correctly
+ * handle the timestamping feature for all attached ports.
+ *
+ * @aux_driver: the structure carring the auxiliary driver information
+ * @ports: list of porst handled by this port owner
+ * @lock: protect access to ports list
+ */
+struct ice_ptp_port_owner {
+ struct auxiliary_driver aux_driver;
+ struct list_head ports;
+ struct mutex lock;
+};
+
#define GLTSYN_TGT_H_IDX_MAX 4
/**
* struct ice_ptp - data used for integrating with CONFIG_PTP_1588_CLOCK
+ * @tx_interrupt_mode: the TX interrupt mode for the PTP clock
* @port: data for the PHY port initialization procedure
+ * @ports_owner: data for the auxiliary driver owner
* @work: delayed work function for periodic tasks
* @cached_phc_time: a cached copy of the PHC time for timestamp extension
* @cached_phc_jiffies: jiffies when cached_phc_time was last updated
@@ -197,7 +225,9 @@ struct ice_ptp_port {
* @late_cached_phc_updates: number of times cached PHC update is late
*/
struct ice_ptp {
+ enum ice_ptp_tx_interrupt tx_interrupt_mode;
struct ice_ptp_port port;
+ struct ice_ptp_port_owner ports_owner;
struct kthread_delayed_work work;
u64 cached_phc_time;
unsigned long cached_phc_jiffies;
--
2.43.0

95
kernel-rt/debian/patches/ice-VDF/0002-ice-introduce-ice_pf_src_tmr_owned.patch Normal file
View File

@ -0,0 +1,95 @@
From f6af978ef435067b4c9f5ff5e159f8b65d969268 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 8 Sep 2023 14:37:14 -0700
Subject: [PATCH 02/36] ice: introduce ice_pf_src_tmr_owned
Add ice_pf_src_tmr_owned() macro to check the function capability bit
indicating if the current function owns the PTP hardware clock. This is
slightly shorter than the more verbose access via
hw.func_caps.ts_func_info.src_tmr_owned. Use this where possible rather
than open coding its equivalent.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 42d40bb21e332151da6fb689bf7d4af8195866ed)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 ++
drivers/net/ethernet/intel/ice/ice_lib.c | 2 +-
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 6 +++---
4 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 0a3d76d184ba..54a98c4032b7 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -197,6 +197,8 @@ extern const char ice_drv_ver[];
#define ice_pf_to_dev(pf) (&((pf)->pdev->dev))
+#define ice_pf_src_tmr_owned(pf) ((pf)->hw.func_caps.ts_func_info.src_tmr_owned)
+
enum ice_feature {
ICE_F_DSCP,
ICE_F_PHY_RCLK,
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 632091487413..106ef843f4b5 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -4010,7 +4010,7 @@ void ice_init_feature_support(struct ice_pf *pf)
if (ice_is_phy_rclk_in_netlist(&pf->hw))
ice_set_feature_support(pf, ICE_F_PHY_RCLK);
/* If we don't own the timer - don't enable other caps */
- if (!pf->hw.func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
break;
if (ice_is_cgu_in_netlist(&pf->hw))
ice_set_feature_support(pf, ICE_F_CGU);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 39cb6ee52abe..e957529b3fd6 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3200,7 +3200,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
ena_mask &= ~PFINT_OICR_TSYN_EVNT_M;
- if (hw->func_caps.ts_func_info.src_tmr_owned) {
+ if (ice_pf_src_tmr_owned(pf)) {
/* Save EVENTs from GLTSYN register */
pf->ptp.ext_ts_irq |= gltsyn_stat &
(GLTSYN_STAT_EVENT0_M |
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index e3012608c9dd..b1951357ba9f 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -448,7 +448,7 @@ static void ice_clear_ptp_clock_index(struct ice_pf *pf)
int err;
/* Do not clear the index if we don't own the timer */
- if (!hw->func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
return;
tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
@@ -2538,7 +2538,7 @@ void ice_ptp_reset(struct ice_pf *pf)
if (test_bit(ICE_PFR_REQ, pf->state))
goto pfr;
- if (!hw->func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
goto reset_ts;
err = ice_ptp_init_phc(hw);
@@ -3091,7 +3091,7 @@ void ice_ptp_init(struct ice_pf *pf)
/* If this function owns the clock hardware, it must allocate and
* configure the PTP clock device to represent it.
*/
- if (hw->func_caps.ts_func_info.src_tmr_owned) {
+ if (ice_pf_src_tmr_owned(pf)) {
err = ice_ptp_init_owner(pf);
if (err)
goto err;
--
2.43.0

93
kernel-rt/debian/patches/ice-VDF/0003-ice-Re-enable-timestamping-correctly-after-reset.patch Normal file
View File

@ -0,0 +1,93 @@
From 3c155fbf8e2a0546302a01cc06e8ece18468148e Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Fri, 1 Dec 2023 10:08:42 -0800
Subject: [PATCH 03/36] ice: Re-enable timestamping correctly after reset
During reset, TX_TSYN interrupt should be processed as it may process
timestamps in brief moments before and after reset.
Timestamping should be enabled on VSIs at the end of reset procedure.
On ice_get_phy_tx_tstamp_ready error, interrupt should not be rearmed
because error only happens on resets.
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 1cc5b6eaad92d69fe4d84bbee5c12ee297d56296)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 19 ++++++++++---------
2 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index e957529b3fd6..d2f3b4374d14 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3190,7 +3190,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (!hw->reset_ongoing && ice_ptp_pf_handles_tx_interrupt(pf))
+ if (ice_ptp_pf_handles_tx_interrupt(pf))
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index b1951357ba9f..92459589f6ce 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -809,7 +809,9 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
/* Read the Tx ready status first */
err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
- if (err || tstamp_ready)
+ if (err)
+ break;
+ else if (tstamp_ready)
return ICE_TX_TSTAMP_WORK_PENDING;
}
@@ -2535,12 +2537,10 @@ void ice_ptp_reset(struct ice_pf *pf)
int err, itr = 1;
u64 time_diff;
- if (test_bit(ICE_PFR_REQ, pf->state))
+ if (test_bit(ICE_PFR_REQ, pf->state) ||
+ !ice_pf_src_tmr_owned(pf))
goto pfr;
- if (!ice_pf_src_tmr_owned(pf))
- goto reset_ts;
-
err = ice_ptp_init_phc(hw);
if (err)
goto err;
@@ -2584,10 +2584,6 @@ void ice_ptp_reset(struct ice_pf *pf)
goto err;
}
-reset_ts:
- /* Restart the PHY timestamping block */
- ice_ptp_reset_phy_timestamping(pf);
-
pfr:
/* Init Tx structures */
if (ice_is_e810(&pf->hw)) {
@@ -2603,6 +2599,11 @@ void ice_ptp_reset(struct ice_pf *pf)
set_bit(ICE_FLAG_PTP, pf->flags);
+ /* Restart the PHY timestamping block */
+ if (!test_bit(ICE_PFR_REQ, pf->state) &&
+ ice_pf_src_tmr_owned(pf))
+ ice_ptp_restart_all_phy(pf);
+
/* Start periodic work going */
kthread_queue_delayed_work(ptp->kworker, &ptp->work, 0);
--
2.43.0

121
kernel-rt/debian/patches/ice-VDF/0004-ice-periodically-kick-Tx-timestamp-interrupt.patch Normal file
View File

@ -0,0 +1,121 @@
From 214f06259ade960e3790b62f96bc1b75e5b76e79 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 1 Dec 2023 10:08:43 -0800
Subject: [PATCH 04/36] ice: periodically kick Tx timestamp interrupt
The E822 hardware for Tx timestamping keeps track of how many
outstanding timestamps are still in the PHY memory block. It will not
generate a new interrupt to the MAC until all of the timestamps in the
region have been read.
If somehow all the available data is not read, but the driver has exited
its interrupt routine already, the PHY will not generate a new interrupt
even if new timestamp data is captured. Because no interrupt is
generated, the driver never processes the timestamp data. This state
results in a permanent failure for all future Tx timestamps.
It is not clear how the driver and hardware could enter this state.
However, if it does, there is currently no recovery mechanism.
Add a recovery mechanism via the periodic PTP work thread which invokes
ice_ptp_periodic_work(). Introduce a new check,
ice_ptp_maybe_trigger_tx_interrupt() which checks the PHY timestamp
ready bitmask. If any bits are set, trigger a software interrupt by
writing to PFINT_OICR.
Once triggered, the main timestamp processing thread will read through
the PHY data and clear the outstanding timestamp data. Once cleared, new
data should trigger interrupts as expected.
This should allow recovery from such a state rather than leaving the
device in a state where we cannot process Tx timestamps.
It is possible that this function checks for timestamp data
simultaneously with the interrupt, and it might trigger additional
unnecessary interrupts. This will cause a small amount of additional
processing.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Andrii Staikov <andrii.staikov@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 712e876371f8350c446a33577cf4a0aedcd4742a)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 50 ++++++++++++++++++++++++
1 file changed, 50 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 92459589f6ce..0d6c7215e0c1 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2509,6 +2509,54 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf)
}
}
+/**
+ * ice_ptp_maybe_trigger_tx_interrupt - Trigger Tx timstamp interrupt
+ * @pf: Board private structure
+ *
+ * The device PHY issues Tx timestamp interrupts to the driver for processing
+ * timestamp data from the PHY. It will not interrupt again until all
+ * current timestamp data is read. In rare circumstances, it is possible that
+ * the driver fails to read all outstanding data.
+ *
+ * To avoid getting permanently stuck, periodically check if the PHY has
+ * outstanding timestamp data. If so, trigger an interrupt from software to
+ * process this data.
+ */
+static void ice_ptp_maybe_trigger_tx_interrupt(struct ice_pf *pf)
+{
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ bool trigger_oicr = false;
+ unsigned int i;
+
+ if (ice_is_e810(hw))
+ return;
+
+ if (!ice_pf_src_tmr_owned(pf))
+ return;
+
+ for (i = 0; i < ICE_MAX_QUAD; i++) {
+ u64 tstamp_ready;
+ int err;
+
+ err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
+ if (!err && tstamp_ready) {
+ trigger_oicr = true;
+ break;
+ }
+ }
+
+ if (trigger_oicr) {
+ /* Trigger a software interrupt, to ensure this data
+ * gets processed.
+ */
+ dev_dbg(dev, "PTP periodic task detected waiting timestamps. Triggering Tx timestamp interrupt now.\n");
+
+ wr32(hw, PFINT_OICR, PFINT_OICR_TSYN_TX_M);
+ ice_flush(hw);
+ }
+}
+
static void ice_ptp_periodic_work(struct kthread_work *work)
{
struct ice_ptp *ptp = container_of(work, struct ice_ptp, work.work);
@@ -2520,6 +2568,8 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
err = ice_ptp_update_cached_phctime(pf);
+ ice_ptp_maybe_trigger_tx_interrupt(pf);
+
/* Run twice a second or reschedule if phc update failed */
kthread_queue_delayed_work(ptp->kworker, &ptp->work,
msecs_to_jiffies(err ? 10 : 500));
--
2.43.0

81
kernel-rt/debian/patches/ice-VDF/0005-ice-PTP-Rename-macros-used-for-PHY-QUAD-port-definit.patch Normal file
View File

@ -0,0 +1,81 @@
From c25fc364d599195403ed9ba51ef8fa6ed3b642ff Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:44 -0700
Subject: [PATCH 05/36] ice: PTP: Rename macros used for PHY/QUAD port
definitions
The ice_fill_phy_msg_e822 function uses several macros to specify the
correct address when sending a sideband message to the PHY block in
hardware.
The names of these macros are fairly generic and confusing. Future
development is going to extend the driver to support new hardware families
which have different relationships between PHY and QUAD. Rename the macros
for clarity and to indicate that they are E822 specific. This also matches
closer to the hardware specification in the data sheet.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 64fd7de2469dd52a7f1517ce95ae22fcb391a8a1)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 8 ++++----
drivers/net/ethernet/intel/ice/ice_type.h | 14 +++++++-------
2 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index a299af39a7c4..03c4aa995e8d 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -294,9 +294,9 @@ ice_fill_phy_msg_e822(struct ice_sbq_msg_input *msg, u8 port, u16 offset)
{
int phy_port, phy, quadtype;
- phy_port = port % ICE_PORTS_PER_PHY;
- phy = port / ICE_PORTS_PER_PHY;
- quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_NUM_QUAD_TYPE;
+ phy_port = port % ICE_PORTS_PER_PHY_E822;
+ phy = port / ICE_PORTS_PER_PHY_E822;
+ quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_QUADS_PER_PHY_E822;
if (quadtype == 0) {
msg->msg_addr_low = P_Q0_L(P_0_BASE + offset, phy_port);
@@ -628,7 +628,7 @@ ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
msg->dest_dev = rmn_0;
- if ((quad % ICE_NUM_QUAD_TYPE) == 0)
+ if ((quad % ICE_QUADS_PER_PHY_E822) == 0)
addr = Q_0_BASE + offset;
else
addr = Q_1_BASE + offset;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 4cd131546aa9..bb5d8b681bc2 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -909,13 +909,13 @@ struct ice_hw {
/* INTRL granularity in 1 us */
u8 intrl_gran;
-#define ICE_PHY_PER_NAC 1
-#define ICE_MAX_QUAD 2
-#define ICE_NUM_QUAD_TYPE 2
-#define ICE_PORTS_PER_QUAD 4
-#define ICE_PHY_0_LAST_QUAD 1
-#define ICE_PORTS_PER_PHY 8
-#define ICE_NUM_EXTERNAL_PORTS ICE_PORTS_PER_PHY
+#define ICE_PHY_PER_NAC_E822 1
+#define ICE_MAX_QUAD 2
+#define ICE_QUADS_PER_PHY_E822 2
+#define ICE_PORTS_PER_PHY_E822 8
+#define ICE_PORTS_PER_QUAD 4
+#define ICE_PORTS_PER_PHY_E810 4
+#define ICE_NUM_EXTERNAL_PORTS (ICE_MAX_QUAD * ICE_PORTS_PER_QUAD)
/* Active package version (currently active) */
struct ice_pkg_ver active_pkg_ver;
--
2.43.0

80
kernel-rt/debian/patches/ice-VDF/0006-ice-PTP-move-quad-value-check-inside-ice_fill_phy_ms.patch Normal file
View File

@ -0,0 +1,80 @@
From 13f48f4c94ad4d317e7c7ccaa188a11850a8aa32 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:45 -0700
Subject: [PATCH 06/36] ice: PTP: move quad value check inside
ice_fill_phy_msg_e822
The callers of ice_fill_phy_msg_e822 check for whether the quad number is
within the expected range. Move this check inside the ice_fill_phy_msg_e822
function instead of duplicating it twice.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit dd84744cf5ea967c8d53aae6b6a45703dbc5c5c4)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 19 ++++++++++++-------
1 file changed, 12 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index 03c4aa995e8d..e024b88ce32b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -621,11 +621,14 @@ ice_write_64b_phy_reg_e822(struct ice_hw *hw, u8 port, u16 low_addr, u64 val)
* Fill a message buffer for accessing a register in a quad shared between
* multiple PHYs.
*/
-static void
+static int
ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
{
u32 addr;
+ if (quad >= ICE_MAX_QUAD)
+ return -EINVAL;
+
msg->dest_dev = rmn_0;
if ((quad % ICE_QUADS_PER_PHY_E822) == 0)
@@ -635,6 +638,8 @@ ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
msg->msg_addr_low = lower_16_bits(addr);
msg->msg_addr_high = upper_16_bits(addr);
+
+ return 0;
}
/**
@@ -653,10 +658,10 @@ ice_read_quad_reg_e822(struct ice_hw *hw, u8 quad, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- if (quad >= ICE_MAX_QUAD)
- return -EINVAL;
+ err = ice_fill_quad_msg_e822(&msg, quad, offset);
+ if (err)
+ return err;
- ice_fill_quad_msg_e822(&msg, quad, offset);
msg.opcode = ice_sbq_msg_rd;
err = ice_sbq_rw_reg(hw, &msg);
@@ -687,10 +692,10 @@ ice_write_quad_reg_e822(struct ice_hw *hw, u8 quad, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- if (quad >= ICE_MAX_QUAD)
- return -EINVAL;
+ err = ice_fill_quad_msg_e822(&msg, quad, offset);
+ if (err)
+ return err;
- ice_fill_quad_msg_e822(&msg, quad, offset);
msg.opcode = ice_sbq_msg_wr;
msg.data = val;
--
2.43.0

97
kernel-rt/debian/patches/ice-VDF/0007-ice-remove-ptp_tx-ring-parameter-flag.patch Normal file
View File

@ -0,0 +1,97 @@
From 7dae9333af82f6c9e2db1940c3a10ae38dabea7b Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:55 -0800
Subject: [PATCH 07/36] ice: remove ptp_tx ring parameter flag
Before performing a Tx timestamp in ice_stamp(), the driver checks a ptp_tx
ring variable to see if timestamping is enabled on that ring. This value is
set for all rings whenever userspace configures Tx timestamping.
Ostensibly this was done to avoid wasting cycles checking other fields when
timestamping has not been enabled. However, for Tx timestamps we already
get an individual per-SKB flag indicating whether userspace wants to
request a timestamp on that packet. We do not gain much by also having
a separate flag to check for whether timestamping was enabled.
In fact, the driver currently fails to restore the field after a PF reset.
Because of this, if a PF reset occurs, timestamps will be disabled.
Since this flag doesn't add value in the hotpath, remove it and always
provide a timestamp if the SKB flag has been set.
A following change will fix the reset path to properly restore user
timestamping configuration completely.
This went unnoticed for some time because one of the most common
applications using Tx timestamps, ptp4l, will reconfigure the socket as
part of its fault recovery logic.
Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 0ffb08b1a45bd6b7694e01da0e1d9e3e788418fb)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 14 --------------
drivers/net/ethernet/intel/ice/ice_txrx.c | 3 ---
drivers/net/ethernet/intel/ice/ice_txrx.h | 1 -
3 files changed, 18 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 0d6c7215e0c1..c03153bdb7c3 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -280,20 +280,6 @@ static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
*/
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
- struct ice_vsi *vsi;
- u16 i;
-
- vsi = ice_get_main_vsi(pf);
- if (!vsi)
- return;
-
- /* Set the timestamp enable flag for all the Tx rings */
- ice_for_each_txq(vsi, i) {
- if (!vsi->tx_rings[i])
- continue;
- vsi->tx_rings[i]->ptp_tx = on;
- }
-
if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
ice_ptp_configure_tx_tstamp(pf, on);
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index 24c914015973..9170a3e8f088 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -2305,9 +2305,6 @@ ice_tstamp(struct ice_tx_ring *tx_ring, struct sk_buff *skb,
if (likely(!(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)))
return;
- if (!tx_ring->ptp_tx)
- return;
-
/* Tx timestamps cannot be sampled when doing TSO */
if (first->tx_flags & ICE_TX_FLAGS_TSO)
return;
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.h b/drivers/net/ethernet/intel/ice/ice_txrx.h
index 407d4c320097..b28b9826bbcd 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.h
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.h
@@ -381,7 +381,6 @@ struct ice_tx_ring {
#define ICE_TX_FLAGS_RING_VLAN_L2TAG2 BIT(2)
u8 flags;
u8 dcb_tc; /* Traffic class of ring */
- u8 ptp_tx;
} ____cacheline_internodealigned_in_smp;
static inline bool ice_ring_uses_build_skb(struct ice_rx_ring *ring)
--
2.43.0

160
kernel-rt/debian/patches/ice-VDF/0008-ice-unify-logic-for-programming-PFINT_TSYN_MSK.patch Normal file
View File

@ -0,0 +1,160 @@
From 99007ca6255e2c35256bd97fa141705d301eb934 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:56 -0800
Subject: [PATCH 08/36] ice: unify logic for programming PFINT_TSYN_MSK
Commit d938a8cca88a ("ice: Auxbus devices & driver for E822 TS") modified
how Tx timestamps are handled for E822 devices. On these devices, only the
clock owner handles reading the Tx timestamp data from firmware. To do
this, the PFINT_TSYN_MSK register is modified from the default value to one
which enables reacting to a Tx timestamp on all PHY ports.
The driver currently programs PFINT_TSYN_MSK in different places depending
on whether the port is the clock owner or not. For the clock owner, the
PFINT_TSYN_MSK value is programmed during ice_ptp_init_owner just before
calling ice_ptp_tx_ena_intr to program the PHY ports.
For the non-clock owner ports, the PFINT_TSYN_MSK is programmed during
ice_ptp_init_port.
If a large enough device reset occurs, the PFINT_TSYN_MSK register will be
reset to the default value in which only the PHY associated directly with
the PF will cause the Tx timestamp interrupt to trigger.
The driver lacks logic to reprogram the PFINT_TSYN_MSK register after a
device reset. For the E822 device, this results in the PF no longer
responding to interrupts for other ports. This results in failure to
deliver Tx timestamps to user space applications.
Rename ice_ptp_configure_tx_tstamp to ice_ptp_cfg_tx_interrupt, and unify
the logic for programming PFINT_TSYN_MSK and PFINT_OICR_ENA into one place.
This function will program both registers according to the combination of
user configuration and device requirements.
This ensures that PFINT_TSYN_MSK is always restored when we configure the
Tx timestamp interrupt.
Fixes: d938a8cca88a ("ice: Auxbus devices & driver for E822 TS")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7d606a1e2d0575b6c3a2600f43f90d1e409f9661)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 60 ++++++++++++++----------
1 file changed, 34 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index c03153bdb7c3..b0bba866e8a2 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -256,21 +256,42 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
}
/**
- * ice_ptp_configure_tx_tstamp - Enable or disable Tx timestamp interrupt
- * @pf: The PF pointer to search in
- * @on: bool value for whether timestamp interrupt is enabled or disabled
+ * ice_ptp_cfg_tx_interrupt - Configure Tx timestamp interrupt for the device
+ * @pf: Board private structure
+ *
+ * Program the device to respond appropriately to the Tx timestamp interrupt
+ * cause.
*/
-static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
+static void ice_ptp_cfg_tx_interrupt(struct ice_pf *pf)
{
+ struct ice_hw *hw = &pf->hw;
+ bool enable;
u32 val;
+ switch (pf->ptp.tx_interrupt_mode) {
+ case ICE_PTP_TX_INTERRUPT_ALL:
+ /* React to interrupts across all quads. */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
+ enable = true;
+ break;
+ case ICE_PTP_TX_INTERRUPT_NONE:
+ /* Do not react to interrupts on any quad. */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
+ enable = false;
+ break;
+ case ICE_PTP_TX_INTERRUPT_SELF:
+ default:
+ enable = pf->ptp.tstamp_config.tx_type == HWTSTAMP_TX_ON;
+ break;
+ }
+
/* Configure the Tx timestamp interrupt */
- val = rd32(&pf->hw, PFINT_OICR_ENA);
- if (on)
+ val = rd32(hw, PFINT_OICR_ENA);
+ if (enable)
val |= PFINT_OICR_TSYN_TX_M;
else
val &= ~PFINT_OICR_TSYN_TX_M;
- wr32(&pf->hw, PFINT_OICR_ENA, val);
+ wr32(hw, PFINT_OICR_ENA, val);
}
/**
@@ -280,10 +301,9 @@ static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
*/
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
- ice_ptp_configure_tx_tstamp(pf, on);
-
pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
+
+ ice_ptp_cfg_tx_interrupt(pf);
}
/**
@@ -2909,15 +2929,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL) {
- /* The clock owner for this device type handles the timestamp
- * interrupt for all ports.
- */
- ice_ptp_configure_tx_tstamp(pf, true);
-
- /* React on all quads interrupts for E82x */
- wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
-
+ if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_tx_ena_intr(pf, true, itr);
if (err)
@@ -2991,13 +3003,6 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E822:
- /* Non-owner PFs don't react to any interrupts on E82x,
- * neither on own quad nor on others
- */
- if (!ice_ptp_pf_handles_tx_interrupt(pf)) {
- ice_ptp_configure_tx_tstamp(pf, false);
- wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
- }
kthread_init_delayed_work(&ptp_port->ov_work,
ice_ptp_wait_for_offsets);
@@ -3142,6 +3147,9 @@ void ice_ptp_init(struct ice_pf *pf)
/* Start the PHY timestamping block */
ice_ptp_reset_phy_timestamping(pf);
+ /* Configure initial Tx interrupt settings */
+ ice_ptp_cfg_tx_interrupt(pf);
+
set_bit(ICE_FLAG_PTP, pf->flags);
err = ice_ptp_init_work(pf, ptp);
if (err)
--
2.43.0

147
kernel-rt/debian/patches/ice-VDF/0009-ice-PTP-Clean-up-timestamp-registers-correctly.patch Normal file
View File

@ -0,0 +1,147 @@
From e5a65377977e338a8f7baf92892481acf1c62403 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:43 -0700
Subject: [PATCH 09/36] ice: PTP: Clean up timestamp registers correctly
E822 PHY TS registers should not be written and the only way to clean up
them is to reset QUAD memory.
To ensure that the status bit for the timestamp index is cleared, ensure
that ice_clear_phy_tstamp implementations first read the timestamp out.
Implementations which can write the register continue to do so.
Add a note to indicate this function should only be called on timestamps
which have their valid bit set. Update the dynamic debug messages to
reflect the actual action taken.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit be65a1a33bdee3912daac50aa6c5270ec9c37010)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 70 +++++++++++++--------
1 file changed, 45 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index e024b88ce32b..cd28430cfdda 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -759,29 +759,32 @@ ice_read_phy_tstamp_e822(struct ice_hw *hw, u8 quad, u8 idx, u64 *tstamp)
* @quad: the quad to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the PHY quad block that is
- * shared between the internal PHYs on the E822 devices.
+ * Read the timestamp out of the quad to clear its timestamp status bit from
+ * the PHY quad block that is shared between the internal PHYs of the E822
+ * devices.
+ *
+ * Note that unlike E810, software cannot directly write to the quad memory
+ * bank registers. E822 relies on the ice_get_phy_tx_tstamp_ready() function
+ * to determine which timestamps are valid. Reading a timestamp auto-clears
+ * the valid bit.
+ *
+ * To directly clear the contents of the timestamp block entirely, discarding
+ * all timestamp data at once, software should instead use
+ * ice_ptp_reset_ts_memory_quad_e822().
+ *
+ * This function should only be called on an idx whose bit is set according to
+ * ice_get_phy_tx_tstamp_ready().
*/
static int
ice_clear_phy_tstamp_e822(struct ice_hw *hw, u8 quad, u8 idx)
{
- u16 lo_addr, hi_addr;
+ u64 unused_tstamp;
int err;
- lo_addr = (u16)TS_L(Q_REG_TX_MEMORY_BANK_START, idx);
- hi_addr = (u16)TS_H(Q_REG_TX_MEMORY_BANK_START, idx);
-
- err = ice_write_quad_reg_e822(hw, quad, lo_addr, 0);
- if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register, err %d\n",
- err);
- return err;
- }
-
- err = ice_write_quad_reg_e822(hw, quad, hi_addr, 0);
+ err = ice_read_phy_tstamp_e822(hw, quad, idx, &unused_tstamp);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to read the timestamp register for quad %u, idx %u, err %d\n",
+ quad, idx, err);
return err;
}
@@ -2816,28 +2819,39 @@ ice_read_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx, u64 *tstamp)
* @lport: the lport to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the timestamp block of the
- * external PHY on the E810 device.
+ * Read the timestamp and then forcibly overwrite its value to clear the valid
+ * bit from the timestamp block of the external PHY on the E810 device.
+ *
+ * This function should only be called on an idx whose bit is set according to
+ * ice_get_phy_tx_tstamp_ready().
*/
static int ice_clear_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx)
{
u32 lo_addr, hi_addr;
+ u64 unused_tstamp;
int err;
+ err = ice_read_phy_tstamp_e810(hw, lport, idx, &unused_tstamp);
+ if (err) {
+ ice_debug(hw, ICE_DBG_PTP, "Failed to read the timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
+ return err;
+ }
+
lo_addr = TS_EXT(LOW_TX_MEMORY_BANK_START, lport, idx);
hi_addr = TS_EXT(HIGH_TX_MEMORY_BANK_START, lport, idx);
err = ice_write_phy_reg_e810(hw, lo_addr, 0);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
return err;
}
err = ice_write_phy_reg_e810(hw, hi_addr, 0);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
return err;
}
@@ -3519,9 +3533,15 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
* @block: the block to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the timestamp block. For
- * E822 devices, the block is the quad to clear from. For E810 devices, the
- * block is the logical port to clear from.
+ * Clear a timestamp from the timestamp block, discarding its value without
+ * returning it. This resets the memory status bit for the timestamp index
+ * allowing it to be reused for another timestamp in the future.
+ *
+ * For E822 devices, the block number is the PHY quad to clear from. For E810
+ * devices, the block number is the logical port to clear from.
+ *
+ * This function must only be called on a timestamp index whose valid bit is
+ * set according to ice_get_phy_tx_tstamp_ready().
*/
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx)
{
--
2.43.0

65
kernel-rt/debian/patches/ice-VDF/0010-ice-Use-PTP-auxbus-for-all-PHYs-restart-in-E822.patch Normal file
View File

@ -0,0 +1,65 @@
From e2a74a0a7dd399b0ee2ddd4889c609dedb85bfb5 Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:35 +0200
Subject: [PATCH 10/36] ice: Use PTP auxbus for all PHYs restart in E822
The E822 (and other devices based on the same PHY) is having issue while
setting the PHC timer - the PHY timers are drifting from the PHC. After
such a set all PHYs need to be restarted and resynchronised - do it
using auxiliary bus.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit af3c5c8748e6d286d4f2dd9800f9d27f29b8e2ef)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 24 +++++++++++++++++++++---
1 file changed, 21 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index b0bba866e8a2..42eb1418eb90 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1496,6 +1496,24 @@ static void ice_ptp_reset_phy_timestamping(struct ice_pf *pf)
ice_ptp_port_phy_restart(&pf->ptp.port);
}
+/**
+ * ice_ptp_restart_all_phy - Restart all PHYs to recalibrate timestamping
+ * @pf: Board private structure
+ */
+static void ice_ptp_restart_all_phy(struct ice_pf *pf)
+{
+ struct list_head *entry;
+
+ list_for_each(entry, &pf->ptp.ports_owner.ports) {
+ struct ice_ptp_port *port = list_entry(entry,
+ struct ice_ptp_port,
+ list_member);
+
+ if (port->link_up)
+ ice_ptp_port_phy_restart(port);
+ }
+}
+
/**
* ice_ptp_adjfine - Adjust clock increment rate
* @info: the driver's PTP info structure
@@ -1933,9 +1951,9 @@ ice_ptp_settime64(struct ptp_clock_info *info, const struct timespec64 *ts)
/* Reenable periodic outputs */
ice_ptp_enable_all_clkout(pf);
- /* Recalibrate and re-enable timestamp block */
- if (pf->ptp.port.link_up)
- ice_ptp_port_phy_restart(&pf->ptp.port);
+ /* Recalibrate and re-enable timestamp blocks for E822/E823 */
+ if (hw->phy_model == ICE_PHY_E822)
+ ice_ptp_restart_all_phy(pf);
exit:
if (err) {
dev_err(ice_pf_to_dev(pf), "PTP failed to set time %d\n", err);
--
2.43.0

2177
kernel-rt/debian/patches/ice-VDF/0011-ice-Rename-E822-to-E82X.patch Normal file
View File

File diff suppressed because it is too large Load Diff

104
kernel-rt/debian/patches/ice-VDF/0012-ice-Schedule-service-task-in-IRQ-top-half.patch Normal file
View File

@ -0,0 +1,104 @@
From 3b37119a08ffe4be182ade746a6b1fe3bcf65921 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 29 Nov 2023 13:40:22 +0100
Subject: [PATCH 12/36] ice: Schedule service task in IRQ top half
Schedule service task and EXTTS in the top half to avoid bottom half
scheduling if possible, which significantly reduces timestamping delay.
Co-developed-by: Michal Michalik <michal.michalik@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 00d50001444ef5c75c8ab476a6674708f3ff613b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 -
drivers/net/ethernet/intel/ice/ice_main.c | 20 +++++++++++---------
2 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 54a98c4032b7..efe78d5e4da1 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -517,7 +517,6 @@ enum ice_pf_flags {
};
enum ice_misc_thread_tasks {
- ICE_MISC_THREAD_EXTTS_EVENT,
ICE_MISC_THREAD_TX_TSTAMP,
ICE_MISC_THREAD_NBITS /* must be last */
};
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d2f3b4374d14..2acaa17a12bf 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3109,6 +3109,7 @@ static void ice_ena_misc_vector(struct ice_pf *pf)
static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
{
struct ice_pf *pf = (struct ice_pf *)data;
+ irqreturn_t ret = IRQ_HANDLED;
struct ice_hw *hw = &pf->hw;
struct device *dev;
u32 oicr, ena_mask;
@@ -3190,8 +3191,10 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (ice_ptp_pf_handles_tx_interrupt(pf))
+ if (ice_ptp_pf_handles_tx_interrupt(pf)) {
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
+ ret = IRQ_WAKE_THREAD;
+ }
}
if (oicr & PFINT_OICR_TSYN_EVNT_M) {
@@ -3207,7 +3210,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
GLTSYN_STAT_EVENT1_M |
GLTSYN_STAT_EVENT2_M);
- set_bit(ICE_MISC_THREAD_EXTTS_EVENT, pf->misc_thread);
+ ice_ptp_extts_event(pf);
}
}
@@ -3230,8 +3233,11 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
set_bit(ICE_PFR_REQ, pf->state);
}
}
+ ice_service_task_schedule(pf);
+ if (ret == IRQ_HANDLED)
+ ice_irq_dynamic_ena(hw, NULL, NULL);
- return IRQ_WAKE_THREAD;
+ return ret;
}
/**
@@ -3247,12 +3253,7 @@ static irqreturn_t ice_misc_intr_thread_fn(int __always_unused irq, void *data)
hw = &pf->hw;
if (ice_is_reset_in_progress(pf->state))
- return IRQ_HANDLED;
-
- ice_service_task_schedule(pf);
-
- if (test_and_clear_bit(ICE_MISC_THREAD_EXTTS_EVENT, pf->misc_thread))
- ice_ptp_extts_event(pf);
+ goto skip_irq;
if (test_and_clear_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread)) {
/* Process outstanding Tx timestamps. If there is more work,
@@ -3264,6 +3265,7 @@ static irqreturn_t ice_misc_intr_thread_fn(int __always_unused irq, void *data)
}
}
+skip_irq:
ice_irq_dynamic_ena(hw, NULL, NULL);
return IRQ_HANDLED;
--
2.43.0

660
kernel-rt/debian/patches/ice-VDF/0013-ice-Enable-SW-interrupt-from-FW-for-LL-TS.patch Normal file
View File

@ -0,0 +1,660 @@
From c4ab92eb3ee89178a012702f2a98477d683fad31 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 29 Nov 2023 13:40:23 +0100
Subject: [PATCH 13/36] ice: Enable SW interrupt from FW for LL TS
Introduce new capability - Low Latency Timestamping with Interrupt.
On supported devices, driver can request a single timestamp from FW
without polling the register afterwards. Instead, FW can issue
a dedicated interrupt when the timestamp was read from the PHY register
and its value is available to read from the register.
This eliminates the need of bottom half scheduling, which results in
minimal delay for timestamping.
For this mode, allocate TS indices sequentially, so that timestamps are
always completed in FIFO manner.
Co-developed-by: Yochai Hagvi <yochai.hagvi@intel.com>
Signed-off-by: Yochai Hagvi <yochai.hagvi@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 82e71b226e0ef770d7bc143701c8b4960b4eb3d5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 +
drivers/net/ethernet/intel/ice/ice_common.c | 3 +
.../net/ethernet/intel/ice/ice_hw_autogen.h | 2 +
drivers/net/ethernet/intel/ice/ice_main.c | 120 +++++++++++--
drivers/net/ethernet/intel/ice/ice_ptp.c | 163 ++++++++++++++++--
drivers/net/ethernet/intel/ice/ice_ptp.h | 9 +
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 1 +
drivers/net/ethernet/intel/ice/ice_type.h | 2 +
8 files changed, 274 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index efe78d5e4da1..ee42a504c2f4 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -594,6 +594,7 @@ struct ice_pf {
u32 hw_csum_rx_error;
u32 oicr_err_reg;
struct msi_map oicr_irq; /* Other interrupt cause MSIX vector */
+ struct msi_map ll_ts_irq; /* LL_TS interrupt MSIX vector */
u16 max_pf_txqs; /* Total Tx queues PF wide */
u16 max_pf_rxqs; /* Total Rx queues PF wide */
u16 num_lan_msix; /* Total MSIX vectors for base driver */
@@ -618,6 +619,7 @@ struct ice_pf {
unsigned long tx_timeout_last_recovery;
u32 tx_timeout_recovery_level;
char int_name[ICE_INT_NAME_STR_LEN];
+ char int_name_ll_ts[ICE_INT_NAME_STR_LEN];
struct auxiliary_device *adev;
int aux_idx;
u32 sw_int_count;
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 7674267a2d90..acf6ac00f804 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -2624,6 +2624,7 @@ ice_parse_1588_dev_caps(struct ice_hw *hw, struct ice_hw_dev_caps *dev_p,
info->tmr1_ena = ((number & ICE_TS_TMR1_ENA_M) != 0);
info->ts_ll_read = ((number & ICE_TS_LL_TX_TS_READ_M) != 0);
+ info->ts_ll_int_read = ((number & ICE_TS_LL_TX_TS_INT_READ_M) != 0);
info->ena_ports = logical_id;
info->tmr_own_map = phys_id;
@@ -2644,6 +2645,8 @@ ice_parse_1588_dev_caps(struct ice_hw *hw, struct ice_hw_dev_caps *dev_p,
info->tmr1_ena);
ice_debug(hw, ICE_DBG_INIT, "dev caps: ts_ll_read = %u\n",
info->ts_ll_read);
+ ice_debug(hw, ICE_DBG_INIT, "dev caps: ts_ll_int_read = %u\n",
+ info->ts_ll_int_read);
ice_debug(hw, ICE_DBG_INIT, "dev caps: ieee_1588 ena_ports = %u\n",
info->ena_ports);
ice_debug(hw, ICE_DBG_INIT, "dev caps: tmr_own_map = %u\n",
diff --git a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
index 6756f3d51d14..fa730bca7f15 100644
--- a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
+++ b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
@@ -200,6 +200,8 @@
#define GLINT_VECT2FUNC_PF_NUM_M ICE_M(0x7, 12)
#define GLINT_VECT2FUNC_IS_PF_S 16
#define GLINT_VECT2FUNC_IS_PF_M BIT(16)
+#define PFINT_ALLOC 0x001D2600
+#define PFINT_ALLOC_FIRST ICE_M(0x7FF, 0)
#define PFINT_FW_CTL 0x0016C800
#define PFINT_FW_CTL_MSIX_INDX_M ICE_M(0x7FF, 0)
#define PFINT_FW_CTL_ITR_INDX_S 11
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 2acaa17a12bf..9163a72368b3 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3071,6 +3071,7 @@ static int ice_xdp(struct net_device *dev, struct netdev_bpf *xdp)
static void ice_ena_misc_vector(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
+ u32 pf_intr_start_offset;
u32 val;
/* Disable anti-spoof detection interrupt to prevent spurious event
@@ -3099,6 +3100,47 @@ static void ice_ena_misc_vector(struct ice_pf *pf)
/* SW_ITR_IDX = 0, but don't change INTENA */
wr32(hw, GLINT_DYN_CTL(pf->oicr_irq.index),
GLINT_DYN_CTL_SW_ITR_INDX_M | GLINT_DYN_CTL_INTENA_MSK_M);
+
+ if (!pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ return;
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ wr32(hw, GLINT_DYN_CTL(pf->ll_ts_irq.index + pf_intr_start_offset),
+ GLINT_DYN_CTL_SW_ITR_INDX_M | GLINT_DYN_CTL_INTENA_MSK_M);
+}
+
+/**
+ * ice_ll_ts_intr - ll_ts interrupt handler
+ * @irq: interrupt number
+ * @data: pointer to a q_vector
+ */
+static irqreturn_t ice_ll_ts_intr(int __always_unused irq, void *data)
+{
+ struct ice_pf *pf = data;
+ u32 pf_intr_start_offset;
+ struct ice_ptp_tx *tx;
+ unsigned long flags;
+ struct ice_hw *hw;
+ u32 val;
+ u8 idx;
+
+ hw = &pf->hw;
+ tx = &pf->ptp.port.tx;
+ spin_lock_irqsave(&tx->lock, flags);
+ ice_ptp_complete_tx_single_tstamp(tx);
+
+ idx = find_next_bit_wrap(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx != tx->len)
+ ice_ptp_req_tx_single_tstamp(tx, idx);
+ spin_unlock_irqrestore(&tx->lock, flags);
+
+ val = GLINT_DYN_CTL_INTENA_M | GLINT_DYN_CTL_CLEARPBA_M |
+ (ICE_ITR_NONE << GLINT_DYN_CTL_ITR_INDX_S);
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ wr32(hw, GLINT_DYN_CTL(pf->ll_ts_irq.index + pf_intr_start_offset),
+ val);
+
+ return IRQ_HANDLED;
}
/**
@@ -3191,7 +3233,19 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (ice_ptp_pf_handles_tx_interrupt(pf)) {
+ if (ice_pf_state_is_nominal(pf) &&
+ pf->hw.dev_caps.ts_dev_info.ts_ll_int_read) {
+ struct ice_ptp_tx *tx = &pf->ptp.port.tx;
+ unsigned long flags;
+ u8 idx;
+
+ spin_lock_irqsave(&tx->lock, flags);
+ idx = find_next_bit_wrap(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx != tx->len)
+ ice_ptp_req_tx_single_tstamp(tx, idx);
+ spin_unlock_irqrestore(&tx->lock, flags);
+ } else if (ice_ptp_pf_handles_tx_interrupt(pf)) {
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
ret = IRQ_WAKE_THREAD;
}
@@ -3295,6 +3349,20 @@ static void ice_dis_ctrlq_interrupts(struct ice_hw *hw)
ice_flush(hw);
}
+/**
+ * ice_free_irq_msix_ll_ts- Unroll ll_ts vector setup
+ * @pf: board private structure
+ */
+static void ice_free_irq_msix_ll_ts(struct ice_pf *pf)
+{
+ int irq_num = pf->ll_ts_irq.virq;
+
+ synchronize_irq(irq_num);
+ devm_free_irq(ice_pf_to_dev(pf), irq_num, pf);
+
+ ice_free_irq(pf, pf->ll_ts_irq);
+}
+
/**
* ice_free_irq_msix_misc - Unroll misc vector setup
* @pf: board private structure
@@ -3314,6 +3382,8 @@ static void ice_free_irq_msix_misc(struct ice_pf *pf)
devm_free_irq(ice_pf_to_dev(pf), misc_irq_num, pf);
ice_free_irq(pf, pf->oicr_irq);
+ if (pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ ice_free_irq_msix_ll_ts(pf);
}
/**
@@ -3339,10 +3409,12 @@ static void ice_ena_ctrlq_interrupts(struct ice_hw *hw, u16 reg_idx)
PFINT_MBX_CTL_CAUSE_ENA_M);
wr32(hw, PFINT_MBX_CTL, val);
- /* This enables Sideband queue Interrupt causes */
- val = ((reg_idx & PFINT_SB_CTL_MSIX_INDX_M) |
- PFINT_SB_CTL_CAUSE_ENA_M);
- wr32(hw, PFINT_SB_CTL, val);
+ if (!hw->dev_caps.ts_dev_info.ts_ll_int_read) {
+ /* enable Sideband queue Interrupt causes */
+ val = ((reg_idx & PFINT_SB_CTL_MSIX_INDX_M) |
+ PFINT_SB_CTL_CAUSE_ENA_M);
+ wr32(hw, PFINT_SB_CTL, val);
+ }
ice_flush(hw);
}
@@ -3359,13 +3431,17 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
- struct msi_map oicr_irq;
+ u32 pf_intr_start_offset;
+ struct msi_map irq;
int err = 0;
if (!pf->int_name[0])
snprintf(pf->int_name, sizeof(pf->int_name) - 1, "%s-%s:misc",
dev_driver_string(dev), dev_name(dev));
+ if (!pf->int_name_ll_ts[0])
+ snprintf(pf->int_name_ll_ts, sizeof(pf->int_name_ll_ts) - 1,
+ "%s-%s:ll_ts", dev_driver_string(dev), dev_name(dev));
/* Do not request IRQ but do enable OICR interrupt since settings are
* lost during reset. Note that this function is called only during
* rebuild path and not while reset is in progress.
@@ -3374,11 +3450,11 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
goto skip_req_irq;
/* reserve one vector in irq_tracker for misc interrupts */
- oicr_irq = ice_alloc_irq(pf, false);
- if (oicr_irq.index < 0)
- return oicr_irq.index;
+ irq = ice_alloc_irq(pf, false);
+ if (irq.index < 0)
+ return irq.index;
- pf->oicr_irq = oicr_irq;
+ pf->oicr_irq = irq;
err = devm_request_threaded_irq(dev, pf->oicr_irq.virq, ice_misc_intr,
ice_misc_intr_thread_fn, 0,
pf->int_name, pf);
@@ -3389,10 +3465,34 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
return err;
}
+ /* reserve one vector in irq_tracker for ll_ts interrupt */
+ if (!pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ goto skip_req_irq;
+
+ irq = ice_alloc_irq(pf, false);
+ if (irq.index < 0)
+ return irq.index;
+
+ pf->ll_ts_irq = irq;
+ err = devm_request_irq(dev, pf->ll_ts_irq.virq, ice_ll_ts_intr, 0,
+ pf->int_name_ll_ts, pf);
+ if (err) {
+ dev_err(dev, "devm_request_irq for %s failed: %d\n",
+ pf->int_name_ll_ts, err);
+ ice_free_irq(pf, pf->ll_ts_irq);
+ return err;
+ }
+
skip_req_irq:
ice_ena_misc_vector(pf);
ice_ena_ctrlq_interrupts(hw, pf->oicr_irq.index);
+ /* This enables LL TS interrupt */
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ if (pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ wr32(hw, PFINT_SB_CTL,
+ ((pf->ll_ts_irq.index + pf_intr_start_offset) &
+ PFINT_SB_CTL_MSIX_INDX_M) | PFINT_SB_CTL_CAUSE_ENA_M);
wr32(hw, GLINT_ITR(ICE_RX_ITR, pf->oicr_irq.index),
ITR_REG_ALIGN(ICE_ITR_8K) >> ICE_ITR_GRAN_S);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 2e6e1fc84d11..75038d826f71 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -634,6 +634,119 @@ ice_ptp_is_tx_tracker_up(struct ice_ptp_tx *tx)
return tx->init && !tx->calibrating;
}
+/**
+ * ice_ptp_req_tx_single_tstamp - Request Tx timestamp for a port from FW
+ * @tx: the PTP Tx timestamp tracker
+ * @idx: index of the timestamp to request
+ */
+void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx)
+{
+ struct ice_ptp_port *ptp_port;
+ struct sk_buff *skb;
+ struct ice_pf *pf;
+
+ if (!tx->init)
+ return;
+
+ ptp_port = container_of(tx, struct ice_ptp_port, tx);
+ pf = ptp_port_to_pf(ptp_port);
+
+ /* Drop packets which have waited for more than 2 seconds */
+ if (time_is_before_jiffies(tx->tstamps[idx].start + 2 * HZ)) {
+ /* Count the number of Tx timestamps that timed out */
+ pf->ptp.tx_hwtstamp_timeouts++;
+
+ skb = tx->tstamps[idx].skb;
+ tx->tstamps[idx].skb = NULL;
+ clear_bit(idx, tx->in_use);
+
+ dev_kfree_skb_any(skb);
+ return;
+ }
+
+ ice_trace(tx_tstamp_fw_req, tx->tstamps[idx].skb, idx);
+
+ /* Write TS index to read to the PF register so the FW can read it */
+ wr32(&pf->hw, PF_SB_ATQBAL,
+ TS_LL_READ_TS_INTR | FIELD_PREP(TS_LL_READ_TS_IDX, idx) |
+ TS_LL_READ_TS);
+ tx->last_ll_ts_idx_read = idx;
+}
+
+/**
+ * ice_ptp_complete_tx_single_tstamp - Complete Tx timestamp for a port
+ * @tx: the PTP Tx timestamp tracker
+ */
+void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
+{
+ struct skb_shared_hwtstamps shhwtstamps = {};
+ u8 idx = tx->last_ll_ts_idx_read;
+ struct ice_ptp_port *ptp_port;
+ u64 raw_tstamp, tstamp;
+ bool drop_ts = false;
+ struct sk_buff *skb;
+ struct ice_pf *pf;
+ u32 val;
+
+ if (!tx->init || tx->last_ll_ts_idx_read < 0)
+ return;
+
+ ptp_port = container_of(tx, struct ice_ptp_port, tx);
+ pf = ptp_port_to_pf(ptp_port);
+
+ ice_trace(tx_tstamp_fw_done, tx->tstamps[idx].skb, idx);
+
+ val = rd32(&pf->hw, PF_SB_ATQBAL);
+
+ /* When the bit is cleared, the TS is ready in the register */
+ if (val & TS_LL_READ_TS) {
+ dev_err(ice_pf_to_dev(pf), "Failed to get the Tx tstamp - FW not ready");
+ return;
+ }
+
+ /* High 8 bit value of the TS is on the bits 16:23 */
+ raw_tstamp = FIELD_GET(TS_LL_READ_TS_HIGH, val);
+ raw_tstamp <<= 32;
+
+ /* Read the low 32 bit value */
+ raw_tstamp |= (u64)rd32(&pf->hw, PF_SB_ATQBAH);
+
+ /* For PHYs which don't implement a proper timestamp ready bitmap,
+ * verify that the timestamp value is different from the last cached
+ * timestamp. If it is not, skip this for now assuming it hasn't yet
+ * been captured by hardware.
+ */
+ if (!drop_ts && tx->verify_cached &&
+ raw_tstamp == tx->tstamps[idx].cached_tstamp)
+ return;
+
+ if (tx->verify_cached && raw_tstamp)
+ tx->tstamps[idx].cached_tstamp = raw_tstamp;
+ clear_bit(idx, tx->in_use);
+ skb = tx->tstamps[idx].skb;
+ tx->tstamps[idx].skb = NULL;
+ if (test_and_clear_bit(idx, tx->stale))
+ drop_ts = true;
+
+ if (!skb)
+ return;
+
+ if (drop_ts) {
+ dev_kfree_skb_any(skb);
+ return;
+ }
+
+ /* Extend the timestamp using cached PHC time */
+ tstamp = ice_ptp_extend_40b_ts(pf, raw_tstamp);
+ if (tstamp) {
+ shhwtstamps.hwtstamp = ns_to_ktime(tstamp);
+ ice_trace(tx_tstamp_complete, skb, idx);
+ }
+
+ skb_tstamp_tx(skb, &shhwtstamps);
+ dev_kfree_skb_any(skb);
+}
+
/**
* ice_ptp_process_tx_tstamp - Process Tx timestamps for a port
* @tx: the PTP Tx timestamp tracker
@@ -685,6 +798,7 @@ ice_ptp_is_tx_tracker_up(struct ice_ptp_tx *tx)
static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
{
struct ice_ptp_port *ptp_port;
+ unsigned long flags;
struct ice_pf *pf;
struct ice_hw *hw;
u64 tstamp_ready;
@@ -756,7 +870,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
drop_ts = true;
skip_ts_read:
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
if (tx->verify_cached && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
@@ -764,7 +878,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
tx->tstamps[idx].skb = NULL;
if (test_and_clear_bit(idx, tx->stale))
drop_ts = true;
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* It is unlikely but possible that the SKB will have been
* flushed at this point due to link change or teardown.
@@ -834,6 +948,7 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
static enum ice_tx_tstamp_work ice_ptp_tx_tstamp(struct ice_ptp_tx *tx)
{
bool more_timestamps;
+ unsigned long flags;
if (!tx->init)
return ICE_TX_TSTAMP_WORK_DONE;
@@ -842,9 +957,9 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp(struct ice_ptp_tx *tx)
ice_ptp_process_tx_tstamp(tx);
/* Check if there are outstanding Tx timestamps */
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
more_timestamps = tx->init && !bitmap_empty(tx->in_use, tx->len);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
if (more_timestamps)
return ICE_TX_TSTAMP_WORK_PENDING;
@@ -881,6 +996,7 @@ ice_ptp_alloc_tx_tracker(struct ice_ptp_tx *tx)
tx->in_use = in_use;
tx->stale = stale;
tx->init = 1;
+ tx->last_ll_ts_idx_read = -1;
spin_lock_init(&tx->lock);
@@ -898,6 +1014,7 @@ static void
ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
{
struct ice_hw *hw = &pf->hw;
+ unsigned long flags;
u64 tstamp_ready;
int err;
u8 idx;
@@ -921,12 +1038,12 @@ ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
if (!hw->reset_ongoing && (tstamp_ready & BIT_ULL(phy_idx)))
ice_clear_phy_tstamp(hw, tx->block, phy_idx);
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
skb = tx->tstamps[idx].skb;
tx->tstamps[idx].skb = NULL;
clear_bit(idx, tx->in_use);
clear_bit(idx, tx->stale);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* Count the number of Tx timestamps flushed */
pf->ptp.tx_hwtstamp_flushed++;
@@ -950,9 +1067,11 @@ ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
static void
ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
{
- spin_lock(&tx->lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&tx->lock, flags);
bitmap_or(tx->stale, tx->stale, tx->in_use, tx->len);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
}
/**
@@ -965,9 +1084,11 @@ ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
static void
ice_ptp_release_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
{
- spin_lock(&tx->lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&tx->lock, flags);
tx->init = 0;
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* wait for potentially outstanding interrupt to complete */
synchronize_irq(pf->oicr_irq.virq);
@@ -1367,6 +1488,7 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
struct ice_pf *pf = ptp_port_to_pf(ptp_port);
u8 port = ptp_port->port_num;
struct ice_hw *hw = &pf->hw;
+ unsigned long flags;
int err;
if (ice_is_e810(hw))
@@ -1380,9 +1502,9 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
kthread_cancel_delayed_work_sync(&ptp_port->ov_work);
/* temporarily disable Tx timestamps while calibrating PHY offset */
- spin_lock(&ptp_port->tx.lock);
+ spin_lock_irqsave(&ptp_port->tx.lock, flags);
ptp_port->tx.calibrating = true;
- spin_unlock(&ptp_port->tx.lock);
+ spin_unlock_irqrestore(&ptp_port->tx.lock, flags);
ptp_port->tx_fifo_busy_cnt = 0;
/* Start the PHY timer in Vernier mode */
@@ -1391,9 +1513,9 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
goto out_unlock;
/* Enable Tx timestamps right away */
- spin_lock(&ptp_port->tx.lock);
+ spin_lock_irqsave(&ptp_port->tx.lock, flags);
ptp_port->tx.calibrating = false;
- spin_unlock(&ptp_port->tx.lock);
+ spin_unlock_irqrestore(&ptp_port->tx.lock, flags);
kthread_queue_delayed_work(pf->ptp.kworker, &ptp_port->ov_work, 0);
@@ -2471,18 +2593,23 @@ static long ice_ptp_create_clock(struct ice_pf *pf)
*/
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
{
+ unsigned long flags;
u8 idx;
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
/* Check that this tracker is accepting new timestamp requests */
if (!ice_ptp_is_tx_tracker_up(tx)) {
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
return -1;
}
/* Find and set the first available index */
- idx = find_first_zero_bit(tx->in_use, tx->len);
+ idx = find_next_zero_bit(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx == tx->len)
+ idx = find_first_zero_bit(tx->in_use, tx->len);
+
if (idx < tx->len) {
/* We got a valid index that no other thread could have set. Store
* a reference to the skb and the start time to allow discarding old
@@ -2496,7 +2623,7 @@ s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
ice_trace(tx_tstamp_request, skb, idx);
}
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* return the appropriate PHY timestamp register index, -1 if no
* indexes were available.
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index a3ae008a3539..64679d3d2c49 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -131,6 +131,7 @@ enum ice_tx_tstamp_work {
* @calibrating: if true, the PHY is calibrating the Tx offset. During this
* window, timestamps are temporarily disabled.
* @verify_cached: if true, verify new timestamp differs from last read value
+ * @last_ll_ts_idx_read: index of the last LL TS read by the FW
*/
struct ice_ptp_tx {
spinlock_t lock; /* lock protecting in_use bitmap */
@@ -143,6 +144,7 @@ struct ice_ptp_tx {
u8 init : 1;
u8 calibrating : 1;
u8 verify_cached : 1;
+ s8 last_ll_ts_idx_read;
};
/* Quad and port information for initializing timestamp blocks */
@@ -296,6 +298,8 @@ int ice_get_ptp_clock_index(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
+void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx);
+void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx);
enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
@@ -330,6 +334,11 @@ ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
return -1;
}
+static inline void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx)
+{ }
+
+static inline void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx) { }
+
static inline bool ice_ptp_process_ts(struct ice_pf *pf)
{
return true;
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index 0cc285614c72..7e8fd369ef7c 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -509,6 +509,7 @@ int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
#define TS_LL_READ_RETRIES 200
#define TS_LL_READ_TS_HIGH GENMASK(23, 16)
#define TS_LL_READ_TS_IDX GENMASK(29, 24)
+#define TS_LL_READ_TS_INTR BIT(30)
#define TS_LL_READ_TS BIT(31)
/* Internal PHY timestamp address */
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index f8b658386552..b0f1f4db1d8b 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -350,6 +350,7 @@ struct ice_ts_func_info {
#define ICE_TS_TMR0_ENA_M BIT(25)
#define ICE_TS_TMR1_ENA_M BIT(26)
#define ICE_TS_LL_TX_TS_READ_M BIT(28)
+#define ICE_TS_LL_TX_TS_INT_READ_M BIT(29)
struct ice_ts_dev_info {
/* Device specific info */
@@ -363,6 +364,7 @@ struct ice_ts_dev_info {
u8 tmr0_ena;
u8 tmr1_ena;
u8 ts_ll_read;
+ u8 ts_ll_int_read;
};
/* Function specific capabilities */
--
2.43.0

310
kernel-rt/debian/patches/ice-VDF/0014-ice-PTP-add-clock-domain-number-to-auxiliary-interfa.patch Normal file
View File

@ -0,0 +1,310 @@
From f267daca86600496d536f85c4d1945558b982427 Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:36 +0200
Subject: [PATCH 14/36] ice: PTP: add clock domain number to auxiliary
interface
The PHC clock id used to be moved between PFs using FW admin queue
shared parameters - move the implementation to auxiliary bus.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit fcd2c1e3139a27766ef263bd2011195dbc8a79f5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 5 -
drivers/net/ethernet/intel/ice/ice_ethtool.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 163 +++---------------
drivers/net/ethernet/intel/ice/ice_ptp.h | 11 +-
4 files changed, 34 insertions(+), 147 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 353ac55bdb9d..9bacb69ead8c 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2360,11 +2360,6 @@ struct ice_aqc_driver_shared_params {
};
enum ice_aqc_driver_params {
- /* OS clock index for PTP timer Domain 0 */
- ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0 = 0,
- /* OS clock index for PTP timer Domain 1 */
- ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1,
-
/* Add new parameters above */
ICE_AQC_DRIVER_PARAM_MAX = 16,
};
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 456cf4785c74..057453d589d5 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -3286,7 +3286,7 @@ ice_get_ts_info(struct net_device *dev, struct ethtool_ts_info *info)
SOF_TIMESTAMPING_RX_HARDWARE |
SOF_TIMESTAMPING_RAW_HARDWARE;
- info->phc_index = ice_get_ptp_clock_index(pf);
+ info->phc_index = ice_ptp_clock_index(pf);
info->tx_types = BIT(HWTSTAMP_TX_OFF) | BIT(HWTSTAMP_TX_ON);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 75038d826f71..a2d0da7dfe83 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -345,131 +345,6 @@ void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena)
ice_set_rx_tstamp(pf, ena);
}
-/**
- * ice_get_ptp_clock_index - Get the PTP clock index
- * @pf: the PF pointer
- *
- * Determine the clock index of the PTP clock associated with this device. If
- * this is the PF controlling the clock, just use the local access to the
- * clock device pointer.
- *
- * Otherwise, read from the driver shared parameters to determine the clock
- * index value.
- *
- * Returns: the index of the PTP clock associated with this device, or -1 if
- * there is no associated clock.
- */
-int ice_get_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- u32 value;
- int err;
-
- /* Use the ptp_clock structure if we're the main PF */
- if (pf->ptp.clock)
- return ptp_clock_index(pf->ptp.clock);
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- err = ice_aq_get_driver_param(hw, param_idx, &value, NULL);
- if (err) {
- dev_err(dev, "Failed to read PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- return -1;
- }
-
- /* The PTP clock index is an integer, and will be between 0 and
- * INT_MAX. The highest bit of the driver shared parameter is used to
- * indicate whether or not the currently stored clock index is valid.
- */
- if (!(value & PTP_SHARED_CLK_IDX_VALID))
- return -1;
-
- return value & ~PTP_SHARED_CLK_IDX_VALID;
-}
-
-/**
- * ice_set_ptp_clock_index - Set the PTP clock index
- * @pf: the PF pointer
- *
- * Set the PTP clock index for this device into the shared driver parameters,
- * so that other PFs associated with this device can read it.
- *
- * If the PF is unable to store the clock index, it will log an error, but
- * will continue operating PTP.
- */
-static void ice_set_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- u32 value;
- int err;
-
- if (!pf->ptp.clock)
- return;
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- value = (u32)ptp_clock_index(pf->ptp.clock);
- if (value > INT_MAX) {
- dev_err(dev, "PTP Clock index is too large to store\n");
- return;
- }
- value |= PTP_SHARED_CLK_IDX_VALID;
-
- err = ice_aq_set_driver_param(hw, param_idx, value, NULL);
- if (err) {
- dev_err(dev, "Failed to set PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- }
-}
-
-/**
- * ice_clear_ptp_clock_index - Clear the PTP clock index
- * @pf: the PF pointer
- *
- * Clear the PTP clock index for this device. Must be called when
- * unregistering the PTP clock, in order to ensure other PFs stop reporting
- * a clock object that no longer exists.
- */
-static void ice_clear_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- int err;
-
- /* Do not clear the index if we don't own the timer */
- if (!ice_pf_src_tmr_owned(pf))
- return;
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- err = ice_aq_set_driver_param(hw, param_idx, 0, NULL);
- if (err) {
- dev_dbg(dev, "Failed to clear PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- }
-}
-
/**
* ice_ptp_read_src_clk_reg - Read the source clock register
* @pf: Board private structure
@@ -2564,7 +2439,6 @@ static void ice_ptp_set_caps(struct ice_pf *pf)
static long ice_ptp_create_clock(struct ice_pf *pf)
{
struct ptp_clock_info *info;
- struct ptp_clock *clock;
struct device *dev;
/* No need to create a clock device if we already have one */
@@ -2577,11 +2451,11 @@ static long ice_ptp_create_clock(struct ice_pf *pf)
dev = ice_pf_to_dev(pf);
/* Attempt to register the clock before enabling the hardware. */
- clock = ptp_clock_register(info, dev);
- if (IS_ERR(clock))
- return PTR_ERR(clock);
-
- pf->ptp.clock = clock;
+ pf->ptp.clock = ptp_clock_register(info, dev);
+ if (IS_ERR(pf->ptp.clock)) {
+ dev_err(ice_pf_to_dev(pf), "Failed to register PTP clock device");
+ return PTR_ERR(pf->ptp.clock);
+ }
return 0;
}
@@ -2996,6 +2870,28 @@ static void ice_ptp_unregister_auxbus_driver(struct ice_pf *pf)
mutex_destroy(&pf->ptp.ports_owner.lock);
}
+/**
+ * ice_ptp_clock_index - Get the PTP clock index for this device
+ * @pf: Board private structure
+ *
+ * Returns: the PTP clock index associated with this PF, or -1 if no PTP clock
+ * is associated.
+ */
+int ice_ptp_clock_index(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev;
+ struct ice_pf *owner_pf;
+ struct ptp_clock *clock;
+
+ aux_dev = &pf->ptp.port.aux_dev;
+ owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ if (!owner_pf)
+ return -1;
+ clock = owner_pf->ptp.clock;
+
+ return clock ? ptp_clock_index(clock) : -1;
+}
+
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
@@ -3086,9 +2982,6 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
if (err)
goto err_clk;
- /* Store the PTP clock index for other PFs */
- ice_set_ptp_clock_index(pf);
-
err = ice_ptp_register_auxbus_driver(pf);
if (err) {
dev_err(ice_pf_to_dev(pf), "Failed to register PTP auxbus driver");
@@ -3097,7 +2990,6 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
return 0;
err_aux:
- ice_clear_ptp_clock_index(pf);
ptp_clock_unregister(pf->ptp.clock);
err_clk:
pf->ptp.clock = NULL;
@@ -3353,7 +3245,6 @@ void ice_ptp_release(struct ice_pf *pf)
/* Disable periodic outputs */
ice_ptp_disable_all_clkout(pf);
- ice_clear_ptp_clock_index(pf);
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 64679d3d2c49..95ebd7a048ec 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -290,11 +290,11 @@ struct ice_ptp {
#define ETH_GLTSYN_ENA(_i) (0x03000348 + ((_i) * 4))
#if IS_ENABLED(CONFIG_PTP_1588_CLOCK)
+int ice_ptp_clock_index(struct ice_pf *pf);
struct ice_pf;
int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr);
int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr);
void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena);
-int ice_get_ptp_clock_index(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
@@ -322,10 +322,6 @@ static inline int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
}
static inline void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena) { }
-static inline int ice_get_ptp_clock_index(struct ice_pf *pf)
-{
- return -1;
-}
static inline void ice_ptp_extts_event(struct ice_pf *pf) { }
static inline s8
@@ -353,5 +349,10 @@ static inline void ice_ptp_release(struct ice_pf *pf) { }
static inline void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
{
}
+
+static inline int ice_ptp_clock_index(struct ice_pf *pf)
+{
+ return -1;
+}
#endif /* IS_ENABLED(CONFIG_PTP_1588_CLOCK) */
#endif /* _ICE_PTP_H_ */
--
2.43.0

266
kernel-rt/debian/patches/ice-VDF/0015-ice-restore-timestamp-configuration-after-device-res.patch Normal file
View File

@ -0,0 +1,266 @@
From eb63973adae478fdcc324f5490d6803646f0cc76 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:57 -0800
Subject: [PATCH 15/36] ice: restore timestamp configuration after device reset
The driver calls ice_ptp_cfg_timestamp() during ice_ptp_prepare_for_reset()
to disable timestamping while the device is resetting. This operation
destroys the user requested configuration. While the driver does call
ice_ptp_cfg_timestamp in ice_rebuild() to restore some hardware settings
after a reset, it unconditionally passes true or false, resulting in
failure to restore previous user space configuration.
This results in a device reset forcibly disabling timestamp configuration
regardless of current user settings.
This was not detected previously due to a quirk of the LinuxPTP ptp4l
application. If ptp4l detects a missing timestamp, it enters a fault state
and performs recovery logic which includes executing SIOCSHWTSTAMP again,
restoring the now accidentally cleared configuration.
Not every application does this, and for these applications, timestamps
will mysteriously stop after a PF reset, without being restored until an
application restart.
Fix this by replacing ice_ptp_cfg_timestamp() with two new functions:
1) ice_ptp_disable_timestamp_mode() which unconditionally disables the
timestamping logic in ice_ptp_prepare_for_reset() and ice_ptp_release()
2) ice_ptp_restore_timestamp_mode() which calls
ice_ptp_restore_tx_interrupt() to restore Tx timestamping configuration,
calls ice_set_rx_tstamp() to restore Rx timestamping configuration, and
issues an immediate TSYN_TX interrupt to ensure that timestamps which
may have occurred during the device reset get processed.
Modify the ice_ptp_set_timestamp_mode to directly save the user
configuration and then call ice_ptp_restore_timestamp_mode. This way, reset
no longer destroys the saved user configuration.
This obsoletes the ice_set_tx_tstamp() function which can now be safely
removed.
With this change, all devices should now restore Tx and Rx timestamping
functionality correctly after a PF reset without application intervention.
Fixes: 77a781155a65 ("ice: enable receive hardware timestamping")
Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7758017911a4f2578d54c318e8fe77bcb5899054)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 12 +---
drivers/net/ethernet/intel/ice/ice_ptp.c | 74 ++++++++++++++---------
drivers/net/ethernet/intel/ice/ice_ptp.h | 5 +-
3 files changed, 51 insertions(+), 40 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 9163a72368b3..8cfb923198e9 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7545,15 +7545,6 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
goto err_vsi_rebuild;
}
- /* configure PTP timestamping after VSI rebuild */
- if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) {
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
- ice_ptp_cfg_timestamp(pf, false);
- else if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL)
- /* for E82x PHC owner always need to have interrupts */
- ice_ptp_cfg_timestamp(pf, true);
- }
-
err = ice_vsi_rebuild_by_type(pf, ICE_VSI_SWITCHDEV_CTRL);
if (err) {
dev_err(dev, "Switchdev CTRL VSI rebuild failed: %d\n", err);
@@ -7605,6 +7596,9 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
ice_plug_aux_dev(pf);
if (ice_is_feature_supported(pf, ICE_F_SRIOV_LAG))
ice_lag_rebuild(pf);
+
+ /* Restore timestamp mode settings after VSI rebuild */
+ ice_ptp_restore_timestamp_mode(pf);
return;
err_vsi_rebuild:
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index a2d0da7dfe83..8fc6905b0f79 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -294,18 +294,6 @@ static void ice_ptp_cfg_tx_interrupt(struct ice_pf *pf)
wr32(hw, PFINT_OICR_ENA, val);
}
-/**
- * ice_set_tx_tstamp - Enable or disable Tx timestamping
- * @pf: The PF pointer to search in
- * @on: bool value for whether timestamps are enabled or disabled
- */
-static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
-{
- pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
-
- ice_ptp_cfg_tx_interrupt(pf);
-}
-
/**
* ice_set_rx_tstamp - Enable or disable Rx timestamping
* @pf: The PF pointer to search in
@@ -317,7 +305,7 @@ static void ice_set_rx_tstamp(struct ice_pf *pf, bool on)
u16 i;
vsi = ice_get_main_vsi(pf);
- if (!vsi)
+ if (!vsi || !vsi->rx_rings)
return;
/* Set the timestamp flag for all the Rx rings */
@@ -326,23 +314,50 @@ static void ice_set_rx_tstamp(struct ice_pf *pf, bool on)
continue;
vsi->rx_rings[i]->ptp_rx = on;
}
+}
+
+/**
+ * ice_ptp_disable_timestamp_mode - Disable current timestamp mode
+ * @pf: Board private structure
+ *
+ * Called during preparation for reset to temporarily disable timestamping on
+ * the device. Called during remove to disable timestamping while cleaning up
+ * driver resources.
+ */
+static void ice_ptp_disable_timestamp_mode(struct ice_pf *pf)
+{
+ struct ice_hw *hw = &pf->hw;
+ u32 val;
+
+ val = rd32(hw, PFINT_OICR_ENA);
+ val &= ~PFINT_OICR_TSYN_TX_M;
+ wr32(hw, PFINT_OICR_ENA, val);
- pf->ptp.tstamp_config.rx_filter = on ? HWTSTAMP_FILTER_ALL :
- HWTSTAMP_FILTER_NONE;
+ ice_set_rx_tstamp(pf, false);
}
/**
- * ice_ptp_cfg_timestamp - Configure timestamp for init/deinit
+ * ice_ptp_restore_timestamp_mode - Restore timestamp configuration
* @pf: Board private structure
- * @ena: bool value to enable or disable time stamp
*
- * This function will configure timestamping during PTP initialization
- * and deinitialization
+ * Called at the end of rebuild to restore timestamp configuration after
+ * a device reset.
*/
-void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena)
+void ice_ptp_restore_timestamp_mode(struct ice_pf *pf)
{
- ice_set_tx_tstamp(pf, ena);
- ice_set_rx_tstamp(pf, ena);
+ struct ice_hw *hw = &pf->hw;
+ bool enable_rx;
+
+ ice_ptp_cfg_tx_interrupt(pf);
+
+ enable_rx = pf->ptp.tstamp_config.rx_filter == HWTSTAMP_FILTER_ALL;
+ ice_set_rx_tstamp(pf, enable_rx);
+
+ /* Trigger an immediate software interrupt to ensure that timestamps
+ * which occurred during reset are handled now.
+ */
+ wr32(hw, PFINT_OICR, PFINT_OICR_TSYN_TX_M);
+ ice_flush(hw);
}
/**
@@ -2152,10 +2167,10 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
{
switch (config->tx_type) {
case HWTSTAMP_TX_OFF:
- ice_set_tx_tstamp(pf, false);
+ pf->ptp.tstamp_config.tx_type = HWTSTAMP_TX_OFF;
break;
case HWTSTAMP_TX_ON:
- ice_set_tx_tstamp(pf, true);
+ pf->ptp.tstamp_config.tx_type = HWTSTAMP_TX_ON;
break;
default:
return -ERANGE;
@@ -2163,7 +2178,7 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
switch (config->rx_filter) {
case HWTSTAMP_FILTER_NONE:
- ice_set_rx_tstamp(pf, false);
+ pf->ptp.tstamp_config.rx_filter = HWTSTAMP_FILTER_NONE;
break;
case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
@@ -2179,12 +2194,15 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
case HWTSTAMP_FILTER_NTP_ALL:
case HWTSTAMP_FILTER_ALL:
- ice_set_rx_tstamp(pf, true);
+ pf->ptp.tstamp_config.rx_filter = HWTSTAMP_FILTER_ALL;
break;
default:
return -ERANGE;
}
+ /* Immediately update the device timestamping mode */
+ ice_ptp_restore_timestamp_mode(pf);
+
return 0;
}
@@ -2904,7 +2922,7 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
clear_bit(ICE_FLAG_PTP, pf->flags);
/* Disable timestamping for both Tx and Rx */
- ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_disable_timestamp_mode(pf);
kthread_cancel_delayed_work_sync(&ptp->work);
@@ -3222,7 +3240,7 @@ void ice_ptp_release(struct ice_pf *pf)
return;
/* Disable timestamping for both Tx and Rx */
- ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_disable_timestamp_mode(pf);
ice_ptp_remove_auxbus_device(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 95ebd7a048ec..130e6d2ae9a5 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -294,7 +294,7 @@ int ice_ptp_clock_index(struct ice_pf *pf);
struct ice_pf;
int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr);
int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr);
-void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena);
+void ice_ptp_restore_timestamp_mode(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
@@ -321,8 +321,7 @@ static inline int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
return -EOPNOTSUPP;
}
-static inline void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena) { }
-
+static inline void ice_ptp_restore_timestamp_mode(struct ice_pf *pf) { }
static inline void ice_ptp_extts_event(struct ice_pf *pf) { }
static inline s8
ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
--
2.43.0

316
kernel-rt/debian/patches/ice-VDF/0016-ice-introduce-PTP-state-machine.patch Normal file
View File

@ -0,0 +1,316 @@
From 5c6115d27a377927d6392b3bfbe9739188c8153c Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:49 -0800
Subject: [PATCH 16/36] ice: introduce PTP state machine
Add PTP state machine so that the driver can correctly identify PTP
state around resets.
When the driver got information about ungraceful reset, PTP was not
prepared for reset and it returned error. When this situation occurs,
prepare PTP before rebuilding its structures.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Co-developed-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 8293e4cb2ff54b1ec4f7206dcb74c908f62a3fb8)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 -
drivers/net/ethernet/intel/ice/ice_ethtool.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 110 +++++++++++--------
drivers/net/ethernet/intel/ice/ice_ptp.h | 10 ++
4 files changed, 74 insertions(+), 49 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index ee42a504c2f4..3278d032a2bd 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -492,7 +492,6 @@ enum ice_pf_flags {
ICE_FLAG_DCB_ENA,
ICE_FLAG_FD_ENA,
ICE_FLAG_PTP_SUPPORTED, /* PTP is supported by NVM */
- ICE_FLAG_PTP, /* PTP is enabled by software */
ICE_FLAG_ADV_FEATURES,
ICE_FLAG_TC_MQPRIO, /* support for Multi queue TC */
ICE_FLAG_CLS_FLOWER,
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 057453d589d5..9e949c493c38 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -3276,7 +3276,7 @@ ice_get_ts_info(struct net_device *dev, struct ethtool_ts_info *info)
struct ice_pf *pf = ice_netdev_to_pf(dev);
/* only report timestamping if PTP is enabled */
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return ethtool_op_get_ts_info(dev, info);
info->so_timestamping = SOF_TIMESTAMPING_TX_SOFTWARE |
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 8fc6905b0f79..36c81c5ee83b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1430,7 +1430,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
struct ice_ptp_port *ptp_port;
struct ice_hw *hw = &pf->hw;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
if (WARN_ON_ONCE(port >= ICE_NUM_EXTERNAL_PORTS))
@@ -2148,7 +2148,7 @@ int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
{
struct hwtstamp_config *config;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return -EIO;
config = &pf->ptp.tstamp_config;
@@ -2218,7 +2218,7 @@ int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr)
struct hwtstamp_config config;
int err;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return -EAGAIN;
if (copy_from_user(&config, ifr->ifr_data, sizeof(config)))
@@ -2606,7 +2606,7 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
struct ice_pf *pf = container_of(ptp, struct ice_pf, ptp);
int err;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
err = ice_ptp_update_cached_phctime(pf);
@@ -2618,6 +2618,42 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
msecs_to_jiffies(err ? 10 : 500));
}
+/**
+ * ice_ptp_prepare_for_reset - Prepare PTP for reset
+ * @pf: Board private structure
+ */
+void ice_ptp_prepare_for_reset(struct ice_pf *pf)
+{
+ struct ice_ptp *ptp = &pf->ptp;
+ u8 src_tmr;
+
+ if (ptp->state != ICE_PTP_READY)
+ return;
+
+ ptp->state = ICE_PTP_RESETTING;
+
+ /* Disable timestamping for both Tx and Rx */
+ ice_ptp_disable_timestamp_mode(pf);
+
+ kthread_cancel_delayed_work_sync(&ptp->work);
+
+ if (test_bit(ICE_PFR_REQ, pf->state))
+ return;
+
+ ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
+
+ /* Disable periodic outputs */
+ ice_ptp_disable_all_clkout(pf);
+
+ src_tmr = ice_get_ptp_src_clock_index(&pf->hw);
+
+ /* Disable source clock */
+ wr32(&pf->hw, GLTSYN_ENA(src_tmr), (u32)~GLTSYN_ENA_TSYN_ENA_M);
+
+ /* Acquire PHC and system timer to restore after reset */
+ ptp->reset_time = ktime_get_real_ns();
+}
+
/**
* ice_ptp_reset - Initialize PTP hardware clock support after reset
* @pf: Board private structure
@@ -2630,6 +2666,14 @@ void ice_ptp_reset(struct ice_pf *pf)
int err, itr = 1;
u64 time_diff;
+ if (ptp->state == ICE_PTP_READY) {
+ ice_ptp_prepare_for_reset(pf);
+ } else if (ptp->state != ICE_PTP_RESETTING) {
+ err = -EINVAL;
+ dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
+ goto err;
+ }
+
if (test_bit(ICE_PFR_REQ, pf->state) ||
!ice_pf_src_tmr_owned(pf))
goto pfr;
@@ -2690,7 +2734,7 @@ void ice_ptp_reset(struct ice_pf *pf)
if (err)
goto err;
- set_bit(ICE_FLAG_PTP, pf->flags);
+ ptp->state = ICE_PTP_READY;
/* Restart the PHY timestamping block */
if (!test_bit(ICE_PFR_REQ, pf->state) &&
@@ -2704,6 +2748,7 @@ void ice_ptp_reset(struct ice_pf *pf)
return;
err:
+ ptp->state = ICE_PTP_ERROR;
dev_err(ice_pf_to_dev(pf), "PTP reset failed %d\n", err);
}
@@ -2910,39 +2955,6 @@ int ice_ptp_clock_index(struct ice_pf *pf)
return clock ? ptp_clock_index(clock) : -1;
}
-/**
- * ice_ptp_prepare_for_reset - Prepare PTP for reset
- * @pf: Board private structure
- */
-void ice_ptp_prepare_for_reset(struct ice_pf *pf)
-{
- struct ice_ptp *ptp = &pf->ptp;
- u8 src_tmr;
-
- clear_bit(ICE_FLAG_PTP, pf->flags);
-
- /* Disable timestamping for both Tx and Rx */
- ice_ptp_disable_timestamp_mode(pf);
-
- kthread_cancel_delayed_work_sync(&ptp->work);
-
- if (test_bit(ICE_PFR_REQ, pf->state))
- return;
-
- ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
-
- /* Disable periodic outputs */
- ice_ptp_disable_all_clkout(pf);
-
- src_tmr = ice_get_ptp_src_clock_index(&pf->hw);
-
- /* Disable source clock */
- wr32(&pf->hw, GLTSYN_ENA(src_tmr), (u32)~GLTSYN_ENA_TSYN_ENA_M);
-
- /* Acquire PHC and system timer to restore after reset */
- ptp->reset_time = ktime_get_real_ns();
-}
-
/**
* ice_ptp_init_owner - Initialize PTP_1588_CLOCK device
* @pf: Board private structure
@@ -3181,6 +3193,8 @@ void ice_ptp_init(struct ice_pf *pf)
struct ice_hw *hw = &pf->hw;
int err;
+ ptp->state = ICE_PTP_INITIALIZING;
+
ice_ptp_init_phy_model(hw);
ice_ptp_init_tx_interrupt_mode(pf);
@@ -3205,12 +3219,13 @@ void ice_ptp_init(struct ice_pf *pf)
/* Configure initial Tx interrupt settings */
ice_ptp_cfg_tx_interrupt(pf);
- set_bit(ICE_FLAG_PTP, pf->flags);
- err = ice_ptp_init_work(pf, ptp);
+ err = ice_ptp_create_auxbus_device(pf);
if (err)
goto err;
- err = ice_ptp_create_auxbus_device(pf);
+ ptp->state = ICE_PTP_READY;
+
+ err = ice_ptp_init_work(pf, ptp);
if (err)
goto err;
@@ -3223,7 +3238,7 @@ void ice_ptp_init(struct ice_pf *pf)
ptp_clock_unregister(ptp->clock);
pf->ptp.clock = NULL;
}
- clear_bit(ICE_FLAG_PTP, pf->flags);
+ ptp->state = ICE_PTP_ERROR;
dev_err(ice_pf_to_dev(pf), "PTP failed %d\n", err);
}
@@ -3236,9 +3251,11 @@ void ice_ptp_init(struct ice_pf *pf)
*/
void ice_ptp_release(struct ice_pf *pf)
{
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
+ pf->ptp.state = ICE_PTP_UNINIT;
+
/* Disable timestamping for both Tx and Rx */
ice_ptp_disable_timestamp_mode(pf);
@@ -3246,8 +3263,6 @@ void ice_ptp_release(struct ice_pf *pf)
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
- clear_bit(ICE_FLAG_PTP, pf->flags);
-
kthread_cancel_delayed_work_sync(&pf->ptp.work);
ice_ptp_port_phy_stop(&pf->ptp.port);
@@ -3257,6 +3272,9 @@ void ice_ptp_release(struct ice_pf *pf)
pf->ptp.kworker = NULL;
}
+ if (ice_pf_src_tmr_owned(pf))
+ ice_ptp_unregister_auxbus_driver(pf);
+
if (!pf->ptp.clock)
return;
@@ -3266,7 +3284,5 @@ void ice_ptp_release(struct ice_pf *pf)
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
- ice_ptp_unregister_auxbus_driver(pf);
-
dev_info(ice_pf_to_dev(pf), "Removed PTP clock\n");
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 130e6d2ae9a5..e3cc69692405 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -203,8 +203,17 @@ struct ice_ptp_port_owner {
#define GLTSYN_TGT_H_IDX_MAX 4
+enum ice_ptp_state {
+ ICE_PTP_UNINIT = 0,
+ ICE_PTP_INITIALIZING,
+ ICE_PTP_READY,
+ ICE_PTP_RESETTING,
+ ICE_PTP_ERROR,
+};
+
/**
* struct ice_ptp - data used for integrating with CONFIG_PTP_1588_CLOCK
+ * @state: current state of PTP state machine
* @tx_interrupt_mode: the TX interrupt mode for the PTP clock
* @port: data for the PHY port initialization procedure
* @ports_owner: data for the auxiliary driver owner
@@ -227,6 +236,7 @@ struct ice_ptp_port_owner {
* @late_cached_phc_updates: number of times cached PHC update is late
*/
struct ice_ptp {
+ enum ice_ptp_state state;
enum ice_ptp_tx_interrupt tx_interrupt_mode;
struct ice_ptp_port port;
struct ice_ptp_port_owner ports_owner;
--
2.43.0

150
kernel-rt/debian/patches/ice-VDF/0017-ice-pass-reset-type-to-PTP-reset-functions.patch Normal file
View File

@ -0,0 +1,150 @@
From 68d481b41ee5c177a1376fb82a98c09c148d982a Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:50 -0800
Subject: [PATCH 17/36] ice: pass reset type to PTP reset functions
The ice_ptp_prepare_for_reset() and ice_ptp_reset() functions currently
check the pf->flags ICE_FLAG_PFR_REQ bit to determine if the current
reset is a PF reset or not.
This is problematic, because it is possible that a PF reset and a higher
level reset (CORE reset, GLOBAL reset, EMP reset) are requested
simultaneously. In that case, the driver performs the highest level
reset requested. However, the ICE_FLAG_PFR_REQ flag will still be set.
The main driver reset functions take an enum ice_reset_req indicating
which reset is actually being performed. Pass this data into the PTP
functions and rely on this instead of relying on the driver flags.
This ensures that the PTP code performs the proper level of reset that
the driver is actually undergoing.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit c75d5e675a8542274fa0f7e52f3c4db1d4859a0c)
[Adjust the ice_ptp.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 4 ++--
drivers/net/ethernet/intel/ice/ice_ptp.c | 13 +++++++------
drivers/net/ethernet/intel/ice/ice_ptp.h | 17 +++++++++++++----
3 files changed, 22 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 8cfb923198e9..d5321410f2d7 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -615,7 +615,7 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
ice_pf_dis_all_vsi(pf, false);
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_prepare_for_reset(pf);
+ ice_ptp_prepare_for_reset(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_exit(pf);
@@ -7533,7 +7533,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
* fail.
*/
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_reset(pf);
+ ice_ptp_reset(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_init(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 36c81c5ee83b..20d1d22235d3 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2621,8 +2621,9 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
+ * @reset_type: the reset type being performed
*/
-void ice_ptp_prepare_for_reset(struct ice_pf *pf)
+void ice_ptp_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
{
struct ice_ptp *ptp = &pf->ptp;
u8 src_tmr;
@@ -2637,7 +2638,7 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
kthread_cancel_delayed_work_sync(&ptp->work);
- if (test_bit(ICE_PFR_REQ, pf->state))
+ if (reset_type == ICE_RESET_PFR)
return;
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
@@ -2657,8 +2658,9 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
/**
* ice_ptp_reset - Initialize PTP hardware clock support after reset
* @pf: Board private structure
+ * @reset_type: the reset type being performed
*/
-void ice_ptp_reset(struct ice_pf *pf)
+void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
{
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
@@ -2667,15 +2669,14 @@ void ice_ptp_reset(struct ice_pf *pf)
u64 time_diff;
if (ptp->state == ICE_PTP_READY) {
- ice_ptp_prepare_for_reset(pf);
+ ice_ptp_prepare_for_reset(pf, reset_type);
} else if (ptp->state != ICE_PTP_RESETTING) {
err = -EINVAL;
dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
goto err;
}
- if (test_bit(ICE_PFR_REQ, pf->state) ||
- !ice_pf_src_tmr_owned(pf))
+ if (reset_type == ICE_RESET_PFR || !ice_pf_src_tmr_owned(pf))
goto pfr;
err = ice_ptp_init_phc(hw);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index e3cc69692405..cd74712a17a1 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -315,8 +315,9 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb);
-void ice_ptp_reset(struct ice_pf *pf);
-void ice_ptp_prepare_for_reset(struct ice_pf *pf);
+void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type);
+void ice_ptp_prepare_for_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type);
void ice_ptp_init(struct ice_pf *pf);
void ice_ptp_release(struct ice_pf *pf);
void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup);
@@ -351,8 +352,16 @@ static inline bool ice_ptp_process_ts(struct ice_pf *pf)
static inline void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb) { }
-static inline void ice_ptp_reset(struct ice_pf *pf) { }
-static inline void ice_ptp_prepare_for_reset(struct ice_pf *pf) { }
+
+static inline void ice_ptp_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
+{
+}
+
+static inline void ice_ptp_prepare_for_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
+{
+}
static inline void ice_ptp_init(struct ice_pf *pf) { }
static inline void ice_ptp_release(struct ice_pf *pf) { }
static inline void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
--
2.43.0

119
kernel-rt/debian/patches/ice-VDF/0018-ice-rename-verify_cached-to-has_ready_bitmap.patch Normal file
View File

@ -0,0 +1,119 @@
From 084497314e63f3d92178bc44500a27a277abc378 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:51 -0800
Subject: [PATCH 18/36] ice: rename verify_cached to has_ready_bitmap
The tx->verify_cached flag is used to inform the Tx timestamp tracking
code whether it needs to verify the cached Tx timestamp value against
a previous captured value. This is necessary on E810 hardware which does
not have a Tx timestamp ready bitmap.
In addition, we currently rely on the fact that the
ice_get_phy_tx_tstamp_ready() function returns all 1s for E810 hardware.
Instead of introducing a brand new flag, rename and verify_cached to
has_ready_bitmap, inverting the relevant checks.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 3f2216e8dbce04da5376ea7df410541f7b687cb0)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 12 ++++++------
drivers/net/ethernet/intel/ice/ice_ptp.h | 8 +++++---
2 files changed, 11 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 20d1d22235d3..a8c6b83579e6 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -606,11 +606,11 @@ void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
* timestamp. If it is not, skip this for now assuming it hasn't yet
* been captured by hardware.
*/
- if (!drop_ts && tx->verify_cached &&
+ if (!drop_ts && !tx->has_ready_bitmap &&
raw_tstamp == tx->tstamps[idx].cached_tstamp)
return;
- if (tx->verify_cached && raw_tstamp)
+ if (!tx->has_ready_bitmap && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
@@ -751,7 +751,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
* from the last cached timestamp. If it is not, skip this for
* now assuming it hasn't yet been captured by hardware.
*/
- if (!drop_ts && tx->verify_cached &&
+ if (!drop_ts && !tx->has_ready_bitmap &&
raw_tstamp == tx->tstamps[idx].cached_tstamp)
continue;
@@ -761,7 +761,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
skip_ts_read:
spin_lock_irqsave(&tx->lock, flags);
- if (tx->verify_cached && raw_tstamp)
+ if (!tx->has_ready_bitmap && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
@@ -1014,7 +1014,7 @@ ice_ptp_init_tx_e82x(struct ice_pf *pf, struct ice_ptp_tx *tx, u8 port)
tx->block = port / ICE_PORTS_PER_QUAD;
tx->offset = (port % ICE_PORTS_PER_QUAD) * INDEX_PER_PORT_E82X;
tx->len = INDEX_PER_PORT_E82X;
- tx->verify_cached = 0;
+ tx->has_ready_bitmap = 1;
return ice_ptp_alloc_tx_tracker(tx);
}
@@ -1037,7 +1037,7 @@ ice_ptp_init_tx_e810(struct ice_pf *pf, struct ice_ptp_tx *tx)
* verify new timestamps against cached copy of the last read
* timestamp.
*/
- tx->verify_cached = 1;
+ tx->has_ready_bitmap = 0;
return ice_ptp_alloc_tx_tracker(tx);
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index cd74712a17a1..1486a0b3b016 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -100,7 +100,7 @@ struct ice_perout_channel {
* the last timestamp we read for a given index. If the current timestamp
* value is the same as the cached value, we assume a new timestamp hasn't
* been captured. This avoids reporting stale timestamps to the stack. This is
- * only done if the verify_cached flag is set in ice_ptp_tx structure.
+ * only done if the has_ready_bitmap flag is not set in ice_ptp_tx structure.
*/
struct ice_tx_tstamp {
struct sk_buff *skb;
@@ -130,7 +130,9 @@ enum ice_tx_tstamp_work {
* @init: if true, the tracker is initialized;
* @calibrating: if true, the PHY is calibrating the Tx offset. During this
* window, timestamps are temporarily disabled.
- * @verify_cached: if true, verify new timestamp differs from last read value
+ * @has_ready_bitmap: if true, the hardware has a valid Tx timestamp ready
+ * bitmap register. If false, fall back to verifying new
+ * timestamp values against previously cached copy.
* @last_ll_ts_idx_read: index of the last LL TS read by the FW
*/
struct ice_ptp_tx {
@@ -143,7 +145,7 @@ struct ice_ptp_tx {
u8 len;
u8 init : 1;
u8 calibrating : 1;
- u8 verify_cached : 1;
+ u8 has_ready_bitmap : 1;
s8 last_ll_ts_idx_read;
};
--
2.43.0

77
kernel-rt/debian/patches/ice-VDF/0019-ice-don-t-check-has_ready_bitmap-in-E810-functions.patch Normal file
View File

@ -0,0 +1,77 @@
From 375bced6b51243a8c8708204dd32960d076d5b83 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:52 -0800
Subject: [PATCH 19/36] ice: don't check has_ready_bitmap in E810 functions
E810 hardware does not have a Tx timestamp ready bitmap. Don't check
has_ready_bitmap in E810-specific functions.
Add has_ready_bitmap check in ice_ptp_process_tx_tstamp() to stop
relying on the fact that ice_get_phy_tx_tstamp_ready() returns all 1s.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit fea82915fca626eaa83f36d8a23194e8593ef4b4)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 23 +++++++++++------------
1 file changed, 11 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index a8c6b83579e6..ddc2dd0b2a28 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -601,17 +601,13 @@ void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
/* Read the low 32 bit value */
raw_tstamp |= (u64)rd32(&pf->hw, PF_SB_ATQBAH);
- /* For PHYs which don't implement a proper timestamp ready bitmap,
- * verify that the timestamp value is different from the last cached
- * timestamp. If it is not, skip this for now assuming it hasn't yet
- * been captured by hardware.
+ /* Devices using this interface always verify the timestamp differs
+ * relative to the last cached timestamp value.
*/
- if (!drop_ts && !tx->has_ready_bitmap &&
- raw_tstamp == tx->tstamps[idx].cached_tstamp)
+ if (raw_tstamp == tx->tstamps[idx].cached_tstamp)
return;
- if (!tx->has_ready_bitmap && raw_tstamp)
- tx->tstamps[idx].cached_tstamp = raw_tstamp;
+ tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
tx->tstamps[idx].skb = NULL;
@@ -701,9 +697,11 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
hw = &pf->hw;
/* Read the Tx ready status first */
- err = ice_get_phy_tx_tstamp_ready(hw, tx->block, &tstamp_ready);
- if (err)
- return;
+ if (tx->has_ready_bitmap) {
+ err = ice_get_phy_tx_tstamp_ready(hw, tx->block, &tstamp_ready);
+ if (err)
+ return;
+ }
/* Drop packets if the link went down */
link_up = ptp_port->link_up;
@@ -731,7 +729,8 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
* If we do not, the hardware logic for generating a new
* interrupt can get stuck on some devices.
*/
- if (!(tstamp_ready & BIT_ULL(phy_idx))) {
+ if (tx->has_ready_bitmap &&
+ !(tstamp_ready & BIT_ULL(phy_idx))) {
if (drop_ts)
goto skip_ts_read;
--
2.43.0

88
kernel-rt/debian/patches/ice-VDF/0020-ice-rename-ice_ptp_tx_cfg_intr.patch Normal file
View File

@ -0,0 +1,88 @@
From a5318a3a04ed9535ab18ef0f0537b3d33862bee9 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:53 -0800
Subject: [PATCH 20/36] ice: rename ice_ptp_tx_cfg_intr
The ice_ptp_tx_cfg_intr() function sends a control queue message to
configure the PHY timestamp interrupt block. This is a very similar name
to a function which is used to configure the MAC Other Interrupt Cause
Enable register.
Rename this function to ice_ptp_cfg_phy_interrupt in order to make it
more obvious to the reader what action it performs, and distinguish it
from other similarly named functions.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 1abefdca85e8664374f53c7bc80d5f5f827ce711)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index ddc2dd0b2a28..c6e9d77fc59b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1455,14 +1455,14 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
}
/**
- * ice_ptp_tx_ena_intr - Enable or disable the Tx timestamp interrupt
+ * ice_ptp_cfg_phy_interrupt - Configure PHY interrupt settings
* @pf: PF private structure
* @ena: bool value to enable or disable interrupt
* @threshold: Minimum number of packets at which intr is triggered
*
* Utility function to enable or disable Tx timestamp interrupt and threshold
*/
-static int ice_ptp_tx_ena_intr(struct ice_pf *pf, bool ena, u32 threshold)
+static int ice_ptp_cfg_phy_interrupt(struct ice_pf *pf, bool ena, u32 threshold)
{
struct ice_hw *hw = &pf->hw;
int err = 0;
@@ -2664,8 +2664,8 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
struct timespec64 ts;
- int err, itr = 1;
u64 time_diff;
+ int err;
if (ptp->state == ICE_PTP_READY) {
ice_ptp_prepare_for_reset(pf, reset_type);
@@ -2716,7 +2716,7 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
- err = ice_ptp_tx_ena_intr(pf, true, itr);
+ err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
if (err)
goto err;
}
@@ -2967,7 +2967,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
struct timespec64 ts;
- int err, itr = 1;
+ int err;
err = ice_ptp_init_phc(hw);
if (err) {
@@ -3002,7 +3002,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
- err = ice_ptp_tx_ena_intr(pf, true, itr);
+ err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
if (err)
goto err_exit;
}
--
2.43.0

191
kernel-rt/debian/patches/ice-VDF/0021-ice-factor-out-ice_ptp_rebuild_owner.patch Normal file
View File

@ -0,0 +1,191 @@
From 9411c5b82a7196b9712488631fd14e67e2d919fa Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:54 -0800
Subject: [PATCH 21/36] ice: factor out ice_ptp_rebuild_owner()
The ice_ptp_reset() function uses a goto to skip past clock owner
operations if performing a PF reset or if the device is not the clock
owner. This is a bit confusing. Factor this out into
ice_ptp_rebuild_owner() instead.
The ice_ptp_reset() function is called by ice_rebuild() to restore PTP
functionality after a device reset. Follow the convention set by the
ice_main.c file and rename this function to ice_ptp_rebuild(), in the
same way that we have ice_prepare_for_reset() and
ice_ptp_prepare_for_reset().
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 803bef817807d2d36c930dada20c96fffae0dd19)
[Adjust ice_ptp.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 62 ++++++++++++++---------
drivers/net/ethernet/intel/ice/ice_ptp.h | 6 +--
3 files changed, 42 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d5321410f2d7..a04dcc89c35d 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7533,7 +7533,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
* fail.
*/
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_reset(pf, reset_type);
+ ice_ptp_rebuild(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_init(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index c6e9d77fc59b..780aa242c86b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2655,11 +2655,13 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
}
/**
- * ice_ptp_reset - Initialize PTP hardware clock support after reset
+ * ice_ptp_rebuild_owner - Initialize PTP clock owner after reset
* @pf: Board private structure
- * @reset_type: the reset type being performed
+ *
+ * Companion function for ice_ptp_rebuild() which handles tasks that only the
+ * PTP clock owner instance should perform.
*/
-void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
+static int ice_ptp_rebuild_owner(struct ice_pf *pf)
{
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
@@ -2667,32 +2669,21 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
u64 time_diff;
int err;
- if (ptp->state == ICE_PTP_READY) {
- ice_ptp_prepare_for_reset(pf, reset_type);
- } else if (ptp->state != ICE_PTP_RESETTING) {
- err = -EINVAL;
- dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
- goto err;
- }
-
- if (reset_type == ICE_RESET_PFR || !ice_pf_src_tmr_owned(pf))
- goto pfr;
-
err = ice_ptp_init_phc(hw);
if (err)
- goto err;
+ return err;
/* Acquire the global hardware lock */
if (!ice_ptp_lock(hw)) {
err = -EBUSY;
- goto err;
+ return err;
}
/* Write the increment time value to PHY and LAN */
err = ice_ptp_write_incval(hw, ice_base_incval(pf));
if (err) {
ice_ptp_unlock(hw);
- goto err;
+ return err;
}
/* Write the initial Time value to PHY and LAN using the cached PHC
@@ -2708,7 +2699,7 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
err = ice_ptp_write_init(pf, &ts);
if (err) {
ice_ptp_unlock(hw);
- goto err;
+ return err;
}
/* Release the global hardware lock */
@@ -2717,11 +2708,39 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
+ if (err)
+ return err;
+
+ ice_ptp_restart_all_phy(pf);
+ }
+
+ return 0;
+}
+
+/**
+ * ice_ptp_rebuild - Initialize PTP hardware clock support after reset
+ * @pf: Board private structure
+ * @reset_type: the reset type being performed
+ */
+void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
+{
+ struct ice_ptp *ptp = &pf->ptp;
+ int err;
+
+ if (ptp->state == ICE_PTP_READY) {
+ ice_ptp_prepare_for_reset(pf, reset_type);
+ } else if (ptp->state != ICE_PTP_RESETTING) {
+ err = -EINVAL;
+ dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
+ goto err;
+ }
+
+ if (ice_pf_src_tmr_owned(pf) && reset_type != ICE_RESET_PFR) {
+ err = ice_ptp_rebuild_owner(pf);
if (err)
goto err;
}
-pfr:
/* Init Tx structures */
if (ice_is_e810(&pf->hw)) {
err = ice_ptp_init_tx_e810(pf, &ptp->port.tx);
@@ -2736,11 +2755,6 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
ptp->state = ICE_PTP_READY;
- /* Restart the PHY timestamping block */
- if (!test_bit(ICE_PFR_REQ, pf->state) &&
- ice_pf_src_tmr_owned(pf))
- ice_ptp_restart_all_phy(pf);
-
/* Start periodic work going */
kthread_queue_delayed_work(ptp->kworker, &ptp->work, 0);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 1486a0b3b016..352405a2daf2 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -317,7 +317,7 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb);
-void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type);
+void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type);
void ice_ptp_prepare_for_reset(struct ice_pf *pf,
enum ice_reset_req reset_type);
void ice_ptp_init(struct ice_pf *pf);
@@ -355,8 +355,8 @@ static inline void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb) { }
-static inline void ice_ptp_reset(struct ice_pf *pf,
- enum ice_reset_req reset_type)
+static inline void ice_ptp_rebuild(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
{
}
--
2.43.0

93
kernel-rt/debian/patches/ice-VDF/0022-ice-stop-destroying-and-reinitalizing-Tx-tracker-dur.patch Normal file
View File

@ -0,0 +1,93 @@
From 1c89a9e26f669bead5ebcac38fa98c20c517769c Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:55 -0800
Subject: [PATCH 22/36] ice: stop destroying and reinitalizing Tx tracker
during reset
The ice driver currently attempts to destroy and re-initialize the Tx
timestamp tracker during the reset flow. The release of the Tx tracker
only happened during CORE reset or GLOBAL reset. The ice_ptp_rebuild()
function always calls the ice_ptp_init_tx function which will allocate
a new tracker data structure, resulting in memory leaks during PF reset.
Certainly the driver should not be allocating a new tracker without
removing the old tracker data, as this results in a memory leak.
Additionally, there's no reason to remove the tracker memory during a
reset. Remove this logic from the reset and rebuild flow. Instead of
releasing the Tx tracker, flush outstanding timestamps just before we
reset the PHY timestamp block in ice_ptp_cfg_phy_interrupt().
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7a25fe5cd5fb2265065ac6765c53c0a1f1e874d3)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 33 +++++++++++++++---------
1 file changed, 21 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 780aa242c86b..48ec59fc5d87 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -963,6 +963,22 @@ ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
spin_unlock_irqrestore(&tx->lock, flags);
}
+/**
+ * ice_ptp_flush_all_tx_tracker - Flush all timestamp trackers on this clock
+ * @pf: Board private structure
+ *
+ * Called by the clock owner to flush all the Tx timestamp trackers associated
+ * with the clock.
+ */
+static void
+ice_ptp_flush_all_tx_tracker(struct ice_pf *pf)
+{
+ struct ice_ptp_port *port;
+
+ list_for_each_entry(port, &pf->ptp.ports_owner.ports, list_member)
+ ice_ptp_flush_tx_tracker(ptp_port_to_pf(port), &port->tx);
+}
+
/**
* ice_ptp_release_tx_tracker - Release allocated memory for Tx tracker
* @pf: Board private structure
@@ -2705,6 +2721,11 @@ static int ice_ptp_rebuild_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
+ /* Flush software tracking of any outstanding timestamps since we're
+ * about to flush the PHY timestamp block.
+ */
+ ice_ptp_flush_all_tx_tracker(pf);
+
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
@@ -2741,18 +2762,6 @@ void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
goto err;
}
- /* Init Tx structures */
- if (ice_is_e810(&pf->hw)) {
- err = ice_ptp_init_tx_e810(pf, &ptp->port.tx);
- } else {
- kthread_init_delayed_work(&ptp->port.ov_work,
- ice_ptp_wait_for_offsets);
- err = ice_ptp_init_tx_e82x(pf, &ptp->port.tx,
- ptp->port.port_num);
- }
- if (err)
- goto err;
-
ptp->state = ICE_PTP_READY;
/* Start periodic work going */
--
2.43.0

457
kernel-rt/debian/patches/ice-VDF/0023-ice-Remove-and-readd-netdev-during-devlink-reload.patch Normal file
View File

@ -0,0 +1,457 @@
From 6f1d1fa58f58ff3f6ce61ab502bd29227ca1bb3f Mon Sep 17 00:00:00 2001
From: Wojciech Drewek <wojciech.drewek@intel.com>
Date: Mon, 5 Feb 2024 14:03:56 +0100
Subject: [PATCH 23/36] ice: Remove and readd netdev during devlink reload
Recent changes to the devlink reload (commit 9b2348e2d6c9
("devlink: warn about existing entities during reload-reinit"))
force the drivers to destroy devlink ports during reinit.
Adjust ice driver to this requirement, unregister netdvice, destroy
devlink port. ice_init_eth() was removed and all the common code
between probe and reload was moved to ice_load().
During devlink reload we can't take devl_lock (it's already taken)
and in ice_probe() we have to lock it. Use devl_* variant of the API
which does not acquire and release devl_lock. Guard ice_load()
with devl_lock only in case of probe.
Suggested-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 41cc4e53934c30f1cf7745c257154e538c78a1f5)
[Adjust ice.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 +
drivers/net/ethernet/intel/ice/ice_devlink.c | 68 ++++++-
drivers/net/ethernet/intel/ice/ice_main.c | 186 ++++++-------------
3 files changed, 125 insertions(+), 131 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 3278d032a2bd..d3f72f9fbcd7 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -978,6 +978,8 @@ int ice_stop(struct net_device *netdev);
void ice_service_task_schedule(struct ice_pf *pf);
int ice_load(struct ice_pf *pf);
void ice_unload(struct ice_pf *pf);
+int ice_init_dev(struct ice_pf *pf);
+void ice_deinit_dev(struct ice_pf *pf);
/**
* ice_set_rdma_cap - enable RDMA support
diff --git a/drivers/net/ethernet/intel/ice/ice_devlink.c b/drivers/net/ethernet/intel/ice/ice_devlink.c
index 3a2261823d93..43007e3674c4 100644
--- a/drivers/net/ethernet/intel/ice/ice_devlink.c
+++ b/drivers/net/ethernet/intel/ice/ice_devlink.c
@@ -444,6 +444,20 @@ ice_devlink_reload_empr_start(struct ice_pf *pf,
return 0;
}
+/**
+ * ice_devlink_reinit_down - unload given PF
+ * @pf: pointer to the PF struct
+ */
+static void ice_devlink_reinit_down(struct ice_pf *pf)
+{
+ /* No need to take devl_lock, it's already taken by devlink API */
+ ice_unload(pf);
+ rtnl_lock();
+ ice_vsi_decfg(ice_get_main_vsi(pf));
+ rtnl_unlock();
+ ice_deinit_dev(pf);
+}
+
/**
* ice_devlink_reload_down - prepare for reload
* @devlink: pointer to the devlink instance to reload
@@ -477,7 +491,7 @@ ice_devlink_reload_down(struct devlink *devlink, bool netns_change,
"Remove all VFs before doing reinit\n");
return -EOPNOTSUPP;
}
- ice_unload(pf);
+ ice_devlink_reinit_down(pf);
return 0;
case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
return ice_devlink_reload_empr_start(pf, extack);
@@ -1240,6 +1254,45 @@ static int ice_devlink_set_parent(struct devlink_rate *devlink_rate,
return status;
}
+/**
+ * ice_devlink_reinit_up - do reinit of the given PF
+ * @pf: pointer to the PF struct
+ */
+static int ice_devlink_reinit_up(struct ice_pf *pf)
+{
+ struct ice_vsi *vsi = ice_get_main_vsi(pf);
+ struct ice_vsi_cfg_params params;
+ int err;
+
+ err = ice_init_dev(pf);
+ if (err)
+ return err;
+
+ params = ice_vsi_to_params(vsi);
+ params.flags = ICE_VSI_FLAG_INIT;
+
+ rtnl_lock();
+ err = ice_vsi_cfg(vsi, &params);
+ rtnl_unlock();
+ if (err)
+ goto err_vsi_cfg;
+
+ /* No need to take devl_lock, it's already taken by devlink API */
+ err = ice_load(pf);
+ if (err)
+ goto err_load;
+
+ return 0;
+
+err_load:
+ rtnl_lock();
+ ice_vsi_decfg(vsi);
+ rtnl_unlock();
+err_vsi_cfg:
+ ice_deinit_dev(pf);
+ return err;
+}
+
/**
* ice_devlink_reload_up - do reload up after reinit
* @devlink: pointer to the devlink instance reloading
@@ -1260,7 +1313,7 @@ ice_devlink_reload_up(struct devlink *devlink,
switch (action) {
case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
- return ice_load(pf);
+ return ice_devlink_reinit_up(pf);
case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
*actions_performed = BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE);
return ice_devlink_reload_empr_finish(pf, extack);
@@ -1540,6 +1593,7 @@ static const struct devlink_port_ops ice_devlink_port_ops = {
* @pf: the PF to create a devlink port for
*
* Create and register a devlink_port for this PF.
+ * This function has to be called under devl_lock.
*
* Return: zero on success or an error code on failure.
*/
@@ -1552,6 +1606,8 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
struct device *dev;
int err;
+ devlink = priv_to_devlink(pf);
+
dev = ice_pf_to_dev(pf);
devlink_port = &pf->devlink_port;
@@ -1572,10 +1628,9 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
ice_devlink_set_switch_id(pf, &attrs.switch_id);
devlink_port_attrs_set(devlink_port, &attrs);
- devlink = priv_to_devlink(pf);
- err = devlink_port_register_with_ops(devlink, devlink_port, vsi->idx,
- &ice_devlink_port_ops);
+ err = devl_port_register_with_ops(devlink, devlink_port, vsi->idx,
+ &ice_devlink_port_ops);
if (err) {
dev_err(dev, "Failed to create devlink port for PF %d, error %d\n",
pf->hw.pf_id, err);
@@ -1590,10 +1645,11 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
* @pf: the PF to cleanup
*
* Unregisters the devlink_port structure associated with this PF.
+ * This function has to be called under devl_lock.
*/
void ice_devlink_destroy_pf_port(struct ice_pf *pf)
{
- devlink_port_unregister(&pf->devlink_port);
+ devl_port_unregister(&pf->devlink_port);
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index a04dcc89c35d..d3340114297a 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -4588,90 +4588,6 @@ static void ice_decfg_netdev(struct ice_vsi *vsi)
vsi->netdev = NULL;
}
-static int ice_start_eth(struct ice_vsi *vsi)
-{
- int err;
-
- err = ice_init_mac_fltr(vsi->back);
- if (err)
- return err;
-
- err = ice_vsi_open(vsi);
- if (err)
- ice_fltr_remove_all(vsi);
-
- return err;
-}
-
-static void ice_stop_eth(struct ice_vsi *vsi)
-{
- ice_fltr_remove_all(vsi);
- ice_vsi_close(vsi);
-}
-
-static int ice_init_eth(struct ice_pf *pf)
-{
- struct ice_vsi *vsi = ice_get_main_vsi(pf);
- int err;
-
- if (!vsi)
- return -EINVAL;
-
- /* init channel list */
- INIT_LIST_HEAD(&vsi->ch_list);
-
- err = ice_cfg_netdev(vsi);
- if (err)
- return err;
- /* Setup DCB netlink interface */
- ice_dcbnl_setup(vsi);
-
- err = ice_init_mac_fltr(pf);
- if (err)
- goto err_init_mac_fltr;
-
- err = ice_devlink_create_pf_port(pf);
- if (err)
- goto err_devlink_create_pf_port;
-
- SET_NETDEV_DEVLINK_PORT(vsi->netdev, &pf->devlink_port);
-
- err = ice_register_netdev(vsi);
- if (err)
- goto err_register_netdev;
-
- err = ice_tc_indir_block_register(vsi);
- if (err)
- goto err_tc_indir_block_register;
-
- ice_napi_add(vsi);
-
- return 0;
-
-err_tc_indir_block_register:
- ice_unregister_netdev(vsi);
-err_register_netdev:
- ice_devlink_destroy_pf_port(pf);
-err_devlink_create_pf_port:
-err_init_mac_fltr:
- ice_decfg_netdev(vsi);
- return err;
-}
-
-static void ice_deinit_eth(struct ice_pf *pf)
-{
- struct ice_vsi *vsi = ice_get_main_vsi(pf);
-
- if (!vsi)
- return;
-
- ice_vsi_close(vsi);
- ice_unregister_netdev(vsi);
- ice_devlink_destroy_pf_port(pf);
- ice_tc_indir_block_unregister(vsi);
- ice_decfg_netdev(vsi);
-}
-
/**
* ice_wait_for_fw - wait for full FW readiness
* @hw: pointer to the hardware structure
@@ -4697,7 +4613,7 @@ static int ice_wait_for_fw(struct ice_hw *hw, u32 timeout)
return -ETIMEDOUT;
}
-static int ice_init_dev(struct ice_pf *pf)
+int ice_init_dev(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
@@ -4790,7 +4706,7 @@ static int ice_init_dev(struct ice_pf *pf)
return err;
}
-static void ice_deinit_dev(struct ice_pf *pf)
+void ice_deinit_dev(struct ice_pf *pf)
{
ice_free_irq_msix_misc(pf);
ice_deinit_pf(pf);
@@ -5091,31 +5007,47 @@ static void ice_deinit(struct ice_pf *pf)
/**
* ice_load - load pf by init hw and starting VSI
* @pf: pointer to the pf instance
+ *
+ * This function has to be called under devl_lock.
*/
int ice_load(struct ice_pf *pf)
{
- struct ice_vsi_cfg_params params = {};
struct ice_vsi *vsi;
int err;
- err = ice_init_dev(pf);
+ devl_assert_locked(priv_to_devlink(pf));
+
+ vsi = ice_get_main_vsi(pf);
+
+ /* init channel list */
+ INIT_LIST_HEAD(&vsi->ch_list);
+
+ err = ice_cfg_netdev(vsi);
if (err)
return err;
- vsi = ice_get_main_vsi(pf);
+ /* Setup DCB netlink interface */
+ ice_dcbnl_setup(vsi);
- params = ice_vsi_to_params(vsi);
- params.flags = ICE_VSI_FLAG_INIT;
+ err = ice_init_mac_fltr(pf);
+ if (err)
+ goto err_init_mac_fltr;
- rtnl_lock();
- err = ice_vsi_cfg(vsi, &params);
+ err = ice_devlink_create_pf_port(pf);
if (err)
- goto err_vsi_cfg;
+ goto err_devlink_create_pf_port;
- err = ice_start_eth(ice_get_main_vsi(pf));
+ SET_NETDEV_DEVLINK_PORT(vsi->netdev, &pf->devlink_port);
+
+ err = ice_register_netdev(vsi);
+ if (err)
+ goto err_register_netdev;
+
+ err = ice_tc_indir_block_register(vsi);
if (err)
- goto err_start_eth;
- rtnl_unlock();
+ goto err_tc_indir_block_register;
+
+ ice_napi_add(vsi);
err = ice_init_rdma(pf);
if (err)
@@ -5129,29 +5061,35 @@ int ice_load(struct ice_pf *pf)
return 0;
err_init_rdma:
- ice_vsi_close(ice_get_main_vsi(pf));
- rtnl_lock();
-err_start_eth:
- ice_vsi_decfg(ice_get_main_vsi(pf));
-err_vsi_cfg:
- rtnl_unlock();
- ice_deinit_dev(pf);
+ ice_tc_indir_block_unregister(vsi);
+err_tc_indir_block_register:
+ ice_unregister_netdev(vsi);
+err_register_netdev:
+ ice_devlink_destroy_pf_port(pf);
+err_devlink_create_pf_port:
+err_init_mac_fltr:
+ ice_decfg_netdev(vsi);
return err;
}
/**
* ice_unload - unload pf by stopping VSI and deinit hw
* @pf: pointer to the pf instance
+ *
+ * This function has to be called under devl_lock.
*/
void ice_unload(struct ice_pf *pf)
{
+ struct ice_vsi *vsi = ice_get_main_vsi(pf);
+
+ devl_assert_locked(priv_to_devlink(pf));
+
ice_deinit_features(pf);
ice_deinit_rdma(pf);
- rtnl_lock();
- ice_stop_eth(ice_get_main_vsi(pf));
- ice_vsi_decfg(ice_get_main_vsi(pf));
- rtnl_unlock();
- ice_deinit_dev(pf);
+ ice_tc_indir_block_unregister(vsi);
+ ice_unregister_netdev(vsi);
+ ice_devlink_destroy_pf_port(pf);
+ ice_decfg_netdev(vsi);
}
/**
@@ -5249,27 +5187,23 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent)
if (err)
goto err_init;
- err = ice_init_eth(pf);
+ devl_lock(priv_to_devlink(pf));
+ err = ice_load(pf);
+ devl_unlock(priv_to_devlink(pf));
if (err)
- goto err_init_eth;
-
- err = ice_init_rdma(pf);
- if (err)
- goto err_init_rdma;
+ goto err_load;
err = ice_init_devlink(pf);
if (err)
goto err_init_devlink;
- ice_init_features(pf);
-
return 0;
err_init_devlink:
- ice_deinit_rdma(pf);
-err_init_rdma:
- ice_deinit_eth(pf);
-err_init_eth:
+ devl_lock(priv_to_devlink(pf));
+ ice_unload(pf);
+ devl_unlock(priv_to_devlink(pf));
+err_load:
ice_deinit(pf);
err_init:
pci_disable_device(pdev);
@@ -5363,12 +5297,14 @@ static void ice_remove(struct pci_dev *pdev)
if (!ice_is_safe_mode(pf))
ice_remove_arfs(pf);
- ice_deinit_features(pf);
+
ice_deinit_devlink(pf);
- ice_deinit_rdma(pf);
- ice_deinit_eth(pf);
- ice_deinit(pf);
+ devl_lock(priv_to_devlink(pf));
+ ice_unload(pf);
+ devl_unlock(priv_to_devlink(pf));
+
+ ice_deinit(pf);
ice_vsi_release_all(pf);
ice_setup_mc_magic_wake(pf);
--
2.43.0

434
kernel-rt/debian/patches/ice-VDF/0024-ice-remove-FW-logging-code.patch Normal file
View File

@ -0,0 +1,434 @@
From e110839c4d9bfa4c885877a69573f48c008d3edd Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:11 -0800
Subject: [PATCH 24/36] ice: remove FW logging code
The FW logging code doesn't work because there is no way to set
cq_ena or uart_ena so remove the code. This code is the original
(v1) way of FW logging so it should be replaced with the v2 way.
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1953fc720e603721764f31daae216a2851664167)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 78 -------
drivers/net/ethernet/intel/ice/ice_common.c | 217 ------------------
drivers/net/ethernet/intel/ice/ice_common.h | 1 -
drivers/net/ethernet/intel/ice/ice_main.c | 3 -
drivers/net/ethernet/intel/ice/ice_type.h | 20 --
5 files changed, 319 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 9bacb69ead8c..3b289e6a225b 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2032,78 +2032,6 @@ struct ice_aqc_add_rdma_qset_data {
struct ice_aqc_add_tx_rdma_qset_entry rdma_qsets[];
};
-/* Configure Firmware Logging Command (indirect 0xFF09)
- * Logging Information Read Response (indirect 0xFF10)
- * Note: The 0xFF10 command has no input parameters.
- */
-struct ice_aqc_fw_logging {
- u8 log_ctrl;
-#define ICE_AQC_FW_LOG_AQ_EN BIT(0)
-#define ICE_AQC_FW_LOG_UART_EN BIT(1)
- u8 rsvd0;
- u8 log_ctrl_valid; /* Not used by 0xFF10 Response */
-#define ICE_AQC_FW_LOG_AQ_VALID BIT(0)
-#define ICE_AQC_FW_LOG_UART_VALID BIT(1)
- u8 rsvd1[5];
- __le32 addr_high;
- __le32 addr_low;
-};
-
-enum ice_aqc_fw_logging_mod {
- ICE_AQC_FW_LOG_ID_GENERAL = 0,
- ICE_AQC_FW_LOG_ID_CTRL,
- ICE_AQC_FW_LOG_ID_LINK,
- ICE_AQC_FW_LOG_ID_LINK_TOPO,
- ICE_AQC_FW_LOG_ID_DNL,
- ICE_AQC_FW_LOG_ID_I2C,
- ICE_AQC_FW_LOG_ID_SDP,
- ICE_AQC_FW_LOG_ID_MDIO,
- ICE_AQC_FW_LOG_ID_ADMINQ,
- ICE_AQC_FW_LOG_ID_HDMA,
- ICE_AQC_FW_LOG_ID_LLDP,
- ICE_AQC_FW_LOG_ID_DCBX,
- ICE_AQC_FW_LOG_ID_DCB,
- ICE_AQC_FW_LOG_ID_NETPROXY,
- ICE_AQC_FW_LOG_ID_NVM,
- ICE_AQC_FW_LOG_ID_AUTH,
- ICE_AQC_FW_LOG_ID_VPD,
- ICE_AQC_FW_LOG_ID_IOSF,
- ICE_AQC_FW_LOG_ID_PARSER,
- ICE_AQC_FW_LOG_ID_SW,
- ICE_AQC_FW_LOG_ID_SCHEDULER,
- ICE_AQC_FW_LOG_ID_TXQ,
- ICE_AQC_FW_LOG_ID_RSVD,
- ICE_AQC_FW_LOG_ID_POST,
- ICE_AQC_FW_LOG_ID_WATCHDOG,
- ICE_AQC_FW_LOG_ID_TASK_DISPATCH,
- ICE_AQC_FW_LOG_ID_MNG,
- ICE_AQC_FW_LOG_ID_MAX,
-};
-
-/* Defines for both above FW logging command/response buffers */
-#define ICE_AQC_FW_LOG_ID_S 0
-#define ICE_AQC_FW_LOG_ID_M (0xFFF << ICE_AQC_FW_LOG_ID_S)
-
-#define ICE_AQC_FW_LOG_CONF_SUCCESS 0 /* Used by response */
-#define ICE_AQC_FW_LOG_CONF_BAD_INDX BIT(12) /* Used by response */
-
-#define ICE_AQC_FW_LOG_EN_S 12
-#define ICE_AQC_FW_LOG_EN_M (0xF << ICE_AQC_FW_LOG_EN_S)
-#define ICE_AQC_FW_LOG_INFO_EN BIT(12) /* Used by command */
-#define ICE_AQC_FW_LOG_INIT_EN BIT(13) /* Used by command */
-#define ICE_AQC_FW_LOG_FLOW_EN BIT(14) /* Used by command */
-#define ICE_AQC_FW_LOG_ERR_EN BIT(15) /* Used by command */
-
-/* Get/Clear FW Log (indirect 0xFF11) */
-struct ice_aqc_get_clear_fw_log {
- u8 flags;
-#define ICE_AQC_FW_LOG_CLEAR BIT(0)
-#define ICE_AQC_FW_LOG_MORE_DATA_AVAIL BIT(1)
- u8 rsvd1[7];
- __le32 addr_high;
- __le32 addr_low;
-};
-
/* Download Package (indirect 0x0C40) */
/* Also used for Update Package (indirect 0x0C41 and 0x0C42) */
struct ice_aqc_download_pkg {
@@ -2448,8 +2376,6 @@ struct ice_aq_desc {
struct ice_aqc_add_rdma_qset add_rdma_qset;
struct ice_aqc_add_get_update_free_vsi vsi_cmd;
struct ice_aqc_add_update_free_vsi_resp add_update_free_vsi_res;
- struct ice_aqc_fw_logging fw_logging;
- struct ice_aqc_get_clear_fw_log get_clear_fw_log;
struct ice_aqc_download_pkg download_pkg;
struct ice_aqc_set_cgu_input_config set_cgu_input_config;
struct ice_aqc_get_cgu_input_config get_cgu_input_config;
@@ -2657,10 +2583,6 @@ enum ice_adminq_opc {
/* Standalone Commands/Events */
ice_aqc_opc_event_lan_overflow = 0x1001,
-
- /* debug commands */
- ice_aqc_opc_fw_logging = 0xFF09,
- ice_aqc_opc_fw_logging_info = 0xFF10,
};
#endif /* _ICE_ADMINQ_CMD_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index acf6ac00f804..a5c4b7ad6a20 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -897,216 +897,6 @@ static void ice_cleanup_fltr_mgmt_struct(struct ice_hw *hw)
devm_kfree(ice_hw_to_dev(hw), sw);
}
-/**
- * ice_get_fw_log_cfg - get FW logging configuration
- * @hw: pointer to the HW struct
- */
-static int ice_get_fw_log_cfg(struct ice_hw *hw)
-{
- struct ice_aq_desc desc;
- __le16 *config;
- int status;
- u16 size;
-
- size = sizeof(*config) * ICE_AQC_FW_LOG_ID_MAX;
- config = kzalloc(size, GFP_KERNEL);
- if (!config)
- return -ENOMEM;
-
- ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logging_info);
-
- status = ice_aq_send_cmd(hw, &desc, config, size, NULL);
- if (!status) {
- u16 i;
-
- /* Save FW logging information into the HW structure */
- for (i = 0; i < ICE_AQC_FW_LOG_ID_MAX; i++) {
- u16 v, m, flgs;
-
- v = le16_to_cpu(config[i]);
- m = (v & ICE_AQC_FW_LOG_ID_M) >> ICE_AQC_FW_LOG_ID_S;
- flgs = (v & ICE_AQC_FW_LOG_EN_M) >> ICE_AQC_FW_LOG_EN_S;
-
- if (m < ICE_AQC_FW_LOG_ID_MAX)
- hw->fw_log.evnts[m].cur = flgs;
- }
- }
-
- kfree(config);
-
- return status;
-}
-
-/**
- * ice_cfg_fw_log - configure FW logging
- * @hw: pointer to the HW struct
- * @enable: enable certain FW logging events if true, disable all if false
- *
- * This function enables/disables the FW logging via Rx CQ events and a UART
- * port based on predetermined configurations. FW logging via the Rx CQ can be
- * enabled/disabled for individual PF's. However, FW logging via the UART can
- * only be enabled/disabled for all PFs on the same device.
- *
- * To enable overall FW logging, the "cq_en" and "uart_en" enable bits in
- * hw->fw_log need to be set accordingly, e.g. based on user-provided input,
- * before initializing the device.
- *
- * When re/configuring FW logging, callers need to update the "cfg" elements of
- * the hw->fw_log.evnts array with the desired logging event configurations for
- * modules of interest. When disabling FW logging completely, the callers can
- * just pass false in the "enable" parameter. On completion, the function will
- * update the "cur" element of the hw->fw_log.evnts array with the resulting
- * logging event configurations of the modules that are being re/configured. FW
- * logging modules that are not part of a reconfiguration operation retain their
- * previous states.
- *
- * Before resetting the device, it is recommended that the driver disables FW
- * logging before shutting down the control queue. When disabling FW logging
- * ("enable" = false), the latest configurations of FW logging events stored in
- * hw->fw_log.evnts[] are not overridden to allow them to be reconfigured after
- * a device reset.
- *
- * When enabling FW logging to emit log messages via the Rx CQ during the
- * device's initialization phase, a mechanism alternative to interrupt handlers
- * needs to be used to extract FW log messages from the Rx CQ periodically and
- * to prevent the Rx CQ from being full and stalling other types of control
- * messages from FW to SW. Interrupts are typically disabled during the device's
- * initialization phase.
- */
-static int ice_cfg_fw_log(struct ice_hw *hw, bool enable)
-{
- struct ice_aqc_fw_logging *cmd;
- u16 i, chgs = 0, len = 0;
- struct ice_aq_desc desc;
- __le16 *data = NULL;
- u8 actv_evnts = 0;
- void *buf = NULL;
- int status = 0;
-
- if (!hw->fw_log.cq_en && !hw->fw_log.uart_en)
- return 0;
-
- /* Disable FW logging only when the control queue is still responsive */
- if (!enable &&
- (!hw->fw_log.actv_evnts || !ice_check_sq_alive(hw, &hw->adminq)))
- return 0;
-
- /* Get current FW log settings */
- status = ice_get_fw_log_cfg(hw);
- if (status)
- return status;
-
- ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logging);
- cmd = &desc.params.fw_logging;
-
- /* Indicate which controls are valid */
- if (hw->fw_log.cq_en)
- cmd->log_ctrl_valid |= ICE_AQC_FW_LOG_AQ_VALID;
-
- if (hw->fw_log.uart_en)
- cmd->log_ctrl_valid |= ICE_AQC_FW_LOG_UART_VALID;
-
- if (enable) {
- /* Fill in an array of entries with FW logging modules and
- * logging events being reconfigured.
- */
- for (i = 0; i < ICE_AQC_FW_LOG_ID_MAX; i++) {
- u16 val;
-
- /* Keep track of enabled event types */
- actv_evnts |= hw->fw_log.evnts[i].cfg;
-
- if (hw->fw_log.evnts[i].cfg == hw->fw_log.evnts[i].cur)
- continue;
-
- if (!data) {
- data = devm_kcalloc(ice_hw_to_dev(hw),
- ICE_AQC_FW_LOG_ID_MAX,
- sizeof(*data),
- GFP_KERNEL);
- if (!data)
- return -ENOMEM;
- }
-
- val = i << ICE_AQC_FW_LOG_ID_S;
- val |= hw->fw_log.evnts[i].cfg << ICE_AQC_FW_LOG_EN_S;
- data[chgs++] = cpu_to_le16(val);
- }
-
- /* Only enable FW logging if at least one module is specified.
- * If FW logging is currently enabled but all modules are not
- * enabled to emit log messages, disable FW logging altogether.
- */
- if (actv_evnts) {
- /* Leave if there is effectively no change */
- if (!chgs)
- goto out;
-
- if (hw->fw_log.cq_en)
- cmd->log_ctrl |= ICE_AQC_FW_LOG_AQ_EN;
-
- if (hw->fw_log.uart_en)
- cmd->log_ctrl |= ICE_AQC_FW_LOG_UART_EN;
-
- buf = data;
- len = sizeof(*data) * chgs;
- desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD);
- }
- }
-
- status = ice_aq_send_cmd(hw, &desc, buf, len, NULL);
- if (!status) {
- /* Update the current configuration to reflect events enabled.
- * hw->fw_log.cq_en and hw->fw_log.uart_en indicate if the FW
- * logging mode is enabled for the device. They do not reflect
- * actual modules being enabled to emit log messages. So, their
- * values remain unchanged even when all modules are disabled.
- */
- u16 cnt = enable ? chgs : (u16)ICE_AQC_FW_LOG_ID_MAX;
-
- hw->fw_log.actv_evnts = actv_evnts;
- for (i = 0; i < cnt; i++) {
- u16 v, m;
-
- if (!enable) {
- /* When disabling all FW logging events as part
- * of device's de-initialization, the original
- * configurations are retained, and can be used
- * to reconfigure FW logging later if the device
- * is re-initialized.
- */
- hw->fw_log.evnts[i].cur = 0;
- continue;
- }
-
- v = le16_to_cpu(data[i]);
- m = (v & ICE_AQC_FW_LOG_ID_M) >> ICE_AQC_FW_LOG_ID_S;
- hw->fw_log.evnts[m].cur = hw->fw_log.evnts[m].cfg;
- }
- }
-
-out:
- devm_kfree(ice_hw_to_dev(hw), data);
-
- return status;
-}
-
-/**
- * ice_output_fw_log
- * @hw: pointer to the HW struct
- * @desc: pointer to the AQ message descriptor
- * @buf: pointer to the buffer accompanying the AQ message
- *
- * Formats a FW Log message and outputs it via the standard driver logs.
- */
-void ice_output_fw_log(struct ice_hw *hw, struct ice_aq_desc *desc, void *buf)
-{
- ice_debug(hw, ICE_DBG_FW_LOG, "[ FW Log Msg Start ]\n");
- ice_debug_array(hw, ICE_DBG_FW_LOG, 16, 1, (u8 *)buf,
- le16_to_cpu(desc->datalen));
- ice_debug(hw, ICE_DBG_FW_LOG, "[ FW Log Msg End ]\n");
-}
-
/**
* ice_get_itr_intrl_gran
* @hw: pointer to the HW struct
@@ -1164,11 +954,6 @@ int ice_init_hw(struct ice_hw *hw)
if (status)
goto err_unroll_cqinit;
- /* Enable FW logging. Not fatal if this fails. */
- status = ice_cfg_fw_log(hw, true);
- if (status)
- ice_debug(hw, ICE_DBG_INIT, "Failed to enable FW logging.\n");
-
status = ice_clear_pf_cfg(hw);
if (status)
goto err_unroll_cqinit;
@@ -1318,8 +1103,6 @@ void ice_deinit_hw(struct ice_hw *hw)
ice_free_hw_tbls(hw);
mutex_destroy(&hw->tnl_lock);
- /* Attempt to disable FW logging before shutting down control queues */
- ice_cfg_fw_log(hw, false);
ice_destroy_all_ctrlq(hw);
/* Clear VSI contexts if not already cleared */
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index 7a966a0c224f..d47e5400351f 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -199,7 +199,6 @@ ice_aq_cfg_lan_txq(struct ice_hw *hw, struct ice_aqc_cfg_txqs_buf *buf,
struct ice_sq_cd *cd);
int ice_replay_vsi(struct ice_hw *hw, u16 vsi_handle);
void ice_replay_post(struct ice_hw *hw);
-void ice_output_fw_log(struct ice_hw *hw, struct ice_aq_desc *desc, void *buf);
struct ice_q_ctx *
ice_get_lan_q_ctx(struct ice_hw *hw, u16 vsi_handle, u8 tc, u16 q_handle);
int ice_sbq_rw_reg(struct ice_hw *hw, struct ice_sbq_msg_input *in);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d3340114297a..e5cc9790969c 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -1535,9 +1535,6 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
ice_vc_process_vf_msg(pf, &event, &data);
break;
- case ice_aqc_opc_fw_logging:
- ice_output_fw_log(hw, &event.desc, event.msg_buf);
- break;
case ice_aqc_opc_lldp_set_mib_change:
ice_dcb_process_lldp_set_mib_change(pf, &event);
break;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index b0f1f4db1d8b..6e1fed0d7384 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -731,24 +731,6 @@ struct ice_switch_info {
DECLARE_BITMAP(prof_res_bm[ICE_MAX_NUM_PROFILES], ICE_MAX_FV_WORDS);
};
-/* FW logging configuration */
-struct ice_fw_log_evnt {
- u8 cfg : 4; /* New event enables to configure */
- u8 cur : 4; /* Current/active event enables */
-};
-
-struct ice_fw_log_cfg {
- u8 cq_en : 1; /* FW logging is enabled via the control queue */
- u8 uart_en : 1; /* FW logging is enabled via UART for all PFs */
- u8 actv_evnts; /* Cumulation of currently enabled log events */
-
-#define ICE_FW_LOG_EVNT_INFO (ICE_AQC_FW_LOG_INFO_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_INIT (ICE_AQC_FW_LOG_INIT_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_FLOW (ICE_AQC_FW_LOG_FLOW_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_ERR (ICE_AQC_FW_LOG_ERR_EN >> ICE_AQC_FW_LOG_EN_S)
- struct ice_fw_log_evnt evnts[ICE_AQC_FW_LOG_ID_MAX];
-};
-
/* Enum defining the different states of the mailbox snapshot in the
* PF-VF mailbox overflow detection algorithm. The snapshot can be in
* states:
@@ -890,8 +872,6 @@ struct ice_hw {
u8 fw_patch; /* firmware patch version */
u32 fw_build; /* firmware build number */
- struct ice_fw_log_cfg fw_log;
-
/* Device max aggregate bandwidths corresponding to the GL_PWR_MODE_CTL
* register. Used for determining the ITR/INTRL granularity during
* initialization.
--
2.43.0

1083
kernel-rt/debian/patches/ice-VDF/0025-ice-configure-FW-logging.patch Normal file
View File

File diff suppressed because it is too large Load Diff

281
kernel-rt/debian/patches/ice-VDF/0026-ice-enable-FW-logging.patch Normal file
View File

@ -0,0 +1,281 @@
From 189d58473481cf01b493fca4e9dd2ab8380d0ce5 Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:13 -0800
Subject: [PATCH 26/36] ice: enable FW logging
Once users have configured the FW logging then allow them to enable it
by writing to the 'fwlog/enable' file. The file accepts a boolean value
(0 or 1) where 1 means enable FW logging and 0 means disable FW logging.
# echo <value> > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/enable
Where <value> is 0 or 1.
The user can read the 'fwlog/enable' file to see whether logging is
enabled or not. Reading the actual data is a separate patch. To see the
current value then:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/enable
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 73671c3162c83a689342fd57f00b5f261682e49b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 3 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 98 +++++++++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 67 +++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.h | 2 +
4 files changed, 170 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 347e4fed5e0d..11391be4efc2 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2336,6 +2336,7 @@ enum ice_aqc_fw_logging_mod {
};
/* Set FW Logging configuration (indirect 0xFF30)
+ * Register for FW Logging (indirect 0xFF31)
* Query FW Logging (indirect 0xFF32)
*/
struct ice_aqc_fw_log {
@@ -2344,6 +2345,7 @@ struct ice_aqc_fw_log {
#define ICE_AQC_FW_LOG_CONF_AQ_EN BIT(1)
#define ICE_AQC_FW_LOG_QUERY_REGISTERED BIT(2)
#define ICE_AQC_FW_LOG_CONF_SET_VALID BIT(3)
+#define ICE_AQC_FW_LOG_AQ_REGISTER BIT(0)
#define ICE_AQC_FW_LOG_AQ_QUERY BIT(2)
u8 rsp_flag;
@@ -2662,6 +2664,7 @@ enum ice_adminq_opc {
/* FW Logging Commands */
ice_aqc_opc_fw_logs_config = 0xFF30,
+ ice_aqc_opc_fw_logs_register = 0xFF31,
ice_aqc_opc_fw_logs_query = 0xFF32,
};
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index 3b0d9b214fd1..3dde99969132 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -281,6 +281,101 @@ static const struct file_operations ice_debugfs_nr_messages_fops = {
.write = ice_debugfs_nr_messages_write,
};
+/**
+ * ice_debugfs_enable_read - read from 'enable' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_enable_read(struct file *filp,
+ char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char buff[32] = {};
+
+ snprintf(buff, sizeof(buff), "%u\n",
+ (u16)(hw->fwlog_cfg.options &
+ ICE_FWLOG_OPTION_IS_REGISTERED) >> 3);
+
+ return simple_read_from_buffer(buffer, count, ppos, buff, strlen(buff));
+}
+
+/**
+ * ice_debugfs_enable_write - write into 'enable' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_enable_write(struct file *filp, const char __user *buf,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char user_val[8], *cmd_buf;
+ bool enable;
+ ssize_t ret;
+
+ /* don't allow partial writes or invalid input */
+ if (*ppos != 0 || count > 2)
+ return -EINVAL;
+
+ cmd_buf = memdup_user(buf, count);
+ if (IS_ERR(cmd_buf))
+ return PTR_ERR(cmd_buf);
+
+ ret = sscanf(cmd_buf, "%s", user_val);
+ if (ret != 1)
+ return -EINVAL;
+
+ ret = kstrtobool(user_val, &enable);
+ if (ret)
+ goto enable_write_error;
+
+ if (enable)
+ hw->fwlog_cfg.options |= ICE_FWLOG_OPTION_ARQ_ENA;
+ else
+ hw->fwlog_cfg.options &= ~ICE_FWLOG_OPTION_ARQ_ENA;
+
+ ret = ice_fwlog_set(hw, &hw->fwlog_cfg);
+ if (ret)
+ goto enable_write_error;
+
+ if (enable)
+ ret = ice_fwlog_register(hw);
+ else
+ ret = ice_fwlog_unregister(hw);
+
+ if (ret)
+ goto enable_write_error;
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+enable_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_enable_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_enable_read,
+ .write = ice_debugfs_enable_write,
+};
+
/**
* ice_debugfs_fwlog_init - setup the debugfs directory
* @pf: the ice that is starting up
@@ -332,6 +427,9 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
pf->ice_debugfs_pf_fwlog_modules = fw_modules;
+ debugfs_create_file("enable", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_enable_fops);
+
return;
err_create_module_files:
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 307e0d04f3fe..25a17cbc1d34 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -63,6 +63,11 @@ void ice_fwlog_deinit(struct ice_hw *hw)
kfree(pf->ice_debugfs_pf_fwlog_modules);
pf->ice_debugfs_pf_fwlog_modules = NULL;
+
+ status = ice_fwlog_unregister(hw);
+ if (status)
+ dev_warn(ice_hw_to_dev(hw), "Unable to unregister FW logging, status: %d\n",
+ status);
}
/**
@@ -197,6 +202,8 @@ static int ice_aq_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg)
cfg->options |= ICE_FWLOG_OPTION_ARQ_ENA;
if (cmd->cmd_flags & ICE_AQC_FW_LOG_CONF_UART_EN)
cfg->options |= ICE_FWLOG_OPTION_UART_ENA;
+ if (cmd->cmd_flags & ICE_AQC_FW_LOG_QUERY_REGISTERED)
+ cfg->options |= ICE_FWLOG_OPTION_IS_REGISTERED;
fw_modules = (struct ice_aqc_fw_log_cfg_resp *)buf;
@@ -226,6 +233,66 @@ int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg)
return ice_aq_fwlog_get(hw, cfg);
}
+/**
+ * ice_aq_fwlog_register - Register PF for firmware logging events (0xFF31)
+ * @hw: pointer to the HW structure
+ * @reg: true to register and false to unregister
+ */
+static int ice_aq_fwlog_register(struct ice_hw *hw, bool reg)
+{
+ struct ice_aq_desc desc;
+
+ ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logs_register);
+
+ if (reg)
+ desc.params.fw_log.cmd_flags = ICE_AQC_FW_LOG_AQ_REGISTER;
+
+ return ice_aq_send_cmd(hw, &desc, NULL, 0, NULL);
+}
+
+/**
+ * ice_fwlog_register - Register the PF for firmware logging
+ * @hw: pointer to the HW structure
+ *
+ * After this call the PF will start to receive firmware logging based on the
+ * configuration set in ice_fwlog_set.
+ */
+int ice_fwlog_register(struct ice_hw *hw)
+{
+ int status;
+
+ if (!ice_fwlog_supported(hw))
+ return -EOPNOTSUPP;
+
+ status = ice_aq_fwlog_register(hw, true);
+ if (status)
+ ice_debug(hw, ICE_DBG_FW_LOG, "Failed to register for firmware logging events over ARQ\n");
+ else
+ hw->fwlog_cfg.options |= ICE_FWLOG_OPTION_IS_REGISTERED;
+
+ return status;
+}
+
+/**
+ * ice_fwlog_unregister - Unregister the PF from firmware logging
+ * @hw: pointer to the HW structure
+ */
+int ice_fwlog_unregister(struct ice_hw *hw)
+{
+ int status;
+
+ if (!ice_fwlog_supported(hw))
+ return -EOPNOTSUPP;
+
+ status = ice_aq_fwlog_register(hw, false);
+ if (status)
+ ice_debug(hw, ICE_DBG_FW_LOG, "Failed to unregister from firmware logging events over ARQ\n");
+ else
+ hw->fwlog_cfg.options &= ~ICE_FWLOG_OPTION_IS_REGISTERED;
+
+ return status;
+}
+
/**
* ice_fwlog_set_supported - Set if FW logging is supported by FW
* @hw: pointer to the HW struct
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.h b/drivers/net/ethernet/intel/ice/ice_fwlog.h
index 8e68ee02713b..45865558425d 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.h
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.h
@@ -53,4 +53,6 @@ int ice_fwlog_init(struct ice_hw *hw);
void ice_fwlog_deinit(struct ice_hw *hw);
int ice_fwlog_set(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
+int ice_fwlog_register(struct ice_hw *hw);
+int ice_fwlog_unregister(struct ice_hw *hw);
#endif /* _ICE_FWLOG_H_ */
--
2.43.0

572
kernel-rt/debian/patches/ice-VDF/0027-ice-add-ability-to-read-and-configure-FW-log-data.patch Normal file
View File

@ -0,0 +1,572 @@
From a584ea88cfdc8ac3f782be1d5d67fa92c3423290 Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:14 -0800
Subject: [PATCH 27/36] ice: add ability to read and configure FW log data
Once logging is enabled the user should read the data from the 'data'
file. The data is in the form of a binary blob that can be sent to Intel
for decoding. To read the data use a command like:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/data > log_data.bin
If the user wants to clear the FW log data that has been stored in the
driver then they can write any value to the 'data' file and that will clear
the data. An example is:
# echo 34 > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/data
In addition to being able to read the data the user can configure how
much memory is used to store FW log data. This allows the user to
increase/decrease the amount of memory based on the users situation.
The data is stored such that if the memory fills up then the oldest data
will get overwritten in a circular manner. To change the amount of
memory the user can write to the 'log_size' file like this:
# echo <value> > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/log_size
Where <value> is one of 128K, 256K, 512K, 1M, and 2M. The default value
is 1M.
The user can see the current value of 'log_size' by reading the file:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/log_size
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 9d3535e71985beb738c4ad2b772c6f0efdce0202)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 2 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 210 ++++++++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 142 ++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.h | 21 ++
drivers/net/ethernet/intel/ice/ice_main.c | 29 +++
drivers/net/ethernet/intel/ice/ice_type.h | 1 +
6 files changed, 405 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 11391be4efc2..f63b57ff2a3d 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2338,6 +2338,7 @@ enum ice_aqc_fw_logging_mod {
/* Set FW Logging configuration (indirect 0xFF30)
* Register for FW Logging (indirect 0xFF31)
* Query FW Logging (indirect 0xFF32)
+ * FW Log Event (indirect 0xFF33)
*/
struct ice_aqc_fw_log {
u8 cmd_flags;
@@ -2666,6 +2667,7 @@ enum ice_adminq_opc {
ice_aqc_opc_fw_logs_config = 0xFF30,
ice_aqc_opc_fw_logs_register = 0xFF31,
ice_aqc_opc_fw_logs_query = 0xFF32,
+ ice_aqc_opc_fw_logs_event = 0xFF33,
};
#endif /* _ICE_ADMINQ_CMD_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index 3dde99969132..c2bfba6b9ead 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -64,6 +64,17 @@ static const char * const ice_fwlog_level_string[] = {
"verbose",
};
+/* the order in this array is important. it matches the ordering of the
+ * values in the FW so the index is the same value as in ice_fwlog_level
+ */
+static const char * const ice_fwlog_log_size[] = {
+ "128K",
+ "256K",
+ "512K",
+ "1M",
+ "2M",
+};
+
/**
* ice_fwlog_print_module_cfg - print current FW logging module configuration
* @hw: pointer to the HW structure
@@ -376,6 +387,199 @@ static const struct file_operations ice_debugfs_enable_fops = {
.write = ice_debugfs_enable_write,
};
+/**
+ * ice_debugfs_log_size_read - read from 'log_size' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_log_size_read(struct file *filp,
+ char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char buff[32] = {};
+ int index;
+
+ index = hw->fwlog_ring.index;
+ snprintf(buff, sizeof(buff), "%s\n", ice_fwlog_log_size[index]);
+
+ return simple_read_from_buffer(buffer, count, ppos, buff, strlen(buff));
+}
+
+/**
+ * ice_debugfs_log_size_write - write into 'log_size' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_log_size_write(struct file *filp, const char __user *buf,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ char user_val[8], *cmd_buf;
+ ssize_t ret;
+ int index;
+
+ /* don't allow partial writes or invalid input */
+ if (*ppos != 0 || count > 5)
+ return -EINVAL;
+
+ cmd_buf = memdup_user(buf, count);
+ if (IS_ERR(cmd_buf))
+ return PTR_ERR(cmd_buf);
+
+ ret = sscanf(cmd_buf, "%s", user_val);
+ if (ret != 1)
+ return -EINVAL;
+
+ index = sysfs_match_string(ice_fwlog_log_size, user_val);
+ if (index < 0) {
+ dev_info(dev, "Invalid log size '%s'. The value must be one of 128K, 256K, 512K, 1M, 2M\n",
+ user_val);
+ ret = -EINVAL;
+ goto log_size_write_error;
+ } else if (hw->fwlog_cfg.options & ICE_FWLOG_OPTION_IS_REGISTERED) {
+ dev_info(dev, "FW logging is currently running. Please disable FW logging to change log_size\n");
+ ret = -EINVAL;
+ goto log_size_write_error;
+ }
+
+ /* free all the buffers and the tracking info and resize */
+ ice_fwlog_realloc_rings(hw, index);
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+log_size_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_log_size_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_log_size_read,
+ .write = ice_debugfs_log_size_write,
+};
+
+/**
+ * ice_debugfs_data_read - read from 'data' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_data_read(struct file *filp, char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ int data_copied = 0;
+ bool done = false;
+
+ if (ice_fwlog_ring_empty(&hw->fwlog_ring))
+ return 0;
+
+ while (!ice_fwlog_ring_empty(&hw->fwlog_ring) && !done) {
+ struct ice_fwlog_data *log;
+ u16 cur_buf_len;
+
+ log = &hw->fwlog_ring.rings[hw->fwlog_ring.head];
+ cur_buf_len = log->data_size;
+ if (cur_buf_len >= count) {
+ done = true;
+ continue;
+ }
+
+ if (copy_to_user(buffer, log->data, cur_buf_len)) {
+ /* if there is an error then bail and return whatever
+ * the driver has copied so far
+ */
+ done = true;
+ continue;
+ }
+
+ data_copied += cur_buf_len;
+ buffer += cur_buf_len;
+ count -= cur_buf_len;
+ *ppos += cur_buf_len;
+ ice_fwlog_ring_increment(&hw->fwlog_ring.head,
+ hw->fwlog_ring.size);
+ }
+
+ return data_copied;
+}
+
+/**
+ * ice_debugfs_data_write - write into 'data' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_data_write(struct file *filp, const char __user *buf, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ ssize_t ret;
+
+ /* don't allow partial writes */
+ if (*ppos != 0)
+ return 0;
+
+ /* any value is allowed to clear the buffer so no need to even look at
+ * what the value is
+ */
+ if (!(hw->fwlog_cfg.options & ICE_FWLOG_OPTION_IS_REGISTERED)) {
+ hw->fwlog_ring.head = 0;
+ hw->fwlog_ring.tail = 0;
+ } else {
+ dev_info(dev, "Can't clear FW log data while FW log running\n");
+ ret = -EINVAL;
+ goto nr_buffs_write_error;
+ }
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+nr_buffs_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_data_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_data_read,
+ .write = ice_debugfs_data_write,
+};
+
/**
* ice_debugfs_fwlog_init - setup the debugfs directory
* @pf: the ice that is starting up
@@ -430,6 +634,12 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
debugfs_create_file("enable", 0600, pf->ice_debugfs_pf_fwlog,
pf, &ice_debugfs_enable_fops);
+ debugfs_create_file("log_size", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_log_size_fops);
+
+ debugfs_create_file("data", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_data_fops);
+
return;
err_create_module_files:
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 25a17cbc1d34..92b5dac481cd 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -1,10 +1,128 @@
// SPDX-License-Identifier: GPL-2.0
/* Copyright (c) 2022, Intel Corporation. */
+#include <linux/vmalloc.h>
#include "ice.h"
#include "ice_common.h"
#include "ice_fwlog.h"
+bool ice_fwlog_ring_full(struct ice_fwlog_ring *rings)
+{
+ u16 head, tail;
+
+ head = rings->head;
+ tail = rings->tail;
+
+ if (head < tail && (tail - head == (rings->size - 1)))
+ return true;
+ else if (head > tail && (tail == (head - 1)))
+ return true;
+
+ return false;
+}
+
+bool ice_fwlog_ring_empty(struct ice_fwlog_ring *rings)
+{
+ return rings->head == rings->tail;
+}
+
+void ice_fwlog_ring_increment(u16 *item, u16 size)
+{
+ *item = (*item + 1) & (size - 1);
+}
+
+static int ice_fwlog_alloc_ring_buffs(struct ice_fwlog_ring *rings)
+{
+ int i, nr_bytes;
+ u8 *mem;
+
+ nr_bytes = rings->size * ICE_AQ_MAX_BUF_LEN;
+ mem = vzalloc(nr_bytes);
+ if (!mem)
+ return -ENOMEM;
+
+ for (i = 0; i < rings->size; i++) {
+ struct ice_fwlog_data *ring = &rings->rings[i];
+
+ ring->data_size = ICE_AQ_MAX_BUF_LEN;
+ ring->data = mem;
+ mem += ICE_AQ_MAX_BUF_LEN;
+ }
+
+ return 0;
+}
+
+static void ice_fwlog_free_ring_buffs(struct ice_fwlog_ring *rings)
+{
+ int i;
+
+ for (i = 0; i < rings->size; i++) {
+ struct ice_fwlog_data *ring = &rings->rings[i];
+
+ /* the first ring is the base memory for the whole range so
+ * free it
+ */
+ if (!i)
+ vfree(ring->data);
+
+ ring->data = NULL;
+ ring->data_size = 0;
+ }
+}
+
+#define ICE_FWLOG_INDEX_TO_BYTES(n) ((128 * 1024) << (n))
+/**
+ * ice_fwlog_realloc_rings - reallocate the FW log rings
+ * @hw: pointer to the HW structure
+ * @index: the new index to use to allocate memory for the log data
+ *
+ */
+void ice_fwlog_realloc_rings(struct ice_hw *hw, int index)
+{
+ struct ice_fwlog_ring ring;
+ int status, ring_size;
+
+ /* convert the number of bytes into a number of 4K buffers. externally
+ * the driver presents the interface to the FW log data as a number of
+ * bytes because that's easy for users to understand. internally the
+ * driver uses a ring of buffers because the driver doesn't know where
+ * the beginning and end of any line of log data is so the driver has
+ * to overwrite data as complete blocks. when the data is returned to
+ * the user the driver knows that the data is correct and the FW log
+ * can be correctly parsed by the tools
+ */
+ ring_size = ICE_FWLOG_INDEX_TO_BYTES(index) / ICE_AQ_MAX_BUF_LEN;
+ if (ring_size == hw->fwlog_ring.size)
+ return;
+
+ /* allocate space for the new rings and buffers then release the
+ * old rings and buffers. that way if we don't have enough
+ * memory then we at least have what we had before
+ */
+ ring.rings = kcalloc(ring_size, sizeof(*ring.rings), GFP_KERNEL);
+ if (!ring.rings)
+ return;
+
+ ring.size = ring_size;
+
+ status = ice_fwlog_alloc_ring_buffs(&ring);
+ if (status) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log ring data buffers\n");
+ ice_fwlog_free_ring_buffs(&ring);
+ kfree(ring.rings);
+ return;
+ }
+
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+
+ hw->fwlog_ring.rings = ring.rings;
+ hw->fwlog_ring.size = ring.size;
+ hw->fwlog_ring.index = index;
+ hw->fwlog_ring.head = 0;
+ hw->fwlog_ring.tail = 0;
+}
+
/**
* ice_fwlog_init - Initialize FW logging configuration
* @hw: pointer to the HW structure
@@ -28,6 +146,25 @@ int ice_fwlog_init(struct ice_hw *hw)
if (status)
return status;
+ hw->fwlog_ring.rings = kcalloc(ICE_FWLOG_RING_SIZE_DFLT,
+ sizeof(*hw->fwlog_ring.rings),
+ GFP_KERNEL);
+ if (!hw->fwlog_ring.rings) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log rings\n");
+ return -ENOMEM;
+ }
+
+ hw->fwlog_ring.size = ICE_FWLOG_RING_SIZE_DFLT;
+ hw->fwlog_ring.index = ICE_FWLOG_RING_SIZE_INDEX_DFLT;
+
+ status = ice_fwlog_alloc_ring_buffs(&hw->fwlog_ring);
+ if (status) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log ring data buffers\n");
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+ return status;
+ }
+
ice_debugfs_fwlog_init(hw->back);
} else {
dev_warn(ice_hw_to_dev(hw), "FW logging is not supported in this NVM image. Please update the NVM to get FW log support\n");
@@ -68,6 +205,11 @@ void ice_fwlog_deinit(struct ice_hw *hw)
if (status)
dev_warn(ice_hw_to_dev(hw), "Unable to unregister FW logging, status: %d\n",
status);
+
+ if (hw->fwlog_ring.rings) {
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+ }
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.h b/drivers/net/ethernet/intel/ice/ice_fwlog.h
index 45865558425d..287e71fa4b86 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.h
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.h
@@ -47,6 +47,26 @@ struct ice_fwlog_cfg {
u16 log_resolution;
};
+struct ice_fwlog_data {
+ u16 data_size;
+ u8 *data;
+};
+
+struct ice_fwlog_ring {
+ struct ice_fwlog_data *rings;
+ u16 index;
+ u16 size;
+ u16 head;
+ u16 tail;
+};
+
+#define ICE_FWLOG_RING_SIZE_INDEX_DFLT 3
+#define ICE_FWLOG_RING_SIZE_DFLT 256
+#define ICE_FWLOG_RING_SIZE_MAX 512
+
+bool ice_fwlog_ring_full(struct ice_fwlog_ring *rings);
+bool ice_fwlog_ring_empty(struct ice_fwlog_ring *rings);
+void ice_fwlog_ring_increment(u16 *item, u16 size);
void ice_fwlog_set_supported(struct ice_hw *hw);
bool ice_fwlog_supported(struct ice_hw *hw);
int ice_fwlog_init(struct ice_hw *hw);
@@ -55,4 +75,5 @@ int ice_fwlog_set(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_register(struct ice_hw *hw);
int ice_fwlog_unregister(struct ice_hw *hw);
+void ice_fwlog_realloc_rings(struct ice_hw *hw, int index);
#endif /* _ICE_FWLOG_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 614e10ab4159..6c6ca5353f28 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -1254,6 +1254,32 @@ ice_handle_link_event(struct ice_pf *pf, struct ice_rq_event_info *event)
return status;
}
+/**
+ * ice_get_fwlog_data - copy the FW log data from ARQ event
+ * @pf: PF that the FW log event is associated with
+ * @event: event structure containing FW log data
+ */
+static void
+ice_get_fwlog_data(struct ice_pf *pf, struct ice_rq_event_info *event)
+{
+ struct ice_fwlog_data *fwlog;
+ struct ice_hw *hw = &pf->hw;
+
+ fwlog = &hw->fwlog_ring.rings[hw->fwlog_ring.tail];
+
+ memset(fwlog->data, 0, PAGE_SIZE);
+ fwlog->data_size = le16_to_cpu(event->desc.datalen);
+
+ memcpy(fwlog->data, event->msg_buf, fwlog->data_size);
+ ice_fwlog_ring_increment(&hw->fwlog_ring.tail, hw->fwlog_ring.size);
+
+ if (ice_fwlog_ring_full(&hw->fwlog_ring)) {
+ /* the rings are full so bump the head to create room */
+ ice_fwlog_ring_increment(&hw->fwlog_ring.head,
+ hw->fwlog_ring.size);
+ }
+}
+
/**
* ice_aq_prep_for_event - Prepare to wait for an AdminQ event from firmware
* @pf: pointer to the PF private structure
@@ -1535,6 +1561,9 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
ice_vc_process_vf_msg(pf, &event, &data);
break;
+ case ice_aqc_opc_fw_logs_event:
+ ice_get_fwlog_data(pf, &event);
+ break;
case ice_aqc_opc_lldp_set_mib_change:
ice_dcb_process_lldp_set_mib_change(pf, &event);
break;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 84bb61aa7409..28e47bb78eaf 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -875,6 +875,7 @@ struct ice_hw {
struct ice_fwlog_cfg fwlog_cfg;
bool fwlog_supported; /* does hardware support FW logging? */
+ struct ice_fwlog_ring fwlog_ring;
/* Device max aggregate bandwidths corresponding to the GL_PWR_MODE_CTL
* register. Used for determining the ITR/INTRL granularity during
--
2.43.0

98
kernel-rt/debian/patches/ice-VDF/0028-ice-Fix-debugfs-with-devlink-reload.patch Normal file
View File

@ -0,0 +1,98 @@
From 8a7f6d8b2105c39f236c51c558e21b787c223861 Mon Sep 17 00:00:00 2001
From: Wojciech Drewek <wojciech.drewek@intel.com>
Date: Mon, 5 Feb 2024 14:03:57 +0100
Subject: [PATCH 28/36] ice: Fix debugfs with devlink reload
During devlink reload it is needed to remove debugfs entries
correlated with only one PF. ice_debugfs_exit() removes all
entries created by ice driver so we can't use it.
Introduce ice_debugfs_pf_deinit() in order to release PF's
debugfs entries. Move ice_debugfs_exit() call to ice_module_exit(),
it makes more sense since ice_debugfs_init() is called in
ice_module_init() and not in ice_probe().
Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 500d0df5b4b2394a06b949bab05f7ed0242b9858)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 10 ++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 2 ++
drivers/net/ethernet/intel/ice/ice_main.c | 3 +--
4 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 7966ac61154c..ed1c6cdedeff 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -895,6 +895,7 @@ static inline bool ice_is_adq_active(struct ice_pf *pf)
}
void ice_debugfs_fwlog_init(struct ice_pf *pf);
+void ice_debugfs_pf_deinit(struct ice_pf *pf);
void ice_debugfs_init(void);
void ice_debugfs_exit(void);
void ice_pf_fwlog_update_module(struct ice_pf *pf, int log_level, int module);
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index c2bfba6b9ead..ba396b22bb7d 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -647,6 +647,16 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
kfree(fw_modules);
}
+/**
+ * ice_debugfs_pf_deinit - cleanup PF's debugfs
+ * @pf: pointer to the PF struct
+ */
+void ice_debugfs_pf_deinit(struct ice_pf *pf)
+{
+ debugfs_remove_recursive(pf->ice_debugfs_pf);
+ pf->ice_debugfs_pf = NULL;
+}
+
/**
* ice_debugfs_init - create root directory for debugfs entries
*/
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 92b5dac481cd..4fd15387a7e5 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -188,6 +188,8 @@ void ice_fwlog_deinit(struct ice_hw *hw)
if (hw->bus.func)
return;
+ ice_debugfs_pf_deinit(hw->back);
+
/* make sure FW logging is disabled to not put the FW in a weird state
* for the next driver load
*/
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 6c6ca5353f28..c882c218281a 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -5325,8 +5325,6 @@ static void ice_remove(struct pci_dev *pdev)
msleep(100);
}
- ice_debugfs_exit();
-
if (test_bit(ICE_FLAG_SRIOV_ENA, pf->flags)) {
set_bit(ICE_VF_RESETS_DISABLED, pf->state);
ice_free_vfs(pf);
@@ -5823,6 +5821,7 @@ module_init(ice_module_init);
static void __exit ice_module_exit(void)
{
pci_unregister_driver(&ice_driver);
+ ice_debugfs_exit();
destroy_workqueue(ice_wq);
destroy_workqueue(ice_lag_wq);
pr_info("module unloaded\n");
--
2.43.0

78
kernel-rt/debian/patches/ice-VDF/0029-ice-remove-vf-lan_vsi_num-field.patch Normal file
View File

@ -0,0 +1,78 @@
From 861015cbb4cf4cb258a1da9e80550fe991be7808 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 16 Feb 2024 14:06:38 -0800
Subject: [PATCH 29/36] ice: remove vf->lan_vsi_num field
The lan_vsi_num field of the VF structure is no longer used for any
purpose. Remove it.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1cf94cbfc61bac89cddeb075fbc100ebd3aea81b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_sriov.c | 1 -
drivers/net/ethernet/intel/ice/ice_vf_lib.c | 4 +---
drivers/net/ethernet/intel/ice/ice_vf_lib.h | 5 -----
3 files changed, 1 insertion(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_sriov.c b/drivers/net/ethernet/intel/ice/ice_sriov.c
index 442162be23ea..3366ac976c44 100644
--- a/drivers/net/ethernet/intel/ice/ice_sriov.c
+++ b/drivers/net/ethernet/intel/ice/ice_sriov.c
@@ -239,7 +239,6 @@ static struct ice_vsi *ice_vf_vsi_setup(struct ice_vf *vf)
}
vf->lan_vsi_idx = vsi->idx;
- vf->lan_vsi_num = vsi->vsi_num;
return vsi;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_vf_lib.c b/drivers/net/ethernet/intel/ice/ice_vf_lib.c
index 03b9d7d74851..303fdf8555cf 100644
--- a/drivers/net/ethernet/intel/ice/ice_vf_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_vf_lib.c
@@ -298,7 +298,6 @@ static int ice_vf_rebuild_vsi(struct ice_vf *vf)
* vf->lan_vsi_idx
*/
vsi->vsi_num = ice_get_hw_vsi_num(&pf->hw, vsi->idx);
- vf->lan_vsi_num = vsi->vsi_num;
return 0;
}
@@ -1299,13 +1298,12 @@ int ice_vf_init_host_cfg(struct ice_vf *vf, struct ice_vsi *vsi)
}
/**
- * ice_vf_invalidate_vsi - invalidate vsi_idx/vsi_num to remove VSI access
+ * ice_vf_invalidate_vsi - invalidate vsi_idx to remove VSI access
* @vf: VF to remove access to VSI for
*/
void ice_vf_invalidate_vsi(struct ice_vf *vf)
{
vf->lan_vsi_idx = ICE_NO_VSI;
- vf->lan_vsi_num = ICE_NO_VSI;
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_vf_lib.h b/drivers/net/ethernet/intel/ice/ice_vf_lib.h
index 48fea6fa0362..1de07accbc5c 100644
--- a/drivers/net/ethernet/intel/ice/ice_vf_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_vf_lib.h
@@ -110,11 +110,6 @@ struct ice_vf {
u8 spoofchk:1;
u8 link_forced:1;
u8 link_up:1; /* only valid if VF link is forced */
- /* VSI indices - actual VSI pointers are maintained in the PF structure
- * When assigned, these will be non-zero, because VSI 0 is always
- * the main LAN VSI for the PF.
- */
- u16 lan_vsi_num; /* ID as used by firmware */
unsigned int min_tx_rate; /* Minimum Tx bandwidth limit in Mbps */
unsigned int max_tx_rate; /* Maximum Tx bandwidth limit in Mbps */
DECLARE_BITMAP(vf_states, ICE_VF_STATES_NBITS); /* VF runtime states */
--
2.43.0

150
kernel-rt/debian/patches/ice-VDF/0030-ice-rename-ice_write_-functions-to-ice_pack_ctx_.patch Normal file
View File

@ -0,0 +1,150 @@
From 6b7fae8669a04943af9f83ef89d39a922ed179fd Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:54 -0800
Subject: [PATCH 30/36] ice: rename ice_write_* functions to ice_pack_ctx_*
In ice_common.c there are 4 functions used for converting the unpacked
software Tx and Rx context structure data into the packed format used by
hardware. These functions have extremely generic names:
* ice_write_byte
* ice_write_word
* ice_write_dword
* ice_write_qword
When I saw these function names my first thought was "write what? to
where?". Understanding what these functions do requires looking at the
implementation details. The functions take bits from an unpacked structure
and copy them into the packed layout used by hardware.
As part of live migration, we will want functions which perform the inverse
operation of reading bits from the packed layout and copying them into the
unpacked format. Naming these as "ice_read_byte", etc would be very
confusing since they appear to write data.
In preparation for adding this new inverse operation, rename the existing
functions to use the prefix "ice_pack_ctx_". This makes it clear that they
perform the bit packing while copying from the unpacked software context
structure to the packed hardware context.
The inverse operations can then neatly be named ice_unpack_ctx_*, clearly
indicating they perform the bit unpacking while copying from the packed
hardware context to the unpacked software context structure.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1260b45dbe2dbc415f3bc1e841c6c098083bcfb8)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 56 ++++++++++-----------
1 file changed, 28 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 6dcba0577633..17f60a98c8ed 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -4267,13 +4267,13 @@ ice_aq_add_rdma_qsets(struct ice_hw *hw, u8 num_qset_grps,
/* End of FW Admin Queue command wrappers */
/**
- * ice_write_byte - write a byte to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_byte - write a byte to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_byte(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_byte(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u8 src_byte, dest_byte, mask;
u8 *from, *dest;
@@ -4306,13 +4306,13 @@ ice_write_byte(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_word - write a word to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_word - write a word to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_word(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_word(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u16 src_word, mask;
__le16 dest_word;
@@ -4349,13 +4349,13 @@ ice_write_word(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_dword - write a dword to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_dword - write a dword to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_dword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_dword(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u32 src_dword, mask;
__le32 dest_dword;
@@ -4400,13 +4400,13 @@ ice_write_dword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_qword - write a qword to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_qword - write a qword to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_qword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u64 src_qword, mask;
__le64 dest_qword;
@@ -4475,16 +4475,16 @@ ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
}
switch (ce_info[f].size_of) {
case sizeof(u8):
- ice_write_byte(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_byte(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u16):
- ice_write_word(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_word(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u32):
- ice_write_dword(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_dword(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u64):
- ice_write_qword(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_qword(src_ctx, dest_ctx, &ce_info[f]);
break;
default:
return -EINVAL;
--
2.43.0

129
kernel-rt/debian/patches/ice-VDF/0031-ice-use-GENMASK-instead-of-BIT-n-1-in-pack-functions.patch Normal file
View File

@ -0,0 +1,129 @@
From 619e0e61b39cf051137613459d36c4fe8f435e57 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:55 -0800
Subject: [PATCH 31/36] ice: use GENMASK instead of BIT(n) - 1 in pack
functions
The functions used to pack the Tx and Rx context into the hardware format
rely on using BIT() and then subtracting 1 to get a bitmask. These
functions even have a comment about how x86 machines can't use this method
for certain widths because the SHL instructions will not work properly.
The Linux kernel already provides the GENMASK macro for generating a
suitable bitmask. Further, GENMASK is capable of generating the mask
including the shift_width. Since width is the total field width, take care
to subtract one to get the final bit position.
Since we now include the shifted bits as part of the mask, shift the source
value first before applying the mask.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit a45d1bf516c097bb7ae4983d3128ebf139be952c)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 44 ++++-----------------
1 file changed, 8 insertions(+), 36 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 17f60a98c8ed..55a2e264dd69 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -4284,14 +4284,11 @@ static void ice_pack_ctx_byte(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
- mask = (u8)(BIT(ce_info->width) - 1);
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
src_byte = *from;
- src_byte &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_byte <<= shift_width;
+ src_byte &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4324,17 +4321,14 @@ static void ice_pack_ctx_word(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
- mask = BIT(ce_info->width) - 1;
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_word = *(u16 *)from;
- src_word &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_word <<= shift_width;
+ src_word &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4367,25 +4361,14 @@ static void ice_pack_ctx_dword(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
-
- /* if the field width is exactly 32 on an x86 machine, then the shift
- * operation will not work because the SHL instructions count is masked
- * to 5 bits so the shift will do nothing
- */
- if (ce_info->width < 32)
- mask = BIT(ce_info->width) - 1;
- else
- mask = (u32)~0;
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_dword = *(u32 *)from;
- src_dword &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_dword <<= shift_width;
+ src_dword &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4418,25 +4401,14 @@ static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
-
- /* if the field width is exactly 64 on an x86 machine, then the shift
- * operation will not work because the SHL instructions count is masked
- * to 6 bits so the shift will do nothing
- */
- if (ce_info->width < 64)
- mask = BIT_ULL(ce_info->width) - 1;
- else
- mask = (u64)~0;
+ mask = GENMASK_ULL(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_qword = *(u64 *)from;
- src_qword &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_qword <<= shift_width;
+ src_qword &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
--
2.43.0

89
kernel-rt/debian/patches/ice-VDF/0032-ice-cleanup-line-splitting-for-context-set-functions.patch Normal file
View File

@ -0,0 +1,89 @@
From 6502dd63a7bd436c72d2ee8b328985b93fa7e2a5 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:56 -0800
Subject: [PATCH 32/36] ice: cleanup line splitting for context set functions
The indentation for ice_set_ctx and ice_write_rxq_ctx breaks the function
name after the return type. This style of breaking is used a lot throughout
the ice driver, even in cases where its not actually helpful for
readability. We no longer prefer this style of line splitting in the
driver, and new code is avoiding it.
Normally, I would leave this alone unless the actual function contents or
description needed updating. However, a future change is going to add
inverse functions for converting packed context to unpacked context
structures. To keep this code uniform with the existing set functions, fix
up the style to the modern format of keeping the type on the same line.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 979c2c049fbea107ce9f8d31f3ba9dba83ddb0a2)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 12 +++++-------
drivers/net/ethernet/intel/ice/ice_common.h | 10 ++++------
2 files changed, 9 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 55a2e264dd69..59ede77a1473 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -1329,9 +1329,8 @@ static const struct ice_ctx_ele ice_rlan_ctx_info[] = {
* it to HW register space and enables the hardware to prefetch descriptors
* instead of only fetching them on demand
*/
-int
-ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
- u32 rxq_index)
+int ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
+ u32 rxq_index)
{
u8 ctx_buf[ICE_RXQ_CTX_SZ] = { 0 };
@@ -4427,11 +4426,10 @@ static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
* @hw: pointer to the hardware structure
* @src_ctx: pointer to a generic non-packed context structure
* @dest_ctx: pointer to memory for the packed structure
- * @ce_info: a description of the structure to be transformed
+ * @ce_info: List of Rx context elements
*/
-int
-ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
- const struct ice_ctx_ele *ce_info)
+int ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
int f;
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index d47e5400351f..1c3c29d30815 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -52,9 +52,8 @@ int ice_get_caps(struct ice_hw *hw);
void ice_set_safe_mode_caps(struct ice_hw *hw);
-int
-ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
- u32 rxq_index);
+int ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
+ u32 rxq_index);
int
ice_aq_get_rss_lut(struct ice_hw *hw, struct ice_aq_get_set_rss_lut_params *get_params);
@@ -71,9 +70,8 @@ bool ice_check_sq_alive(struct ice_hw *hw, struct ice_ctl_q_info *cq);
int ice_aq_q_shutdown(struct ice_hw *hw, bool unloading);
void ice_fill_dflt_direct_cmd_desc(struct ice_aq_desc *desc, u16 opcode);
extern const struct ice_ctx_ele ice_tlan_ctx_info[];
-int
-ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
- const struct ice_ctx_ele *ce_info);
+int ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info);
extern struct mutex ice_global_cfg_lock_sw;
--
2.43.0

161
kernel-rt/debian/patches/ice-VDF/0033-ice-do-not-disable-Tx-queues-twice-in-ice_down.patch Normal file
View File

@ -0,0 +1,161 @@
From 675a8843a0de1411666389eeabeea452161f8cc5 Mon Sep 17 00:00:00 2001
From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Date: Fri, 23 Feb 2024 17:06:27 +0100
Subject: [PATCH 33/36] ice: do not disable Tx queues twice in ice_down()
ice_down() clears QINT_TQCTL_CAUSE_ENA_M bit twice, which is not
necessary. First clearing happens in ice_vsi_dis_irq() and second in
ice_vsi_stop_tx_ring() - remove the first one.
While at it, make ice_vsi_dis_irq() static as ice_down() is the only
current caller of it.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit d5926e01e3739542bb047b77f850d7f641eaa7bc)
[Adjust ice_lib.c with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_lib.c | 55 -----------------------
drivers/net/ethernet/intel/ice/ice_lib.h | 2 -
drivers/net/ethernet/intel/ice/ice_main.c | 44 ++++++++++++++++++
3 files changed, 44 insertions(+), 57 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 106ef843f4b5..f23cb9c8e3dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -2877,61 +2877,6 @@ void ice_dis_vsi(struct ice_vsi *vsi, bool locked)
}
}
-/**
- * ice_vsi_dis_irq - Mask off queue interrupt generation on the VSI
- * @vsi: the VSI being un-configured
- */
-void ice_vsi_dis_irq(struct ice_vsi *vsi)
-{
- struct ice_pf *pf = vsi->back;
- struct ice_hw *hw = &pf->hw;
- u32 val;
- int i;
-
- /* disable interrupt causation from each queue */
- if (vsi->tx_rings) {
- ice_for_each_txq(vsi, i) {
- if (vsi->tx_rings[i]) {
- u16 reg;
-
- reg = vsi->tx_rings[i]->reg_idx;
- val = rd32(hw, QINT_TQCTL(reg));
- val &= ~QINT_TQCTL_CAUSE_ENA_M;
- wr32(hw, QINT_TQCTL(reg), val);
- }
- }
- }
-
- if (vsi->rx_rings) {
- ice_for_each_rxq(vsi, i) {
- if (vsi->rx_rings[i]) {
- u16 reg;
-
- reg = vsi->rx_rings[i]->reg_idx;
- val = rd32(hw, QINT_RQCTL(reg));
- val &= ~QINT_RQCTL_CAUSE_ENA_M;
- wr32(hw, QINT_RQCTL(reg), val);
- }
- }
- }
-
- /* disable each interrupt */
- ice_for_each_q_vector(vsi, i) {
- if (!vsi->q_vectors[i])
- continue;
- wr32(hw, GLINT_DYN_CTL(vsi->q_vectors[i]->reg_idx), 0);
- }
-
- ice_flush(hw);
-
- /* don't call synchronize_irq() for VF's from the host */
- if (vsi->type == ICE_VSI_VF)
- return;
-
- ice_for_each_q_vector(vsi, i)
- synchronize_irq(vsi->q_vectors[i]->irq.virq);
-}
-
/**
* ice_vsi_release - Delete a VSI and free its resources
* @vsi: the VSI being removed
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h
index f24f5d1e6f9c..dbd0f3409323 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_lib.h
@@ -110,8 +110,6 @@ void
ice_write_qrxflxp_cntxt(struct ice_hw *hw, u16 pf_q, u32 rxdid, u32 prio,
bool ena_ts);
-void ice_vsi_dis_irq(struct ice_vsi *vsi);
-
void ice_vsi_free_irq(struct ice_vsi *vsi);
void ice_vsi_free_rx_rings(struct ice_vsi *vsi);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index c882c218281a..685635a22616 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7023,6 +7023,50 @@ static void ice_napi_disable_all(struct ice_vsi *vsi)
}
}
+/**
+ * ice_vsi_dis_irq - Mask off queue interrupt generation on the VSI
+ * @vsi: the VSI being un-configured
+ */
+static void ice_vsi_dis_irq(struct ice_vsi *vsi)
+{
+ struct ice_pf *pf = vsi->back;
+ struct ice_hw *hw = &pf->hw;
+ u32 val;
+ int i;
+
+ /* disable interrupt causation from each Rx queue; Tx queues are
+ * handled in ice_vsi_stop_tx_ring()
+ */
+ if (vsi->rx_rings) {
+ ice_for_each_rxq(vsi, i) {
+ if (vsi->rx_rings[i]) {
+ u16 reg;
+
+ reg = vsi->rx_rings[i]->reg_idx;
+ val = rd32(hw, QINT_RQCTL(reg));
+ val &= ~QINT_RQCTL_CAUSE_ENA_M;
+ wr32(hw, QINT_RQCTL(reg), val);
+ }
+ }
+ }
+
+ /* disable each interrupt */
+ ice_for_each_q_vector(vsi, i) {
+ if (!vsi->q_vectors[i])
+ continue;
+ wr32(hw, GLINT_DYN_CTL(vsi->q_vectors[i]->reg_idx), 0);
+ }
+
+ ice_flush(hw);
+
+ /* don't call synchronize_irq() for VF's from the host */
+ if (vsi->type == ICE_VSI_VF)
+ return;
+
+ ice_for_each_q_vector(vsi, i)
+ synchronize_irq(vsi->q_vectors[i]->irq.virq);
+}
+
/**
* ice_down - Shutdown the connection
* @vsi: The VSI being stopped
--
2.43.0

274
kernel-rt/debian/patches/ice-VDF/0034-ice-Fix-improper-extts-handling.patch Normal file
View File

@ -0,0 +1,274 @@
From e53280c20bbe58015a91178268244d5e831276f4 Mon Sep 17 00:00:00 2001
From: Milena Olech <milena.olech@intel.com>
Date: Tue, 2 Jul 2024 10:14:54 -0700
Subject: [PATCH 34/36] ice: Fix improper extts handling
Extts events are disabled and enabled by the application ts2phc.
However, in case where the driver is removed when the application is
running, a specific extts event remains enabled and can cause a kernel
crash.
As a side effect, when the driver is reloaded and application is started
again, remaining extts event for the channel from a previous run will
keep firing and the message "extts on unexpected channel" might be
printed to the user.
To avoid that, extts events shall be disabled when PTP is released.
Fixes: 172db5f91d5f ("ice: add support for auxiliary input/output pins")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Co-developed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Milena Olech <milena.olech@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20240702171459.2606611-2-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 00d3b4f54582d4e4a02cda5886bb336eeab268cc)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 105 ++++++++++++++++++-----
drivers/net/ethernet/intel/ice/ice_ptp.h | 8 ++
2 files changed, 91 insertions(+), 22 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 48ec59fc5d87..6e06c5d596b9 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1603,27 +1603,24 @@ void ice_ptp_extts_event(struct ice_pf *pf)
/**
* ice_ptp_cfg_extts - Configure EXTTS pin and channel
* @pf: Board private structure
- * @ena: true to enable; false to disable
* @chan: GPIO channel (0-3)
- * @gpio_pin: GPIO pin
- * @extts_flags: request flags from the ptp_extts_request.flags
+ * @config: desired EXTTS configuration.
+ * @store: If set to true, the values will be stored
+ *
+ * Configure an external timestamp event on the requested channel.
*/
-static int
-ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
- unsigned int extts_flags)
+static void ice_ptp_cfg_extts(struct ice_pf *pf, unsigned int chan,
+ struct ice_extts_channel *config, bool store)
{
u32 func, aux_reg, gpio_reg, irq_reg;
struct ice_hw *hw = &pf->hw;
u8 tmr_idx;
- if (chan > (unsigned int)pf->ptp.info.n_ext_ts)
- return -EINVAL;
-
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
irq_reg = rd32(hw, PFINT_OICR_ENA);
- if (ena) {
+ if (config->ena) {
/* Enable the interrupt */
irq_reg |= PFINT_OICR_TSYN_EVNT_M;
aux_reg = GLTSYN_AUX_IN_0_INT_ENA_M;
@@ -1632,9 +1629,9 @@ ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
#define GLTSYN_AUX_IN_0_EVNTLVL_FALLING_EDGE BIT(1)
/* set event level to requested edge */
- if (extts_flags & PTP_FALLING_EDGE)
+ if (config->flags & PTP_FALLING_EDGE)
aux_reg |= GLTSYN_AUX_IN_0_EVNTLVL_FALLING_EDGE;
- if (extts_flags & PTP_RISING_EDGE)
+ if (config->flags & PTP_RISING_EDGE)
aux_reg |= GLTSYN_AUX_IN_0_EVNTLVL_RISING_EDGE;
/* Write GPIO CTL reg.
@@ -1656,9 +1653,47 @@ ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
wr32(hw, PFINT_OICR_ENA, irq_reg);
wr32(hw, GLTSYN_AUX_IN(chan, tmr_idx), aux_reg);
- wr32(hw, GLGEN_GPIO_CTL(gpio_pin), gpio_reg);
+ wr32(hw, GLGEN_GPIO_CTL(config->gpio_pin), gpio_reg);
- return 0;
+ if (store)
+ memcpy(&pf->ptp.extts_channels[chan], config, sizeof(*config));
+}
+
+/**
+ * ice_ptp_disable_all_extts - Disable all EXTTS channels
+ * @pf: Board private structure
+ */
+static void ice_ptp_disable_all_extts(struct ice_pf *pf)
+{
+ struct ice_extts_channel extts_cfg = {};
+ int i;
+
+ for (i = 0; i < pf->ptp.info.n_ext_ts; i++) {
+ if (pf->ptp.extts_channels[i].ena) {
+ extts_cfg.gpio_pin = pf->ptp.extts_channels[i].gpio_pin;
+ extts_cfg.ena = false;
+ ice_ptp_cfg_extts(pf, i, &extts_cfg, false);
+ }
+ }
+
+ synchronize_irq(pf->oicr_irq.virq);
+}
+
+/**
+ * ice_ptp_enable_all_extts - Enable all EXTTS channels
+ * @pf: Board private structure
+ *
+ * Called during reset to restore user configuration.
+ */
+static void ice_ptp_enable_all_extts(struct ice_pf *pf)
+{
+ int i;
+
+ for (i = 0; i < pf->ptp.info.n_ext_ts; i++) {
+ if (pf->ptp.extts_channels[i].ena)
+ ice_ptp_cfg_extts(pf, i, &pf->ptp.extts_channels[i],
+ false);
+ }
}
/**
@@ -1815,7 +1850,6 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
struct ptp_clock_request *rq, int on)
{
struct ice_pf *pf = ptp_info_to_pf(info);
- struct ice_perout_channel clk_cfg = {0};
bool sma_pres = false;
unsigned int chan;
u32 gpio_pin;
@@ -1826,6 +1860,9 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
switch (rq->type) {
case PTP_CLK_REQ_PEROUT:
+ {
+ struct ice_perout_channel clk_cfg = {};
+
chan = rq->perout.index;
if (sma_pres) {
if (chan == ice_pin_desc_e810t[SMA1].chan)
@@ -1853,7 +1890,11 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
err = ice_ptp_cfg_clkout(pf, chan, &clk_cfg, true);
break;
+ }
case PTP_CLK_REQ_EXTTS:
+ {
+ struct ice_extts_channel extts_cfg = {};
+
chan = rq->extts.index;
if (sma_pres) {
if (chan < ice_pin_desc_e810t[SMA2].chan)
@@ -1869,9 +1910,13 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
gpio_pin = chan;
}
- err = ice_ptp_cfg_extts(pf, !!on, chan, gpio_pin,
- rq->extts.flags);
- break;
+ extts_cfg.flags = rq->extts.flags;
+ extts_cfg.gpio_pin = gpio_pin;
+ extts_cfg.ena = !!on;
+
+ ice_ptp_cfg_extts(pf, chan, &extts_cfg, true);
+ return 0;
+ }
default:
return -EOPNOTSUPP;
}
@@ -1889,21 +1934,31 @@ static int ice_ptp_gpio_enable_e823(struct ptp_clock_info *info,
struct ptp_clock_request *rq, int on)
{
struct ice_pf *pf = ptp_info_to_pf(info);
- struct ice_perout_channel clk_cfg = {0};
int err;
switch (rq->type) {
case PTP_CLK_REQ_PPS:
+ {
+ struct ice_perout_channel clk_cfg = {};
+
clk_cfg.gpio_pin = PPS_PIN_INDEX;
clk_cfg.period = NSEC_PER_SEC;
clk_cfg.ena = !!on;
err = ice_ptp_cfg_clkout(pf, PPS_CLK_GEN_CHAN, &clk_cfg, true);
break;
+ }
case PTP_CLK_REQ_EXTTS:
- err = ice_ptp_cfg_extts(pf, !!on, rq->extts.index,
- TIME_SYNC_PIN_INDEX, rq->extts.flags);
- break;
+ {
+ struct ice_extts_channel extts_cfg = {};
+
+ extts_cfg.flags = rq->extts.flags;
+ extts_cfg.gpio_pin = TIME_SYNC_PIN_INDEX;
+ extts_cfg.ena = !!on;
+
+ ice_ptp_cfg_extts(pf, rq->extts.index, &extts_cfg, true);
+ return 0;
+ }
default:
return -EOPNOTSUPP;
}
@@ -2735,6 +2790,10 @@ static int ice_ptp_rebuild_owner(struct ice_pf *pf)
ice_ptp_restart_all_phy(pf);
}
+ /* Re-enable all periodic outputs and external timestamp events */
+ ice_ptp_enable_all_clkout(pf);
+ ice_ptp_enable_all_extts(pf);
+
return 0;
}
@@ -3286,6 +3345,8 @@ void ice_ptp_release(struct ice_pf *pf)
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
+ ice_ptp_disable_all_extts(pf);
+
kthread_cancel_delayed_work_sync(&pf->ptp.work);
ice_ptp_port_phy_stop(&pf->ptp.port);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 352405a2daf2..c6469a5a7afb 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -33,6 +33,12 @@ struct ice_perout_channel {
u64 start_time;
};
+struct ice_extts_channel {
+ bool ena;
+ u32 gpio_pin;
+ u32 flags;
+};
+
/* The ice hardware captures Tx hardware timestamps in the PHY. The timestamp
* is stored in a buffer of registers. Depending on the specific hardware,
* this buffer might be shared across multiple PHY ports.
@@ -226,6 +232,7 @@ enum ice_ptp_state {
* @ext_ts_irq: the external timestamp IRQ in use
* @kworker: kwork thread for handling periodic work
* @perout_channels: periodic output data
+ * @extts_channels: channels for external timestamps
* @info: structure defining PTP hardware capabilities
* @clock: pointer to registered PTP clock device
* @tstamp_config: hardware timestamping configuration
@@ -249,6 +256,7 @@ struct ice_ptp {
u8 ext_ts_irq;
struct kthread_worker *kworker;
struct ice_perout_channel perout_channels[GLTSYN_TGT_H_IDX_MAX];
+ struct ice_extts_channel extts_channels[GLTSYN_TGT_H_IDX_MAX];
struct ptp_clock_info info;
struct ptp_clock *clock;
struct hwtstamp_config tstamp_config;
--
2.43.0

49
kernel-rt/debian/patches/ice-VDF/0035-ice-Don-t-process-extts-if-PTP-is-disabled.patch Normal file
View File

@ -0,0 +1,49 @@
From 6c24a32820031f9713d0c0cf7ac6f4ead6b58052 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 2 Jul 2024 10:14:55 -0700
Subject: [PATCH 35/36] ice: Don't process extts if PTP is disabled
The ice_ptp_extts_event() function can race with ice_ptp_release() and
result in a NULL pointer dereference which leads to a kernel panic.
Panic occurs because the ice_ptp_extts_event() function calls
ptp_clock_event() with a NULL pointer. The ice driver has already
released the PTP clock by the time the interrupt for the next external
timestamp event occurs.
To fix this, modify the ice_ptp_extts_event() function to check the
PTP state and bail early if PTP is not ready.
Fixes: 172db5f91d5f ("ice: add support for auxiliary input/output pins")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20240702171459.2606611-3-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 996422e3230e41468f652d754fefd1bdbcd4604e)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 6e06c5d596b9..ceb4ba19c511 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1578,6 +1578,10 @@ void ice_ptp_extts_event(struct ice_pf *pf)
u8 chan, tmr_idx;
u32 hi, lo;
+ /* Don't process timestamp events if PTP is not ready */
+ if (pf->ptp.state != ICE_PTP_READY)
+ return;
+
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
/* Event time is captured by one of the two matched registers
* GLTSYN_EVNT_L: 32 LSB of sampled time event
--
2.43.0

628
kernel-rt/debian/patches/ice-VDF/0036-ice-Introduce-ice_ptp_hw-struct.patch Normal file
View File

@ -0,0 +1,628 @@
From 1ce01cb7cdb0bf4c18a546a62f224c8032d75ebd Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Tue, 28 May 2024 16:03:51 -0700
Subject: [PATCH 36/36] ice: Introduce ice_ptp_hw struct
Create new ice_ptp_hw struct and use it for all HW and PTP-related
fields from struct ice_hw.
Replace definitions with struct fields, which values are set accordingly
to a specific device.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528-next-2024-05-28-ptp-refactors-v1-1-c082739bb6f6@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit d551d075b043821880b8afc0010ef70d050716d0)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 24 ++++
drivers/net/ethernet/intel/ice/ice_common.h | 1 +
drivers/net/ethernet/intel/ice/ice_ptp.c | 22 ++--
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 134 ++++++++++++--------
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 4 +-
drivers/net/ethernet/intel/ice/ice_type.h | 17 ++-
6 files changed, 126 insertions(+), 76 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 59ede77a1473..147004e0170b 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -209,6 +209,30 @@ bool ice_is_e810t(struct ice_hw *hw)
return false;
}
+/**
+ * ice_is_e822 - Check if a device is E822 family device
+ * @hw: pointer to the hardware structure
+ *
+ * Return: true if the device is E822 based, false if not.
+ */
+bool ice_is_e822(struct ice_hw *hw)
+{
+ switch (hw->device_id) {
+ case ICE_DEV_ID_E822C_BACKPLANE:
+ case ICE_DEV_ID_E822C_QSFP:
+ case ICE_DEV_ID_E822C_SFP:
+ case ICE_DEV_ID_E822C_10G_BASE_T:
+ case ICE_DEV_ID_E822C_SGMII:
+ case ICE_DEV_ID_E822L_BACKPLANE:
+ case ICE_DEV_ID_E822L_SFP:
+ case ICE_DEV_ID_E822L_10G_BASE_T:
+ case ICE_DEV_ID_E822L_SGMII:
+ return true;
+ default:
+ return false;
+ }
+}
+
/**
* ice_is_e823
* @hw: pointer to the hardware structure
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index 1c3c29d30815..9d38777310e5 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -245,6 +245,7 @@ void
ice_stat_update32(struct ice_hw *hw, u32 reg, bool prev_stat_loaded,
u64 *prev_stat, u64 *cur_stat);
bool ice_is_e810t(struct ice_hw *hw);
+bool ice_is_e822(struct ice_hw *hw);
bool ice_is_e823(struct ice_hw *hw);
int
ice_sched_query_elem(struct ice_hw *hw, u32 node_teid,
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index ceb4ba19c511..bb1572a353d0 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -812,7 +812,7 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
}
mutex_unlock(&pf->ptp.ports_owner.lock);
- for (i = 0; i < ICE_MAX_QUAD; i++) {
+ for (i = 0; i < ICE_GET_QUAD_NUM(pf->hw.ptp.num_lports); i++) {
u64 tstamp_ready;
int err;
@@ -1026,7 +1026,7 @@ ice_ptp_release_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
static int
ice_ptp_init_tx_e82x(struct ice_pf *pf, struct ice_ptp_tx *tx, u8 port)
{
- tx->block = port / ICE_PORTS_PER_QUAD;
+ tx->block = ICE_GET_QUAD_NUM(port);
tx->offset = (port % ICE_PORTS_PER_QUAD) * INDEX_PER_PORT_E82X;
tx->len = INDEX_PER_PORT_E82X;
tx->has_ready_bitmap = 1;
@@ -1248,8 +1248,8 @@ static u64 ice_base_incval(struct ice_pf *pf)
*/
static int ice_ptp_check_tx_fifo(struct ice_ptp_port *port)
{
- int quad = port->port_num / ICE_PORTS_PER_QUAD;
int offs = port->port_num % ICE_PORTS_PER_QUAD;
+ int quad = ICE_GET_QUAD_NUM(port->port_num);
struct ice_pf *pf;
struct ice_hw *hw;
u32 val, phy_sts;
@@ -1448,7 +1448,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
if (pf->ptp.state != ICE_PTP_READY)
return;
- if (WARN_ON_ONCE(port >= ICE_NUM_EXTERNAL_PORTS))
+ if (WARN_ON_ONCE(port >= hw->ptp.num_lports))
return;
ptp_port = &pf->ptp.port;
@@ -1458,7 +1458,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
/* Update cached link status for this port immediately */
ptp_port->link_up = linkup;
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
/* Do not reconfigure E810 PHY */
return;
@@ -1487,7 +1487,7 @@ static int ice_ptp_cfg_phy_interrupt(struct ice_pf *pf, bool ena, u32 threshold)
ice_ptp_reset_ts_memory(hw);
- for (quad = 0; quad < ICE_MAX_QUAD; quad++) {
+ for (quad = 0; quad < ICE_GET_QUAD_NUM(hw->ptp.num_lports); quad++) {
err = ice_read_quad_reg_e82x(hw, quad, Q_REG_TX_MEM_GBL_CFG,
&val);
if (err)
@@ -2038,7 +2038,7 @@ ice_ptp_settime64(struct ptp_clock_info *info, const struct timespec64 *ts)
ice_ptp_enable_all_clkout(pf);
/* Recalibrate and re-enable timestamp blocks for E822/E823 */
- if (hw->phy_model == ICE_PHY_E82X)
+ if (hw->ptp.phy_model == ICE_PHY_E82X)
ice_ptp_restart_all_phy(pf);
exit:
if (err) {
@@ -2652,7 +2652,7 @@ static void ice_ptp_maybe_trigger_tx_interrupt(struct ice_pf *pf)
if (!ice_pf_src_tmr_owned(pf))
return;
- for (i = 0; i < ICE_MAX_QUAD; i++) {
+ for (i = 0; i < ICE_GET_QUAD_NUM(hw->ptp.num_lports); i++) {
u64 tstamp_ready;
int err;
@@ -3152,7 +3152,7 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
mutex_init(&ptp_port->ps_lock);
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E82X:
@@ -3245,7 +3245,7 @@ static void ice_ptp_remove_auxbus_device(struct ice_pf *pf)
*/
static void ice_ptp_init_tx_interrupt_mode(struct ice_pf *pf)
{
- switch (pf->hw.phy_model) {
+ switch (pf->hw.ptp.phy_model) {
case ICE_PHY_E82X:
/* E822 based PHY has the clock owner process the interrupt
* for all ports.
@@ -3281,7 +3281,7 @@ void ice_ptp_init(struct ice_pf *pf)
ptp->state = ICE_PTP_INITIALIZING;
- ice_ptp_init_phy_model(hw);
+ ice_ptp_init_hw(hw);
ice_ptp_init_tx_interrupt_mode(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index 7337e7e710ed..313a72dad813 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -285,18 +285,21 @@ static void ice_ptp_exec_tmr_cmd(struct ice_hw *hw)
/**
* ice_fill_phy_msg_e82x - Fill message data for a PHY register access
+ * @hw: pointer to the HW struct
* @msg: the PHY message buffer to fill in
* @port: the port to access
* @offset: the register offset
*/
-static void
-ice_fill_phy_msg_e82x(struct ice_sbq_msg_input *msg, u8 port, u16 offset)
+static void ice_fill_phy_msg_e82x(struct ice_hw *hw,
+ struct ice_sbq_msg_input *msg, u8 port,
+ u16 offset)
{
int phy_port, phy, quadtype;
- phy_port = port % ICE_PORTS_PER_PHY_E82X;
- phy = port / ICE_PORTS_PER_PHY_E82X;
- quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_QUADS_PER_PHY_E82X;
+ phy_port = port % hw->ptp.ports_per_phy;
+ phy = port / hw->ptp.ports_per_phy;
+ quadtype = ICE_GET_QUAD_NUM(port) %
+ ICE_GET_QUAD_NUM(hw->ptp.ports_per_phy);
if (quadtype == 0) {
msg->msg_addr_low = P_Q0_L(P_0_BASE + offset, phy_port);
@@ -427,7 +430,7 @@ ice_read_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- ice_fill_phy_msg_e82x(&msg, port, offset);
+ ice_fill_phy_msg_e82x(hw, &msg, port, offset);
msg.opcode = ice_sbq_msg_rd;
err = ice_sbq_rw_reg(hw, &msg);
@@ -504,7 +507,7 @@ ice_write_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- ice_fill_phy_msg_e82x(&msg, port, offset);
+ ice_fill_phy_msg_e82x(hw, &msg, port, offset);
msg.opcode = ice_sbq_msg_wr;
msg.data = val;
@@ -614,24 +617,30 @@ ice_write_64b_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 low_addr, u64 val)
/**
* ice_fill_quad_msg_e82x - Fill message data for quad register access
+ * @hw: pointer to the HW struct
* @msg: the PHY message buffer to fill in
* @quad: the quad to access
* @offset: the register offset
*
* Fill a message buffer for accessing a register in a quad shared between
* multiple PHYs.
+ *
+ * Return:
+ * * %0 - OK
+ * * %-EINVAL - invalid quad number
*/
-static int
-ice_fill_quad_msg_e82x(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
+static int ice_fill_quad_msg_e82x(struct ice_hw *hw,
+ struct ice_sbq_msg_input *msg, u8 quad,
+ u16 offset)
{
u32 addr;
- if (quad >= ICE_MAX_QUAD)
+ if (quad >= ICE_GET_QUAD_NUM(hw->ptp.num_lports))
return -EINVAL;
msg->dest_dev = rmn_0;
- if ((quad % ICE_QUADS_PER_PHY_E82X) == 0)
+ if (!(quad % ICE_GET_QUAD_NUM(hw->ptp.ports_per_phy)))
addr = Q_0_BASE + offset;
else
addr = Q_1_BASE + offset;
@@ -658,7 +667,7 @@ ice_read_quad_reg_e82x(struct ice_hw *hw, u8 quad, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- err = ice_fill_quad_msg_e82x(&msg, quad, offset);
+ err = ice_fill_quad_msg_e82x(hw, &msg, quad, offset);
if (err)
return err;
@@ -692,7 +701,7 @@ ice_write_quad_reg_e82x(struct ice_hw *hw, u8 quad, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- err = ice_fill_quad_msg_e82x(&msg, quad, offset);
+ err = ice_fill_quad_msg_e82x(hw, &msg, quad, offset);
if (err)
return err;
@@ -813,7 +822,7 @@ static void ice_ptp_reset_ts_memory_e82x(struct ice_hw *hw)
{
unsigned int quad;
- for (quad = 0; quad < ICE_MAX_QUAD; quad++)
+ for (quad = 0; quad < ICE_GET_QUAD_NUM(hw->ptp.num_lports); quad++)
ice_ptp_reset_ts_memory_quad_e82x(hw, quad);
}
@@ -1110,7 +1119,7 @@ static int ice_ptp_set_vernier_wl(struct ice_hw *hw)
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_write_phy_reg_e82x(hw, port, P_REG_WL,
@@ -1175,7 +1184,7 @@ ice_ptp_prep_phy_time_e82x(struct ice_hw *hw, u32 time)
*/
phy_time = (u64)time << 32;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
/* Tx case */
err = ice_write_64b_phy_reg_e82x(hw, port,
P_REG_TX_TIMER_INC_PRE_L,
@@ -1278,7 +1287,7 @@ ice_ptp_prep_phy_adj_e82x(struct ice_hw *hw, s32 adj)
else
cycles = -(((s64)-adj) << 32);
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_ptp_prep_port_adj_e82x(hw, port, cycles);
@@ -1304,7 +1313,7 @@ ice_ptp_prep_phy_incval_e82x(struct ice_hw *hw, u64 incval)
int err;
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
err = ice_write_40b_phy_reg_e82x(hw, port, P_REG_TIMETUS_L,
incval);
if (err)
@@ -1460,7 +1469,7 @@ ice_ptp_one_port_cmd(struct ice_hw *hw, u8 configured_port,
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
enum ice_ptp_tmr_cmd cmd;
int err;
@@ -1490,7 +1499,7 @@ ice_ptp_port_cmd_e82x(struct ice_hw *hw, enum ice_ptp_tmr_cmd cmd)
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_ptp_write_port_cmd_e82x(hw, port, cmd);
@@ -1603,7 +1612,7 @@ static void ice_phy_cfg_lane_e82x(struct ice_hw *hw, u8 port)
return;
}
- quad = port / ICE_PORTS_PER_QUAD;
+ quad = ICE_GET_QUAD_NUM(port);
err = ice_read_quad_reg_e82x(hw, quad, Q_REG_TX_MEM_GBL_CFG, &val);
if (err) {
@@ -2632,6 +2641,17 @@ ice_get_phy_tx_tstamp_ready_e82x(struct ice_hw *hw, u8 quad, u64 *tstamp_ready)
return 0;
}
+/**
+ * ice_ptp_init_phy_e82x - initialize PHY parameters
+ * @ptp: pointer to the PTP HW struct
+ */
+static void ice_ptp_init_phy_e82x(struct ice_ptp_hw *ptp)
+{
+ ptp->phy_model = ICE_PHY_E82X;
+ ptp->num_lports = 8;
+ ptp->ports_per_phy = 8;
+}
+
/* E810 functions
*
* The following functions operate on the E810 series devices which use
@@ -2859,17 +2879,21 @@ static int ice_clear_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx)
}
/**
- * ice_ptp_init_phy_e810 - Enable PTP function on the external PHY
+ * ice_ptp_init_phc_e810 - Perform E810 specific PHC initialization
* @hw: pointer to HW struct
*
- * Enable the timesync PTP functionality for the external PHY connected to
- * this function.
+ * Perform E810-specific PTP hardware clock initialization steps.
+ *
+ * Return: 0 on success, other error codes when failed to initialize TimeSync
*/
-int ice_ptp_init_phy_e810(struct ice_hw *hw)
+static int ice_ptp_init_phc_e810(struct ice_hw *hw)
{
u8 tmr_idx;
int err;
+ /* Ensure synchronization delay is zero */
+ wr32(hw, GLTSYN_SYNC_DLAY, 0);
+
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
err = ice_write_phy_reg_e810(hw, ETH_GLTSYN_ENA(tmr_idx),
GLTSYN_ENA_TSYN_ENA_M);
@@ -2880,21 +2904,6 @@ int ice_ptp_init_phy_e810(struct ice_hw *hw)
return err;
}
-/**
- * ice_ptp_init_phc_e810 - Perform E810 specific PHC initialization
- * @hw: pointer to HW struct
- *
- * Perform E810-specific PTP hardware clock initialization steps.
- */
-static int ice_ptp_init_phc_e810(struct ice_hw *hw)
-{
- /* Ensure synchronization delay is zero */
- wr32(hw, GLTSYN_SYNC_DLAY, 0);
-
- /* Initialize the PHY */
- return ice_ptp_init_phy_e810(hw);
-}
-
/**
* ice_ptp_prep_phy_time_e810 - Prepare PHY port with initial time
* @hw: Board private structure
@@ -3238,6 +3247,17 @@ int ice_read_pca9575_reg_e810t(struct ice_hw *hw, u8 offset, u8 *data)
return ice_aq_read_i2c(hw, link_topo, 0, addr, 1, data, NULL);
}
+/**
+ * ice_ptp_init_phy_e810 - initialize PHY parameters
+ * @ptp: pointer to the PTP HW struct
+ */
+static void ice_ptp_init_phy_e810(struct ice_ptp_hw *ptp)
+{
+ ptp->phy_model = ICE_PHY_E810;
+ ptp->num_lports = 8;
+ ptp->ports_per_phy = 4;
+}
+
/* Device agnostic functions
*
* The following functions implement shared behavior common to both E822 and
@@ -3295,18 +3315,22 @@ void ice_ptp_unlock(struct ice_hw *hw)
}
/**
- * ice_ptp_init_phy_model - Initialize hw->phy_model based on device type
+ * ice_ptp_init_hw - Initialize hw based on device type
* @hw: pointer to the HW structure
*
- * Determine the PHY model for the device, and initialize hw->phy_model
+ * Determine the PHY model for the device, and initialize hw
* for use by other functions.
*/
-void ice_ptp_init_phy_model(struct ice_hw *hw)
+void ice_ptp_init_hw(struct ice_hw *hw)
{
- if (ice_is_e810(hw))
- hw->phy_model = ICE_PHY_E810;
+ struct ice_ptp_hw *ptp = &hw->ptp;
+
+ if (ice_is_e822(hw) || ice_is_e823(hw))
+ ice_ptp_init_phy_e82x(ptp);
+ else if (ice_is_e810(hw))
+ ice_ptp_init_phy_e810(ptp);
else
- hw->phy_model = ICE_PHY_E82X;
+ ptp->phy_model = ICE_PHY_UNSUP;
}
/**
@@ -3327,7 +3351,7 @@ static int ice_ptp_tmr_cmd(struct ice_hw *hw, enum ice_ptp_tmr_cmd cmd)
ice_ptp_src_cmd(hw, cmd);
/* Next, prepare the ports */
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_port_cmd_e810(hw, cmd);
break;
@@ -3379,7 +3403,7 @@ int ice_ptp_init_time(struct ice_hw *hw, u64 time)
/* PHY timers */
/* Fill Rx and Tx ports and send msg to PHY */
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_time_e810(hw, time & 0xFFFFFFFF);
break;
@@ -3421,7 +3445,7 @@ int ice_ptp_write_incval(struct ice_hw *hw, u64 incval)
wr32(hw, GLTSYN_SHADJ_L(tmr_idx), lower_32_bits(incval));
wr32(hw, GLTSYN_SHADJ_H(tmr_idx), upper_32_bits(incval));
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_incval_e810(hw, incval);
break;
@@ -3487,7 +3511,7 @@ int ice_ptp_adj_clock(struct ice_hw *hw, s32 adj)
wr32(hw, GLTSYN_SHADJ_L(tmr_idx), 0);
wr32(hw, GLTSYN_SHADJ_H(tmr_idx), adj);
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_adj_e810(hw, adj);
break;
@@ -3517,7 +3541,7 @@ int ice_ptp_adj_clock(struct ice_hw *hw, s32 adj)
*/
int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_read_phy_tstamp_e810(hw, block, idx, tstamp);
case ICE_PHY_E82X:
@@ -3545,7 +3569,7 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
*/
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_clear_phy_tstamp_e810(hw, block, idx);
case ICE_PHY_E82X:
@@ -3606,7 +3630,7 @@ int ice_get_pf_c827_idx(struct ice_hw *hw, u8 *idx)
*/
void ice_ptp_reset_ts_memory(struct ice_hw *hw)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E82X:
ice_ptp_reset_ts_memory_e82x(hw);
break;
@@ -3632,7 +3656,7 @@ int ice_ptp_init_phc(struct ice_hw *hw)
/* Clear event err indications for auxiliary pins */
(void)rd32(hw, GLTSYN_STAT(src_idx));
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_ptp_init_phc_e810(hw);
case ICE_PHY_E82X:
@@ -3655,7 +3679,7 @@ int ice_ptp_init_phc(struct ice_hw *hw)
*/
int ice_get_phy_tx_tstamp_ready(struct ice_hw *hw, u8 block, u64 *tstamp_ready)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_get_phy_tx_tstamp_ready_e810(hw, block,
tstamp_ready);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index 7e8fd369ef7c..d788221eba57 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -211,6 +211,7 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp);
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx);
void ice_ptp_reset_ts_memory(struct ice_hw *hw);
int ice_ptp_init_phc(struct ice_hw *hw);
+void ice_ptp_init_hw(struct ice_hw *hw);
int ice_get_phy_tx_tstamp_ready(struct ice_hw *hw, u8 block, u64 *tstamp_ready);
/* E822 family functions */
@@ -265,7 +266,6 @@ int ice_phy_cfg_tx_offset_e82x(struct ice_hw *hw, u8 port);
int ice_phy_cfg_rx_offset_e82x(struct ice_hw *hw, u8 port);
/* E810 family functions */
-int ice_ptp_init_phy_e810(struct ice_hw *hw);
int ice_read_sma_ctrl_e810t(struct ice_hw *hw, u8 *data);
int ice_write_sma_ctrl_e810t(struct ice_hw *hw, u8 data);
int ice_read_pca9575_reg_e810t(struct ice_hw *hw, u8 offset, u8 *data);
@@ -280,8 +280,6 @@ int ice_get_cgu_state(struct ice_hw *hw, u8 dpll_idx,
u8 *ref_state, u8 *eec_mode, s64 *phase_offset,
enum dpll_lock_status *dpll_state);
int ice_get_cgu_rclk_pin_info(struct ice_hw *hw, u8 *base_idx, u8 *pin_num);
-
-void ice_ptp_init_phy_model(struct ice_hw *hw);
int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
unsigned long *caps);
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 28e47bb78eaf..6fc4cd1030d0 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -807,6 +807,9 @@ struct ice_mbx_data {
u16 async_watermark_val;
};
+#define ICE_PORTS_PER_QUAD 4
+#define ICE_GET_QUAD_NUM(port) ((port) / ICE_PORTS_PER_QUAD)
+
/* PHY model */
enum ice_phy_model {
ICE_PHY_UNSUP = -1,
@@ -814,6 +817,12 @@ enum ice_phy_model {
ICE_PHY_E82X,
};
+struct ice_ptp_hw {
+ enum ice_phy_model phy_model;
+ u8 num_lports;
+ u8 ports_per_phy;
+};
+
/* Port hardware description */
struct ice_hw {
u8 __iomem *hw_addr;
@@ -835,7 +844,6 @@ struct ice_hw {
u8 revision_id;
u8 pf_id; /* device profile info */
- enum ice_phy_model phy_model;
u16 max_burst_size; /* driver sets this value */
@@ -896,12 +904,7 @@ struct ice_hw {
/* INTRL granularity in 1 us */
u8 intrl_gran;
-#define ICE_MAX_QUAD 2
-#define ICE_QUADS_PER_PHY_E82X 2
-#define ICE_PORTS_PER_PHY_E82X 8
-#define ICE_PORTS_PER_QUAD 4
-#define ICE_PORTS_PER_PHY_E810 4
-#define ICE_NUM_EXTERNAL_PORTS (ICE_MAX_QUAD * ICE_PORTS_PER_QUAD)
+ struct ice_ptp_hw ptp;
/* Active package version (currently active) */
struct ice_pkg_ver active_pkg_ver;
--
2.43.0

80
kernel-rt/debian/patches/ice-VDF/0037-ice-Introduce-ice_get_base_incval-helper.patch Normal file
View File

@ -0,0 +1,80 @@
From cd12b5c8239993e395436ff9a01b524103aa0641 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 28 May 2024 16:03:56 -0700
Subject: [PATCH] ice: Introduce ice_get_base_incval() helper
Add a new helper for getting base clock increment value for specific HW.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528-next-2024-05-28-ptp-refactors-v1-6-c082739bb6f6@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 1f374d57c39386520586539641cafc999d0f3ef5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 9 +--------
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 18 ++++++++++++++++++
2 files changed, 19 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index bb1572a353d0..44b8fc8021cd 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -7,8 +7,6 @@
#define E810_OUT_PROP_DELAY_NS 1
-#define UNKNOWN_INCVAL_E82X 0x100000000ULL
-
static const struct ptp_pin_desc ice_pin_desc_e810t[] = {
/* name idx func chan */
{ "GNSS", GNSS, PTP_PF_EXTTS, 0, { 0, } },
@@ -1229,12 +1227,7 @@ static u64 ice_base_incval(struct ice_pf *pf)
struct ice_hw *hw = &pf->hw;
u64 incval;
- if (ice_is_e810(hw))
- incval = ICE_PTP_NOMINAL_INCVAL_E810;
- else if (ice_e82x_time_ref(hw) < NUM_ICE_TIME_REF_FREQ)
- incval = ice_e82x_nominal_incval(ice_e82x_time_ref(hw));
- else
- incval = UNKNOWN_INCVAL_E82X;
+ incval = ice_get_base_incval(hw);
dev_dbg(ice_pf_to_dev(pf), "PTP: using base increment value of 0x%016llx\n",
incval);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index d788221eba57..749a3f2d8293 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -283,6 +283,24 @@ int ice_get_cgu_rclk_pin_info(struct ice_hw *hw, u8 *base_idx, u8 *pin_num);
int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
unsigned long *caps);
+/**
+ * ice_get_base_incval - Get base clock increment value
+ * @hw: pointer to the HW struct
+ *
+ * Return: base clock increment value for supported PHYs, 0 otherwise
+ */
+static inline u64 ice_get_base_incval(struct ice_hw *hw)
+{
+ switch (hw->ptp.phy_model) {
+ case ICE_PHY_E810:
+ return ICE_PTP_NOMINAL_INCVAL_E810;
+ case ICE_PHY_E82X:
+ return ice_e82x_nominal_incval(ice_e82x_time_ref(hw));
+ default:
+ return 0;
+ }
+}
+
#define PFTSYN_SEM_BYTES 4
#define ICE_PTP_CLOCK_INDEX_0 0x00
--
2.43.0

32
kernel-rt/debian/patches/ice-VDF/0038-ice-modify-the-ice-driver-version-to-stx.4.patch Normal file
View File

@ -0,0 +1,32 @@
From 437206483113743a4ef40c2f7e14f09705049672 Mon Sep 17 00:00:00 2001
From: Jiping Ma <jiping.ma2@windriver.com>
Date: Mon, 2 Sep 2024 03:18:08 +0000
Subject: [PATCH] ice:modify the ice driver version to stx.4
Change the ice driver min version to stx.4 because we back ported
the upstream 36 commits to our code base to support the customer's
requirement.
The ice driver version should be ice-6.6.40-stx.4.
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Makefile b/Makefile
index 5807b310bdca..7163d25405f8 100644
--- a/Makefile
+++ b/Makefile
@@ -1227,7 +1227,7 @@ uapi-asm-generic:
# KERNELRELEASE can change from a few different places, meaning version.h
# needs to be updated, so this check is forced on all builds
-ICE_STX = "-stx.3"
+ICE_STX = "-stx.4"
I40E_STX = "-stx.0"
IAVF_STX = "-stx.0"
--
2.43.0

38
kernel-rt/debian/patches/series
View File

@ -64,3 +64,41 @@ ice-dpll/0046-dpll-fix-dpll_xa_ref_-_del-for-multiple-registration.patch
ice-dpll/0047-ice-modify-the-ice-driver-min-version-to-stx.2.patch
ice-mdd/0001-ice-Add-automatic-VF-reset-on-Tx-MDD-events.patch
ice-mdd/0002-ice-modify-the-ice-driver-min-version-to-stx.3.patch
ice-VDF/0001-ice-Auxbus-devices-driver-for-E822-TS.patch
ice-VDF/0002-ice-introduce-ice_pf_src_tmr_owned.patch
ice-VDF/0003-ice-Re-enable-timestamping-correctly-after-reset.patch
ice-VDF/0004-ice-periodically-kick-Tx-timestamp-interrupt.patch
ice-VDF/0005-ice-PTP-Rename-macros-used-for-PHY-QUAD-port-definit.patch
ice-VDF/0006-ice-PTP-move-quad-value-check-inside-ice_fill_phy_ms.patch
ice-VDF/0007-ice-remove-ptp_tx-ring-parameter-flag.patch
ice-VDF/0008-ice-unify-logic-for-programming-PFINT_TSYN_MSK.patch
ice-VDF/0009-ice-PTP-Clean-up-timestamp-registers-correctly.patch
ice-VDF/0010-ice-Use-PTP-auxbus-for-all-PHYs-restart-in-E822.patch
ice-VDF/0011-ice-Rename-E822-to-E82X.patch
ice-VDF/0012-ice-Schedule-service-task-in-IRQ-top-half.patch
ice-VDF/0013-ice-Enable-SW-interrupt-from-FW-for-LL-TS.patch
ice-VDF/0014-ice-PTP-add-clock-domain-number-to-auxiliary-interfa.patch
ice-VDF/0015-ice-restore-timestamp-configuration-after-device-res.patch
ice-VDF/0016-ice-introduce-PTP-state-machine.patch
ice-VDF/0017-ice-pass-reset-type-to-PTP-reset-functions.patch
ice-VDF/0018-ice-rename-verify_cached-to-has_ready_bitmap.patch
ice-VDF/0019-ice-don-t-check-has_ready_bitmap-in-E810-functions.patch
ice-VDF/0020-ice-rename-ice_ptp_tx_cfg_intr.patch
ice-VDF/0021-ice-factor-out-ice_ptp_rebuild_owner.patch
ice-VDF/0022-ice-stop-destroying-and-reinitalizing-Tx-tracker-dur.patch
ice-VDF/0023-ice-Remove-and-readd-netdev-during-devlink-reload.patch
ice-VDF/0024-ice-remove-FW-logging-code.patch
ice-VDF/0025-ice-configure-FW-logging.patch
ice-VDF/0026-ice-enable-FW-logging.patch
ice-VDF/0027-ice-add-ability-to-read-and-configure-FW-log-data.patch
ice-VDF/0028-ice-Fix-debugfs-with-devlink-reload.patch
ice-VDF/0029-ice-remove-vf-lan_vsi_num-field.patch
ice-VDF/0030-ice-rename-ice_write_-functions-to-ice_pack_ctx_.patch
ice-VDF/0031-ice-use-GENMASK-instead-of-BIT-n-1-in-pack-functions.patch
ice-VDF/0032-ice-cleanup-line-splitting-for-context-set-functions.patch
ice-VDF/0033-ice-do-not-disable-Tx-queues-twice-in-ice_down.patch
ice-VDF/0034-ice-Fix-improper-extts-handling.patch
ice-VDF/0035-ice-Don-t-process-extts-if-PTP-is-disabled.patch
ice-VDF/0036-ice-Introduce-ice_ptp_hw-struct.patch
ice-VDF/0037-ice-Introduce-ice_get_base_incval-helper.patch
ice-VDF/0038-ice-modify-the-ice-driver-version-to-stx.4.patch

684
kernel-std/debian/patches/ice-VDF/0001-ice-Auxbus-devices-driver-for-E822-TS.patch Normal file
View File

@ -0,0 +1,684 @@
From ba0d88d4ff54805aac7aec77cc5b05d0df9114da Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:34 +0200
Subject: [PATCH 01/36] ice: Auxbus devices & driver for E822 TS
There is a problem in HW in E822-based devices leading to race
condition.
It might happen that, in order:
- PF0 (which owns the PHC) requests few timestamps,
- PF1 requests a timestamp,
- interrupt is being triggered and both PF0 and PF1 threads are woken
up,
- PF0 got one timestamp, still waiting for others so not going to sleep,
- PF1 gets it's timestamp, process it and go to sleep,
- PF1 requests a timestamp again,
- just before PF0 goes to sleep timestamp of PF1 appear,
- PF0 finishes all it's timestamps and go to sleep (PF1 also sleeping).
That leaves PF1 timestamp memory not read, which lead to blocking the
next interrupt from arriving.
Fix it by adding auxiliary devices and only one driver to handle all the
timestamps for all PF's by PHC owner. In the past each PF requested it's
own timestamps and process it from the start till the end which causes
problem described above. Currently each PF requests the timestamps as
before, but the actual reading of the completed timestamps is being done
by the PTP auxiliary driver, which is registered by the PF which owns PHC.
Additionally, the newly introduced auxiliary driver/devices for PTP clock
owner will be used for other features in all products (including E810).
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit d938a8cca88a5f02f523f95fe3d2d1214f4b4a8d)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 12 +
.../net/ethernet/intel/ice/ice_hw_autogen.h | 1 +
drivers/net/ethernet/intel/ice/ice_main.c | 11 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 393 +++++++++++++++++-
drivers/net/ethernet/intel/ice/ice_ptp.h | 30 ++
5 files changed, 430 insertions(+), 17 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index b9cd0113b859..0a3d76d184ba 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -671,6 +671,18 @@ static inline bool ice_vector_ch_enabled(struct ice_q_vector *qv)
return !!qv->ch; /* Enable it to run with TC */
}
+/**
+ * ice_ptp_pf_handles_tx_interrupt - Check if PF handles Tx interrupt
+ * @pf: Board private structure
+ *
+ * Return true if this PF should respond to the Tx timestamp interrupt
+ * indication in the miscellaneous OICR interrupt handler.
+ */
+static inline bool ice_ptp_pf_handles_tx_interrupt(struct ice_pf *pf)
+{
+ return pf->ptp.tx_interrupt_mode != ICE_PTP_TX_INTERRUPT_NONE;
+}
+
/**
* ice_irq_dynamic_ena - Enable default interrupt generation settings
* @hw: pointer to HW struct
diff --git a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
index 531cc2194741..6756f3d51d14 100644
--- a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
+++ b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
@@ -231,6 +231,7 @@
#define PFINT_SB_CTL 0x0016B600
#define PFINT_SB_CTL_MSIX_INDX_M ICE_M(0x7FF, 0)
#define PFINT_SB_CTL_CAUSE_ENA_M BIT(30)
+#define PFINT_TSYN_MSK 0x0016C980
#define QINT_RQCTL(_QRX) (0x00150000 + ((_QRX) * 4))
#define QINT_RQCTL_MSIX_INDX_S 0
#define QINT_RQCTL_MSIX_INDX_M ICE_M(0x7FF, 0)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 8a6acb5a722e..39cb6ee52abe 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3190,7 +3190,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (!hw->reset_ongoing)
+ if (!hw->reset_ongoing && ice_ptp_pf_handles_tx_interrupt(pf))
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
}
@@ -7444,8 +7444,13 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
}
/* configure PTP timestamping after VSI rebuild */
- if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_cfg_timestamp(pf, false);
+ if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) {
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
+ ice_ptp_cfg_timestamp(pf, false);
+ else if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL)
+ /* for E82x PHC owner always need to have interrupts */
+ ice_ptp_cfg_timestamp(pf, true);
+ }
err = ice_vsi_rebuild_by_type(pf, ICE_VSI_SWITCHDEV_CTRL);
if (err) {
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 3648d3cccacc..e3012608c9dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -255,6 +255,24 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
return ice_ptp_set_sma_e810t(info, pin, func);
}
+/**
+ * ice_ptp_configure_tx_tstamp - Enable or disable Tx timestamp interrupt
+ * @pf: The PF pointer to search in
+ * @on: bool value for whether timestamp interrupt is enabled or disabled
+ */
+static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
+{
+ u32 val;
+
+ /* Configure the Tx timestamp interrupt */
+ val = rd32(&pf->hw, PFINT_OICR_ENA);
+ if (on)
+ val |= PFINT_OICR_TSYN_TX_M;
+ else
+ val &= ~PFINT_OICR_TSYN_TX_M;
+ wr32(&pf->hw, PFINT_OICR_ENA, val);
+}
+
/**
* ice_set_tx_tstamp - Enable or disable Tx timestamping
* @pf: The PF pointer to search in
@@ -263,7 +281,6 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
struct ice_vsi *vsi;
- u32 val;
u16 i;
vsi = ice_get_main_vsi(pf);
@@ -277,13 +294,8 @@ static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
vsi->tx_rings[i]->ptp_tx = on;
}
- /* Configure the Tx timestamp interrupt */
- val = rd32(&pf->hw, PFINT_OICR_ENA);
- if (on)
- val |= PFINT_OICR_TSYN_TX_M;
- else
- val &= ~PFINT_OICR_TSYN_TX_M;
- wr32(&pf->hw, PFINT_OICR_ENA, val);
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
+ ice_ptp_configure_tx_tstamp(pf, on);
pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
}
@@ -674,9 +686,6 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
int err;
u8 idx;
- if (!tx->init)
- return;
-
ptp_port = container_of(tx, struct ice_ptp_port, tx);
pf = ptp_port_to_pf(ptp_port);
hw = &pf->hw;
@@ -774,6 +783,39 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
}
}
+/**
+ * ice_ptp_tx_tstamp_owner - Process Tx timestamps for all ports on the device
+ * @pf: Board private structure
+ */
+static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
+{
+ struct ice_ptp_port *port;
+ unsigned int i;
+
+ mutex_lock(&pf->ptp.ports_owner.lock);
+ list_for_each_entry(port, &pf->ptp.ports_owner.ports, list_member) {
+ struct ice_ptp_tx *tx = &port->tx;
+
+ if (!tx || !tx->init)
+ continue;
+
+ ice_ptp_process_tx_tstamp(tx);
+ }
+ mutex_unlock(&pf->ptp.ports_owner.lock);
+
+ for (i = 0; i < ICE_MAX_QUAD; i++) {
+ u64 tstamp_ready;
+ int err;
+
+ /* Read the Tx ready status first */
+ err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
+ if (err || tstamp_ready)
+ return ICE_TX_TSTAMP_WORK_PENDING;
+ }
+
+ return ICE_TX_TSTAMP_WORK_DONE;
+}
+
/**
* ice_ptp_tx_tstamp - Process Tx timestamps for this function.
* @tx: Tx tracking structure to initialize
@@ -2448,7 +2490,21 @@ s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
*/
enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf)
{
- return ice_ptp_tx_tstamp(&pf->ptp.port.tx);
+ switch (pf->ptp.tx_interrupt_mode) {
+ case ICE_PTP_TX_INTERRUPT_NONE:
+ /* This device has the clock owner handle timestamps for it */
+ return ICE_TX_TSTAMP_WORK_DONE;
+ case ICE_PTP_TX_INTERRUPT_SELF:
+ /* This device handles its own timestamps */
+ return ice_ptp_tx_tstamp(&pf->ptp.port.tx);
+ case ICE_PTP_TX_INTERRUPT_ALL:
+ /* This device handles timestamps for all ports */
+ return ice_ptp_tx_tstamp_owner(pf);
+ default:
+ WARN_ONCE(1, "Unexpected Tx timestamp interrupt mode %u\n",
+ pf->ptp.tx_interrupt_mode);
+ return ICE_TX_TSTAMP_WORK_DONE;
+ }
}
static void ice_ptp_periodic_work(struct kthread_work *work)
@@ -2557,6 +2613,187 @@ void ice_ptp_reset(struct ice_pf *pf)
dev_err(ice_pf_to_dev(pf), "PTP reset failed %d\n", err);
}
+/**
+ * ice_ptp_aux_dev_to_aux_pf - Get auxiliary PF handle for the auxiliary device
+ * @aux_dev: auxiliary device to get the auxiliary PF for
+ */
+static struct ice_pf *
+ice_ptp_aux_dev_to_aux_pf(struct auxiliary_device *aux_dev)
+{
+ struct ice_ptp_port *aux_port;
+ struct ice_ptp *aux_ptp;
+
+ aux_port = container_of(aux_dev, struct ice_ptp_port, aux_dev);
+ aux_ptp = container_of(aux_port, struct ice_ptp, port);
+
+ return container_of(aux_ptp, struct ice_pf, ptp);
+}
+
+/**
+ * ice_ptp_aux_dev_to_owner_pf - Get PF handle for the auxiliary device
+ * @aux_dev: auxiliary device to get the PF for
+ */
+static struct ice_pf *
+ice_ptp_aux_dev_to_owner_pf(struct auxiliary_device *aux_dev)
+{
+ struct ice_ptp_port_owner *ports_owner;
+ struct auxiliary_driver *aux_drv;
+ struct ice_ptp *owner_ptp;
+
+ if (!aux_dev->dev.driver)
+ return NULL;
+
+ aux_drv = to_auxiliary_drv(aux_dev->dev.driver);
+ ports_owner = container_of(aux_drv, struct ice_ptp_port_owner,
+ aux_driver);
+ owner_ptp = container_of(ports_owner, struct ice_ptp, ports_owner);
+ return container_of(owner_ptp, struct ice_pf, ptp);
+}
+
+/**
+ * ice_ptp_auxbus_probe - Probe auxiliary devices
+ * @aux_dev: PF's auxiliary device
+ * @id: Auxiliary device ID
+ */
+static int ice_ptp_auxbus_probe(struct auxiliary_device *aux_dev,
+ const struct auxiliary_device_id *id)
+{
+ struct ice_pf *owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ struct ice_pf *aux_pf = ice_ptp_aux_dev_to_aux_pf(aux_dev);
+
+ if (WARN_ON(!owner_pf))
+ return -ENODEV;
+
+ INIT_LIST_HEAD(&aux_pf->ptp.port.list_member);
+ mutex_lock(&owner_pf->ptp.ports_owner.lock);
+ list_add(&aux_pf->ptp.port.list_member,
+ &owner_pf->ptp.ports_owner.ports);
+ mutex_unlock(&owner_pf->ptp.ports_owner.lock);
+
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_remove - Remove auxiliary devices from the bus
+ * @aux_dev: PF's auxiliary device
+ */
+static void ice_ptp_auxbus_remove(struct auxiliary_device *aux_dev)
+{
+ struct ice_pf *owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ struct ice_pf *aux_pf = ice_ptp_aux_dev_to_aux_pf(aux_dev);
+
+ mutex_lock(&owner_pf->ptp.ports_owner.lock);
+ list_del(&aux_pf->ptp.port.list_member);
+ mutex_unlock(&owner_pf->ptp.ports_owner.lock);
+}
+
+/**
+ * ice_ptp_auxbus_shutdown
+ * @aux_dev: PF's auxiliary device
+ */
+static void ice_ptp_auxbus_shutdown(struct auxiliary_device *aux_dev)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+}
+
+/**
+ * ice_ptp_auxbus_suspend
+ * @aux_dev: PF's auxiliary device
+ * @state: power management state indicator
+ */
+static int
+ice_ptp_auxbus_suspend(struct auxiliary_device *aux_dev, pm_message_t state)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_resume
+ * @aux_dev: PF's auxiliary device
+ */
+static int ice_ptp_auxbus_resume(struct auxiliary_device *aux_dev)
+{
+ /* Doing nothing here, but handle to auxbus driver must be satisfied */
+ return 0;
+}
+
+/**
+ * ice_ptp_auxbus_create_id_table - Create auxiliary device ID table
+ * @pf: Board private structure
+ * @name: auxiliary bus driver name
+ */
+static struct auxiliary_device_id *
+ice_ptp_auxbus_create_id_table(struct ice_pf *pf, const char *name)
+{
+ struct auxiliary_device_id *ids;
+
+ /* Second id left empty to terminate the array */
+ ids = devm_kcalloc(ice_pf_to_dev(pf), 2,
+ sizeof(struct auxiliary_device_id), GFP_KERNEL);
+ if (!ids)
+ return NULL;
+
+ snprintf(ids[0].name, sizeof(ids[0].name), "ice.%s", name);
+
+ return ids;
+}
+
+/**
+ * ice_ptp_register_auxbus_driver - Register PTP auxiliary bus driver
+ * @pf: Board private structure
+ */
+static int ice_ptp_register_auxbus_driver(struct ice_pf *pf)
+{
+ struct auxiliary_driver *aux_driver;
+ struct ice_ptp *ptp;
+ struct device *dev;
+ char *name;
+ int err;
+
+ ptp = &pf->ptp;
+ dev = ice_pf_to_dev(pf);
+ aux_driver = &ptp->ports_owner.aux_driver;
+ INIT_LIST_HEAD(&ptp->ports_owner.ports);
+ mutex_init(&ptp->ports_owner.lock);
+ name = devm_kasprintf(dev, GFP_KERNEL, "ptp_aux_dev_%u_%u_clk%u",
+ pf->pdev->bus->number, PCI_SLOT(pf->pdev->devfn),
+ ice_get_ptp_src_clock_index(&pf->hw));
+
+ aux_driver->name = name;
+ aux_driver->shutdown = ice_ptp_auxbus_shutdown;
+ aux_driver->suspend = ice_ptp_auxbus_suspend;
+ aux_driver->remove = ice_ptp_auxbus_remove;
+ aux_driver->resume = ice_ptp_auxbus_resume;
+ aux_driver->probe = ice_ptp_auxbus_probe;
+ aux_driver->id_table = ice_ptp_auxbus_create_id_table(pf, name);
+ if (!aux_driver->id_table)
+ return -ENOMEM;
+
+ err = auxiliary_driver_register(aux_driver);
+ if (err) {
+ devm_kfree(dev, aux_driver->id_table);
+ dev_err(dev, "Failed registering aux_driver, name <%s>\n",
+ name);
+ }
+
+ return err;
+}
+
+/**
+ * ice_ptp_unregister_auxbus_driver - Unregister PTP auxiliary bus driver
+ * @pf: Board private structure
+ */
+static void ice_ptp_unregister_auxbus_driver(struct ice_pf *pf)
+{
+ struct auxiliary_driver *aux_driver = &pf->ptp.ports_owner.aux_driver;
+
+ auxiliary_driver_unregister(aux_driver);
+ devm_kfree(ice_pf_to_dev(pf), aux_driver->id_table);
+
+ mutex_destroy(&pf->ptp.ports_owner.lock);
+}
+
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
@@ -2635,7 +2872,15 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
- if (!ice_is_e810(hw)) {
+ if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL) {
+ /* The clock owner for this device type handles the timestamp
+ * interrupt for all ports.
+ */
+ ice_ptp_configure_tx_tstamp(pf, true);
+
+ /* React on all quads interrupts for E82x */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
+
/* Enable quad interrupts */
err = ice_ptp_tx_ena_intr(pf, true, itr);
if (err)
@@ -2650,8 +2895,16 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Store the PTP clock index for other PFs */
ice_set_ptp_clock_index(pf);
- return 0;
+ err = ice_ptp_register_auxbus_driver(pf);
+ if (err) {
+ dev_err(ice_pf_to_dev(pf), "Failed to register PTP auxbus driver");
+ goto err_aux;
+ }
+ return 0;
+err_aux:
+ ice_clear_ptp_clock_index(pf);
+ ptp_clock_unregister(pf->ptp.clock);
err_clk:
pf->ptp.clock = NULL;
err_exit:
@@ -2701,6 +2954,13 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E822:
+ /* Non-owner PFs don't react to any interrupts on E82x,
+ * neither on own quad nor on others
+ */
+ if (!ice_ptp_pf_handles_tx_interrupt(pf)) {
+ ice_ptp_configure_tx_tstamp(pf, false);
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
+ }
kthread_init_delayed_work(&ptp_port->ov_work,
ice_ptp_wait_for_offsets);
@@ -2711,6 +2971,101 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
}
}
+/**
+ * ice_ptp_release_auxbus_device
+ * @dev: device that utilizes the auxbus
+ */
+static void ice_ptp_release_auxbus_device(struct device *dev)
+{
+ /* Doing nothing here, but handle to auxbux device must be satisfied */
+}
+
+/**
+ * ice_ptp_create_auxbus_device - Create PTP auxiliary bus device
+ * @pf: Board private structure
+ */
+static int ice_ptp_create_auxbus_device(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev;
+ struct ice_ptp *ptp;
+ struct device *dev;
+ char *name;
+ int err;
+ u32 id;
+
+ ptp = &pf->ptp;
+ id = ptp->port.port_num;
+ dev = ice_pf_to_dev(pf);
+
+ aux_dev = &ptp->port.aux_dev;
+
+ name = devm_kasprintf(dev, GFP_KERNEL, "ptp_aux_dev_%u_%u_clk%u",
+ pf->pdev->bus->number, PCI_SLOT(pf->pdev->devfn),
+ ice_get_ptp_src_clock_index(&pf->hw));
+
+ aux_dev->name = name;
+ aux_dev->id = id;
+ aux_dev->dev.release = ice_ptp_release_auxbus_device;
+ aux_dev->dev.parent = dev;
+
+ err = auxiliary_device_init(aux_dev);
+ if (err)
+ goto aux_err;
+
+ err = auxiliary_device_add(aux_dev);
+ if (err) {
+ auxiliary_device_uninit(aux_dev);
+ goto aux_err;
+ }
+
+ return 0;
+aux_err:
+ dev_err(dev, "Failed to create PTP auxiliary bus device <%s>\n", name);
+ devm_kfree(dev, name);
+ return err;
+}
+
+/**
+ * ice_ptp_remove_auxbus_device - Remove PTP auxiliary bus device
+ * @pf: Board private structure
+ */
+static void ice_ptp_remove_auxbus_device(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev = &pf->ptp.port.aux_dev;
+
+ auxiliary_device_delete(aux_dev);
+ auxiliary_device_uninit(aux_dev);
+
+ memset(aux_dev, 0, sizeof(*aux_dev));
+}
+
+/**
+ * ice_ptp_init_tx_interrupt_mode - Initialize device Tx interrupt mode
+ * @pf: Board private structure
+ *
+ * Initialize the Tx timestamp interrupt mode for this device. For most device
+ * types, each PF processes the interrupt and manages its own timestamps. For
+ * E822-based devices, only the clock owner processes the timestamps. Other
+ * PFs disable the interrupt and do not process their own timestamps.
+ */
+static void ice_ptp_init_tx_interrupt_mode(struct ice_pf *pf)
+{
+ switch (pf->hw.phy_model) {
+ case ICE_PHY_E822:
+ /* E822 based PHY has the clock owner process the interrupt
+ * for all ports.
+ */
+ if (ice_pf_src_tmr_owned(pf))
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_ALL;
+ else
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_NONE;
+ break;
+ default:
+ /* other PHY types handle their own Tx interrupt */
+ pf->ptp.tx_interrupt_mode = ICE_PTP_TX_INTERRUPT_SELF;
+ }
+}
+
/**
* ice_ptp_init - Initialize PTP hardware clock support
* @pf: Board private structure
@@ -2731,6 +3086,8 @@ void ice_ptp_init(struct ice_pf *pf)
ice_ptp_init_phy_model(hw);
+ ice_ptp_init_tx_interrupt_mode(pf);
+
/* If this function owns the clock hardware, it must allocate and
* configure the PTP clock device to represent it.
*/
@@ -2753,6 +3110,10 @@ void ice_ptp_init(struct ice_pf *pf)
if (err)
goto err;
+ err = ice_ptp_create_auxbus_device(pf);
+ if (err)
+ goto err;
+
dev_info(ice_pf_to_dev(pf), "PTP init successful\n");
return;
@@ -2781,6 +3142,8 @@ void ice_ptp_release(struct ice_pf *pf)
/* Disable timestamping for both Tx and Rx */
ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_remove_auxbus_device(pf);
+
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
clear_bit(ICE_FLAG_PTP, pf->flags);
@@ -2804,5 +3167,7 @@ void ice_ptp_release(struct ice_pf *pf)
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
+ ice_ptp_unregister_auxbus_driver(pf);
+
dev_info(ice_pf_to_dev(pf), "Removed PTP clock\n");
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 995a57019ba7..d94c22329df0 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -157,7 +157,9 @@ struct ice_ptp_tx {
* ready for PTP functionality. It is used to track the port initialization
* and determine when the port's PHY offset is valid.
*
+ * @list_member: list member structure of auxiliary device
* @tx: Tx timestamp tracking for this port
+ * @aux_dev: auxiliary device associated with this port
* @ov_work: delayed work task for tracking when PHY offset is valid
* @ps_lock: mutex used to protect the overall PTP PHY start procedure
* @link_up: indicates whether the link is up
@@ -165,7 +167,9 @@ struct ice_ptp_tx {
* @port_num: the port number this structure represents
*/
struct ice_ptp_port {
+ struct list_head list_member;
struct ice_ptp_tx tx;
+ struct auxiliary_device aux_dev;
struct kthread_delayed_work ov_work;
struct mutex ps_lock; /* protects overall PTP PHY start procedure */
bool link_up;
@@ -173,11 +177,35 @@ struct ice_ptp_port {
u8 port_num;
};
+enum ice_ptp_tx_interrupt {
+ ICE_PTP_TX_INTERRUPT_NONE = 0,
+ ICE_PTP_TX_INTERRUPT_SELF,
+ ICE_PTP_TX_INTERRUPT_ALL,
+};
+
+/**
+ * struct ice_ptp_port_owner - data used to handle the PTP clock owner info
+ *
+ * This structure contains data necessary for the PTP clock owner to correctly
+ * handle the timestamping feature for all attached ports.
+ *
+ * @aux_driver: the structure carring the auxiliary driver information
+ * @ports: list of porst handled by this port owner
+ * @lock: protect access to ports list
+ */
+struct ice_ptp_port_owner {
+ struct auxiliary_driver aux_driver;
+ struct list_head ports;
+ struct mutex lock;
+};
+
#define GLTSYN_TGT_H_IDX_MAX 4
/**
* struct ice_ptp - data used for integrating with CONFIG_PTP_1588_CLOCK
+ * @tx_interrupt_mode: the TX interrupt mode for the PTP clock
* @port: data for the PHY port initialization procedure
+ * @ports_owner: data for the auxiliary driver owner
* @work: delayed work function for periodic tasks
* @cached_phc_time: a cached copy of the PHC time for timestamp extension
* @cached_phc_jiffies: jiffies when cached_phc_time was last updated
@@ -197,7 +225,9 @@ struct ice_ptp_port {
* @late_cached_phc_updates: number of times cached PHC update is late
*/
struct ice_ptp {
+ enum ice_ptp_tx_interrupt tx_interrupt_mode;
struct ice_ptp_port port;
+ struct ice_ptp_port_owner ports_owner;
struct kthread_delayed_work work;
u64 cached_phc_time;
unsigned long cached_phc_jiffies;
--
2.43.0

95
kernel-std/debian/patches/ice-VDF/0002-ice-introduce-ice_pf_src_tmr_owned.patch Normal file
View File

@ -0,0 +1,95 @@
From f6af978ef435067b4c9f5ff5e159f8b65d969268 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 8 Sep 2023 14:37:14 -0700
Subject: [PATCH 02/36] ice: introduce ice_pf_src_tmr_owned
Add ice_pf_src_tmr_owned() macro to check the function capability bit
indicating if the current function owns the PTP hardware clock. This is
slightly shorter than the more verbose access via
hw.func_caps.ts_func_info.src_tmr_owned. Use this where possible rather
than open coding its equivalent.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 42d40bb21e332151da6fb689bf7d4af8195866ed)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 ++
drivers/net/ethernet/intel/ice/ice_lib.c | 2 +-
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 6 +++---
4 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 0a3d76d184ba..54a98c4032b7 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -197,6 +197,8 @@ extern const char ice_drv_ver[];
#define ice_pf_to_dev(pf) (&((pf)->pdev->dev))
+#define ice_pf_src_tmr_owned(pf) ((pf)->hw.func_caps.ts_func_info.src_tmr_owned)
+
enum ice_feature {
ICE_F_DSCP,
ICE_F_PHY_RCLK,
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 632091487413..106ef843f4b5 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -4010,7 +4010,7 @@ void ice_init_feature_support(struct ice_pf *pf)
if (ice_is_phy_rclk_in_netlist(&pf->hw))
ice_set_feature_support(pf, ICE_F_PHY_RCLK);
/* If we don't own the timer - don't enable other caps */
- if (!pf->hw.func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
break;
if (ice_is_cgu_in_netlist(&pf->hw))
ice_set_feature_support(pf, ICE_F_CGU);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 39cb6ee52abe..e957529b3fd6 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3200,7 +3200,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
ena_mask &= ~PFINT_OICR_TSYN_EVNT_M;
- if (hw->func_caps.ts_func_info.src_tmr_owned) {
+ if (ice_pf_src_tmr_owned(pf)) {
/* Save EVENTs from GLTSYN register */
pf->ptp.ext_ts_irq |= gltsyn_stat &
(GLTSYN_STAT_EVENT0_M |
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index e3012608c9dd..b1951357ba9f 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -448,7 +448,7 @@ static void ice_clear_ptp_clock_index(struct ice_pf *pf)
int err;
/* Do not clear the index if we don't own the timer */
- if (!hw->func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
return;
tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
@@ -2538,7 +2538,7 @@ void ice_ptp_reset(struct ice_pf *pf)
if (test_bit(ICE_PFR_REQ, pf->state))
goto pfr;
- if (!hw->func_caps.ts_func_info.src_tmr_owned)
+ if (!ice_pf_src_tmr_owned(pf))
goto reset_ts;
err = ice_ptp_init_phc(hw);
@@ -3091,7 +3091,7 @@ void ice_ptp_init(struct ice_pf *pf)
/* If this function owns the clock hardware, it must allocate and
* configure the PTP clock device to represent it.
*/
- if (hw->func_caps.ts_func_info.src_tmr_owned) {
+ if (ice_pf_src_tmr_owned(pf)) {
err = ice_ptp_init_owner(pf);
if (err)
goto err;
--
2.43.0

93
kernel-std/debian/patches/ice-VDF/0003-ice-Re-enable-timestamping-correctly-after-reset.patch Normal file
View File

@ -0,0 +1,93 @@
From 3c155fbf8e2a0546302a01cc06e8ece18468148e Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Fri, 1 Dec 2023 10:08:42 -0800
Subject: [PATCH 03/36] ice: Re-enable timestamping correctly after reset
During reset, TX_TSYN interrupt should be processed as it may process
timestamps in brief moments before and after reset.
Timestamping should be enabled on VSIs at the end of reset procedure.
On ice_get_phy_tx_tstamp_ready error, interrupt should not be rearmed
because error only happens on resets.
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 1cc5b6eaad92d69fe4d84bbee5c12ee297d56296)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 19 ++++++++++---------
2 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index e957529b3fd6..d2f3b4374d14 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3190,7 +3190,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (!hw->reset_ongoing && ice_ptp_pf_handles_tx_interrupt(pf))
+ if (ice_ptp_pf_handles_tx_interrupt(pf))
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index b1951357ba9f..92459589f6ce 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -809,7 +809,9 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
/* Read the Tx ready status first */
err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
- if (err || tstamp_ready)
+ if (err)
+ break;
+ else if (tstamp_ready)
return ICE_TX_TSTAMP_WORK_PENDING;
}
@@ -2535,12 +2537,10 @@ void ice_ptp_reset(struct ice_pf *pf)
int err, itr = 1;
u64 time_diff;
- if (test_bit(ICE_PFR_REQ, pf->state))
+ if (test_bit(ICE_PFR_REQ, pf->state) ||
+ !ice_pf_src_tmr_owned(pf))
goto pfr;
- if (!ice_pf_src_tmr_owned(pf))
- goto reset_ts;
-
err = ice_ptp_init_phc(hw);
if (err)
goto err;
@@ -2584,10 +2584,6 @@ void ice_ptp_reset(struct ice_pf *pf)
goto err;
}
-reset_ts:
- /* Restart the PHY timestamping block */
- ice_ptp_reset_phy_timestamping(pf);
-
pfr:
/* Init Tx structures */
if (ice_is_e810(&pf->hw)) {
@@ -2603,6 +2599,11 @@ void ice_ptp_reset(struct ice_pf *pf)
set_bit(ICE_FLAG_PTP, pf->flags);
+ /* Restart the PHY timestamping block */
+ if (!test_bit(ICE_PFR_REQ, pf->state) &&
+ ice_pf_src_tmr_owned(pf))
+ ice_ptp_restart_all_phy(pf);
+
/* Start periodic work going */
kthread_queue_delayed_work(ptp->kworker, &ptp->work, 0);
--
2.43.0

121
kernel-std/debian/patches/ice-VDF/0004-ice-periodically-kick-Tx-timestamp-interrupt.patch Normal file
View File

@ -0,0 +1,121 @@
From 214f06259ade960e3790b62f96bc1b75e5b76e79 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 1 Dec 2023 10:08:43 -0800
Subject: [PATCH 04/36] ice: periodically kick Tx timestamp interrupt
The E822 hardware for Tx timestamping keeps track of how many
outstanding timestamps are still in the PHY memory block. It will not
generate a new interrupt to the MAC until all of the timestamps in the
region have been read.
If somehow all the available data is not read, but the driver has exited
its interrupt routine already, the PHY will not generate a new interrupt
even if new timestamp data is captured. Because no interrupt is
generated, the driver never processes the timestamp data. This state
results in a permanent failure for all future Tx timestamps.
It is not clear how the driver and hardware could enter this state.
However, if it does, there is currently no recovery mechanism.
Add a recovery mechanism via the periodic PTP work thread which invokes
ice_ptp_periodic_work(). Introduce a new check,
ice_ptp_maybe_trigger_tx_interrupt() which checks the PHY timestamp
ready bitmask. If any bits are set, trigger a software interrupt by
writing to PFINT_OICR.
Once triggered, the main timestamp processing thread will read through
the PHY data and clear the outstanding timestamp data. Once cleared, new
data should trigger interrupts as expected.
This should allow recovery from such a state rather than leaving the
device in a state where we cannot process Tx timestamps.
It is possible that this function checks for timestamp data
simultaneously with the interrupt, and it might trigger additional
unnecessary interrupts. This will cause a small amount of additional
processing.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Andrii Staikov <andrii.staikov@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 712e876371f8350c446a33577cf4a0aedcd4742a)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 50 ++++++++++++++++++++++++
1 file changed, 50 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 92459589f6ce..0d6c7215e0c1 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2509,6 +2509,54 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf)
}
}
+/**
+ * ice_ptp_maybe_trigger_tx_interrupt - Trigger Tx timstamp interrupt
+ * @pf: Board private structure
+ *
+ * The device PHY issues Tx timestamp interrupts to the driver for processing
+ * timestamp data from the PHY. It will not interrupt again until all
+ * current timestamp data is read. In rare circumstances, it is possible that
+ * the driver fails to read all outstanding data.
+ *
+ * To avoid getting permanently stuck, periodically check if the PHY has
+ * outstanding timestamp data. If so, trigger an interrupt from software to
+ * process this data.
+ */
+static void ice_ptp_maybe_trigger_tx_interrupt(struct ice_pf *pf)
+{
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ bool trigger_oicr = false;
+ unsigned int i;
+
+ if (ice_is_e810(hw))
+ return;
+
+ if (!ice_pf_src_tmr_owned(pf))
+ return;
+
+ for (i = 0; i < ICE_MAX_QUAD; i++) {
+ u64 tstamp_ready;
+ int err;
+
+ err = ice_get_phy_tx_tstamp_ready(&pf->hw, i, &tstamp_ready);
+ if (!err && tstamp_ready) {
+ trigger_oicr = true;
+ break;
+ }
+ }
+
+ if (trigger_oicr) {
+ /* Trigger a software interrupt, to ensure this data
+ * gets processed.
+ */
+ dev_dbg(dev, "PTP periodic task detected waiting timestamps. Triggering Tx timestamp interrupt now.\n");
+
+ wr32(hw, PFINT_OICR, PFINT_OICR_TSYN_TX_M);
+ ice_flush(hw);
+ }
+}
+
static void ice_ptp_periodic_work(struct kthread_work *work)
{
struct ice_ptp *ptp = container_of(work, struct ice_ptp, work.work);
@@ -2520,6 +2568,8 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
err = ice_ptp_update_cached_phctime(pf);
+ ice_ptp_maybe_trigger_tx_interrupt(pf);
+
/* Run twice a second or reschedule if phc update failed */
kthread_queue_delayed_work(ptp->kworker, &ptp->work,
msecs_to_jiffies(err ? 10 : 500));
--
2.43.0

81
kernel-std/debian/patches/ice-VDF/0005-ice-PTP-Rename-macros-used-for-PHY-QUAD-port-definit.patch Normal file
View File

@ -0,0 +1,81 @@
From c25fc364d599195403ed9ba51ef8fa6ed3b642ff Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:44 -0700
Subject: [PATCH 05/36] ice: PTP: Rename macros used for PHY/QUAD port
definitions
The ice_fill_phy_msg_e822 function uses several macros to specify the
correct address when sending a sideband message to the PHY block in
hardware.
The names of these macros are fairly generic and confusing. Future
development is going to extend the driver to support new hardware families
which have different relationships between PHY and QUAD. Rename the macros
for clarity and to indicate that they are E822 specific. This also matches
closer to the hardware specification in the data sheet.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 64fd7de2469dd52a7f1517ce95ae22fcb391a8a1)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 8 ++++----
drivers/net/ethernet/intel/ice/ice_type.h | 14 +++++++-------
2 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index a299af39a7c4..03c4aa995e8d 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -294,9 +294,9 @@ ice_fill_phy_msg_e822(struct ice_sbq_msg_input *msg, u8 port, u16 offset)
{
int phy_port, phy, quadtype;
- phy_port = port % ICE_PORTS_PER_PHY;
- phy = port / ICE_PORTS_PER_PHY;
- quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_NUM_QUAD_TYPE;
+ phy_port = port % ICE_PORTS_PER_PHY_E822;
+ phy = port / ICE_PORTS_PER_PHY_E822;
+ quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_QUADS_PER_PHY_E822;
if (quadtype == 0) {
msg->msg_addr_low = P_Q0_L(P_0_BASE + offset, phy_port);
@@ -628,7 +628,7 @@ ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
msg->dest_dev = rmn_0;
- if ((quad % ICE_NUM_QUAD_TYPE) == 0)
+ if ((quad % ICE_QUADS_PER_PHY_E822) == 0)
addr = Q_0_BASE + offset;
else
addr = Q_1_BASE + offset;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 4cd131546aa9..bb5d8b681bc2 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -909,13 +909,13 @@ struct ice_hw {
/* INTRL granularity in 1 us */
u8 intrl_gran;
-#define ICE_PHY_PER_NAC 1
-#define ICE_MAX_QUAD 2
-#define ICE_NUM_QUAD_TYPE 2
-#define ICE_PORTS_PER_QUAD 4
-#define ICE_PHY_0_LAST_QUAD 1
-#define ICE_PORTS_PER_PHY 8
-#define ICE_NUM_EXTERNAL_PORTS ICE_PORTS_PER_PHY
+#define ICE_PHY_PER_NAC_E822 1
+#define ICE_MAX_QUAD 2
+#define ICE_QUADS_PER_PHY_E822 2
+#define ICE_PORTS_PER_PHY_E822 8
+#define ICE_PORTS_PER_QUAD 4
+#define ICE_PORTS_PER_PHY_E810 4
+#define ICE_NUM_EXTERNAL_PORTS (ICE_MAX_QUAD * ICE_PORTS_PER_QUAD)
/* Active package version (currently active) */
struct ice_pkg_ver active_pkg_ver;
--
2.43.0

80
kernel-std/debian/patches/ice-VDF/0006-ice-PTP-move-quad-value-check-inside-ice_fill_phy_ms.patch Normal file
View File

@ -0,0 +1,80 @@
From 13f48f4c94ad4d317e7c7ccaa188a11850a8aa32 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:45 -0700
Subject: [PATCH 06/36] ice: PTP: move quad value check inside
ice_fill_phy_msg_e822
The callers of ice_fill_phy_msg_e822 check for whether the quad number is
within the expected range. Move this check inside the ice_fill_phy_msg_e822
function instead of duplicating it twice.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit dd84744cf5ea967c8d53aae6b6a45703dbc5c5c4)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 19 ++++++++++++-------
1 file changed, 12 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index 03c4aa995e8d..e024b88ce32b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -621,11 +621,14 @@ ice_write_64b_phy_reg_e822(struct ice_hw *hw, u8 port, u16 low_addr, u64 val)
* Fill a message buffer for accessing a register in a quad shared between
* multiple PHYs.
*/
-static void
+static int
ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
{
u32 addr;
+ if (quad >= ICE_MAX_QUAD)
+ return -EINVAL;
+
msg->dest_dev = rmn_0;
if ((quad % ICE_QUADS_PER_PHY_E822) == 0)
@@ -635,6 +638,8 @@ ice_fill_quad_msg_e822(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
msg->msg_addr_low = lower_16_bits(addr);
msg->msg_addr_high = upper_16_bits(addr);
+
+ return 0;
}
/**
@@ -653,10 +658,10 @@ ice_read_quad_reg_e822(struct ice_hw *hw, u8 quad, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- if (quad >= ICE_MAX_QUAD)
- return -EINVAL;
+ err = ice_fill_quad_msg_e822(&msg, quad, offset);
+ if (err)
+ return err;
- ice_fill_quad_msg_e822(&msg, quad, offset);
msg.opcode = ice_sbq_msg_rd;
err = ice_sbq_rw_reg(hw, &msg);
@@ -687,10 +692,10 @@ ice_write_quad_reg_e822(struct ice_hw *hw, u8 quad, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- if (quad >= ICE_MAX_QUAD)
- return -EINVAL;
+ err = ice_fill_quad_msg_e822(&msg, quad, offset);
+ if (err)
+ return err;
- ice_fill_quad_msg_e822(&msg, quad, offset);
msg.opcode = ice_sbq_msg_wr;
msg.data = val;
--
2.43.0

97
kernel-std/debian/patches/ice-VDF/0007-ice-remove-ptp_tx-ring-parameter-flag.patch Normal file
View File

@ -0,0 +1,97 @@
From 7dae9333af82f6c9e2db1940c3a10ae38dabea7b Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:55 -0800
Subject: [PATCH 07/36] ice: remove ptp_tx ring parameter flag
Before performing a Tx timestamp in ice_stamp(), the driver checks a ptp_tx
ring variable to see if timestamping is enabled on that ring. This value is
set for all rings whenever userspace configures Tx timestamping.
Ostensibly this was done to avoid wasting cycles checking other fields when
timestamping has not been enabled. However, for Tx timestamps we already
get an individual per-SKB flag indicating whether userspace wants to
request a timestamp on that packet. We do not gain much by also having
a separate flag to check for whether timestamping was enabled.
In fact, the driver currently fails to restore the field after a PF reset.
Because of this, if a PF reset occurs, timestamps will be disabled.
Since this flag doesn't add value in the hotpath, remove it and always
provide a timestamp if the SKB flag has been set.
A following change will fix the reset path to properly restore user
timestamping configuration completely.
This went unnoticed for some time because one of the most common
applications using Tx timestamps, ptp4l, will reconfigure the socket as
part of its fault recovery logic.
Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 0ffb08b1a45bd6b7694e01da0e1d9e3e788418fb)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 14 --------------
drivers/net/ethernet/intel/ice/ice_txrx.c | 3 ---
drivers/net/ethernet/intel/ice/ice_txrx.h | 1 -
3 files changed, 18 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 0d6c7215e0c1..c03153bdb7c3 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -280,20 +280,6 @@ static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
*/
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
- struct ice_vsi *vsi;
- u16 i;
-
- vsi = ice_get_main_vsi(pf);
- if (!vsi)
- return;
-
- /* Set the timestamp enable flag for all the Tx rings */
- ice_for_each_txq(vsi, i) {
- if (!vsi->tx_rings[i])
- continue;
- vsi->tx_rings[i]->ptp_tx = on;
- }
-
if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
ice_ptp_configure_tx_tstamp(pf, on);
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index 24c914015973..9170a3e8f088 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -2305,9 +2305,6 @@ ice_tstamp(struct ice_tx_ring *tx_ring, struct sk_buff *skb,
if (likely(!(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)))
return;
- if (!tx_ring->ptp_tx)
- return;
-
/* Tx timestamps cannot be sampled when doing TSO */
if (first->tx_flags & ICE_TX_FLAGS_TSO)
return;
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.h b/drivers/net/ethernet/intel/ice/ice_txrx.h
index 407d4c320097..b28b9826bbcd 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.h
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.h
@@ -381,7 +381,6 @@ struct ice_tx_ring {
#define ICE_TX_FLAGS_RING_VLAN_L2TAG2 BIT(2)
u8 flags;
u8 dcb_tc; /* Traffic class of ring */
- u8 ptp_tx;
} ____cacheline_internodealigned_in_smp;
static inline bool ice_ring_uses_build_skb(struct ice_rx_ring *ring)
--
2.43.0

160
kernel-std/debian/patches/ice-VDF/0008-ice-unify-logic-for-programming-PFINT_TSYN_MSK.patch Normal file
View File

@ -0,0 +1,160 @@
From 99007ca6255e2c35256bd97fa141705d301eb934 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:56 -0800
Subject: [PATCH 08/36] ice: unify logic for programming PFINT_TSYN_MSK
Commit d938a8cca88a ("ice: Auxbus devices & driver for E822 TS") modified
how Tx timestamps are handled for E822 devices. On these devices, only the
clock owner handles reading the Tx timestamp data from firmware. To do
this, the PFINT_TSYN_MSK register is modified from the default value to one
which enables reacting to a Tx timestamp on all PHY ports.
The driver currently programs PFINT_TSYN_MSK in different places depending
on whether the port is the clock owner or not. For the clock owner, the
PFINT_TSYN_MSK value is programmed during ice_ptp_init_owner just before
calling ice_ptp_tx_ena_intr to program the PHY ports.
For the non-clock owner ports, the PFINT_TSYN_MSK is programmed during
ice_ptp_init_port.
If a large enough device reset occurs, the PFINT_TSYN_MSK register will be
reset to the default value in which only the PHY associated directly with
the PF will cause the Tx timestamp interrupt to trigger.
The driver lacks logic to reprogram the PFINT_TSYN_MSK register after a
device reset. For the E822 device, this results in the PF no longer
responding to interrupts for other ports. This results in failure to
deliver Tx timestamps to user space applications.
Rename ice_ptp_configure_tx_tstamp to ice_ptp_cfg_tx_interrupt, and unify
the logic for programming PFINT_TSYN_MSK and PFINT_OICR_ENA into one place.
This function will program both registers according to the combination of
user configuration and device requirements.
This ensures that PFINT_TSYN_MSK is always restored when we configure the
Tx timestamp interrupt.
Fixes: d938a8cca88a ("ice: Auxbus devices & driver for E822 TS")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7d606a1e2d0575b6c3a2600f43f90d1e409f9661)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 60 ++++++++++++++----------
1 file changed, 34 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index c03153bdb7c3..b0bba866e8a2 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -256,21 +256,42 @@ ice_verify_pin_e810t(struct ptp_clock_info *info, unsigned int pin,
}
/**
- * ice_ptp_configure_tx_tstamp - Enable or disable Tx timestamp interrupt
- * @pf: The PF pointer to search in
- * @on: bool value for whether timestamp interrupt is enabled or disabled
+ * ice_ptp_cfg_tx_interrupt - Configure Tx timestamp interrupt for the device
+ * @pf: Board private structure
+ *
+ * Program the device to respond appropriately to the Tx timestamp interrupt
+ * cause.
*/
-static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
+static void ice_ptp_cfg_tx_interrupt(struct ice_pf *pf)
{
+ struct ice_hw *hw = &pf->hw;
+ bool enable;
u32 val;
+ switch (pf->ptp.tx_interrupt_mode) {
+ case ICE_PTP_TX_INTERRUPT_ALL:
+ /* React to interrupts across all quads. */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
+ enable = true;
+ break;
+ case ICE_PTP_TX_INTERRUPT_NONE:
+ /* Do not react to interrupts on any quad. */
+ wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
+ enable = false;
+ break;
+ case ICE_PTP_TX_INTERRUPT_SELF:
+ default:
+ enable = pf->ptp.tstamp_config.tx_type == HWTSTAMP_TX_ON;
+ break;
+ }
+
/* Configure the Tx timestamp interrupt */
- val = rd32(&pf->hw, PFINT_OICR_ENA);
- if (on)
+ val = rd32(hw, PFINT_OICR_ENA);
+ if (enable)
val |= PFINT_OICR_TSYN_TX_M;
else
val &= ~PFINT_OICR_TSYN_TX_M;
- wr32(&pf->hw, PFINT_OICR_ENA, val);
+ wr32(hw, PFINT_OICR_ENA, val);
}
/**
@@ -280,10 +301,9 @@ static void ice_ptp_configure_tx_tstamp(struct ice_pf *pf, bool on)
*/
static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
{
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
- ice_ptp_configure_tx_tstamp(pf, on);
-
pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
+
+ ice_ptp_cfg_tx_interrupt(pf);
}
/**
@@ -2909,15 +2929,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL) {
- /* The clock owner for this device type handles the timestamp
- * interrupt for all ports.
- */
- ice_ptp_configure_tx_tstamp(pf, true);
-
- /* React on all quads interrupts for E82x */
- wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x1f);
-
+ if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_tx_ena_intr(pf, true, itr);
if (err)
@@ -2991,13 +3003,6 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E822:
- /* Non-owner PFs don't react to any interrupts on E82x,
- * neither on own quad nor on others
- */
- if (!ice_ptp_pf_handles_tx_interrupt(pf)) {
- ice_ptp_configure_tx_tstamp(pf, false);
- wr32(hw, PFINT_TSYN_MSK + (0x4 * hw->pf_id), (u32)0x0);
- }
kthread_init_delayed_work(&ptp_port->ov_work,
ice_ptp_wait_for_offsets);
@@ -3142,6 +3147,9 @@ void ice_ptp_init(struct ice_pf *pf)
/* Start the PHY timestamping block */
ice_ptp_reset_phy_timestamping(pf);
+ /* Configure initial Tx interrupt settings */
+ ice_ptp_cfg_tx_interrupt(pf);
+
set_bit(ICE_FLAG_PTP, pf->flags);
err = ice_ptp_init_work(pf, ptp);
if (err)
--
2.43.0

147
kernel-std/debian/patches/ice-VDF/0009-ice-PTP-Clean-up-timestamp-registers-correctly.patch Normal file
View File

@ -0,0 +1,147 @@
From e5a65377977e338a8f7baf92892481acf1c62403 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 26 Jul 2023 11:27:43 -0700
Subject: [PATCH 09/36] ice: PTP: Clean up timestamp registers correctly
E822 PHY TS registers should not be written and the only way to clean up
them is to reset QUAD memory.
To ensure that the status bit for the timestamp index is cleared, ensure
that ice_clear_phy_tstamp implementations first read the timestamp out.
Implementations which can write the register continue to do so.
Add a note to indicate this function should only be called on timestamps
which have their valid bit set. Update the dynamic debug messages to
reflect the actual action taken.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit be65a1a33bdee3912daac50aa6c5270ec9c37010)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 70 +++++++++++++--------
1 file changed, 45 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index e024b88ce32b..cd28430cfdda 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -759,29 +759,32 @@ ice_read_phy_tstamp_e822(struct ice_hw *hw, u8 quad, u8 idx, u64 *tstamp)
* @quad: the quad to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the PHY quad block that is
- * shared between the internal PHYs on the E822 devices.
+ * Read the timestamp out of the quad to clear its timestamp status bit from
+ * the PHY quad block that is shared between the internal PHYs of the E822
+ * devices.
+ *
+ * Note that unlike E810, software cannot directly write to the quad memory
+ * bank registers. E822 relies on the ice_get_phy_tx_tstamp_ready() function
+ * to determine which timestamps are valid. Reading a timestamp auto-clears
+ * the valid bit.
+ *
+ * To directly clear the contents of the timestamp block entirely, discarding
+ * all timestamp data at once, software should instead use
+ * ice_ptp_reset_ts_memory_quad_e822().
+ *
+ * This function should only be called on an idx whose bit is set according to
+ * ice_get_phy_tx_tstamp_ready().
*/
static int
ice_clear_phy_tstamp_e822(struct ice_hw *hw, u8 quad, u8 idx)
{
- u16 lo_addr, hi_addr;
+ u64 unused_tstamp;
int err;
- lo_addr = (u16)TS_L(Q_REG_TX_MEMORY_BANK_START, idx);
- hi_addr = (u16)TS_H(Q_REG_TX_MEMORY_BANK_START, idx);
-
- err = ice_write_quad_reg_e822(hw, quad, lo_addr, 0);
- if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register, err %d\n",
- err);
- return err;
- }
-
- err = ice_write_quad_reg_e822(hw, quad, hi_addr, 0);
+ err = ice_read_phy_tstamp_e822(hw, quad, idx, &unused_tstamp);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to read the timestamp register for quad %u, idx %u, err %d\n",
+ quad, idx, err);
return err;
}
@@ -2816,28 +2819,39 @@ ice_read_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx, u64 *tstamp)
* @lport: the lport to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the timestamp block of the
- * external PHY on the E810 device.
+ * Read the timestamp and then forcibly overwrite its value to clear the valid
+ * bit from the timestamp block of the external PHY on the E810 device.
+ *
+ * This function should only be called on an idx whose bit is set according to
+ * ice_get_phy_tx_tstamp_ready().
*/
static int ice_clear_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx)
{
u32 lo_addr, hi_addr;
+ u64 unused_tstamp;
int err;
+ err = ice_read_phy_tstamp_e810(hw, lport, idx, &unused_tstamp);
+ if (err) {
+ ice_debug(hw, ICE_DBG_PTP, "Failed to read the timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
+ return err;
+ }
+
lo_addr = TS_EXT(LOW_TX_MEMORY_BANK_START, lport, idx);
hi_addr = TS_EXT(HIGH_TX_MEMORY_BANK_START, lport, idx);
err = ice_write_phy_reg_e810(hw, lo_addr, 0);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to clear low PTP timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
return err;
}
err = ice_write_phy_reg_e810(hw, hi_addr, 0);
if (err) {
- ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register, err %d\n",
- err);
+ ice_debug(hw, ICE_DBG_PTP, "Failed to clear high PTP timestamp register for lport %u, idx %u, err %d\n",
+ lport, idx, err);
return err;
}
@@ -3519,9 +3533,15 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
* @block: the block to read from
* @idx: the timestamp index to reset
*
- * Clear a timestamp, resetting its valid bit, from the timestamp block. For
- * E822 devices, the block is the quad to clear from. For E810 devices, the
- * block is the logical port to clear from.
+ * Clear a timestamp from the timestamp block, discarding its value without
+ * returning it. This resets the memory status bit for the timestamp index
+ * allowing it to be reused for another timestamp in the future.
+ *
+ * For E822 devices, the block number is the PHY quad to clear from. For E810
+ * devices, the block number is the logical port to clear from.
+ *
+ * This function must only be called on a timestamp index whose valid bit is
+ * set according to ice_get_phy_tx_tstamp_ready().
*/
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx)
{
--
2.43.0

65
kernel-std/debian/patches/ice-VDF/0010-ice-Use-PTP-auxbus-for-all-PHYs-restart-in-E822.patch Normal file
View File

@ -0,0 +1,65 @@
From e2a74a0a7dd399b0ee2ddd4889c609dedb85bfb5 Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:35 +0200
Subject: [PATCH 10/36] ice: Use PTP auxbus for all PHYs restart in E822
The E822 (and other devices based on the same PHY) is having issue while
setting the PHC timer - the PHY timers are drifting from the PHC. After
such a set all PHYs need to be restarted and resynchronised - do it
using auxiliary bus.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit af3c5c8748e6d286d4f2dd9800f9d27f29b8e2ef)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 24 +++++++++++++++++++++---
1 file changed, 21 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index b0bba866e8a2..42eb1418eb90 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1496,6 +1496,24 @@ static void ice_ptp_reset_phy_timestamping(struct ice_pf *pf)
ice_ptp_port_phy_restart(&pf->ptp.port);
}
+/**
+ * ice_ptp_restart_all_phy - Restart all PHYs to recalibrate timestamping
+ * @pf: Board private structure
+ */
+static void ice_ptp_restart_all_phy(struct ice_pf *pf)
+{
+ struct list_head *entry;
+
+ list_for_each(entry, &pf->ptp.ports_owner.ports) {
+ struct ice_ptp_port *port = list_entry(entry,
+ struct ice_ptp_port,
+ list_member);
+
+ if (port->link_up)
+ ice_ptp_port_phy_restart(port);
+ }
+}
+
/**
* ice_ptp_adjfine - Adjust clock increment rate
* @info: the driver's PTP info structure
@@ -1933,9 +1951,9 @@ ice_ptp_settime64(struct ptp_clock_info *info, const struct timespec64 *ts)
/* Reenable periodic outputs */
ice_ptp_enable_all_clkout(pf);
- /* Recalibrate and re-enable timestamp block */
- if (pf->ptp.port.link_up)
- ice_ptp_port_phy_restart(&pf->ptp.port);
+ /* Recalibrate and re-enable timestamp blocks for E822/E823 */
+ if (hw->phy_model == ICE_PHY_E822)
+ ice_ptp_restart_all_phy(pf);
exit:
if (err) {
dev_err(ice_pf_to_dev(pf), "PTP failed to set time %d\n", err);
--
2.43.0

2177
kernel-std/debian/patches/ice-VDF/0011-ice-Rename-E822-to-E82X.patch Normal file
View File

File diff suppressed because it is too large Load Diff

104
kernel-std/debian/patches/ice-VDF/0012-ice-Schedule-service-task-in-IRQ-top-half.patch Normal file
View File

@ -0,0 +1,104 @@
From 3b37119a08ffe4be182ade746a6b1fe3bcf65921 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 29 Nov 2023 13:40:22 +0100
Subject: [PATCH 12/36] ice: Schedule service task in IRQ top half
Schedule service task and EXTTS in the top half to avoid bottom half
scheduling if possible, which significantly reduces timestamping delay.
Co-developed-by: Michal Michalik <michal.michalik@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 00d50001444ef5c75c8ab476a6674708f3ff613b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 -
drivers/net/ethernet/intel/ice/ice_main.c | 20 +++++++++++---------
2 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 54a98c4032b7..efe78d5e4da1 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -517,7 +517,6 @@ enum ice_pf_flags {
};
enum ice_misc_thread_tasks {
- ICE_MISC_THREAD_EXTTS_EVENT,
ICE_MISC_THREAD_TX_TSTAMP,
ICE_MISC_THREAD_NBITS /* must be last */
};
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d2f3b4374d14..2acaa17a12bf 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3109,6 +3109,7 @@ static void ice_ena_misc_vector(struct ice_pf *pf)
static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
{
struct ice_pf *pf = (struct ice_pf *)data;
+ irqreturn_t ret = IRQ_HANDLED;
struct ice_hw *hw = &pf->hw;
struct device *dev;
u32 oicr, ena_mask;
@@ -3190,8 +3191,10 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (ice_ptp_pf_handles_tx_interrupt(pf))
+ if (ice_ptp_pf_handles_tx_interrupt(pf)) {
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
+ ret = IRQ_WAKE_THREAD;
+ }
}
if (oicr & PFINT_OICR_TSYN_EVNT_M) {
@@ -3207,7 +3210,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
GLTSYN_STAT_EVENT1_M |
GLTSYN_STAT_EVENT2_M);
- set_bit(ICE_MISC_THREAD_EXTTS_EVENT, pf->misc_thread);
+ ice_ptp_extts_event(pf);
}
}
@@ -3230,8 +3233,11 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
set_bit(ICE_PFR_REQ, pf->state);
}
}
+ ice_service_task_schedule(pf);
+ if (ret == IRQ_HANDLED)
+ ice_irq_dynamic_ena(hw, NULL, NULL);
- return IRQ_WAKE_THREAD;
+ return ret;
}
/**
@@ -3247,12 +3253,7 @@ static irqreturn_t ice_misc_intr_thread_fn(int __always_unused irq, void *data)
hw = &pf->hw;
if (ice_is_reset_in_progress(pf->state))
- return IRQ_HANDLED;
-
- ice_service_task_schedule(pf);
-
- if (test_and_clear_bit(ICE_MISC_THREAD_EXTTS_EVENT, pf->misc_thread))
- ice_ptp_extts_event(pf);
+ goto skip_irq;
if (test_and_clear_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread)) {
/* Process outstanding Tx timestamps. If there is more work,
@@ -3264,6 +3265,7 @@ static irqreturn_t ice_misc_intr_thread_fn(int __always_unused irq, void *data)
}
}
+skip_irq:
ice_irq_dynamic_ena(hw, NULL, NULL);
return IRQ_HANDLED;
--
2.43.0

660
kernel-std/debian/patches/ice-VDF/0013-ice-Enable-SW-interrupt-from-FW-for-LL-TS.patch Normal file
View File

@ -0,0 +1,660 @@
From c4ab92eb3ee89178a012702f2a98477d683fad31 Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Wed, 29 Nov 2023 13:40:23 +0100
Subject: [PATCH 13/36] ice: Enable SW interrupt from FW for LL TS
Introduce new capability - Low Latency Timestamping with Interrupt.
On supported devices, driver can request a single timestamp from FW
without polling the register afterwards. Instead, FW can issue
a dedicated interrupt when the timestamp was read from the PHY register
and its value is available to read from the register.
This eliminates the need of bottom half scheduling, which results in
minimal delay for timestamping.
For this mode, allocate TS indices sequentially, so that timestamps are
always completed in FIFO manner.
Co-developed-by: Yochai Hagvi <yochai.hagvi@intel.com>
Signed-off-by: Yochai Hagvi <yochai.hagvi@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 82e71b226e0ef770d7bc143701c8b4960b4eb3d5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 +
drivers/net/ethernet/intel/ice/ice_common.c | 3 +
.../net/ethernet/intel/ice/ice_hw_autogen.h | 2 +
drivers/net/ethernet/intel/ice/ice_main.c | 120 +++++++++++--
drivers/net/ethernet/intel/ice/ice_ptp.c | 163 ++++++++++++++++--
drivers/net/ethernet/intel/ice/ice_ptp.h | 9 +
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 1 +
drivers/net/ethernet/intel/ice/ice_type.h | 2 +
8 files changed, 274 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index efe78d5e4da1..ee42a504c2f4 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -594,6 +594,7 @@ struct ice_pf {
u32 hw_csum_rx_error;
u32 oicr_err_reg;
struct msi_map oicr_irq; /* Other interrupt cause MSIX vector */
+ struct msi_map ll_ts_irq; /* LL_TS interrupt MSIX vector */
u16 max_pf_txqs; /* Total Tx queues PF wide */
u16 max_pf_rxqs; /* Total Rx queues PF wide */
u16 num_lan_msix; /* Total MSIX vectors for base driver */
@@ -618,6 +619,7 @@ struct ice_pf {
unsigned long tx_timeout_last_recovery;
u32 tx_timeout_recovery_level;
char int_name[ICE_INT_NAME_STR_LEN];
+ char int_name_ll_ts[ICE_INT_NAME_STR_LEN];
struct auxiliary_device *adev;
int aux_idx;
u32 sw_int_count;
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 7674267a2d90..acf6ac00f804 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -2624,6 +2624,7 @@ ice_parse_1588_dev_caps(struct ice_hw *hw, struct ice_hw_dev_caps *dev_p,
info->tmr1_ena = ((number & ICE_TS_TMR1_ENA_M) != 0);
info->ts_ll_read = ((number & ICE_TS_LL_TX_TS_READ_M) != 0);
+ info->ts_ll_int_read = ((number & ICE_TS_LL_TX_TS_INT_READ_M) != 0);
info->ena_ports = logical_id;
info->tmr_own_map = phys_id;
@@ -2644,6 +2645,8 @@ ice_parse_1588_dev_caps(struct ice_hw *hw, struct ice_hw_dev_caps *dev_p,
info->tmr1_ena);
ice_debug(hw, ICE_DBG_INIT, "dev caps: ts_ll_read = %u\n",
info->ts_ll_read);
+ ice_debug(hw, ICE_DBG_INIT, "dev caps: ts_ll_int_read = %u\n",
+ info->ts_ll_int_read);
ice_debug(hw, ICE_DBG_INIT, "dev caps: ieee_1588 ena_ports = %u\n",
info->ena_ports);
ice_debug(hw, ICE_DBG_INIT, "dev caps: tmr_own_map = %u\n",
diff --git a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
index 6756f3d51d14..fa730bca7f15 100644
--- a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
+++ b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h
@@ -200,6 +200,8 @@
#define GLINT_VECT2FUNC_PF_NUM_M ICE_M(0x7, 12)
#define GLINT_VECT2FUNC_IS_PF_S 16
#define GLINT_VECT2FUNC_IS_PF_M BIT(16)
+#define PFINT_ALLOC 0x001D2600
+#define PFINT_ALLOC_FIRST ICE_M(0x7FF, 0)
#define PFINT_FW_CTL 0x0016C800
#define PFINT_FW_CTL_MSIX_INDX_M ICE_M(0x7FF, 0)
#define PFINT_FW_CTL_ITR_INDX_S 11
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 2acaa17a12bf..9163a72368b3 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -3071,6 +3071,7 @@ static int ice_xdp(struct net_device *dev, struct netdev_bpf *xdp)
static void ice_ena_misc_vector(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
+ u32 pf_intr_start_offset;
u32 val;
/* Disable anti-spoof detection interrupt to prevent spurious event
@@ -3099,6 +3100,47 @@ static void ice_ena_misc_vector(struct ice_pf *pf)
/* SW_ITR_IDX = 0, but don't change INTENA */
wr32(hw, GLINT_DYN_CTL(pf->oicr_irq.index),
GLINT_DYN_CTL_SW_ITR_INDX_M | GLINT_DYN_CTL_INTENA_MSK_M);
+
+ if (!pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ return;
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ wr32(hw, GLINT_DYN_CTL(pf->ll_ts_irq.index + pf_intr_start_offset),
+ GLINT_DYN_CTL_SW_ITR_INDX_M | GLINT_DYN_CTL_INTENA_MSK_M);
+}
+
+/**
+ * ice_ll_ts_intr - ll_ts interrupt handler
+ * @irq: interrupt number
+ * @data: pointer to a q_vector
+ */
+static irqreturn_t ice_ll_ts_intr(int __always_unused irq, void *data)
+{
+ struct ice_pf *pf = data;
+ u32 pf_intr_start_offset;
+ struct ice_ptp_tx *tx;
+ unsigned long flags;
+ struct ice_hw *hw;
+ u32 val;
+ u8 idx;
+
+ hw = &pf->hw;
+ tx = &pf->ptp.port.tx;
+ spin_lock_irqsave(&tx->lock, flags);
+ ice_ptp_complete_tx_single_tstamp(tx);
+
+ idx = find_next_bit_wrap(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx != tx->len)
+ ice_ptp_req_tx_single_tstamp(tx, idx);
+ spin_unlock_irqrestore(&tx->lock, flags);
+
+ val = GLINT_DYN_CTL_INTENA_M | GLINT_DYN_CTL_CLEARPBA_M |
+ (ICE_ITR_NONE << GLINT_DYN_CTL_ITR_INDX_S);
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ wr32(hw, GLINT_DYN_CTL(pf->ll_ts_irq.index + pf_intr_start_offset),
+ val);
+
+ return IRQ_HANDLED;
}
/**
@@ -3191,7 +3233,19 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
if (oicr & PFINT_OICR_TSYN_TX_M) {
ena_mask &= ~PFINT_OICR_TSYN_TX_M;
- if (ice_ptp_pf_handles_tx_interrupt(pf)) {
+ if (ice_pf_state_is_nominal(pf) &&
+ pf->hw.dev_caps.ts_dev_info.ts_ll_int_read) {
+ struct ice_ptp_tx *tx = &pf->ptp.port.tx;
+ unsigned long flags;
+ u8 idx;
+
+ spin_lock_irqsave(&tx->lock, flags);
+ idx = find_next_bit_wrap(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx != tx->len)
+ ice_ptp_req_tx_single_tstamp(tx, idx);
+ spin_unlock_irqrestore(&tx->lock, flags);
+ } else if (ice_ptp_pf_handles_tx_interrupt(pf)) {
set_bit(ICE_MISC_THREAD_TX_TSTAMP, pf->misc_thread);
ret = IRQ_WAKE_THREAD;
}
@@ -3295,6 +3349,20 @@ static void ice_dis_ctrlq_interrupts(struct ice_hw *hw)
ice_flush(hw);
}
+/**
+ * ice_free_irq_msix_ll_ts- Unroll ll_ts vector setup
+ * @pf: board private structure
+ */
+static void ice_free_irq_msix_ll_ts(struct ice_pf *pf)
+{
+ int irq_num = pf->ll_ts_irq.virq;
+
+ synchronize_irq(irq_num);
+ devm_free_irq(ice_pf_to_dev(pf), irq_num, pf);
+
+ ice_free_irq(pf, pf->ll_ts_irq);
+}
+
/**
* ice_free_irq_msix_misc - Unroll misc vector setup
* @pf: board private structure
@@ -3314,6 +3382,8 @@ static void ice_free_irq_msix_misc(struct ice_pf *pf)
devm_free_irq(ice_pf_to_dev(pf), misc_irq_num, pf);
ice_free_irq(pf, pf->oicr_irq);
+ if (pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ ice_free_irq_msix_ll_ts(pf);
}
/**
@@ -3339,10 +3409,12 @@ static void ice_ena_ctrlq_interrupts(struct ice_hw *hw, u16 reg_idx)
PFINT_MBX_CTL_CAUSE_ENA_M);
wr32(hw, PFINT_MBX_CTL, val);
- /* This enables Sideband queue Interrupt causes */
- val = ((reg_idx & PFINT_SB_CTL_MSIX_INDX_M) |
- PFINT_SB_CTL_CAUSE_ENA_M);
- wr32(hw, PFINT_SB_CTL, val);
+ if (!hw->dev_caps.ts_dev_info.ts_ll_int_read) {
+ /* enable Sideband queue Interrupt causes */
+ val = ((reg_idx & PFINT_SB_CTL_MSIX_INDX_M) |
+ PFINT_SB_CTL_CAUSE_ENA_M);
+ wr32(hw, PFINT_SB_CTL, val);
+ }
ice_flush(hw);
}
@@ -3359,13 +3431,17 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
- struct msi_map oicr_irq;
+ u32 pf_intr_start_offset;
+ struct msi_map irq;
int err = 0;
if (!pf->int_name[0])
snprintf(pf->int_name, sizeof(pf->int_name) - 1, "%s-%s:misc",
dev_driver_string(dev), dev_name(dev));
+ if (!pf->int_name_ll_ts[0])
+ snprintf(pf->int_name_ll_ts, sizeof(pf->int_name_ll_ts) - 1,
+ "%s-%s:ll_ts", dev_driver_string(dev), dev_name(dev));
/* Do not request IRQ but do enable OICR interrupt since settings are
* lost during reset. Note that this function is called only during
* rebuild path and not while reset is in progress.
@@ -3374,11 +3450,11 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
goto skip_req_irq;
/* reserve one vector in irq_tracker for misc interrupts */
- oicr_irq = ice_alloc_irq(pf, false);
- if (oicr_irq.index < 0)
- return oicr_irq.index;
+ irq = ice_alloc_irq(pf, false);
+ if (irq.index < 0)
+ return irq.index;
- pf->oicr_irq = oicr_irq;
+ pf->oicr_irq = irq;
err = devm_request_threaded_irq(dev, pf->oicr_irq.virq, ice_misc_intr,
ice_misc_intr_thread_fn, 0,
pf->int_name, pf);
@@ -3389,10 +3465,34 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
return err;
}
+ /* reserve one vector in irq_tracker for ll_ts interrupt */
+ if (!pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ goto skip_req_irq;
+
+ irq = ice_alloc_irq(pf, false);
+ if (irq.index < 0)
+ return irq.index;
+
+ pf->ll_ts_irq = irq;
+ err = devm_request_irq(dev, pf->ll_ts_irq.virq, ice_ll_ts_intr, 0,
+ pf->int_name_ll_ts, pf);
+ if (err) {
+ dev_err(dev, "devm_request_irq for %s failed: %d\n",
+ pf->int_name_ll_ts, err);
+ ice_free_irq(pf, pf->ll_ts_irq);
+ return err;
+ }
+
skip_req_irq:
ice_ena_misc_vector(pf);
ice_ena_ctrlq_interrupts(hw, pf->oicr_irq.index);
+ /* This enables LL TS interrupt */
+ pf_intr_start_offset = rd32(hw, PFINT_ALLOC) & PFINT_ALLOC_FIRST;
+ if (pf->hw.dev_caps.ts_dev_info.ts_ll_int_read)
+ wr32(hw, PFINT_SB_CTL,
+ ((pf->ll_ts_irq.index + pf_intr_start_offset) &
+ PFINT_SB_CTL_MSIX_INDX_M) | PFINT_SB_CTL_CAUSE_ENA_M);
wr32(hw, GLINT_ITR(ICE_RX_ITR, pf->oicr_irq.index),
ITR_REG_ALIGN(ICE_ITR_8K) >> ICE_ITR_GRAN_S);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 2e6e1fc84d11..75038d826f71 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -634,6 +634,119 @@ ice_ptp_is_tx_tracker_up(struct ice_ptp_tx *tx)
return tx->init && !tx->calibrating;
}
+/**
+ * ice_ptp_req_tx_single_tstamp - Request Tx timestamp for a port from FW
+ * @tx: the PTP Tx timestamp tracker
+ * @idx: index of the timestamp to request
+ */
+void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx)
+{
+ struct ice_ptp_port *ptp_port;
+ struct sk_buff *skb;
+ struct ice_pf *pf;
+
+ if (!tx->init)
+ return;
+
+ ptp_port = container_of(tx, struct ice_ptp_port, tx);
+ pf = ptp_port_to_pf(ptp_port);
+
+ /* Drop packets which have waited for more than 2 seconds */
+ if (time_is_before_jiffies(tx->tstamps[idx].start + 2 * HZ)) {
+ /* Count the number of Tx timestamps that timed out */
+ pf->ptp.tx_hwtstamp_timeouts++;
+
+ skb = tx->tstamps[idx].skb;
+ tx->tstamps[idx].skb = NULL;
+ clear_bit(idx, tx->in_use);
+
+ dev_kfree_skb_any(skb);
+ return;
+ }
+
+ ice_trace(tx_tstamp_fw_req, tx->tstamps[idx].skb, idx);
+
+ /* Write TS index to read to the PF register so the FW can read it */
+ wr32(&pf->hw, PF_SB_ATQBAL,
+ TS_LL_READ_TS_INTR | FIELD_PREP(TS_LL_READ_TS_IDX, idx) |
+ TS_LL_READ_TS);
+ tx->last_ll_ts_idx_read = idx;
+}
+
+/**
+ * ice_ptp_complete_tx_single_tstamp - Complete Tx timestamp for a port
+ * @tx: the PTP Tx timestamp tracker
+ */
+void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
+{
+ struct skb_shared_hwtstamps shhwtstamps = {};
+ u8 idx = tx->last_ll_ts_idx_read;
+ struct ice_ptp_port *ptp_port;
+ u64 raw_tstamp, tstamp;
+ bool drop_ts = false;
+ struct sk_buff *skb;
+ struct ice_pf *pf;
+ u32 val;
+
+ if (!tx->init || tx->last_ll_ts_idx_read < 0)
+ return;
+
+ ptp_port = container_of(tx, struct ice_ptp_port, tx);
+ pf = ptp_port_to_pf(ptp_port);
+
+ ice_trace(tx_tstamp_fw_done, tx->tstamps[idx].skb, idx);
+
+ val = rd32(&pf->hw, PF_SB_ATQBAL);
+
+ /* When the bit is cleared, the TS is ready in the register */
+ if (val & TS_LL_READ_TS) {
+ dev_err(ice_pf_to_dev(pf), "Failed to get the Tx tstamp - FW not ready");
+ return;
+ }
+
+ /* High 8 bit value of the TS is on the bits 16:23 */
+ raw_tstamp = FIELD_GET(TS_LL_READ_TS_HIGH, val);
+ raw_tstamp <<= 32;
+
+ /* Read the low 32 bit value */
+ raw_tstamp |= (u64)rd32(&pf->hw, PF_SB_ATQBAH);
+
+ /* For PHYs which don't implement a proper timestamp ready bitmap,
+ * verify that the timestamp value is different from the last cached
+ * timestamp. If it is not, skip this for now assuming it hasn't yet
+ * been captured by hardware.
+ */
+ if (!drop_ts && tx->verify_cached &&
+ raw_tstamp == tx->tstamps[idx].cached_tstamp)
+ return;
+
+ if (tx->verify_cached && raw_tstamp)
+ tx->tstamps[idx].cached_tstamp = raw_tstamp;
+ clear_bit(idx, tx->in_use);
+ skb = tx->tstamps[idx].skb;
+ tx->tstamps[idx].skb = NULL;
+ if (test_and_clear_bit(idx, tx->stale))
+ drop_ts = true;
+
+ if (!skb)
+ return;
+
+ if (drop_ts) {
+ dev_kfree_skb_any(skb);
+ return;
+ }
+
+ /* Extend the timestamp using cached PHC time */
+ tstamp = ice_ptp_extend_40b_ts(pf, raw_tstamp);
+ if (tstamp) {
+ shhwtstamps.hwtstamp = ns_to_ktime(tstamp);
+ ice_trace(tx_tstamp_complete, skb, idx);
+ }
+
+ skb_tstamp_tx(skb, &shhwtstamps);
+ dev_kfree_skb_any(skb);
+}
+
/**
* ice_ptp_process_tx_tstamp - Process Tx timestamps for a port
* @tx: the PTP Tx timestamp tracker
@@ -685,6 +798,7 @@ ice_ptp_is_tx_tracker_up(struct ice_ptp_tx *tx)
static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
{
struct ice_ptp_port *ptp_port;
+ unsigned long flags;
struct ice_pf *pf;
struct ice_hw *hw;
u64 tstamp_ready;
@@ -756,7 +870,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
drop_ts = true;
skip_ts_read:
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
if (tx->verify_cached && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
@@ -764,7 +878,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
tx->tstamps[idx].skb = NULL;
if (test_and_clear_bit(idx, tx->stale))
drop_ts = true;
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* It is unlikely but possible that the SKB will have been
* flushed at this point due to link change or teardown.
@@ -834,6 +948,7 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
static enum ice_tx_tstamp_work ice_ptp_tx_tstamp(struct ice_ptp_tx *tx)
{
bool more_timestamps;
+ unsigned long flags;
if (!tx->init)
return ICE_TX_TSTAMP_WORK_DONE;
@@ -842,9 +957,9 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp(struct ice_ptp_tx *tx)
ice_ptp_process_tx_tstamp(tx);
/* Check if there are outstanding Tx timestamps */
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
more_timestamps = tx->init && !bitmap_empty(tx->in_use, tx->len);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
if (more_timestamps)
return ICE_TX_TSTAMP_WORK_PENDING;
@@ -881,6 +996,7 @@ ice_ptp_alloc_tx_tracker(struct ice_ptp_tx *tx)
tx->in_use = in_use;
tx->stale = stale;
tx->init = 1;
+ tx->last_ll_ts_idx_read = -1;
spin_lock_init(&tx->lock);
@@ -898,6 +1014,7 @@ static void
ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
{
struct ice_hw *hw = &pf->hw;
+ unsigned long flags;
u64 tstamp_ready;
int err;
u8 idx;
@@ -921,12 +1038,12 @@ ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
if (!hw->reset_ongoing && (tstamp_ready & BIT_ULL(phy_idx)))
ice_clear_phy_tstamp(hw, tx->block, phy_idx);
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
skb = tx->tstamps[idx].skb;
tx->tstamps[idx].skb = NULL;
clear_bit(idx, tx->in_use);
clear_bit(idx, tx->stale);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* Count the number of Tx timestamps flushed */
pf->ptp.tx_hwtstamp_flushed++;
@@ -950,9 +1067,11 @@ ice_ptp_flush_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
static void
ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
{
- spin_lock(&tx->lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&tx->lock, flags);
bitmap_or(tx->stale, tx->stale, tx->in_use, tx->len);
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
}
/**
@@ -965,9 +1084,11 @@ ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
static void
ice_ptp_release_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
{
- spin_lock(&tx->lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&tx->lock, flags);
tx->init = 0;
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* wait for potentially outstanding interrupt to complete */
synchronize_irq(pf->oicr_irq.virq);
@@ -1367,6 +1488,7 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
struct ice_pf *pf = ptp_port_to_pf(ptp_port);
u8 port = ptp_port->port_num;
struct ice_hw *hw = &pf->hw;
+ unsigned long flags;
int err;
if (ice_is_e810(hw))
@@ -1380,9 +1502,9 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
kthread_cancel_delayed_work_sync(&ptp_port->ov_work);
/* temporarily disable Tx timestamps while calibrating PHY offset */
- spin_lock(&ptp_port->tx.lock);
+ spin_lock_irqsave(&ptp_port->tx.lock, flags);
ptp_port->tx.calibrating = true;
- spin_unlock(&ptp_port->tx.lock);
+ spin_unlock_irqrestore(&ptp_port->tx.lock, flags);
ptp_port->tx_fifo_busy_cnt = 0;
/* Start the PHY timer in Vernier mode */
@@ -1391,9 +1513,9 @@ ice_ptp_port_phy_restart(struct ice_ptp_port *ptp_port)
goto out_unlock;
/* Enable Tx timestamps right away */
- spin_lock(&ptp_port->tx.lock);
+ spin_lock_irqsave(&ptp_port->tx.lock, flags);
ptp_port->tx.calibrating = false;
- spin_unlock(&ptp_port->tx.lock);
+ spin_unlock_irqrestore(&ptp_port->tx.lock, flags);
kthread_queue_delayed_work(pf->ptp.kworker, &ptp_port->ov_work, 0);
@@ -2471,18 +2593,23 @@ static long ice_ptp_create_clock(struct ice_pf *pf)
*/
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
{
+ unsigned long flags;
u8 idx;
- spin_lock(&tx->lock);
+ spin_lock_irqsave(&tx->lock, flags);
/* Check that this tracker is accepting new timestamp requests */
if (!ice_ptp_is_tx_tracker_up(tx)) {
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
return -1;
}
/* Find and set the first available index */
- idx = find_first_zero_bit(tx->in_use, tx->len);
+ idx = find_next_zero_bit(tx->in_use, tx->len,
+ tx->last_ll_ts_idx_read + 1);
+ if (idx == tx->len)
+ idx = find_first_zero_bit(tx->in_use, tx->len);
+
if (idx < tx->len) {
/* We got a valid index that no other thread could have set. Store
* a reference to the skb and the start time to allow discarding old
@@ -2496,7 +2623,7 @@ s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
ice_trace(tx_tstamp_request, skb, idx);
}
- spin_unlock(&tx->lock);
+ spin_unlock_irqrestore(&tx->lock, flags);
/* return the appropriate PHY timestamp register index, -1 if no
* indexes were available.
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index a3ae008a3539..64679d3d2c49 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -131,6 +131,7 @@ enum ice_tx_tstamp_work {
* @calibrating: if true, the PHY is calibrating the Tx offset. During this
* window, timestamps are temporarily disabled.
* @verify_cached: if true, verify new timestamp differs from last read value
+ * @last_ll_ts_idx_read: index of the last LL TS read by the FW
*/
struct ice_ptp_tx {
spinlock_t lock; /* lock protecting in_use bitmap */
@@ -143,6 +144,7 @@ struct ice_ptp_tx {
u8 init : 1;
u8 calibrating : 1;
u8 verify_cached : 1;
+ s8 last_ll_ts_idx_read;
};
/* Quad and port information for initializing timestamp blocks */
@@ -296,6 +298,8 @@ int ice_get_ptp_clock_index(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
+void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx);
+void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx);
enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
@@ -330,6 +334,11 @@ ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
return -1;
}
+static inline void ice_ptp_req_tx_single_tstamp(struct ice_ptp_tx *tx, u8 idx)
+{ }
+
+static inline void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx) { }
+
static inline bool ice_ptp_process_ts(struct ice_pf *pf)
{
return true;
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index 0cc285614c72..7e8fd369ef7c 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -509,6 +509,7 @@ int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
#define TS_LL_READ_RETRIES 200
#define TS_LL_READ_TS_HIGH GENMASK(23, 16)
#define TS_LL_READ_TS_IDX GENMASK(29, 24)
+#define TS_LL_READ_TS_INTR BIT(30)
#define TS_LL_READ_TS BIT(31)
/* Internal PHY timestamp address */
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index f8b658386552..b0f1f4db1d8b 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -350,6 +350,7 @@ struct ice_ts_func_info {
#define ICE_TS_TMR0_ENA_M BIT(25)
#define ICE_TS_TMR1_ENA_M BIT(26)
#define ICE_TS_LL_TX_TS_READ_M BIT(28)
+#define ICE_TS_LL_TX_TS_INT_READ_M BIT(29)
struct ice_ts_dev_info {
/* Device specific info */
@@ -363,6 +364,7 @@ struct ice_ts_dev_info {
u8 tmr0_ena;
u8 tmr1_ena;
u8 ts_ll_read;
+ u8 ts_ll_int_read;
};
/* Function specific capabilities */
--
2.43.0

310
kernel-std/debian/patches/ice-VDF/0014-ice-PTP-add-clock-domain-number-to-auxiliary-interfa.patch Normal file
View File

@ -0,0 +1,310 @@
From f267daca86600496d536f85c4d1945558b982427 Mon Sep 17 00:00:00 2001
From: Michal Michalik <michal.michalik@intel.com>
Date: Thu, 27 Jul 2023 15:50:36 +0200
Subject: [PATCH 14/36] ice: PTP: add clock domain number to auxiliary
interface
The PHC clock id used to be moved between PFs using FW admin queue
shared parameters - move the implementation to auxiliary bus.
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Michal Michalik <michal.michalik@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit fcd2c1e3139a27766ef263bd2011195dbc8a79f5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 5 -
drivers/net/ethernet/intel/ice/ice_ethtool.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 163 +++---------------
drivers/net/ethernet/intel/ice/ice_ptp.h | 11 +-
4 files changed, 34 insertions(+), 147 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 353ac55bdb9d..9bacb69ead8c 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2360,11 +2360,6 @@ struct ice_aqc_driver_shared_params {
};
enum ice_aqc_driver_params {
- /* OS clock index for PTP timer Domain 0 */
- ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0 = 0,
- /* OS clock index for PTP timer Domain 1 */
- ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1,
-
/* Add new parameters above */
ICE_AQC_DRIVER_PARAM_MAX = 16,
};
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 456cf4785c74..057453d589d5 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -3286,7 +3286,7 @@ ice_get_ts_info(struct net_device *dev, struct ethtool_ts_info *info)
SOF_TIMESTAMPING_RX_HARDWARE |
SOF_TIMESTAMPING_RAW_HARDWARE;
- info->phc_index = ice_get_ptp_clock_index(pf);
+ info->phc_index = ice_ptp_clock_index(pf);
info->tx_types = BIT(HWTSTAMP_TX_OFF) | BIT(HWTSTAMP_TX_ON);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 75038d826f71..a2d0da7dfe83 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -345,131 +345,6 @@ void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena)
ice_set_rx_tstamp(pf, ena);
}
-/**
- * ice_get_ptp_clock_index - Get the PTP clock index
- * @pf: the PF pointer
- *
- * Determine the clock index of the PTP clock associated with this device. If
- * this is the PF controlling the clock, just use the local access to the
- * clock device pointer.
- *
- * Otherwise, read from the driver shared parameters to determine the clock
- * index value.
- *
- * Returns: the index of the PTP clock associated with this device, or -1 if
- * there is no associated clock.
- */
-int ice_get_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- u32 value;
- int err;
-
- /* Use the ptp_clock structure if we're the main PF */
- if (pf->ptp.clock)
- return ptp_clock_index(pf->ptp.clock);
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- err = ice_aq_get_driver_param(hw, param_idx, &value, NULL);
- if (err) {
- dev_err(dev, "Failed to read PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- return -1;
- }
-
- /* The PTP clock index is an integer, and will be between 0 and
- * INT_MAX. The highest bit of the driver shared parameter is used to
- * indicate whether or not the currently stored clock index is valid.
- */
- if (!(value & PTP_SHARED_CLK_IDX_VALID))
- return -1;
-
- return value & ~PTP_SHARED_CLK_IDX_VALID;
-}
-
-/**
- * ice_set_ptp_clock_index - Set the PTP clock index
- * @pf: the PF pointer
- *
- * Set the PTP clock index for this device into the shared driver parameters,
- * so that other PFs associated with this device can read it.
- *
- * If the PF is unable to store the clock index, it will log an error, but
- * will continue operating PTP.
- */
-static void ice_set_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- u32 value;
- int err;
-
- if (!pf->ptp.clock)
- return;
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- value = (u32)ptp_clock_index(pf->ptp.clock);
- if (value > INT_MAX) {
- dev_err(dev, "PTP Clock index is too large to store\n");
- return;
- }
- value |= PTP_SHARED_CLK_IDX_VALID;
-
- err = ice_aq_set_driver_param(hw, param_idx, value, NULL);
- if (err) {
- dev_err(dev, "Failed to set PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- }
-}
-
-/**
- * ice_clear_ptp_clock_index - Clear the PTP clock index
- * @pf: the PF pointer
- *
- * Clear the PTP clock index for this device. Must be called when
- * unregistering the PTP clock, in order to ensure other PFs stop reporting
- * a clock object that no longer exists.
- */
-static void ice_clear_ptp_clock_index(struct ice_pf *pf)
-{
- struct device *dev = ice_pf_to_dev(pf);
- enum ice_aqc_driver_params param_idx;
- struct ice_hw *hw = &pf->hw;
- u8 tmr_idx;
- int err;
-
- /* Do not clear the index if we don't own the timer */
- if (!ice_pf_src_tmr_owned(pf))
- return;
-
- tmr_idx = hw->func_caps.ts_func_info.tmr_index_assoc;
- if (!tmr_idx)
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR0;
- else
- param_idx = ICE_AQC_DRIVER_PARAM_CLK_IDX_TMR1;
-
- err = ice_aq_set_driver_param(hw, param_idx, 0, NULL);
- if (err) {
- dev_dbg(dev, "Failed to clear PTP clock index parameter, err %d aq_err %s\n",
- err, ice_aq_str(hw->adminq.sq_last_status));
- }
-}
-
/**
* ice_ptp_read_src_clk_reg - Read the source clock register
* @pf: Board private structure
@@ -2564,7 +2439,6 @@ static void ice_ptp_set_caps(struct ice_pf *pf)
static long ice_ptp_create_clock(struct ice_pf *pf)
{
struct ptp_clock_info *info;
- struct ptp_clock *clock;
struct device *dev;
/* No need to create a clock device if we already have one */
@@ -2577,11 +2451,11 @@ static long ice_ptp_create_clock(struct ice_pf *pf)
dev = ice_pf_to_dev(pf);
/* Attempt to register the clock before enabling the hardware. */
- clock = ptp_clock_register(info, dev);
- if (IS_ERR(clock))
- return PTR_ERR(clock);
-
- pf->ptp.clock = clock;
+ pf->ptp.clock = ptp_clock_register(info, dev);
+ if (IS_ERR(pf->ptp.clock)) {
+ dev_err(ice_pf_to_dev(pf), "Failed to register PTP clock device");
+ return PTR_ERR(pf->ptp.clock);
+ }
return 0;
}
@@ -2996,6 +2870,28 @@ static void ice_ptp_unregister_auxbus_driver(struct ice_pf *pf)
mutex_destroy(&pf->ptp.ports_owner.lock);
}
+/**
+ * ice_ptp_clock_index - Get the PTP clock index for this device
+ * @pf: Board private structure
+ *
+ * Returns: the PTP clock index associated with this PF, or -1 if no PTP clock
+ * is associated.
+ */
+int ice_ptp_clock_index(struct ice_pf *pf)
+{
+ struct auxiliary_device *aux_dev;
+ struct ice_pf *owner_pf;
+ struct ptp_clock *clock;
+
+ aux_dev = &pf->ptp.port.aux_dev;
+ owner_pf = ice_ptp_aux_dev_to_owner_pf(aux_dev);
+ if (!owner_pf)
+ return -1;
+ clock = owner_pf->ptp.clock;
+
+ return clock ? ptp_clock_index(clock) : -1;
+}
+
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
@@ -3086,9 +2982,6 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
if (err)
goto err_clk;
- /* Store the PTP clock index for other PFs */
- ice_set_ptp_clock_index(pf);
-
err = ice_ptp_register_auxbus_driver(pf);
if (err) {
dev_err(ice_pf_to_dev(pf), "Failed to register PTP auxbus driver");
@@ -3097,7 +2990,6 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
return 0;
err_aux:
- ice_clear_ptp_clock_index(pf);
ptp_clock_unregister(pf->ptp.clock);
err_clk:
pf->ptp.clock = NULL;
@@ -3353,7 +3245,6 @@ void ice_ptp_release(struct ice_pf *pf)
/* Disable periodic outputs */
ice_ptp_disable_all_clkout(pf);
- ice_clear_ptp_clock_index(pf);
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 64679d3d2c49..95ebd7a048ec 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -290,11 +290,11 @@ struct ice_ptp {
#define ETH_GLTSYN_ENA(_i) (0x03000348 + ((_i) * 4))
#if IS_ENABLED(CONFIG_PTP_1588_CLOCK)
+int ice_ptp_clock_index(struct ice_pf *pf);
struct ice_pf;
int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr);
int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr);
void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena);
-int ice_get_ptp_clock_index(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
@@ -322,10 +322,6 @@ static inline int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
}
static inline void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena) { }
-static inline int ice_get_ptp_clock_index(struct ice_pf *pf)
-{
- return -1;
-}
static inline void ice_ptp_extts_event(struct ice_pf *pf) { }
static inline s8
@@ -353,5 +349,10 @@ static inline void ice_ptp_release(struct ice_pf *pf) { }
static inline void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
{
}
+
+static inline int ice_ptp_clock_index(struct ice_pf *pf)
+{
+ return -1;
+}
#endif /* IS_ENABLED(CONFIG_PTP_1588_CLOCK) */
#endif /* _ICE_PTP_H_ */
--
2.43.0

266
kernel-std/debian/patches/ice-VDF/0015-ice-restore-timestamp-configuration-after-device-res.patch Normal file
View File

@ -0,0 +1,266 @@
From eb63973adae478fdcc324f5490d6803646f0cc76 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 21 Nov 2023 13:12:57 -0800
Subject: [PATCH 15/36] ice: restore timestamp configuration after device reset
The driver calls ice_ptp_cfg_timestamp() during ice_ptp_prepare_for_reset()
to disable timestamping while the device is resetting. This operation
destroys the user requested configuration. While the driver does call
ice_ptp_cfg_timestamp in ice_rebuild() to restore some hardware settings
after a reset, it unconditionally passes true or false, resulting in
failure to restore previous user space configuration.
This results in a device reset forcibly disabling timestamp configuration
regardless of current user settings.
This was not detected previously due to a quirk of the LinuxPTP ptp4l
application. If ptp4l detects a missing timestamp, it enters a fault state
and performs recovery logic which includes executing SIOCSHWTSTAMP again,
restoring the now accidentally cleared configuration.
Not every application does this, and for these applications, timestamps
will mysteriously stop after a PF reset, without being restored until an
application restart.
Fix this by replacing ice_ptp_cfg_timestamp() with two new functions:
1) ice_ptp_disable_timestamp_mode() which unconditionally disables the
timestamping logic in ice_ptp_prepare_for_reset() and ice_ptp_release()
2) ice_ptp_restore_timestamp_mode() which calls
ice_ptp_restore_tx_interrupt() to restore Tx timestamping configuration,
calls ice_set_rx_tstamp() to restore Rx timestamping configuration, and
issues an immediate TSYN_TX interrupt to ensure that timestamps which
may have occurred during the device reset get processed.
Modify the ice_ptp_set_timestamp_mode to directly save the user
configuration and then call ice_ptp_restore_timestamp_mode. This way, reset
no longer destroys the saved user configuration.
This obsoletes the ice_set_tx_tstamp() function which can now be safely
removed.
With this change, all devices should now restore Tx and Rx timestamping
functionality correctly after a PF reset without application intervention.
Fixes: 77a781155a65 ("ice: enable receive hardware timestamping")
Fixes: ea9b847cda64 ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7758017911a4f2578d54c318e8fe77bcb5899054)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 12 +---
drivers/net/ethernet/intel/ice/ice_ptp.c | 74 ++++++++++++++---------
drivers/net/ethernet/intel/ice/ice_ptp.h | 5 +-
3 files changed, 51 insertions(+), 40 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 9163a72368b3..8cfb923198e9 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7545,15 +7545,6 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
goto err_vsi_rebuild;
}
- /* configure PTP timestamping after VSI rebuild */
- if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) {
- if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_SELF)
- ice_ptp_cfg_timestamp(pf, false);
- else if (pf->ptp.tx_interrupt_mode == ICE_PTP_TX_INTERRUPT_ALL)
- /* for E82x PHC owner always need to have interrupts */
- ice_ptp_cfg_timestamp(pf, true);
- }
-
err = ice_vsi_rebuild_by_type(pf, ICE_VSI_SWITCHDEV_CTRL);
if (err) {
dev_err(dev, "Switchdev CTRL VSI rebuild failed: %d\n", err);
@@ -7605,6 +7596,9 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
ice_plug_aux_dev(pf);
if (ice_is_feature_supported(pf, ICE_F_SRIOV_LAG))
ice_lag_rebuild(pf);
+
+ /* Restore timestamp mode settings after VSI rebuild */
+ ice_ptp_restore_timestamp_mode(pf);
return;
err_vsi_rebuild:
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index a2d0da7dfe83..8fc6905b0f79 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -294,18 +294,6 @@ static void ice_ptp_cfg_tx_interrupt(struct ice_pf *pf)
wr32(hw, PFINT_OICR_ENA, val);
}
-/**
- * ice_set_tx_tstamp - Enable or disable Tx timestamping
- * @pf: The PF pointer to search in
- * @on: bool value for whether timestamps are enabled or disabled
- */
-static void ice_set_tx_tstamp(struct ice_pf *pf, bool on)
-{
- pf->ptp.tstamp_config.tx_type = on ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
-
- ice_ptp_cfg_tx_interrupt(pf);
-}
-
/**
* ice_set_rx_tstamp - Enable or disable Rx timestamping
* @pf: The PF pointer to search in
@@ -317,7 +305,7 @@ static void ice_set_rx_tstamp(struct ice_pf *pf, bool on)
u16 i;
vsi = ice_get_main_vsi(pf);
- if (!vsi)
+ if (!vsi || !vsi->rx_rings)
return;
/* Set the timestamp flag for all the Rx rings */
@@ -326,23 +314,50 @@ static void ice_set_rx_tstamp(struct ice_pf *pf, bool on)
continue;
vsi->rx_rings[i]->ptp_rx = on;
}
+}
+
+/**
+ * ice_ptp_disable_timestamp_mode - Disable current timestamp mode
+ * @pf: Board private structure
+ *
+ * Called during preparation for reset to temporarily disable timestamping on
+ * the device. Called during remove to disable timestamping while cleaning up
+ * driver resources.
+ */
+static void ice_ptp_disable_timestamp_mode(struct ice_pf *pf)
+{
+ struct ice_hw *hw = &pf->hw;
+ u32 val;
+
+ val = rd32(hw, PFINT_OICR_ENA);
+ val &= ~PFINT_OICR_TSYN_TX_M;
+ wr32(hw, PFINT_OICR_ENA, val);
- pf->ptp.tstamp_config.rx_filter = on ? HWTSTAMP_FILTER_ALL :
- HWTSTAMP_FILTER_NONE;
+ ice_set_rx_tstamp(pf, false);
}
/**
- * ice_ptp_cfg_timestamp - Configure timestamp for init/deinit
+ * ice_ptp_restore_timestamp_mode - Restore timestamp configuration
* @pf: Board private structure
- * @ena: bool value to enable or disable time stamp
*
- * This function will configure timestamping during PTP initialization
- * and deinitialization
+ * Called at the end of rebuild to restore timestamp configuration after
+ * a device reset.
*/
-void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena)
+void ice_ptp_restore_timestamp_mode(struct ice_pf *pf)
{
- ice_set_tx_tstamp(pf, ena);
- ice_set_rx_tstamp(pf, ena);
+ struct ice_hw *hw = &pf->hw;
+ bool enable_rx;
+
+ ice_ptp_cfg_tx_interrupt(pf);
+
+ enable_rx = pf->ptp.tstamp_config.rx_filter == HWTSTAMP_FILTER_ALL;
+ ice_set_rx_tstamp(pf, enable_rx);
+
+ /* Trigger an immediate software interrupt to ensure that timestamps
+ * which occurred during reset are handled now.
+ */
+ wr32(hw, PFINT_OICR, PFINT_OICR_TSYN_TX_M);
+ ice_flush(hw);
}
/**
@@ -2152,10 +2167,10 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
{
switch (config->tx_type) {
case HWTSTAMP_TX_OFF:
- ice_set_tx_tstamp(pf, false);
+ pf->ptp.tstamp_config.tx_type = HWTSTAMP_TX_OFF;
break;
case HWTSTAMP_TX_ON:
- ice_set_tx_tstamp(pf, true);
+ pf->ptp.tstamp_config.tx_type = HWTSTAMP_TX_ON;
break;
default:
return -ERANGE;
@@ -2163,7 +2178,7 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
switch (config->rx_filter) {
case HWTSTAMP_FILTER_NONE:
- ice_set_rx_tstamp(pf, false);
+ pf->ptp.tstamp_config.rx_filter = HWTSTAMP_FILTER_NONE;
break;
case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
@@ -2179,12 +2194,15 @@ ice_ptp_set_timestamp_mode(struct ice_pf *pf, struct hwtstamp_config *config)
case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
case HWTSTAMP_FILTER_NTP_ALL:
case HWTSTAMP_FILTER_ALL:
- ice_set_rx_tstamp(pf, true);
+ pf->ptp.tstamp_config.rx_filter = HWTSTAMP_FILTER_ALL;
break;
default:
return -ERANGE;
}
+ /* Immediately update the device timestamping mode */
+ ice_ptp_restore_timestamp_mode(pf);
+
return 0;
}
@@ -2904,7 +2922,7 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
clear_bit(ICE_FLAG_PTP, pf->flags);
/* Disable timestamping for both Tx and Rx */
- ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_disable_timestamp_mode(pf);
kthread_cancel_delayed_work_sync(&ptp->work);
@@ -3222,7 +3240,7 @@ void ice_ptp_release(struct ice_pf *pf)
return;
/* Disable timestamping for both Tx and Rx */
- ice_ptp_cfg_timestamp(pf, false);
+ ice_ptp_disable_timestamp_mode(pf);
ice_ptp_remove_auxbus_device(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 95ebd7a048ec..130e6d2ae9a5 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -294,7 +294,7 @@ int ice_ptp_clock_index(struct ice_pf *pf);
struct ice_pf;
int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr);
int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr);
-void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena);
+void ice_ptp_restore_timestamp_mode(struct ice_pf *pf);
void ice_ptp_extts_event(struct ice_pf *pf);
s8 ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb);
@@ -321,8 +321,7 @@ static inline int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
return -EOPNOTSUPP;
}
-static inline void ice_ptp_cfg_timestamp(struct ice_pf *pf, bool ena) { }
-
+static inline void ice_ptp_restore_timestamp_mode(struct ice_pf *pf) { }
static inline void ice_ptp_extts_event(struct ice_pf *pf) { }
static inline s8
ice_ptp_request_ts(struct ice_ptp_tx *tx, struct sk_buff *skb)
--
2.43.0

316
kernel-std/debian/patches/ice-VDF/0016-ice-introduce-PTP-state-machine.patch Normal file
View File

@ -0,0 +1,316 @@
From 5c6115d27a377927d6392b3bfbe9739188c8153c Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:49 -0800
Subject: [PATCH 16/36] ice: introduce PTP state machine
Add PTP state machine so that the driver can correctly identify PTP
state around resets.
When the driver got information about ungraceful reset, PTP was not
prepared for reset and it returned error. When this situation occurs,
prepare PTP before rebuilding its structures.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Co-developed-by: Karol Kolacinski <karol.kolacinski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 8293e4cb2ff54b1ec4f7206dcb74c908f62a3fb8)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 -
drivers/net/ethernet/intel/ice/ice_ethtool.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 110 +++++++++++--------
drivers/net/ethernet/intel/ice/ice_ptp.h | 10 ++
4 files changed, 74 insertions(+), 49 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index ee42a504c2f4..3278d032a2bd 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -492,7 +492,6 @@ enum ice_pf_flags {
ICE_FLAG_DCB_ENA,
ICE_FLAG_FD_ENA,
ICE_FLAG_PTP_SUPPORTED, /* PTP is supported by NVM */
- ICE_FLAG_PTP, /* PTP is enabled by software */
ICE_FLAG_ADV_FEATURES,
ICE_FLAG_TC_MQPRIO, /* support for Multi queue TC */
ICE_FLAG_CLS_FLOWER,
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 057453d589d5..9e949c493c38 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -3276,7 +3276,7 @@ ice_get_ts_info(struct net_device *dev, struct ethtool_ts_info *info)
struct ice_pf *pf = ice_netdev_to_pf(dev);
/* only report timestamping if PTP is enabled */
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return ethtool_op_get_ts_info(dev, info);
info->so_timestamping = SOF_TIMESTAMPING_TX_SOFTWARE |
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 8fc6905b0f79..36c81c5ee83b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1430,7 +1430,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
struct ice_ptp_port *ptp_port;
struct ice_hw *hw = &pf->hw;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
if (WARN_ON_ONCE(port >= ICE_NUM_EXTERNAL_PORTS))
@@ -2148,7 +2148,7 @@ int ice_ptp_get_ts_config(struct ice_pf *pf, struct ifreq *ifr)
{
struct hwtstamp_config *config;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return -EIO;
config = &pf->ptp.tstamp_config;
@@ -2218,7 +2218,7 @@ int ice_ptp_set_ts_config(struct ice_pf *pf, struct ifreq *ifr)
struct hwtstamp_config config;
int err;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return -EAGAIN;
if (copy_from_user(&config, ifr->ifr_data, sizeof(config)))
@@ -2606,7 +2606,7 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
struct ice_pf *pf = container_of(ptp, struct ice_pf, ptp);
int err;
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
err = ice_ptp_update_cached_phctime(pf);
@@ -2618,6 +2618,42 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
msecs_to_jiffies(err ? 10 : 500));
}
+/**
+ * ice_ptp_prepare_for_reset - Prepare PTP for reset
+ * @pf: Board private structure
+ */
+void ice_ptp_prepare_for_reset(struct ice_pf *pf)
+{
+ struct ice_ptp *ptp = &pf->ptp;
+ u8 src_tmr;
+
+ if (ptp->state != ICE_PTP_READY)
+ return;
+
+ ptp->state = ICE_PTP_RESETTING;
+
+ /* Disable timestamping for both Tx and Rx */
+ ice_ptp_disable_timestamp_mode(pf);
+
+ kthread_cancel_delayed_work_sync(&ptp->work);
+
+ if (test_bit(ICE_PFR_REQ, pf->state))
+ return;
+
+ ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
+
+ /* Disable periodic outputs */
+ ice_ptp_disable_all_clkout(pf);
+
+ src_tmr = ice_get_ptp_src_clock_index(&pf->hw);
+
+ /* Disable source clock */
+ wr32(&pf->hw, GLTSYN_ENA(src_tmr), (u32)~GLTSYN_ENA_TSYN_ENA_M);
+
+ /* Acquire PHC and system timer to restore after reset */
+ ptp->reset_time = ktime_get_real_ns();
+}
+
/**
* ice_ptp_reset - Initialize PTP hardware clock support after reset
* @pf: Board private structure
@@ -2630,6 +2666,14 @@ void ice_ptp_reset(struct ice_pf *pf)
int err, itr = 1;
u64 time_diff;
+ if (ptp->state == ICE_PTP_READY) {
+ ice_ptp_prepare_for_reset(pf);
+ } else if (ptp->state != ICE_PTP_RESETTING) {
+ err = -EINVAL;
+ dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
+ goto err;
+ }
+
if (test_bit(ICE_PFR_REQ, pf->state) ||
!ice_pf_src_tmr_owned(pf))
goto pfr;
@@ -2690,7 +2734,7 @@ void ice_ptp_reset(struct ice_pf *pf)
if (err)
goto err;
- set_bit(ICE_FLAG_PTP, pf->flags);
+ ptp->state = ICE_PTP_READY;
/* Restart the PHY timestamping block */
if (!test_bit(ICE_PFR_REQ, pf->state) &&
@@ -2704,6 +2748,7 @@ void ice_ptp_reset(struct ice_pf *pf)
return;
err:
+ ptp->state = ICE_PTP_ERROR;
dev_err(ice_pf_to_dev(pf), "PTP reset failed %d\n", err);
}
@@ -2910,39 +2955,6 @@ int ice_ptp_clock_index(struct ice_pf *pf)
return clock ? ptp_clock_index(clock) : -1;
}
-/**
- * ice_ptp_prepare_for_reset - Prepare PTP for reset
- * @pf: Board private structure
- */
-void ice_ptp_prepare_for_reset(struct ice_pf *pf)
-{
- struct ice_ptp *ptp = &pf->ptp;
- u8 src_tmr;
-
- clear_bit(ICE_FLAG_PTP, pf->flags);
-
- /* Disable timestamping for both Tx and Rx */
- ice_ptp_disable_timestamp_mode(pf);
-
- kthread_cancel_delayed_work_sync(&ptp->work);
-
- if (test_bit(ICE_PFR_REQ, pf->state))
- return;
-
- ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
-
- /* Disable periodic outputs */
- ice_ptp_disable_all_clkout(pf);
-
- src_tmr = ice_get_ptp_src_clock_index(&pf->hw);
-
- /* Disable source clock */
- wr32(&pf->hw, GLTSYN_ENA(src_tmr), (u32)~GLTSYN_ENA_TSYN_ENA_M);
-
- /* Acquire PHC and system timer to restore after reset */
- ptp->reset_time = ktime_get_real_ns();
-}
-
/**
* ice_ptp_init_owner - Initialize PTP_1588_CLOCK device
* @pf: Board private structure
@@ -3181,6 +3193,8 @@ void ice_ptp_init(struct ice_pf *pf)
struct ice_hw *hw = &pf->hw;
int err;
+ ptp->state = ICE_PTP_INITIALIZING;
+
ice_ptp_init_phy_model(hw);
ice_ptp_init_tx_interrupt_mode(pf);
@@ -3205,12 +3219,13 @@ void ice_ptp_init(struct ice_pf *pf)
/* Configure initial Tx interrupt settings */
ice_ptp_cfg_tx_interrupt(pf);
- set_bit(ICE_FLAG_PTP, pf->flags);
- err = ice_ptp_init_work(pf, ptp);
+ err = ice_ptp_create_auxbus_device(pf);
if (err)
goto err;
- err = ice_ptp_create_auxbus_device(pf);
+ ptp->state = ICE_PTP_READY;
+
+ err = ice_ptp_init_work(pf, ptp);
if (err)
goto err;
@@ -3223,7 +3238,7 @@ void ice_ptp_init(struct ice_pf *pf)
ptp_clock_unregister(ptp->clock);
pf->ptp.clock = NULL;
}
- clear_bit(ICE_FLAG_PTP, pf->flags);
+ ptp->state = ICE_PTP_ERROR;
dev_err(ice_pf_to_dev(pf), "PTP failed %d\n", err);
}
@@ -3236,9 +3251,11 @@ void ice_ptp_init(struct ice_pf *pf)
*/
void ice_ptp_release(struct ice_pf *pf)
{
- if (!test_bit(ICE_FLAG_PTP, pf->flags))
+ if (pf->ptp.state != ICE_PTP_READY)
return;
+ pf->ptp.state = ICE_PTP_UNINIT;
+
/* Disable timestamping for both Tx and Rx */
ice_ptp_disable_timestamp_mode(pf);
@@ -3246,8 +3263,6 @@ void ice_ptp_release(struct ice_pf *pf)
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
- clear_bit(ICE_FLAG_PTP, pf->flags);
-
kthread_cancel_delayed_work_sync(&pf->ptp.work);
ice_ptp_port_phy_stop(&pf->ptp.port);
@@ -3257,6 +3272,9 @@ void ice_ptp_release(struct ice_pf *pf)
pf->ptp.kworker = NULL;
}
+ if (ice_pf_src_tmr_owned(pf))
+ ice_ptp_unregister_auxbus_driver(pf);
+
if (!pf->ptp.clock)
return;
@@ -3266,7 +3284,5 @@ void ice_ptp_release(struct ice_pf *pf)
ptp_clock_unregister(pf->ptp.clock);
pf->ptp.clock = NULL;
- ice_ptp_unregister_auxbus_driver(pf);
-
dev_info(ice_pf_to_dev(pf), "Removed PTP clock\n");
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 130e6d2ae9a5..e3cc69692405 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -203,8 +203,17 @@ struct ice_ptp_port_owner {
#define GLTSYN_TGT_H_IDX_MAX 4
+enum ice_ptp_state {
+ ICE_PTP_UNINIT = 0,
+ ICE_PTP_INITIALIZING,
+ ICE_PTP_READY,
+ ICE_PTP_RESETTING,
+ ICE_PTP_ERROR,
+};
+
/**
* struct ice_ptp - data used for integrating with CONFIG_PTP_1588_CLOCK
+ * @state: current state of PTP state machine
* @tx_interrupt_mode: the TX interrupt mode for the PTP clock
* @port: data for the PHY port initialization procedure
* @ports_owner: data for the auxiliary driver owner
@@ -227,6 +236,7 @@ struct ice_ptp_port_owner {
* @late_cached_phc_updates: number of times cached PHC update is late
*/
struct ice_ptp {
+ enum ice_ptp_state state;
enum ice_ptp_tx_interrupt tx_interrupt_mode;
struct ice_ptp_port port;
struct ice_ptp_port_owner ports_owner;
--
2.43.0

150
kernel-std/debian/patches/ice-VDF/0017-ice-pass-reset-type-to-PTP-reset-functions.patch Normal file
View File

@ -0,0 +1,150 @@
From 68d481b41ee5c177a1376fb82a98c09c148d982a Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:50 -0800
Subject: [PATCH 17/36] ice: pass reset type to PTP reset functions
The ice_ptp_prepare_for_reset() and ice_ptp_reset() functions currently
check the pf->flags ICE_FLAG_PFR_REQ bit to determine if the current
reset is a PF reset or not.
This is problematic, because it is possible that a PF reset and a higher
level reset (CORE reset, GLOBAL reset, EMP reset) are requested
simultaneously. In that case, the driver performs the highest level
reset requested. However, the ICE_FLAG_PFR_REQ flag will still be set.
The main driver reset functions take an enum ice_reset_req indicating
which reset is actually being performed. Pass this data into the PTP
functions and rely on this instead of relying on the driver flags.
This ensures that the PTP code performs the proper level of reset that
the driver is actually undergoing.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit c75d5e675a8542274fa0f7e52f3c4db1d4859a0c)
[Adjust the ice_ptp.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 4 ++--
drivers/net/ethernet/intel/ice/ice_ptp.c | 13 +++++++------
drivers/net/ethernet/intel/ice/ice_ptp.h | 17 +++++++++++++----
3 files changed, 22 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 8cfb923198e9..d5321410f2d7 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -615,7 +615,7 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
ice_pf_dis_all_vsi(pf, false);
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_prepare_for_reset(pf);
+ ice_ptp_prepare_for_reset(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_exit(pf);
@@ -7533,7 +7533,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
* fail.
*/
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_reset(pf);
+ ice_ptp_reset(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_init(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 36c81c5ee83b..20d1d22235d3 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2621,8 +2621,9 @@ static void ice_ptp_periodic_work(struct kthread_work *work)
/**
* ice_ptp_prepare_for_reset - Prepare PTP for reset
* @pf: Board private structure
+ * @reset_type: the reset type being performed
*/
-void ice_ptp_prepare_for_reset(struct ice_pf *pf)
+void ice_ptp_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
{
struct ice_ptp *ptp = &pf->ptp;
u8 src_tmr;
@@ -2637,7 +2638,7 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
kthread_cancel_delayed_work_sync(&ptp->work);
- if (test_bit(ICE_PFR_REQ, pf->state))
+ if (reset_type == ICE_RESET_PFR)
return;
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
@@ -2657,8 +2658,9 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf)
/**
* ice_ptp_reset - Initialize PTP hardware clock support after reset
* @pf: Board private structure
+ * @reset_type: the reset type being performed
*/
-void ice_ptp_reset(struct ice_pf *pf)
+void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
{
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
@@ -2667,15 +2669,14 @@ void ice_ptp_reset(struct ice_pf *pf)
u64 time_diff;
if (ptp->state == ICE_PTP_READY) {
- ice_ptp_prepare_for_reset(pf);
+ ice_ptp_prepare_for_reset(pf, reset_type);
} else if (ptp->state != ICE_PTP_RESETTING) {
err = -EINVAL;
dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
goto err;
}
- if (test_bit(ICE_PFR_REQ, pf->state) ||
- !ice_pf_src_tmr_owned(pf))
+ if (reset_type == ICE_RESET_PFR || !ice_pf_src_tmr_owned(pf))
goto pfr;
err = ice_ptp_init_phc(hw);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index e3cc69692405..cd74712a17a1 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -315,8 +315,9 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb);
-void ice_ptp_reset(struct ice_pf *pf);
-void ice_ptp_prepare_for_reset(struct ice_pf *pf);
+void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type);
+void ice_ptp_prepare_for_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type);
void ice_ptp_init(struct ice_pf *pf);
void ice_ptp_release(struct ice_pf *pf);
void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup);
@@ -351,8 +352,16 @@ static inline bool ice_ptp_process_ts(struct ice_pf *pf)
static inline void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb) { }
-static inline void ice_ptp_reset(struct ice_pf *pf) { }
-static inline void ice_ptp_prepare_for_reset(struct ice_pf *pf) { }
+
+static inline void ice_ptp_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
+{
+}
+
+static inline void ice_ptp_prepare_for_reset(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
+{
+}
static inline void ice_ptp_init(struct ice_pf *pf) { }
static inline void ice_ptp_release(struct ice_pf *pf) { }
static inline void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
--
2.43.0

119
kernel-std/debian/patches/ice-VDF/0018-ice-rename-verify_cached-to-has_ready_bitmap.patch Normal file
View File

@ -0,0 +1,119 @@
From 084497314e63f3d92178bc44500a27a277abc378 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:51 -0800
Subject: [PATCH 18/36] ice: rename verify_cached to has_ready_bitmap
The tx->verify_cached flag is used to inform the Tx timestamp tracking
code whether it needs to verify the cached Tx timestamp value against
a previous captured value. This is necessary on E810 hardware which does
not have a Tx timestamp ready bitmap.
In addition, we currently rely on the fact that the
ice_get_phy_tx_tstamp_ready() function returns all 1s for E810 hardware.
Instead of introducing a brand new flag, rename and verify_cached to
has_ready_bitmap, inverting the relevant checks.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 3f2216e8dbce04da5376ea7df410541f7b687cb0)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 12 ++++++------
drivers/net/ethernet/intel/ice/ice_ptp.h | 8 +++++---
2 files changed, 11 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 20d1d22235d3..a8c6b83579e6 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -606,11 +606,11 @@ void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
* timestamp. If it is not, skip this for now assuming it hasn't yet
* been captured by hardware.
*/
- if (!drop_ts && tx->verify_cached &&
+ if (!drop_ts && !tx->has_ready_bitmap &&
raw_tstamp == tx->tstamps[idx].cached_tstamp)
return;
- if (tx->verify_cached && raw_tstamp)
+ if (!tx->has_ready_bitmap && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
@@ -751,7 +751,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
* from the last cached timestamp. If it is not, skip this for
* now assuming it hasn't yet been captured by hardware.
*/
- if (!drop_ts && tx->verify_cached &&
+ if (!drop_ts && !tx->has_ready_bitmap &&
raw_tstamp == tx->tstamps[idx].cached_tstamp)
continue;
@@ -761,7 +761,7 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
skip_ts_read:
spin_lock_irqsave(&tx->lock, flags);
- if (tx->verify_cached && raw_tstamp)
+ if (!tx->has_ready_bitmap && raw_tstamp)
tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
@@ -1014,7 +1014,7 @@ ice_ptp_init_tx_e82x(struct ice_pf *pf, struct ice_ptp_tx *tx, u8 port)
tx->block = port / ICE_PORTS_PER_QUAD;
tx->offset = (port % ICE_PORTS_PER_QUAD) * INDEX_PER_PORT_E82X;
tx->len = INDEX_PER_PORT_E82X;
- tx->verify_cached = 0;
+ tx->has_ready_bitmap = 1;
return ice_ptp_alloc_tx_tracker(tx);
}
@@ -1037,7 +1037,7 @@ ice_ptp_init_tx_e810(struct ice_pf *pf, struct ice_ptp_tx *tx)
* verify new timestamps against cached copy of the last read
* timestamp.
*/
- tx->verify_cached = 1;
+ tx->has_ready_bitmap = 0;
return ice_ptp_alloc_tx_tracker(tx);
}
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index cd74712a17a1..1486a0b3b016 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -100,7 +100,7 @@ struct ice_perout_channel {
* the last timestamp we read for a given index. If the current timestamp
* value is the same as the cached value, we assume a new timestamp hasn't
* been captured. This avoids reporting stale timestamps to the stack. This is
- * only done if the verify_cached flag is set in ice_ptp_tx structure.
+ * only done if the has_ready_bitmap flag is not set in ice_ptp_tx structure.
*/
struct ice_tx_tstamp {
struct sk_buff *skb;
@@ -130,7 +130,9 @@ enum ice_tx_tstamp_work {
* @init: if true, the tracker is initialized;
* @calibrating: if true, the PHY is calibrating the Tx offset. During this
* window, timestamps are temporarily disabled.
- * @verify_cached: if true, verify new timestamp differs from last read value
+ * @has_ready_bitmap: if true, the hardware has a valid Tx timestamp ready
+ * bitmap register. If false, fall back to verifying new
+ * timestamp values against previously cached copy.
* @last_ll_ts_idx_read: index of the last LL TS read by the FW
*/
struct ice_ptp_tx {
@@ -143,7 +145,7 @@ struct ice_ptp_tx {
u8 len;
u8 init : 1;
u8 calibrating : 1;
- u8 verify_cached : 1;
+ u8 has_ready_bitmap : 1;
s8 last_ll_ts_idx_read;
};
--
2.43.0

77
kernel-std/debian/patches/ice-VDF/0019-ice-don-t-check-has_ready_bitmap-in-E810-functions.patch Normal file
View File

@ -0,0 +1,77 @@
From 375bced6b51243a8c8708204dd32960d076d5b83 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:52 -0800
Subject: [PATCH 19/36] ice: don't check has_ready_bitmap in E810 functions
E810 hardware does not have a Tx timestamp ready bitmap. Don't check
has_ready_bitmap in E810-specific functions.
Add has_ready_bitmap check in ice_ptp_process_tx_tstamp() to stop
relying on the fact that ice_get_phy_tx_tstamp_ready() returns all 1s.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit fea82915fca626eaa83f36d8a23194e8593ef4b4)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 23 +++++++++++------------
1 file changed, 11 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index a8c6b83579e6..ddc2dd0b2a28 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -601,17 +601,13 @@ void ice_ptp_complete_tx_single_tstamp(struct ice_ptp_tx *tx)
/* Read the low 32 bit value */
raw_tstamp |= (u64)rd32(&pf->hw, PF_SB_ATQBAH);
- /* For PHYs which don't implement a proper timestamp ready bitmap,
- * verify that the timestamp value is different from the last cached
- * timestamp. If it is not, skip this for now assuming it hasn't yet
- * been captured by hardware.
+ /* Devices using this interface always verify the timestamp differs
+ * relative to the last cached timestamp value.
*/
- if (!drop_ts && !tx->has_ready_bitmap &&
- raw_tstamp == tx->tstamps[idx].cached_tstamp)
+ if (raw_tstamp == tx->tstamps[idx].cached_tstamp)
return;
- if (!tx->has_ready_bitmap && raw_tstamp)
- tx->tstamps[idx].cached_tstamp = raw_tstamp;
+ tx->tstamps[idx].cached_tstamp = raw_tstamp;
clear_bit(idx, tx->in_use);
skb = tx->tstamps[idx].skb;
tx->tstamps[idx].skb = NULL;
@@ -701,9 +697,11 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
hw = &pf->hw;
/* Read the Tx ready status first */
- err = ice_get_phy_tx_tstamp_ready(hw, tx->block, &tstamp_ready);
- if (err)
- return;
+ if (tx->has_ready_bitmap) {
+ err = ice_get_phy_tx_tstamp_ready(hw, tx->block, &tstamp_ready);
+ if (err)
+ return;
+ }
/* Drop packets if the link went down */
link_up = ptp_port->link_up;
@@ -731,7 +729,8 @@ static void ice_ptp_process_tx_tstamp(struct ice_ptp_tx *tx)
* If we do not, the hardware logic for generating a new
* interrupt can get stuck on some devices.
*/
- if (!(tstamp_ready & BIT_ULL(phy_idx))) {
+ if (tx->has_ready_bitmap &&
+ !(tstamp_ready & BIT_ULL(phy_idx))) {
if (drop_ts)
goto skip_ts_read;
--
2.43.0

88
kernel-std/debian/patches/ice-VDF/0020-ice-rename-ice_ptp_tx_cfg_intr.patch Normal file
View File

@ -0,0 +1,88 @@
From a5318a3a04ed9535ab18ef0f0537b3d33862bee9 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:53 -0800
Subject: [PATCH 20/36] ice: rename ice_ptp_tx_cfg_intr
The ice_ptp_tx_cfg_intr() function sends a control queue message to
configure the PHY timestamp interrupt block. This is a very similar name
to a function which is used to configure the MAC Other Interrupt Cause
Enable register.
Rename this function to ice_ptp_cfg_phy_interrupt in order to make it
more obvious to the reader what action it performs, and distinguish it
from other similarly named functions.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 1abefdca85e8664374f53c7bc80d5f5f827ce711)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index ddc2dd0b2a28..c6e9d77fc59b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1455,14 +1455,14 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
}
/**
- * ice_ptp_tx_ena_intr - Enable or disable the Tx timestamp interrupt
+ * ice_ptp_cfg_phy_interrupt - Configure PHY interrupt settings
* @pf: PF private structure
* @ena: bool value to enable or disable interrupt
* @threshold: Minimum number of packets at which intr is triggered
*
* Utility function to enable or disable Tx timestamp interrupt and threshold
*/
-static int ice_ptp_tx_ena_intr(struct ice_pf *pf, bool ena, u32 threshold)
+static int ice_ptp_cfg_phy_interrupt(struct ice_pf *pf, bool ena, u32 threshold)
{
struct ice_hw *hw = &pf->hw;
int err = 0;
@@ -2664,8 +2664,8 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
struct timespec64 ts;
- int err, itr = 1;
u64 time_diff;
+ int err;
if (ptp->state == ICE_PTP_READY) {
ice_ptp_prepare_for_reset(pf, reset_type);
@@ -2716,7 +2716,7 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
- err = ice_ptp_tx_ena_intr(pf, true, itr);
+ err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
if (err)
goto err;
}
@@ -2967,7 +2967,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
{
struct ice_hw *hw = &pf->hw;
struct timespec64 ts;
- int err, itr = 1;
+ int err;
err = ice_ptp_init_phc(hw);
if (err) {
@@ -3002,7 +3002,7 @@ static int ice_ptp_init_owner(struct ice_pf *pf)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
- err = ice_ptp_tx_ena_intr(pf, true, itr);
+ err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
if (err)
goto err_exit;
}
--
2.43.0

191
kernel-std/debian/patches/ice-VDF/0021-ice-factor-out-ice_ptp_rebuild_owner.patch Normal file
View File

@ -0,0 +1,191 @@
From 9411c5b82a7196b9712488631fd14e67e2d919fa Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:54 -0800
Subject: [PATCH 21/36] ice: factor out ice_ptp_rebuild_owner()
The ice_ptp_reset() function uses a goto to skip past clock owner
operations if performing a PF reset or if the device is not the clock
owner. This is a bit confusing. Factor this out into
ice_ptp_rebuild_owner() instead.
The ice_ptp_reset() function is called by ice_rebuild() to restore PTP
functionality after a device reset. Follow the convention set by the
ice_main.c file and rename this function to ice_ptp_rebuild(), in the
same way that we have ice_prepare_for_reset() and
ice_ptp_prepare_for_reset().
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 803bef817807d2d36c930dada20c96fffae0dd19)
[Adjust ice_ptp.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_main.c | 2 +-
drivers/net/ethernet/intel/ice/ice_ptp.c | 62 ++++++++++++++---------
drivers/net/ethernet/intel/ice/ice_ptp.h | 6 +--
3 files changed, 42 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d5321410f2d7..a04dcc89c35d 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7533,7 +7533,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
* fail.
*/
if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags))
- ice_ptp_reset(pf, reset_type);
+ ice_ptp_rebuild(pf, reset_type);
if (ice_is_feature_supported(pf, ICE_F_GNSS))
ice_gnss_init(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index c6e9d77fc59b..780aa242c86b 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -2655,11 +2655,13 @@ void ice_ptp_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
}
/**
- * ice_ptp_reset - Initialize PTP hardware clock support after reset
+ * ice_ptp_rebuild_owner - Initialize PTP clock owner after reset
* @pf: Board private structure
- * @reset_type: the reset type being performed
+ *
+ * Companion function for ice_ptp_rebuild() which handles tasks that only the
+ * PTP clock owner instance should perform.
*/
-void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
+static int ice_ptp_rebuild_owner(struct ice_pf *pf)
{
struct ice_ptp *ptp = &pf->ptp;
struct ice_hw *hw = &pf->hw;
@@ -2667,32 +2669,21 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
u64 time_diff;
int err;
- if (ptp->state == ICE_PTP_READY) {
- ice_ptp_prepare_for_reset(pf, reset_type);
- } else if (ptp->state != ICE_PTP_RESETTING) {
- err = -EINVAL;
- dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
- goto err;
- }
-
- if (reset_type == ICE_RESET_PFR || !ice_pf_src_tmr_owned(pf))
- goto pfr;
-
err = ice_ptp_init_phc(hw);
if (err)
- goto err;
+ return err;
/* Acquire the global hardware lock */
if (!ice_ptp_lock(hw)) {
err = -EBUSY;
- goto err;
+ return err;
}
/* Write the increment time value to PHY and LAN */
err = ice_ptp_write_incval(hw, ice_base_incval(pf));
if (err) {
ice_ptp_unlock(hw);
- goto err;
+ return err;
}
/* Write the initial Time value to PHY and LAN using the cached PHC
@@ -2708,7 +2699,7 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
err = ice_ptp_write_init(pf, &ts);
if (err) {
ice_ptp_unlock(hw);
- goto err;
+ return err;
}
/* Release the global hardware lock */
@@ -2717,11 +2708,39 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
+ if (err)
+ return err;
+
+ ice_ptp_restart_all_phy(pf);
+ }
+
+ return 0;
+}
+
+/**
+ * ice_ptp_rebuild - Initialize PTP hardware clock support after reset
+ * @pf: Board private structure
+ * @reset_type: the reset type being performed
+ */
+void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
+{
+ struct ice_ptp *ptp = &pf->ptp;
+ int err;
+
+ if (ptp->state == ICE_PTP_READY) {
+ ice_ptp_prepare_for_reset(pf, reset_type);
+ } else if (ptp->state != ICE_PTP_RESETTING) {
+ err = -EINVAL;
+ dev_err(ice_pf_to_dev(pf), "PTP was not initialized\n");
+ goto err;
+ }
+
+ if (ice_pf_src_tmr_owned(pf) && reset_type != ICE_RESET_PFR) {
+ err = ice_ptp_rebuild_owner(pf);
if (err)
goto err;
}
-pfr:
/* Init Tx structures */
if (ice_is_e810(&pf->hw)) {
err = ice_ptp_init_tx_e810(pf, &ptp->port.tx);
@@ -2736,11 +2755,6 @@ void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
ptp->state = ICE_PTP_READY;
- /* Restart the PHY timestamping block */
- if (!test_bit(ICE_PFR_REQ, pf->state) &&
- ice_pf_src_tmr_owned(pf))
- ice_ptp_restart_all_phy(pf);
-
/* Start periodic work going */
kthread_queue_delayed_work(ptp->kworker, &ptp->work, 0);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 1486a0b3b016..352405a2daf2 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -317,7 +317,7 @@ enum ice_tx_tstamp_work ice_ptp_process_ts(struct ice_pf *pf);
void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb);
-void ice_ptp_reset(struct ice_pf *pf, enum ice_reset_req reset_type);
+void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type);
void ice_ptp_prepare_for_reset(struct ice_pf *pf,
enum ice_reset_req reset_type);
void ice_ptp_init(struct ice_pf *pf);
@@ -355,8 +355,8 @@ static inline void
ice_ptp_rx_hwtstamp(struct ice_rx_ring *rx_ring,
union ice_32b_rx_flex_desc *rx_desc, struct sk_buff *skb) { }
-static inline void ice_ptp_reset(struct ice_pf *pf,
- enum ice_reset_req reset_type)
+static inline void ice_ptp_rebuild(struct ice_pf *pf,
+ enum ice_reset_req reset_type)
{
}
--
2.43.0

93
kernel-std/debian/patches/ice-VDF/0022-ice-stop-destroying-and-reinitalizing-Tx-tracker-dur.patch Normal file
View File

@ -0,0 +1,93 @@
From 1c89a9e26f669bead5ebcac38fa98c20c517769c Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Thu, 25 Jan 2024 13:57:55 -0800
Subject: [PATCH 22/36] ice: stop destroying and reinitalizing Tx tracker
during reset
The ice driver currently attempts to destroy and re-initialize the Tx
timestamp tracker during the reset flow. The release of the Tx tracker
only happened during CORE reset or GLOBAL reset. The ice_ptp_rebuild()
function always calls the ice_ptp_init_tx function which will allocate
a new tracker data structure, resulting in memory leaks during PF reset.
Certainly the driver should not be allocating a new tracker without
removing the old tracker data, as this results in a memory leak.
Additionally, there's no reason to remove the tracker memory during a
reset. Remove this logic from the reset and rebuild flow. Instead of
releasing the Tx tracker, flush outstanding timestamps just before we
reset the PHY timestamp block in ice_ptp_cfg_phy_interrupt().
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
(cherry picked from commit 7a25fe5cd5fb2265065ac6765c53c0a1f1e874d3)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 33 +++++++++++++++---------
1 file changed, 21 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 780aa242c86b..48ec59fc5d87 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -963,6 +963,22 @@ ice_ptp_mark_tx_tracker_stale(struct ice_ptp_tx *tx)
spin_unlock_irqrestore(&tx->lock, flags);
}
+/**
+ * ice_ptp_flush_all_tx_tracker - Flush all timestamp trackers on this clock
+ * @pf: Board private structure
+ *
+ * Called by the clock owner to flush all the Tx timestamp trackers associated
+ * with the clock.
+ */
+static void
+ice_ptp_flush_all_tx_tracker(struct ice_pf *pf)
+{
+ struct ice_ptp_port *port;
+
+ list_for_each_entry(port, &pf->ptp.ports_owner.ports, list_member)
+ ice_ptp_flush_tx_tracker(ptp_port_to_pf(port), &port->tx);
+}
+
/**
* ice_ptp_release_tx_tracker - Release allocated memory for Tx tracker
* @pf: Board private structure
@@ -2705,6 +2721,11 @@ static int ice_ptp_rebuild_owner(struct ice_pf *pf)
/* Release the global hardware lock */
ice_ptp_unlock(hw);
+ /* Flush software tracking of any outstanding timestamps since we're
+ * about to flush the PHY timestamp block.
+ */
+ ice_ptp_flush_all_tx_tracker(pf);
+
if (!ice_is_e810(hw)) {
/* Enable quad interrupts */
err = ice_ptp_cfg_phy_interrupt(pf, true, 1);
@@ -2741,18 +2762,6 @@ void ice_ptp_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
goto err;
}
- /* Init Tx structures */
- if (ice_is_e810(&pf->hw)) {
- err = ice_ptp_init_tx_e810(pf, &ptp->port.tx);
- } else {
- kthread_init_delayed_work(&ptp->port.ov_work,
- ice_ptp_wait_for_offsets);
- err = ice_ptp_init_tx_e82x(pf, &ptp->port.tx,
- ptp->port.port_num);
- }
- if (err)
- goto err;
-
ptp->state = ICE_PTP_READY;
/* Start periodic work going */
--
2.43.0

457
kernel-std/debian/patches/ice-VDF/0023-ice-Remove-and-readd-netdev-during-devlink-reload.patch Normal file
View File

@ -0,0 +1,457 @@
From 6f1d1fa58f58ff3f6ce61ab502bd29227ca1bb3f Mon Sep 17 00:00:00 2001
From: Wojciech Drewek <wojciech.drewek@intel.com>
Date: Mon, 5 Feb 2024 14:03:56 +0100
Subject: [PATCH 23/36] ice: Remove and readd netdev during devlink reload
Recent changes to the devlink reload (commit 9b2348e2d6c9
("devlink: warn about existing entities during reload-reinit"))
force the drivers to destroy devlink ports during reinit.
Adjust ice driver to this requirement, unregister netdvice, destroy
devlink port. ice_init_eth() was removed and all the common code
between probe and reload was moved to ice_load().
During devlink reload we can't take devl_lock (it's already taken)
and in ice_probe() we have to lock it. Use devl_* variant of the API
which does not acquire and release devl_lock. Guard ice_load()
with devl_lock only in case of probe.
Suggested-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 41cc4e53934c30f1cf7745c257154e538c78a1f5)
[Adjust ice.h with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 2 +
drivers/net/ethernet/intel/ice/ice_devlink.c | 68 ++++++-
drivers/net/ethernet/intel/ice/ice_main.c | 186 ++++++-------------
3 files changed, 125 insertions(+), 131 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 3278d032a2bd..d3f72f9fbcd7 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -978,6 +978,8 @@ int ice_stop(struct net_device *netdev);
void ice_service_task_schedule(struct ice_pf *pf);
int ice_load(struct ice_pf *pf);
void ice_unload(struct ice_pf *pf);
+int ice_init_dev(struct ice_pf *pf);
+void ice_deinit_dev(struct ice_pf *pf);
/**
* ice_set_rdma_cap - enable RDMA support
diff --git a/drivers/net/ethernet/intel/ice/ice_devlink.c b/drivers/net/ethernet/intel/ice/ice_devlink.c
index 3a2261823d93..43007e3674c4 100644
--- a/drivers/net/ethernet/intel/ice/ice_devlink.c
+++ b/drivers/net/ethernet/intel/ice/ice_devlink.c
@@ -444,6 +444,20 @@ ice_devlink_reload_empr_start(struct ice_pf *pf,
return 0;
}
+/**
+ * ice_devlink_reinit_down - unload given PF
+ * @pf: pointer to the PF struct
+ */
+static void ice_devlink_reinit_down(struct ice_pf *pf)
+{
+ /* No need to take devl_lock, it's already taken by devlink API */
+ ice_unload(pf);
+ rtnl_lock();
+ ice_vsi_decfg(ice_get_main_vsi(pf));
+ rtnl_unlock();
+ ice_deinit_dev(pf);
+}
+
/**
* ice_devlink_reload_down - prepare for reload
* @devlink: pointer to the devlink instance to reload
@@ -477,7 +491,7 @@ ice_devlink_reload_down(struct devlink *devlink, bool netns_change,
"Remove all VFs before doing reinit\n");
return -EOPNOTSUPP;
}
- ice_unload(pf);
+ ice_devlink_reinit_down(pf);
return 0;
case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
return ice_devlink_reload_empr_start(pf, extack);
@@ -1240,6 +1254,45 @@ static int ice_devlink_set_parent(struct devlink_rate *devlink_rate,
return status;
}
+/**
+ * ice_devlink_reinit_up - do reinit of the given PF
+ * @pf: pointer to the PF struct
+ */
+static int ice_devlink_reinit_up(struct ice_pf *pf)
+{
+ struct ice_vsi *vsi = ice_get_main_vsi(pf);
+ struct ice_vsi_cfg_params params;
+ int err;
+
+ err = ice_init_dev(pf);
+ if (err)
+ return err;
+
+ params = ice_vsi_to_params(vsi);
+ params.flags = ICE_VSI_FLAG_INIT;
+
+ rtnl_lock();
+ err = ice_vsi_cfg(vsi, &params);
+ rtnl_unlock();
+ if (err)
+ goto err_vsi_cfg;
+
+ /* No need to take devl_lock, it's already taken by devlink API */
+ err = ice_load(pf);
+ if (err)
+ goto err_load;
+
+ return 0;
+
+err_load:
+ rtnl_lock();
+ ice_vsi_decfg(vsi);
+ rtnl_unlock();
+err_vsi_cfg:
+ ice_deinit_dev(pf);
+ return err;
+}
+
/**
* ice_devlink_reload_up - do reload up after reinit
* @devlink: pointer to the devlink instance reloading
@@ -1260,7 +1313,7 @@ ice_devlink_reload_up(struct devlink *devlink,
switch (action) {
case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
- return ice_load(pf);
+ return ice_devlink_reinit_up(pf);
case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
*actions_performed = BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE);
return ice_devlink_reload_empr_finish(pf, extack);
@@ -1540,6 +1593,7 @@ static const struct devlink_port_ops ice_devlink_port_ops = {
* @pf: the PF to create a devlink port for
*
* Create and register a devlink_port for this PF.
+ * This function has to be called under devl_lock.
*
* Return: zero on success or an error code on failure.
*/
@@ -1552,6 +1606,8 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
struct device *dev;
int err;
+ devlink = priv_to_devlink(pf);
+
dev = ice_pf_to_dev(pf);
devlink_port = &pf->devlink_port;
@@ -1572,10 +1628,9 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
ice_devlink_set_switch_id(pf, &attrs.switch_id);
devlink_port_attrs_set(devlink_port, &attrs);
- devlink = priv_to_devlink(pf);
- err = devlink_port_register_with_ops(devlink, devlink_port, vsi->idx,
- &ice_devlink_port_ops);
+ err = devl_port_register_with_ops(devlink, devlink_port, vsi->idx,
+ &ice_devlink_port_ops);
if (err) {
dev_err(dev, "Failed to create devlink port for PF %d, error %d\n",
pf->hw.pf_id, err);
@@ -1590,10 +1645,11 @@ int ice_devlink_create_pf_port(struct ice_pf *pf)
* @pf: the PF to cleanup
*
* Unregisters the devlink_port structure associated with this PF.
+ * This function has to be called under devl_lock.
*/
void ice_devlink_destroy_pf_port(struct ice_pf *pf)
{
- devlink_port_unregister(&pf->devlink_port);
+ devl_port_unregister(&pf->devlink_port);
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index a04dcc89c35d..d3340114297a 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -4588,90 +4588,6 @@ static void ice_decfg_netdev(struct ice_vsi *vsi)
vsi->netdev = NULL;
}
-static int ice_start_eth(struct ice_vsi *vsi)
-{
- int err;
-
- err = ice_init_mac_fltr(vsi->back);
- if (err)
- return err;
-
- err = ice_vsi_open(vsi);
- if (err)
- ice_fltr_remove_all(vsi);
-
- return err;
-}
-
-static void ice_stop_eth(struct ice_vsi *vsi)
-{
- ice_fltr_remove_all(vsi);
- ice_vsi_close(vsi);
-}
-
-static int ice_init_eth(struct ice_pf *pf)
-{
- struct ice_vsi *vsi = ice_get_main_vsi(pf);
- int err;
-
- if (!vsi)
- return -EINVAL;
-
- /* init channel list */
- INIT_LIST_HEAD(&vsi->ch_list);
-
- err = ice_cfg_netdev(vsi);
- if (err)
- return err;
- /* Setup DCB netlink interface */
- ice_dcbnl_setup(vsi);
-
- err = ice_init_mac_fltr(pf);
- if (err)
- goto err_init_mac_fltr;
-
- err = ice_devlink_create_pf_port(pf);
- if (err)
- goto err_devlink_create_pf_port;
-
- SET_NETDEV_DEVLINK_PORT(vsi->netdev, &pf->devlink_port);
-
- err = ice_register_netdev(vsi);
- if (err)
- goto err_register_netdev;
-
- err = ice_tc_indir_block_register(vsi);
- if (err)
- goto err_tc_indir_block_register;
-
- ice_napi_add(vsi);
-
- return 0;
-
-err_tc_indir_block_register:
- ice_unregister_netdev(vsi);
-err_register_netdev:
- ice_devlink_destroy_pf_port(pf);
-err_devlink_create_pf_port:
-err_init_mac_fltr:
- ice_decfg_netdev(vsi);
- return err;
-}
-
-static void ice_deinit_eth(struct ice_pf *pf)
-{
- struct ice_vsi *vsi = ice_get_main_vsi(pf);
-
- if (!vsi)
- return;
-
- ice_vsi_close(vsi);
- ice_unregister_netdev(vsi);
- ice_devlink_destroy_pf_port(pf);
- ice_tc_indir_block_unregister(vsi);
- ice_decfg_netdev(vsi);
-}
-
/**
* ice_wait_for_fw - wait for full FW readiness
* @hw: pointer to the hardware structure
@@ -4697,7 +4613,7 @@ static int ice_wait_for_fw(struct ice_hw *hw, u32 timeout)
return -ETIMEDOUT;
}
-static int ice_init_dev(struct ice_pf *pf)
+int ice_init_dev(struct ice_pf *pf)
{
struct device *dev = ice_pf_to_dev(pf);
struct ice_hw *hw = &pf->hw;
@@ -4790,7 +4706,7 @@ static int ice_init_dev(struct ice_pf *pf)
return err;
}
-static void ice_deinit_dev(struct ice_pf *pf)
+void ice_deinit_dev(struct ice_pf *pf)
{
ice_free_irq_msix_misc(pf);
ice_deinit_pf(pf);
@@ -5091,31 +5007,47 @@ static void ice_deinit(struct ice_pf *pf)
/**
* ice_load - load pf by init hw and starting VSI
* @pf: pointer to the pf instance
+ *
+ * This function has to be called under devl_lock.
*/
int ice_load(struct ice_pf *pf)
{
- struct ice_vsi_cfg_params params = {};
struct ice_vsi *vsi;
int err;
- err = ice_init_dev(pf);
+ devl_assert_locked(priv_to_devlink(pf));
+
+ vsi = ice_get_main_vsi(pf);
+
+ /* init channel list */
+ INIT_LIST_HEAD(&vsi->ch_list);
+
+ err = ice_cfg_netdev(vsi);
if (err)
return err;
- vsi = ice_get_main_vsi(pf);
+ /* Setup DCB netlink interface */
+ ice_dcbnl_setup(vsi);
- params = ice_vsi_to_params(vsi);
- params.flags = ICE_VSI_FLAG_INIT;
+ err = ice_init_mac_fltr(pf);
+ if (err)
+ goto err_init_mac_fltr;
- rtnl_lock();
- err = ice_vsi_cfg(vsi, &params);
+ err = ice_devlink_create_pf_port(pf);
if (err)
- goto err_vsi_cfg;
+ goto err_devlink_create_pf_port;
- err = ice_start_eth(ice_get_main_vsi(pf));
+ SET_NETDEV_DEVLINK_PORT(vsi->netdev, &pf->devlink_port);
+
+ err = ice_register_netdev(vsi);
+ if (err)
+ goto err_register_netdev;
+
+ err = ice_tc_indir_block_register(vsi);
if (err)
- goto err_start_eth;
- rtnl_unlock();
+ goto err_tc_indir_block_register;
+
+ ice_napi_add(vsi);
err = ice_init_rdma(pf);
if (err)
@@ -5129,29 +5061,35 @@ int ice_load(struct ice_pf *pf)
return 0;
err_init_rdma:
- ice_vsi_close(ice_get_main_vsi(pf));
- rtnl_lock();
-err_start_eth:
- ice_vsi_decfg(ice_get_main_vsi(pf));
-err_vsi_cfg:
- rtnl_unlock();
- ice_deinit_dev(pf);
+ ice_tc_indir_block_unregister(vsi);
+err_tc_indir_block_register:
+ ice_unregister_netdev(vsi);
+err_register_netdev:
+ ice_devlink_destroy_pf_port(pf);
+err_devlink_create_pf_port:
+err_init_mac_fltr:
+ ice_decfg_netdev(vsi);
return err;
}
/**
* ice_unload - unload pf by stopping VSI and deinit hw
* @pf: pointer to the pf instance
+ *
+ * This function has to be called under devl_lock.
*/
void ice_unload(struct ice_pf *pf)
{
+ struct ice_vsi *vsi = ice_get_main_vsi(pf);
+
+ devl_assert_locked(priv_to_devlink(pf));
+
ice_deinit_features(pf);
ice_deinit_rdma(pf);
- rtnl_lock();
- ice_stop_eth(ice_get_main_vsi(pf));
- ice_vsi_decfg(ice_get_main_vsi(pf));
- rtnl_unlock();
- ice_deinit_dev(pf);
+ ice_tc_indir_block_unregister(vsi);
+ ice_unregister_netdev(vsi);
+ ice_devlink_destroy_pf_port(pf);
+ ice_decfg_netdev(vsi);
}
/**
@@ -5249,27 +5187,23 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent)
if (err)
goto err_init;
- err = ice_init_eth(pf);
+ devl_lock(priv_to_devlink(pf));
+ err = ice_load(pf);
+ devl_unlock(priv_to_devlink(pf));
if (err)
- goto err_init_eth;
-
- err = ice_init_rdma(pf);
- if (err)
- goto err_init_rdma;
+ goto err_load;
err = ice_init_devlink(pf);
if (err)
goto err_init_devlink;
- ice_init_features(pf);
-
return 0;
err_init_devlink:
- ice_deinit_rdma(pf);
-err_init_rdma:
- ice_deinit_eth(pf);
-err_init_eth:
+ devl_lock(priv_to_devlink(pf));
+ ice_unload(pf);
+ devl_unlock(priv_to_devlink(pf));
+err_load:
ice_deinit(pf);
err_init:
pci_disable_device(pdev);
@@ -5363,12 +5297,14 @@ static void ice_remove(struct pci_dev *pdev)
if (!ice_is_safe_mode(pf))
ice_remove_arfs(pf);
- ice_deinit_features(pf);
+
ice_deinit_devlink(pf);
- ice_deinit_rdma(pf);
- ice_deinit_eth(pf);
- ice_deinit(pf);
+ devl_lock(priv_to_devlink(pf));
+ ice_unload(pf);
+ devl_unlock(priv_to_devlink(pf));
+
+ ice_deinit(pf);
ice_vsi_release_all(pf);
ice_setup_mc_magic_wake(pf);
--
2.43.0

434
kernel-std/debian/patches/ice-VDF/0024-ice-remove-FW-logging-code.patch Normal file
View File

@ -0,0 +1,434 @@
From e110839c4d9bfa4c885877a69573f48c008d3edd Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:11 -0800
Subject: [PATCH 24/36] ice: remove FW logging code
The FW logging code doesn't work because there is no way to set
cq_ena or uart_ena so remove the code. This code is the original
(v1) way of FW logging so it should be replaced with the v2 way.
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1953fc720e603721764f31daae216a2851664167)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 78 -------
drivers/net/ethernet/intel/ice/ice_common.c | 217 ------------------
drivers/net/ethernet/intel/ice/ice_common.h | 1 -
drivers/net/ethernet/intel/ice/ice_main.c | 3 -
drivers/net/ethernet/intel/ice/ice_type.h | 20 --
5 files changed, 319 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 9bacb69ead8c..3b289e6a225b 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2032,78 +2032,6 @@ struct ice_aqc_add_rdma_qset_data {
struct ice_aqc_add_tx_rdma_qset_entry rdma_qsets[];
};
-/* Configure Firmware Logging Command (indirect 0xFF09)
- * Logging Information Read Response (indirect 0xFF10)
- * Note: The 0xFF10 command has no input parameters.
- */
-struct ice_aqc_fw_logging {
- u8 log_ctrl;
-#define ICE_AQC_FW_LOG_AQ_EN BIT(0)
-#define ICE_AQC_FW_LOG_UART_EN BIT(1)
- u8 rsvd0;
- u8 log_ctrl_valid; /* Not used by 0xFF10 Response */
-#define ICE_AQC_FW_LOG_AQ_VALID BIT(0)
-#define ICE_AQC_FW_LOG_UART_VALID BIT(1)
- u8 rsvd1[5];
- __le32 addr_high;
- __le32 addr_low;
-};
-
-enum ice_aqc_fw_logging_mod {
- ICE_AQC_FW_LOG_ID_GENERAL = 0,
- ICE_AQC_FW_LOG_ID_CTRL,
- ICE_AQC_FW_LOG_ID_LINK,
- ICE_AQC_FW_LOG_ID_LINK_TOPO,
- ICE_AQC_FW_LOG_ID_DNL,
- ICE_AQC_FW_LOG_ID_I2C,
- ICE_AQC_FW_LOG_ID_SDP,
- ICE_AQC_FW_LOG_ID_MDIO,
- ICE_AQC_FW_LOG_ID_ADMINQ,
- ICE_AQC_FW_LOG_ID_HDMA,
- ICE_AQC_FW_LOG_ID_LLDP,
- ICE_AQC_FW_LOG_ID_DCBX,
- ICE_AQC_FW_LOG_ID_DCB,
- ICE_AQC_FW_LOG_ID_NETPROXY,
- ICE_AQC_FW_LOG_ID_NVM,
- ICE_AQC_FW_LOG_ID_AUTH,
- ICE_AQC_FW_LOG_ID_VPD,
- ICE_AQC_FW_LOG_ID_IOSF,
- ICE_AQC_FW_LOG_ID_PARSER,
- ICE_AQC_FW_LOG_ID_SW,
- ICE_AQC_FW_LOG_ID_SCHEDULER,
- ICE_AQC_FW_LOG_ID_TXQ,
- ICE_AQC_FW_LOG_ID_RSVD,
- ICE_AQC_FW_LOG_ID_POST,
- ICE_AQC_FW_LOG_ID_WATCHDOG,
- ICE_AQC_FW_LOG_ID_TASK_DISPATCH,
- ICE_AQC_FW_LOG_ID_MNG,
- ICE_AQC_FW_LOG_ID_MAX,
-};
-
-/* Defines for both above FW logging command/response buffers */
-#define ICE_AQC_FW_LOG_ID_S 0
-#define ICE_AQC_FW_LOG_ID_M (0xFFF << ICE_AQC_FW_LOG_ID_S)
-
-#define ICE_AQC_FW_LOG_CONF_SUCCESS 0 /* Used by response */
-#define ICE_AQC_FW_LOG_CONF_BAD_INDX BIT(12) /* Used by response */
-
-#define ICE_AQC_FW_LOG_EN_S 12
-#define ICE_AQC_FW_LOG_EN_M (0xF << ICE_AQC_FW_LOG_EN_S)
-#define ICE_AQC_FW_LOG_INFO_EN BIT(12) /* Used by command */
-#define ICE_AQC_FW_LOG_INIT_EN BIT(13) /* Used by command */
-#define ICE_AQC_FW_LOG_FLOW_EN BIT(14) /* Used by command */
-#define ICE_AQC_FW_LOG_ERR_EN BIT(15) /* Used by command */
-
-/* Get/Clear FW Log (indirect 0xFF11) */
-struct ice_aqc_get_clear_fw_log {
- u8 flags;
-#define ICE_AQC_FW_LOG_CLEAR BIT(0)
-#define ICE_AQC_FW_LOG_MORE_DATA_AVAIL BIT(1)
- u8 rsvd1[7];
- __le32 addr_high;
- __le32 addr_low;
-};
-
/* Download Package (indirect 0x0C40) */
/* Also used for Update Package (indirect 0x0C41 and 0x0C42) */
struct ice_aqc_download_pkg {
@@ -2448,8 +2376,6 @@ struct ice_aq_desc {
struct ice_aqc_add_rdma_qset add_rdma_qset;
struct ice_aqc_add_get_update_free_vsi vsi_cmd;
struct ice_aqc_add_update_free_vsi_resp add_update_free_vsi_res;
- struct ice_aqc_fw_logging fw_logging;
- struct ice_aqc_get_clear_fw_log get_clear_fw_log;
struct ice_aqc_download_pkg download_pkg;
struct ice_aqc_set_cgu_input_config set_cgu_input_config;
struct ice_aqc_get_cgu_input_config get_cgu_input_config;
@@ -2657,10 +2583,6 @@ enum ice_adminq_opc {
/* Standalone Commands/Events */
ice_aqc_opc_event_lan_overflow = 0x1001,
-
- /* debug commands */
- ice_aqc_opc_fw_logging = 0xFF09,
- ice_aqc_opc_fw_logging_info = 0xFF10,
};
#endif /* _ICE_ADMINQ_CMD_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index acf6ac00f804..a5c4b7ad6a20 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -897,216 +897,6 @@ static void ice_cleanup_fltr_mgmt_struct(struct ice_hw *hw)
devm_kfree(ice_hw_to_dev(hw), sw);
}
-/**
- * ice_get_fw_log_cfg - get FW logging configuration
- * @hw: pointer to the HW struct
- */
-static int ice_get_fw_log_cfg(struct ice_hw *hw)
-{
- struct ice_aq_desc desc;
- __le16 *config;
- int status;
- u16 size;
-
- size = sizeof(*config) * ICE_AQC_FW_LOG_ID_MAX;
- config = kzalloc(size, GFP_KERNEL);
- if (!config)
- return -ENOMEM;
-
- ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logging_info);
-
- status = ice_aq_send_cmd(hw, &desc, config, size, NULL);
- if (!status) {
- u16 i;
-
- /* Save FW logging information into the HW structure */
- for (i = 0; i < ICE_AQC_FW_LOG_ID_MAX; i++) {
- u16 v, m, flgs;
-
- v = le16_to_cpu(config[i]);
- m = (v & ICE_AQC_FW_LOG_ID_M) >> ICE_AQC_FW_LOG_ID_S;
- flgs = (v & ICE_AQC_FW_LOG_EN_M) >> ICE_AQC_FW_LOG_EN_S;
-
- if (m < ICE_AQC_FW_LOG_ID_MAX)
- hw->fw_log.evnts[m].cur = flgs;
- }
- }
-
- kfree(config);
-
- return status;
-}
-
-/**
- * ice_cfg_fw_log - configure FW logging
- * @hw: pointer to the HW struct
- * @enable: enable certain FW logging events if true, disable all if false
- *
- * This function enables/disables the FW logging via Rx CQ events and a UART
- * port based on predetermined configurations. FW logging via the Rx CQ can be
- * enabled/disabled for individual PF's. However, FW logging via the UART can
- * only be enabled/disabled for all PFs on the same device.
- *
- * To enable overall FW logging, the "cq_en" and "uart_en" enable bits in
- * hw->fw_log need to be set accordingly, e.g. based on user-provided input,
- * before initializing the device.
- *
- * When re/configuring FW logging, callers need to update the "cfg" elements of
- * the hw->fw_log.evnts array with the desired logging event configurations for
- * modules of interest. When disabling FW logging completely, the callers can
- * just pass false in the "enable" parameter. On completion, the function will
- * update the "cur" element of the hw->fw_log.evnts array with the resulting
- * logging event configurations of the modules that are being re/configured. FW
- * logging modules that are not part of a reconfiguration operation retain their
- * previous states.
- *
- * Before resetting the device, it is recommended that the driver disables FW
- * logging before shutting down the control queue. When disabling FW logging
- * ("enable" = false), the latest configurations of FW logging events stored in
- * hw->fw_log.evnts[] are not overridden to allow them to be reconfigured after
- * a device reset.
- *
- * When enabling FW logging to emit log messages via the Rx CQ during the
- * device's initialization phase, a mechanism alternative to interrupt handlers
- * needs to be used to extract FW log messages from the Rx CQ periodically and
- * to prevent the Rx CQ from being full and stalling other types of control
- * messages from FW to SW. Interrupts are typically disabled during the device's
- * initialization phase.
- */
-static int ice_cfg_fw_log(struct ice_hw *hw, bool enable)
-{
- struct ice_aqc_fw_logging *cmd;
- u16 i, chgs = 0, len = 0;
- struct ice_aq_desc desc;
- __le16 *data = NULL;
- u8 actv_evnts = 0;
- void *buf = NULL;
- int status = 0;
-
- if (!hw->fw_log.cq_en && !hw->fw_log.uart_en)
- return 0;
-
- /* Disable FW logging only when the control queue is still responsive */
- if (!enable &&
- (!hw->fw_log.actv_evnts || !ice_check_sq_alive(hw, &hw->adminq)))
- return 0;
-
- /* Get current FW log settings */
- status = ice_get_fw_log_cfg(hw);
- if (status)
- return status;
-
- ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logging);
- cmd = &desc.params.fw_logging;
-
- /* Indicate which controls are valid */
- if (hw->fw_log.cq_en)
- cmd->log_ctrl_valid |= ICE_AQC_FW_LOG_AQ_VALID;
-
- if (hw->fw_log.uart_en)
- cmd->log_ctrl_valid |= ICE_AQC_FW_LOG_UART_VALID;
-
- if (enable) {
- /* Fill in an array of entries with FW logging modules and
- * logging events being reconfigured.
- */
- for (i = 0; i < ICE_AQC_FW_LOG_ID_MAX; i++) {
- u16 val;
-
- /* Keep track of enabled event types */
- actv_evnts |= hw->fw_log.evnts[i].cfg;
-
- if (hw->fw_log.evnts[i].cfg == hw->fw_log.evnts[i].cur)
- continue;
-
- if (!data) {
- data = devm_kcalloc(ice_hw_to_dev(hw),
- ICE_AQC_FW_LOG_ID_MAX,
- sizeof(*data),
- GFP_KERNEL);
- if (!data)
- return -ENOMEM;
- }
-
- val = i << ICE_AQC_FW_LOG_ID_S;
- val |= hw->fw_log.evnts[i].cfg << ICE_AQC_FW_LOG_EN_S;
- data[chgs++] = cpu_to_le16(val);
- }
-
- /* Only enable FW logging if at least one module is specified.
- * If FW logging is currently enabled but all modules are not
- * enabled to emit log messages, disable FW logging altogether.
- */
- if (actv_evnts) {
- /* Leave if there is effectively no change */
- if (!chgs)
- goto out;
-
- if (hw->fw_log.cq_en)
- cmd->log_ctrl |= ICE_AQC_FW_LOG_AQ_EN;
-
- if (hw->fw_log.uart_en)
- cmd->log_ctrl |= ICE_AQC_FW_LOG_UART_EN;
-
- buf = data;
- len = sizeof(*data) * chgs;
- desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD);
- }
- }
-
- status = ice_aq_send_cmd(hw, &desc, buf, len, NULL);
- if (!status) {
- /* Update the current configuration to reflect events enabled.
- * hw->fw_log.cq_en and hw->fw_log.uart_en indicate if the FW
- * logging mode is enabled for the device. They do not reflect
- * actual modules being enabled to emit log messages. So, their
- * values remain unchanged even when all modules are disabled.
- */
- u16 cnt = enable ? chgs : (u16)ICE_AQC_FW_LOG_ID_MAX;
-
- hw->fw_log.actv_evnts = actv_evnts;
- for (i = 0; i < cnt; i++) {
- u16 v, m;
-
- if (!enable) {
- /* When disabling all FW logging events as part
- * of device's de-initialization, the original
- * configurations are retained, and can be used
- * to reconfigure FW logging later if the device
- * is re-initialized.
- */
- hw->fw_log.evnts[i].cur = 0;
- continue;
- }
-
- v = le16_to_cpu(data[i]);
- m = (v & ICE_AQC_FW_LOG_ID_M) >> ICE_AQC_FW_LOG_ID_S;
- hw->fw_log.evnts[m].cur = hw->fw_log.evnts[m].cfg;
- }
- }
-
-out:
- devm_kfree(ice_hw_to_dev(hw), data);
-
- return status;
-}
-
-/**
- * ice_output_fw_log
- * @hw: pointer to the HW struct
- * @desc: pointer to the AQ message descriptor
- * @buf: pointer to the buffer accompanying the AQ message
- *
- * Formats a FW Log message and outputs it via the standard driver logs.
- */
-void ice_output_fw_log(struct ice_hw *hw, struct ice_aq_desc *desc, void *buf)
-{
- ice_debug(hw, ICE_DBG_FW_LOG, "[ FW Log Msg Start ]\n");
- ice_debug_array(hw, ICE_DBG_FW_LOG, 16, 1, (u8 *)buf,
- le16_to_cpu(desc->datalen));
- ice_debug(hw, ICE_DBG_FW_LOG, "[ FW Log Msg End ]\n");
-}
-
/**
* ice_get_itr_intrl_gran
* @hw: pointer to the HW struct
@@ -1164,11 +954,6 @@ int ice_init_hw(struct ice_hw *hw)
if (status)
goto err_unroll_cqinit;
- /* Enable FW logging. Not fatal if this fails. */
- status = ice_cfg_fw_log(hw, true);
- if (status)
- ice_debug(hw, ICE_DBG_INIT, "Failed to enable FW logging.\n");
-
status = ice_clear_pf_cfg(hw);
if (status)
goto err_unroll_cqinit;
@@ -1318,8 +1103,6 @@ void ice_deinit_hw(struct ice_hw *hw)
ice_free_hw_tbls(hw);
mutex_destroy(&hw->tnl_lock);
- /* Attempt to disable FW logging before shutting down control queues */
- ice_cfg_fw_log(hw, false);
ice_destroy_all_ctrlq(hw);
/* Clear VSI contexts if not already cleared */
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index 7a966a0c224f..d47e5400351f 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -199,7 +199,6 @@ ice_aq_cfg_lan_txq(struct ice_hw *hw, struct ice_aqc_cfg_txqs_buf *buf,
struct ice_sq_cd *cd);
int ice_replay_vsi(struct ice_hw *hw, u16 vsi_handle);
void ice_replay_post(struct ice_hw *hw);
-void ice_output_fw_log(struct ice_hw *hw, struct ice_aq_desc *desc, void *buf);
struct ice_q_ctx *
ice_get_lan_q_ctx(struct ice_hw *hw, u16 vsi_handle, u8 tc, u16 q_handle);
int ice_sbq_rw_reg(struct ice_hw *hw, struct ice_sbq_msg_input *in);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d3340114297a..e5cc9790969c 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -1535,9 +1535,6 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
ice_vc_process_vf_msg(pf, &event, &data);
break;
- case ice_aqc_opc_fw_logging:
- ice_output_fw_log(hw, &event.desc, event.msg_buf);
- break;
case ice_aqc_opc_lldp_set_mib_change:
ice_dcb_process_lldp_set_mib_change(pf, &event);
break;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index b0f1f4db1d8b..6e1fed0d7384 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -731,24 +731,6 @@ struct ice_switch_info {
DECLARE_BITMAP(prof_res_bm[ICE_MAX_NUM_PROFILES], ICE_MAX_FV_WORDS);
};
-/* FW logging configuration */
-struct ice_fw_log_evnt {
- u8 cfg : 4; /* New event enables to configure */
- u8 cur : 4; /* Current/active event enables */
-};
-
-struct ice_fw_log_cfg {
- u8 cq_en : 1; /* FW logging is enabled via the control queue */
- u8 uart_en : 1; /* FW logging is enabled via UART for all PFs */
- u8 actv_evnts; /* Cumulation of currently enabled log events */
-
-#define ICE_FW_LOG_EVNT_INFO (ICE_AQC_FW_LOG_INFO_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_INIT (ICE_AQC_FW_LOG_INIT_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_FLOW (ICE_AQC_FW_LOG_FLOW_EN >> ICE_AQC_FW_LOG_EN_S)
-#define ICE_FW_LOG_EVNT_ERR (ICE_AQC_FW_LOG_ERR_EN >> ICE_AQC_FW_LOG_EN_S)
- struct ice_fw_log_evnt evnts[ICE_AQC_FW_LOG_ID_MAX];
-};
-
/* Enum defining the different states of the mailbox snapshot in the
* PF-VF mailbox overflow detection algorithm. The snapshot can be in
* states:
@@ -890,8 +872,6 @@ struct ice_hw {
u8 fw_patch; /* firmware patch version */
u32 fw_build; /* firmware build number */
- struct ice_fw_log_cfg fw_log;
-
/* Device max aggregate bandwidths corresponding to the GL_PWR_MODE_CTL
* register. Used for determining the ITR/INTRL granularity during
* initialization.
--
2.43.0

1083
kernel-std/debian/patches/ice-VDF/0025-ice-configure-FW-logging.patch Normal file
View File

File diff suppressed because it is too large Load Diff

281
kernel-std/debian/patches/ice-VDF/0026-ice-enable-FW-logging.patch Normal file
View File

@ -0,0 +1,281 @@
From 189d58473481cf01b493fca4e9dd2ab8380d0ce5 Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:13 -0800
Subject: [PATCH 26/36] ice: enable FW logging
Once users have configured the FW logging then allow them to enable it
by writing to the 'fwlog/enable' file. The file accepts a boolean value
(0 or 1) where 1 means enable FW logging and 0 means disable FW logging.
# echo <value> > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/enable
Where <value> is 0 or 1.
The user can read the 'fwlog/enable' file to see whether logging is
enabled or not. Reading the actual data is a separate patch. To see the
current value then:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/enable
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 73671c3162c83a689342fd57f00b5f261682e49b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 3 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 98 +++++++++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 67 +++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.h | 2 +
4 files changed, 170 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 347e4fed5e0d..11391be4efc2 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2336,6 +2336,7 @@ enum ice_aqc_fw_logging_mod {
};
/* Set FW Logging configuration (indirect 0xFF30)
+ * Register for FW Logging (indirect 0xFF31)
* Query FW Logging (indirect 0xFF32)
*/
struct ice_aqc_fw_log {
@@ -2344,6 +2345,7 @@ struct ice_aqc_fw_log {
#define ICE_AQC_FW_LOG_CONF_AQ_EN BIT(1)
#define ICE_AQC_FW_LOG_QUERY_REGISTERED BIT(2)
#define ICE_AQC_FW_LOG_CONF_SET_VALID BIT(3)
+#define ICE_AQC_FW_LOG_AQ_REGISTER BIT(0)
#define ICE_AQC_FW_LOG_AQ_QUERY BIT(2)
u8 rsp_flag;
@@ -2662,6 +2664,7 @@ enum ice_adminq_opc {
/* FW Logging Commands */
ice_aqc_opc_fw_logs_config = 0xFF30,
+ ice_aqc_opc_fw_logs_register = 0xFF31,
ice_aqc_opc_fw_logs_query = 0xFF32,
};
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index 3b0d9b214fd1..3dde99969132 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -281,6 +281,101 @@ static const struct file_operations ice_debugfs_nr_messages_fops = {
.write = ice_debugfs_nr_messages_write,
};
+/**
+ * ice_debugfs_enable_read - read from 'enable' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_enable_read(struct file *filp,
+ char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char buff[32] = {};
+
+ snprintf(buff, sizeof(buff), "%u\n",
+ (u16)(hw->fwlog_cfg.options &
+ ICE_FWLOG_OPTION_IS_REGISTERED) >> 3);
+
+ return simple_read_from_buffer(buffer, count, ppos, buff, strlen(buff));
+}
+
+/**
+ * ice_debugfs_enable_write - write into 'enable' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_enable_write(struct file *filp, const char __user *buf,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char user_val[8], *cmd_buf;
+ bool enable;
+ ssize_t ret;
+
+ /* don't allow partial writes or invalid input */
+ if (*ppos != 0 || count > 2)
+ return -EINVAL;
+
+ cmd_buf = memdup_user(buf, count);
+ if (IS_ERR(cmd_buf))
+ return PTR_ERR(cmd_buf);
+
+ ret = sscanf(cmd_buf, "%s", user_val);
+ if (ret != 1)
+ return -EINVAL;
+
+ ret = kstrtobool(user_val, &enable);
+ if (ret)
+ goto enable_write_error;
+
+ if (enable)
+ hw->fwlog_cfg.options |= ICE_FWLOG_OPTION_ARQ_ENA;
+ else
+ hw->fwlog_cfg.options &= ~ICE_FWLOG_OPTION_ARQ_ENA;
+
+ ret = ice_fwlog_set(hw, &hw->fwlog_cfg);
+ if (ret)
+ goto enable_write_error;
+
+ if (enable)
+ ret = ice_fwlog_register(hw);
+ else
+ ret = ice_fwlog_unregister(hw);
+
+ if (ret)
+ goto enable_write_error;
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+enable_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_enable_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_enable_read,
+ .write = ice_debugfs_enable_write,
+};
+
/**
* ice_debugfs_fwlog_init - setup the debugfs directory
* @pf: the ice that is starting up
@@ -332,6 +427,9 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
pf->ice_debugfs_pf_fwlog_modules = fw_modules;
+ debugfs_create_file("enable", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_enable_fops);
+
return;
err_create_module_files:
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 307e0d04f3fe..25a17cbc1d34 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -63,6 +63,11 @@ void ice_fwlog_deinit(struct ice_hw *hw)
kfree(pf->ice_debugfs_pf_fwlog_modules);
pf->ice_debugfs_pf_fwlog_modules = NULL;
+
+ status = ice_fwlog_unregister(hw);
+ if (status)
+ dev_warn(ice_hw_to_dev(hw), "Unable to unregister FW logging, status: %d\n",
+ status);
}
/**
@@ -197,6 +202,8 @@ static int ice_aq_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg)
cfg->options |= ICE_FWLOG_OPTION_ARQ_ENA;
if (cmd->cmd_flags & ICE_AQC_FW_LOG_CONF_UART_EN)
cfg->options |= ICE_FWLOG_OPTION_UART_ENA;
+ if (cmd->cmd_flags & ICE_AQC_FW_LOG_QUERY_REGISTERED)
+ cfg->options |= ICE_FWLOG_OPTION_IS_REGISTERED;
fw_modules = (struct ice_aqc_fw_log_cfg_resp *)buf;
@@ -226,6 +233,66 @@ int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg)
return ice_aq_fwlog_get(hw, cfg);
}
+/**
+ * ice_aq_fwlog_register - Register PF for firmware logging events (0xFF31)
+ * @hw: pointer to the HW structure
+ * @reg: true to register and false to unregister
+ */
+static int ice_aq_fwlog_register(struct ice_hw *hw, bool reg)
+{
+ struct ice_aq_desc desc;
+
+ ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_fw_logs_register);
+
+ if (reg)
+ desc.params.fw_log.cmd_flags = ICE_AQC_FW_LOG_AQ_REGISTER;
+
+ return ice_aq_send_cmd(hw, &desc, NULL, 0, NULL);
+}
+
+/**
+ * ice_fwlog_register - Register the PF for firmware logging
+ * @hw: pointer to the HW structure
+ *
+ * After this call the PF will start to receive firmware logging based on the
+ * configuration set in ice_fwlog_set.
+ */
+int ice_fwlog_register(struct ice_hw *hw)
+{
+ int status;
+
+ if (!ice_fwlog_supported(hw))
+ return -EOPNOTSUPP;
+
+ status = ice_aq_fwlog_register(hw, true);
+ if (status)
+ ice_debug(hw, ICE_DBG_FW_LOG, "Failed to register for firmware logging events over ARQ\n");
+ else
+ hw->fwlog_cfg.options |= ICE_FWLOG_OPTION_IS_REGISTERED;
+
+ return status;
+}
+
+/**
+ * ice_fwlog_unregister - Unregister the PF from firmware logging
+ * @hw: pointer to the HW structure
+ */
+int ice_fwlog_unregister(struct ice_hw *hw)
+{
+ int status;
+
+ if (!ice_fwlog_supported(hw))
+ return -EOPNOTSUPP;
+
+ status = ice_aq_fwlog_register(hw, false);
+ if (status)
+ ice_debug(hw, ICE_DBG_FW_LOG, "Failed to unregister from firmware logging events over ARQ\n");
+ else
+ hw->fwlog_cfg.options &= ~ICE_FWLOG_OPTION_IS_REGISTERED;
+
+ return status;
+}
+
/**
* ice_fwlog_set_supported - Set if FW logging is supported by FW
* @hw: pointer to the HW struct
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.h b/drivers/net/ethernet/intel/ice/ice_fwlog.h
index 8e68ee02713b..45865558425d 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.h
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.h
@@ -53,4 +53,6 @@ int ice_fwlog_init(struct ice_hw *hw);
void ice_fwlog_deinit(struct ice_hw *hw);
int ice_fwlog_set(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
+int ice_fwlog_register(struct ice_hw *hw);
+int ice_fwlog_unregister(struct ice_hw *hw);
#endif /* _ICE_FWLOG_H_ */
--
2.43.0

572
kernel-std/debian/patches/ice-VDF/0027-ice-add-ability-to-read-and-configure-FW-log-data.patch Normal file
View File

@ -0,0 +1,572 @@
From a584ea88cfdc8ac3f782be1d5d67fa92c3423290 Mon Sep 17 00:00:00 2001
From: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Date: Tue, 12 Dec 2023 21:07:14 -0800
Subject: [PATCH 27/36] ice: add ability to read and configure FW log data
Once logging is enabled the user should read the data from the 'data'
file. The data is in the form of a binary blob that can be sent to Intel
for decoding. To read the data use a command like:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/data > log_data.bin
If the user wants to clear the FW log data that has been stored in the
driver then they can write any value to the 'data' file and that will clear
the data. An example is:
# echo 34 > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/data
In addition to being able to read the data the user can configure how
much memory is used to store FW log data. This allows the user to
increase/decrease the amount of memory based on the users situation.
The data is stored such that if the memory fills up then the oldest data
will get overwritten in a circular manner. To change the amount of
memory the user can write to the 'log_size' file like this:
# echo <value> > /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/log_size
Where <value> is one of 128K, 256K, 512K, 1M, and 2M. The default value
is 1M.
The user can see the current value of 'log_size' by reading the file:
# cat /sys/kernel/debug/ice/0000\:18\:00.0/fwlog/log_size
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 9d3535e71985beb738c4ad2b772c6f0efdce0202)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
.../net/ethernet/intel/ice/ice_adminq_cmd.h | 2 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 210 ++++++++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 142 ++++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.h | 21 ++
drivers/net/ethernet/intel/ice/ice_main.c | 29 +++
drivers/net/ethernet/intel/ice/ice_type.h | 1 +
6 files changed, 405 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
index 11391be4efc2..f63b57ff2a3d 100644
--- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
+++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h
@@ -2338,6 +2338,7 @@ enum ice_aqc_fw_logging_mod {
/* Set FW Logging configuration (indirect 0xFF30)
* Register for FW Logging (indirect 0xFF31)
* Query FW Logging (indirect 0xFF32)
+ * FW Log Event (indirect 0xFF33)
*/
struct ice_aqc_fw_log {
u8 cmd_flags;
@@ -2666,6 +2667,7 @@ enum ice_adminq_opc {
ice_aqc_opc_fw_logs_config = 0xFF30,
ice_aqc_opc_fw_logs_register = 0xFF31,
ice_aqc_opc_fw_logs_query = 0xFF32,
+ ice_aqc_opc_fw_logs_event = 0xFF33,
};
#endif /* _ICE_ADMINQ_CMD_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index 3dde99969132..c2bfba6b9ead 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -64,6 +64,17 @@ static const char * const ice_fwlog_level_string[] = {
"verbose",
};
+/* the order in this array is important. it matches the ordering of the
+ * values in the FW so the index is the same value as in ice_fwlog_level
+ */
+static const char * const ice_fwlog_log_size[] = {
+ "128K",
+ "256K",
+ "512K",
+ "1M",
+ "2M",
+};
+
/**
* ice_fwlog_print_module_cfg - print current FW logging module configuration
* @hw: pointer to the HW structure
@@ -376,6 +387,199 @@ static const struct file_operations ice_debugfs_enable_fops = {
.write = ice_debugfs_enable_write,
};
+/**
+ * ice_debugfs_log_size_read - read from 'log_size' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_log_size_read(struct file *filp,
+ char __user *buffer, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ char buff[32] = {};
+ int index;
+
+ index = hw->fwlog_ring.index;
+ snprintf(buff, sizeof(buff), "%s\n", ice_fwlog_log_size[index]);
+
+ return simple_read_from_buffer(buffer, count, ppos, buff, strlen(buff));
+}
+
+/**
+ * ice_debugfs_log_size_write - write into 'log_size' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_log_size_write(struct file *filp, const char __user *buf,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ char user_val[8], *cmd_buf;
+ ssize_t ret;
+ int index;
+
+ /* don't allow partial writes or invalid input */
+ if (*ppos != 0 || count > 5)
+ return -EINVAL;
+
+ cmd_buf = memdup_user(buf, count);
+ if (IS_ERR(cmd_buf))
+ return PTR_ERR(cmd_buf);
+
+ ret = sscanf(cmd_buf, "%s", user_val);
+ if (ret != 1)
+ return -EINVAL;
+
+ index = sysfs_match_string(ice_fwlog_log_size, user_val);
+ if (index < 0) {
+ dev_info(dev, "Invalid log size '%s'. The value must be one of 128K, 256K, 512K, 1M, 2M\n",
+ user_val);
+ ret = -EINVAL;
+ goto log_size_write_error;
+ } else if (hw->fwlog_cfg.options & ICE_FWLOG_OPTION_IS_REGISTERED) {
+ dev_info(dev, "FW logging is currently running. Please disable FW logging to change log_size\n");
+ ret = -EINVAL;
+ goto log_size_write_error;
+ }
+
+ /* free all the buffers and the tracking info and resize */
+ ice_fwlog_realloc_rings(hw, index);
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+log_size_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_log_size_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_log_size_read,
+ .write = ice_debugfs_log_size_write,
+};
+
+/**
+ * ice_debugfs_data_read - read from 'data' file
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ */
+static ssize_t ice_debugfs_data_read(struct file *filp, char __user *buffer,
+ size_t count, loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct ice_hw *hw = &pf->hw;
+ int data_copied = 0;
+ bool done = false;
+
+ if (ice_fwlog_ring_empty(&hw->fwlog_ring))
+ return 0;
+
+ while (!ice_fwlog_ring_empty(&hw->fwlog_ring) && !done) {
+ struct ice_fwlog_data *log;
+ u16 cur_buf_len;
+
+ log = &hw->fwlog_ring.rings[hw->fwlog_ring.head];
+ cur_buf_len = log->data_size;
+ if (cur_buf_len >= count) {
+ done = true;
+ continue;
+ }
+
+ if (copy_to_user(buffer, log->data, cur_buf_len)) {
+ /* if there is an error then bail and return whatever
+ * the driver has copied so far
+ */
+ done = true;
+ continue;
+ }
+
+ data_copied += cur_buf_len;
+ buffer += cur_buf_len;
+ count -= cur_buf_len;
+ *ppos += cur_buf_len;
+ ice_fwlog_ring_increment(&hw->fwlog_ring.head,
+ hw->fwlog_ring.size);
+ }
+
+ return data_copied;
+}
+
+/**
+ * ice_debugfs_data_write - write into 'data' file
+ * @filp: the opened file
+ * @buf: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ */
+static ssize_t
+ice_debugfs_data_write(struct file *filp, const char __user *buf, size_t count,
+ loff_t *ppos)
+{
+ struct ice_pf *pf = filp->private_data;
+ struct device *dev = ice_pf_to_dev(pf);
+ struct ice_hw *hw = &pf->hw;
+ ssize_t ret;
+
+ /* don't allow partial writes */
+ if (*ppos != 0)
+ return 0;
+
+ /* any value is allowed to clear the buffer so no need to even look at
+ * what the value is
+ */
+ if (!(hw->fwlog_cfg.options & ICE_FWLOG_OPTION_IS_REGISTERED)) {
+ hw->fwlog_ring.head = 0;
+ hw->fwlog_ring.tail = 0;
+ } else {
+ dev_info(dev, "Can't clear FW log data while FW log running\n");
+ ret = -EINVAL;
+ goto nr_buffs_write_error;
+ }
+
+ /* if we get here, nothing went wrong; return count since we didn't
+ * really write anything
+ */
+ ret = (ssize_t)count;
+
+nr_buffs_write_error:
+ /* This function always consumes all of the written input, or produces
+ * an error. Check and enforce this. Otherwise, the write operation
+ * won't complete properly.
+ */
+ if (WARN_ON(ret != (ssize_t)count && ret >= 0))
+ ret = -EIO;
+
+ return ret;
+}
+
+static const struct file_operations ice_debugfs_data_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = ice_debugfs_data_read,
+ .write = ice_debugfs_data_write,
+};
+
/**
* ice_debugfs_fwlog_init - setup the debugfs directory
* @pf: the ice that is starting up
@@ -430,6 +634,12 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
debugfs_create_file("enable", 0600, pf->ice_debugfs_pf_fwlog,
pf, &ice_debugfs_enable_fops);
+ debugfs_create_file("log_size", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_log_size_fops);
+
+ debugfs_create_file("data", 0600, pf->ice_debugfs_pf_fwlog,
+ pf, &ice_debugfs_data_fops);
+
return;
err_create_module_files:
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 25a17cbc1d34..92b5dac481cd 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -1,10 +1,128 @@
// SPDX-License-Identifier: GPL-2.0
/* Copyright (c) 2022, Intel Corporation. */
+#include <linux/vmalloc.h>
#include "ice.h"
#include "ice_common.h"
#include "ice_fwlog.h"
+bool ice_fwlog_ring_full(struct ice_fwlog_ring *rings)
+{
+ u16 head, tail;
+
+ head = rings->head;
+ tail = rings->tail;
+
+ if (head < tail && (tail - head == (rings->size - 1)))
+ return true;
+ else if (head > tail && (tail == (head - 1)))
+ return true;
+
+ return false;
+}
+
+bool ice_fwlog_ring_empty(struct ice_fwlog_ring *rings)
+{
+ return rings->head == rings->tail;
+}
+
+void ice_fwlog_ring_increment(u16 *item, u16 size)
+{
+ *item = (*item + 1) & (size - 1);
+}
+
+static int ice_fwlog_alloc_ring_buffs(struct ice_fwlog_ring *rings)
+{
+ int i, nr_bytes;
+ u8 *mem;
+
+ nr_bytes = rings->size * ICE_AQ_MAX_BUF_LEN;
+ mem = vzalloc(nr_bytes);
+ if (!mem)
+ return -ENOMEM;
+
+ for (i = 0; i < rings->size; i++) {
+ struct ice_fwlog_data *ring = &rings->rings[i];
+
+ ring->data_size = ICE_AQ_MAX_BUF_LEN;
+ ring->data = mem;
+ mem += ICE_AQ_MAX_BUF_LEN;
+ }
+
+ return 0;
+}
+
+static void ice_fwlog_free_ring_buffs(struct ice_fwlog_ring *rings)
+{
+ int i;
+
+ for (i = 0; i < rings->size; i++) {
+ struct ice_fwlog_data *ring = &rings->rings[i];
+
+ /* the first ring is the base memory for the whole range so
+ * free it
+ */
+ if (!i)
+ vfree(ring->data);
+
+ ring->data = NULL;
+ ring->data_size = 0;
+ }
+}
+
+#define ICE_FWLOG_INDEX_TO_BYTES(n) ((128 * 1024) << (n))
+/**
+ * ice_fwlog_realloc_rings - reallocate the FW log rings
+ * @hw: pointer to the HW structure
+ * @index: the new index to use to allocate memory for the log data
+ *
+ */
+void ice_fwlog_realloc_rings(struct ice_hw *hw, int index)
+{
+ struct ice_fwlog_ring ring;
+ int status, ring_size;
+
+ /* convert the number of bytes into a number of 4K buffers. externally
+ * the driver presents the interface to the FW log data as a number of
+ * bytes because that's easy for users to understand. internally the
+ * driver uses a ring of buffers because the driver doesn't know where
+ * the beginning and end of any line of log data is so the driver has
+ * to overwrite data as complete blocks. when the data is returned to
+ * the user the driver knows that the data is correct and the FW log
+ * can be correctly parsed by the tools
+ */
+ ring_size = ICE_FWLOG_INDEX_TO_BYTES(index) / ICE_AQ_MAX_BUF_LEN;
+ if (ring_size == hw->fwlog_ring.size)
+ return;
+
+ /* allocate space for the new rings and buffers then release the
+ * old rings and buffers. that way if we don't have enough
+ * memory then we at least have what we had before
+ */
+ ring.rings = kcalloc(ring_size, sizeof(*ring.rings), GFP_KERNEL);
+ if (!ring.rings)
+ return;
+
+ ring.size = ring_size;
+
+ status = ice_fwlog_alloc_ring_buffs(&ring);
+ if (status) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log ring data buffers\n");
+ ice_fwlog_free_ring_buffs(&ring);
+ kfree(ring.rings);
+ return;
+ }
+
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+
+ hw->fwlog_ring.rings = ring.rings;
+ hw->fwlog_ring.size = ring.size;
+ hw->fwlog_ring.index = index;
+ hw->fwlog_ring.head = 0;
+ hw->fwlog_ring.tail = 0;
+}
+
/**
* ice_fwlog_init - Initialize FW logging configuration
* @hw: pointer to the HW structure
@@ -28,6 +146,25 @@ int ice_fwlog_init(struct ice_hw *hw)
if (status)
return status;
+ hw->fwlog_ring.rings = kcalloc(ICE_FWLOG_RING_SIZE_DFLT,
+ sizeof(*hw->fwlog_ring.rings),
+ GFP_KERNEL);
+ if (!hw->fwlog_ring.rings) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log rings\n");
+ return -ENOMEM;
+ }
+
+ hw->fwlog_ring.size = ICE_FWLOG_RING_SIZE_DFLT;
+ hw->fwlog_ring.index = ICE_FWLOG_RING_SIZE_INDEX_DFLT;
+
+ status = ice_fwlog_alloc_ring_buffs(&hw->fwlog_ring);
+ if (status) {
+ dev_warn(ice_hw_to_dev(hw), "Unable to allocate memory for FW log ring data buffers\n");
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+ return status;
+ }
+
ice_debugfs_fwlog_init(hw->back);
} else {
dev_warn(ice_hw_to_dev(hw), "FW logging is not supported in this NVM image. Please update the NVM to get FW log support\n");
@@ -68,6 +205,11 @@ void ice_fwlog_deinit(struct ice_hw *hw)
if (status)
dev_warn(ice_hw_to_dev(hw), "Unable to unregister FW logging, status: %d\n",
status);
+
+ if (hw->fwlog_ring.rings) {
+ ice_fwlog_free_ring_buffs(&hw->fwlog_ring);
+ kfree(hw->fwlog_ring.rings);
+ }
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.h b/drivers/net/ethernet/intel/ice/ice_fwlog.h
index 45865558425d..287e71fa4b86 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.h
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.h
@@ -47,6 +47,26 @@ struct ice_fwlog_cfg {
u16 log_resolution;
};
+struct ice_fwlog_data {
+ u16 data_size;
+ u8 *data;
+};
+
+struct ice_fwlog_ring {
+ struct ice_fwlog_data *rings;
+ u16 index;
+ u16 size;
+ u16 head;
+ u16 tail;
+};
+
+#define ICE_FWLOG_RING_SIZE_INDEX_DFLT 3
+#define ICE_FWLOG_RING_SIZE_DFLT 256
+#define ICE_FWLOG_RING_SIZE_MAX 512
+
+bool ice_fwlog_ring_full(struct ice_fwlog_ring *rings);
+bool ice_fwlog_ring_empty(struct ice_fwlog_ring *rings);
+void ice_fwlog_ring_increment(u16 *item, u16 size);
void ice_fwlog_set_supported(struct ice_hw *hw);
bool ice_fwlog_supported(struct ice_hw *hw);
int ice_fwlog_init(struct ice_hw *hw);
@@ -55,4 +75,5 @@ int ice_fwlog_set(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_get(struct ice_hw *hw, struct ice_fwlog_cfg *cfg);
int ice_fwlog_register(struct ice_hw *hw);
int ice_fwlog_unregister(struct ice_hw *hw);
+void ice_fwlog_realloc_rings(struct ice_hw *hw, int index);
#endif /* _ICE_FWLOG_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 614e10ab4159..6c6ca5353f28 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -1254,6 +1254,32 @@ ice_handle_link_event(struct ice_pf *pf, struct ice_rq_event_info *event)
return status;
}
+/**
+ * ice_get_fwlog_data - copy the FW log data from ARQ event
+ * @pf: PF that the FW log event is associated with
+ * @event: event structure containing FW log data
+ */
+static void
+ice_get_fwlog_data(struct ice_pf *pf, struct ice_rq_event_info *event)
+{
+ struct ice_fwlog_data *fwlog;
+ struct ice_hw *hw = &pf->hw;
+
+ fwlog = &hw->fwlog_ring.rings[hw->fwlog_ring.tail];
+
+ memset(fwlog->data, 0, PAGE_SIZE);
+ fwlog->data_size = le16_to_cpu(event->desc.datalen);
+
+ memcpy(fwlog->data, event->msg_buf, fwlog->data_size);
+ ice_fwlog_ring_increment(&hw->fwlog_ring.tail, hw->fwlog_ring.size);
+
+ if (ice_fwlog_ring_full(&hw->fwlog_ring)) {
+ /* the rings are full so bump the head to create room */
+ ice_fwlog_ring_increment(&hw->fwlog_ring.head,
+ hw->fwlog_ring.size);
+ }
+}
+
/**
* ice_aq_prep_for_event - Prepare to wait for an AdminQ event from firmware
* @pf: pointer to the PF private structure
@@ -1535,6 +1561,9 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
ice_vc_process_vf_msg(pf, &event, &data);
break;
+ case ice_aqc_opc_fw_logs_event:
+ ice_get_fwlog_data(pf, &event);
+ break;
case ice_aqc_opc_lldp_set_mib_change:
ice_dcb_process_lldp_set_mib_change(pf, &event);
break;
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 84bb61aa7409..28e47bb78eaf 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -875,6 +875,7 @@ struct ice_hw {
struct ice_fwlog_cfg fwlog_cfg;
bool fwlog_supported; /* does hardware support FW logging? */
+ struct ice_fwlog_ring fwlog_ring;
/* Device max aggregate bandwidths corresponding to the GL_PWR_MODE_CTL
* register. Used for determining the ITR/INTRL granularity during
--
2.43.0

98
kernel-std/debian/patches/ice-VDF/0028-ice-Fix-debugfs-with-devlink-reload.patch Normal file
View File

@ -0,0 +1,98 @@
From 8a7f6d8b2105c39f236c51c558e21b787c223861 Mon Sep 17 00:00:00 2001
From: Wojciech Drewek <wojciech.drewek@intel.com>
Date: Mon, 5 Feb 2024 14:03:57 +0100
Subject: [PATCH 28/36] ice: Fix debugfs with devlink reload
During devlink reload it is needed to remove debugfs entries
correlated with only one PF. ice_debugfs_exit() removes all
entries created by ice driver so we can't use it.
Introduce ice_debugfs_pf_deinit() in order to release PF's
debugfs entries. Move ice_debugfs_exit() call to ice_module_exit(),
it makes more sense since ice_debugfs_init() is called in
ice_module_init() and not in ice_probe().
Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Reviewed-by: Brett Creeley <brett.creeley@amd.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 500d0df5b4b2394a06b949bab05f7ed0242b9858)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice.h | 1 +
drivers/net/ethernet/intel/ice/ice_debugfs.c | 10 ++++++++++
drivers/net/ethernet/intel/ice/ice_fwlog.c | 2 ++
drivers/net/ethernet/intel/ice/ice_main.c | 3 +--
4 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index 7966ac61154c..ed1c6cdedeff 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -895,6 +895,7 @@ static inline bool ice_is_adq_active(struct ice_pf *pf)
}
void ice_debugfs_fwlog_init(struct ice_pf *pf);
+void ice_debugfs_pf_deinit(struct ice_pf *pf);
void ice_debugfs_init(void);
void ice_debugfs_exit(void);
void ice_pf_fwlog_update_module(struct ice_pf *pf, int log_level, int module);
diff --git a/drivers/net/ethernet/intel/ice/ice_debugfs.c b/drivers/net/ethernet/intel/ice/ice_debugfs.c
index c2bfba6b9ead..ba396b22bb7d 100644
--- a/drivers/net/ethernet/intel/ice/ice_debugfs.c
+++ b/drivers/net/ethernet/intel/ice/ice_debugfs.c
@@ -647,6 +647,16 @@ void ice_debugfs_fwlog_init(struct ice_pf *pf)
kfree(fw_modules);
}
+/**
+ * ice_debugfs_pf_deinit - cleanup PF's debugfs
+ * @pf: pointer to the PF struct
+ */
+void ice_debugfs_pf_deinit(struct ice_pf *pf)
+{
+ debugfs_remove_recursive(pf->ice_debugfs_pf);
+ pf->ice_debugfs_pf = NULL;
+}
+
/**
* ice_debugfs_init - create root directory for debugfs entries
*/
diff --git a/drivers/net/ethernet/intel/ice/ice_fwlog.c b/drivers/net/ethernet/intel/ice/ice_fwlog.c
index 92b5dac481cd..4fd15387a7e5 100644
--- a/drivers/net/ethernet/intel/ice/ice_fwlog.c
+++ b/drivers/net/ethernet/intel/ice/ice_fwlog.c
@@ -188,6 +188,8 @@ void ice_fwlog_deinit(struct ice_hw *hw)
if (hw->bus.func)
return;
+ ice_debugfs_pf_deinit(hw->back);
+
/* make sure FW logging is disabled to not put the FW in a weird state
* for the next driver load
*/
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 6c6ca5353f28..c882c218281a 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -5325,8 +5325,6 @@ static void ice_remove(struct pci_dev *pdev)
msleep(100);
}
- ice_debugfs_exit();
-
if (test_bit(ICE_FLAG_SRIOV_ENA, pf->flags)) {
set_bit(ICE_VF_RESETS_DISABLED, pf->state);
ice_free_vfs(pf);
@@ -5823,6 +5821,7 @@ module_init(ice_module_init);
static void __exit ice_module_exit(void)
{
pci_unregister_driver(&ice_driver);
+ ice_debugfs_exit();
destroy_workqueue(ice_wq);
destroy_workqueue(ice_lag_wq);
pr_info("module unloaded\n");
--
2.43.0

78
kernel-std/debian/patches/ice-VDF/0029-ice-remove-vf-lan_vsi_num-field.patch Normal file
View File

@ -0,0 +1,78 @@
From 861015cbb4cf4cb258a1da9e80550fe991be7808 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Fri, 16 Feb 2024 14:06:38 -0800
Subject: [PATCH 29/36] ice: remove vf->lan_vsi_num field
The lan_vsi_num field of the VF structure is no longer used for any
purpose. Remove it.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1cf94cbfc61bac89cddeb075fbc100ebd3aea81b)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_sriov.c | 1 -
drivers/net/ethernet/intel/ice/ice_vf_lib.c | 4 +---
drivers/net/ethernet/intel/ice/ice_vf_lib.h | 5 -----
3 files changed, 1 insertion(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_sriov.c b/drivers/net/ethernet/intel/ice/ice_sriov.c
index 442162be23ea..3366ac976c44 100644
--- a/drivers/net/ethernet/intel/ice/ice_sriov.c
+++ b/drivers/net/ethernet/intel/ice/ice_sriov.c
@@ -239,7 +239,6 @@ static struct ice_vsi *ice_vf_vsi_setup(struct ice_vf *vf)
}
vf->lan_vsi_idx = vsi->idx;
- vf->lan_vsi_num = vsi->vsi_num;
return vsi;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_vf_lib.c b/drivers/net/ethernet/intel/ice/ice_vf_lib.c
index 03b9d7d74851..303fdf8555cf 100644
--- a/drivers/net/ethernet/intel/ice/ice_vf_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_vf_lib.c
@@ -298,7 +298,6 @@ static int ice_vf_rebuild_vsi(struct ice_vf *vf)
* vf->lan_vsi_idx
*/
vsi->vsi_num = ice_get_hw_vsi_num(&pf->hw, vsi->idx);
- vf->lan_vsi_num = vsi->vsi_num;
return 0;
}
@@ -1299,13 +1298,12 @@ int ice_vf_init_host_cfg(struct ice_vf *vf, struct ice_vsi *vsi)
}
/**
- * ice_vf_invalidate_vsi - invalidate vsi_idx/vsi_num to remove VSI access
+ * ice_vf_invalidate_vsi - invalidate vsi_idx to remove VSI access
* @vf: VF to remove access to VSI for
*/
void ice_vf_invalidate_vsi(struct ice_vf *vf)
{
vf->lan_vsi_idx = ICE_NO_VSI;
- vf->lan_vsi_num = ICE_NO_VSI;
}
/**
diff --git a/drivers/net/ethernet/intel/ice/ice_vf_lib.h b/drivers/net/ethernet/intel/ice/ice_vf_lib.h
index 48fea6fa0362..1de07accbc5c 100644
--- a/drivers/net/ethernet/intel/ice/ice_vf_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_vf_lib.h
@@ -110,11 +110,6 @@ struct ice_vf {
u8 spoofchk:1;
u8 link_forced:1;
u8 link_up:1; /* only valid if VF link is forced */
- /* VSI indices - actual VSI pointers are maintained in the PF structure
- * When assigned, these will be non-zero, because VSI 0 is always
- * the main LAN VSI for the PF.
- */
- u16 lan_vsi_num; /* ID as used by firmware */
unsigned int min_tx_rate; /* Minimum Tx bandwidth limit in Mbps */
unsigned int max_tx_rate; /* Maximum Tx bandwidth limit in Mbps */
DECLARE_BITMAP(vf_states, ICE_VF_STATES_NBITS); /* VF runtime states */
--
2.43.0

150
kernel-std/debian/patches/ice-VDF/0030-ice-rename-ice_write_-functions-to-ice_pack_ctx_.patch Normal file
View File

@ -0,0 +1,150 @@
From 6b7fae8669a04943af9f83ef89d39a922ed179fd Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:54 -0800
Subject: [PATCH 30/36] ice: rename ice_write_* functions to ice_pack_ctx_*
In ice_common.c there are 4 functions used for converting the unpacked
software Tx and Rx context structure data into the packed format used by
hardware. These functions have extremely generic names:
* ice_write_byte
* ice_write_word
* ice_write_dword
* ice_write_qword
When I saw these function names my first thought was "write what? to
where?". Understanding what these functions do requires looking at the
implementation details. The functions take bits from an unpacked structure
and copy them into the packed layout used by hardware.
As part of live migration, we will want functions which perform the inverse
operation of reading bits from the packed layout and copying them into the
unpacked format. Naming these as "ice_read_byte", etc would be very
confusing since they appear to write data.
In preparation for adding this new inverse operation, rename the existing
functions to use the prefix "ice_pack_ctx_". This makes it clear that they
perform the bit packing while copying from the unpacked software context
structure to the packed hardware context.
The inverse operations can then neatly be named ice_unpack_ctx_*, clearly
indicating they perform the bit unpacking while copying from the packed
hardware context to the unpacked software context structure.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 1260b45dbe2dbc415f3bc1e841c6c098083bcfb8)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 56 ++++++++++-----------
1 file changed, 28 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 6dcba0577633..17f60a98c8ed 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -4267,13 +4267,13 @@ ice_aq_add_rdma_qsets(struct ice_hw *hw, u8 num_qset_grps,
/* End of FW Admin Queue command wrappers */
/**
- * ice_write_byte - write a byte to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_byte - write a byte to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_byte(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_byte(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u8 src_byte, dest_byte, mask;
u8 *from, *dest;
@@ -4306,13 +4306,13 @@ ice_write_byte(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_word - write a word to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_word - write a word to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_word(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_word(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u16 src_word, mask;
__le16 dest_word;
@@ -4349,13 +4349,13 @@ ice_write_word(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_dword - write a dword to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_dword - write a dword to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_dword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_dword(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u32 src_dword, mask;
__le32 dest_dword;
@@ -4400,13 +4400,13 @@ ice_write_dword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
}
/**
- * ice_write_qword - write a qword to a packed context structure
- * @src_ctx: the context structure to read from
- * @dest_ctx: the context to be written to
- * @ce_info: a description of the struct to be filled
+ * ice_pack_ctx_qword - write a qword to a packed context structure
+ * @src_ctx: unpacked source context structure
+ * @dest_ctx: packed destination context data
+ * @ce_info: context element description
*/
-static void
-ice_write_qword(u8 *src_ctx, u8 *dest_ctx, const struct ice_ctx_ele *ce_info)
+static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
u64 src_qword, mask;
__le64 dest_qword;
@@ -4475,16 +4475,16 @@ ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
}
switch (ce_info[f].size_of) {
case sizeof(u8):
- ice_write_byte(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_byte(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u16):
- ice_write_word(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_word(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u32):
- ice_write_dword(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_dword(src_ctx, dest_ctx, &ce_info[f]);
break;
case sizeof(u64):
- ice_write_qword(src_ctx, dest_ctx, &ce_info[f]);
+ ice_pack_ctx_qword(src_ctx, dest_ctx, &ce_info[f]);
break;
default:
return -EINVAL;
--
2.43.0

129
kernel-std/debian/patches/ice-VDF/0031-ice-use-GENMASK-instead-of-BIT-n-1-in-pack-functions.patch Normal file
View File

@ -0,0 +1,129 @@
From 619e0e61b39cf051137613459d36c4fe8f435e57 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:55 -0800
Subject: [PATCH 31/36] ice: use GENMASK instead of BIT(n) - 1 in pack
functions
The functions used to pack the Tx and Rx context into the hardware format
rely on using BIT() and then subtracting 1 to get a bitmask. These
functions even have a comment about how x86 machines can't use this method
for certain widths because the SHL instructions will not work properly.
The Linux kernel already provides the GENMASK macro for generating a
suitable bitmask. Further, GENMASK is capable of generating the mask
including the shift_width. Since width is the total field width, take care
to subtract one to get the final bit position.
Since we now include the shifted bits as part of the mask, shift the source
value first before applying the mask.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit a45d1bf516c097bb7ae4983d3128ebf139be952c)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 44 ++++-----------------
1 file changed, 8 insertions(+), 36 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 17f60a98c8ed..55a2e264dd69 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -4284,14 +4284,11 @@ static void ice_pack_ctx_byte(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
- mask = (u8)(BIT(ce_info->width) - 1);
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
src_byte = *from;
- src_byte &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_byte <<= shift_width;
+ src_byte &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4324,17 +4321,14 @@ static void ice_pack_ctx_word(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
- mask = BIT(ce_info->width) - 1;
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_word = *(u16 *)from;
- src_word &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_word <<= shift_width;
+ src_word &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4367,25 +4361,14 @@ static void ice_pack_ctx_dword(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
-
- /* if the field width is exactly 32 on an x86 machine, then the shift
- * operation will not work because the SHL instructions count is masked
- * to 5 bits so the shift will do nothing
- */
- if (ce_info->width < 32)
- mask = BIT(ce_info->width) - 1;
- else
- mask = (u32)~0;
+ mask = GENMASK(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_dword = *(u32 *)from;
- src_dword &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_dword <<= shift_width;
+ src_dword &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
@@ -4418,25 +4401,14 @@ static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
/* prepare the bits and mask */
shift_width = ce_info->lsb % 8;
-
- /* if the field width is exactly 64 on an x86 machine, then the shift
- * operation will not work because the SHL instructions count is masked
- * to 6 bits so the shift will do nothing
- */
- if (ce_info->width < 64)
- mask = BIT_ULL(ce_info->width) - 1;
- else
- mask = (u64)~0;
+ mask = GENMASK_ULL(ce_info->width - 1 + shift_width, shift_width);
/* don't swizzle the bits until after the mask because the mask bits
* will be in a different bit position on big endian machines
*/
src_qword = *(u64 *)from;
- src_qword &= mask;
-
- /* shift to correct alignment */
- mask <<= shift_width;
src_qword <<= shift_width;
+ src_qword &= mask;
/* get the current bits from the target bit string */
dest = dest_ctx + (ce_info->lsb / 8);
--
2.43.0

89
kernel-std/debian/patches/ice-VDF/0032-ice-cleanup-line-splitting-for-context-set-functions.patch Normal file
View File

@ -0,0 +1,89 @@
From 6502dd63a7bd436c72d2ee8b328985b93fa7e2a5 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Mon, 26 Feb 2024 16:14:56 -0800
Subject: [PATCH 32/36] ice: cleanup line splitting for context set functions
The indentation for ice_set_ctx and ice_write_rxq_ctx breaks the function
name after the return type. This style of breaking is used a lot throughout
the ice driver, even in cases where its not actually helpful for
readability. We no longer prefer this style of line splitting in the
driver, and new code is avoiding it.
Normally, I would leave this alone unless the actual function contents or
description needed updating. However, a future change is going to add
inverse functions for converting packed context to unpacked context
structures. To keep this code uniform with the existing set functions, fix
up the style to the modern format of keeping the type on the same line.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit 979c2c049fbea107ce9f8d31f3ba9dba83ddb0a2)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 12 +++++-------
drivers/net/ethernet/intel/ice/ice_common.h | 10 ++++------
2 files changed, 9 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 55a2e264dd69..59ede77a1473 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -1329,9 +1329,8 @@ static const struct ice_ctx_ele ice_rlan_ctx_info[] = {
* it to HW register space and enables the hardware to prefetch descriptors
* instead of only fetching them on demand
*/
-int
-ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
- u32 rxq_index)
+int ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
+ u32 rxq_index)
{
u8 ctx_buf[ICE_RXQ_CTX_SZ] = { 0 };
@@ -4427,11 +4426,10 @@ static void ice_pack_ctx_qword(u8 *src_ctx, u8 *dest_ctx,
* @hw: pointer to the hardware structure
* @src_ctx: pointer to a generic non-packed context structure
* @dest_ctx: pointer to memory for the packed structure
- * @ce_info: a description of the structure to be transformed
+ * @ce_info: List of Rx context elements
*/
-int
-ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
- const struct ice_ctx_ele *ce_info)
+int ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info)
{
int f;
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index d47e5400351f..1c3c29d30815 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -52,9 +52,8 @@ int ice_get_caps(struct ice_hw *hw);
void ice_set_safe_mode_caps(struct ice_hw *hw);
-int
-ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
- u32 rxq_index);
+int ice_write_rxq_ctx(struct ice_hw *hw, struct ice_rlan_ctx *rlan_ctx,
+ u32 rxq_index);
int
ice_aq_get_rss_lut(struct ice_hw *hw, struct ice_aq_get_set_rss_lut_params *get_params);
@@ -71,9 +70,8 @@ bool ice_check_sq_alive(struct ice_hw *hw, struct ice_ctl_q_info *cq);
int ice_aq_q_shutdown(struct ice_hw *hw, bool unloading);
void ice_fill_dflt_direct_cmd_desc(struct ice_aq_desc *desc, u16 opcode);
extern const struct ice_ctx_ele ice_tlan_ctx_info[];
-int
-ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
- const struct ice_ctx_ele *ce_info);
+int ice_set_ctx(struct ice_hw *hw, u8 *src_ctx, u8 *dest_ctx,
+ const struct ice_ctx_ele *ce_info);
extern struct mutex ice_global_cfg_lock_sw;
--
2.43.0

161
kernel-std/debian/patches/ice-VDF/0033-ice-do-not-disable-Tx-queues-twice-in-ice_down.patch Normal file
View File

@ -0,0 +1,161 @@
From 675a8843a0de1411666389eeabeea452161f8cc5 Mon Sep 17 00:00:00 2001
From: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Date: Fri, 23 Feb 2024 17:06:27 +0100
Subject: [PATCH 33/36] ice: do not disable Tx queues twice in ice_down()
ice_down() clears QINT_TQCTL_CAUSE_ENA_M bit twice, which is not
necessary. First clearing happens in ice_vsi_dis_irq() and second in
ice_vsi_stop_tx_ring() - remove the first one.
While at it, make ice_vsi_dis_irq() static as ice_down() is the only
current caller of it.
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit d5926e01e3739542bb047b77f850d7f641eaa7bc)
[Adjust ice_lib.c with the context change.]
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_lib.c | 55 -----------------------
drivers/net/ethernet/intel/ice/ice_lib.h | 2 -
drivers/net/ethernet/intel/ice/ice_main.c | 44 ++++++++++++++++++
3 files changed, 44 insertions(+), 57 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 106ef843f4b5..f23cb9c8e3dd 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -2877,61 +2877,6 @@ void ice_dis_vsi(struct ice_vsi *vsi, bool locked)
}
}
-/**
- * ice_vsi_dis_irq - Mask off queue interrupt generation on the VSI
- * @vsi: the VSI being un-configured
- */
-void ice_vsi_dis_irq(struct ice_vsi *vsi)
-{
- struct ice_pf *pf = vsi->back;
- struct ice_hw *hw = &pf->hw;
- u32 val;
- int i;
-
- /* disable interrupt causation from each queue */
- if (vsi->tx_rings) {
- ice_for_each_txq(vsi, i) {
- if (vsi->tx_rings[i]) {
- u16 reg;
-
- reg = vsi->tx_rings[i]->reg_idx;
- val = rd32(hw, QINT_TQCTL(reg));
- val &= ~QINT_TQCTL_CAUSE_ENA_M;
- wr32(hw, QINT_TQCTL(reg), val);
- }
- }
- }
-
- if (vsi->rx_rings) {
- ice_for_each_rxq(vsi, i) {
- if (vsi->rx_rings[i]) {
- u16 reg;
-
- reg = vsi->rx_rings[i]->reg_idx;
- val = rd32(hw, QINT_RQCTL(reg));
- val &= ~QINT_RQCTL_CAUSE_ENA_M;
- wr32(hw, QINT_RQCTL(reg), val);
- }
- }
- }
-
- /* disable each interrupt */
- ice_for_each_q_vector(vsi, i) {
- if (!vsi->q_vectors[i])
- continue;
- wr32(hw, GLINT_DYN_CTL(vsi->q_vectors[i]->reg_idx), 0);
- }
-
- ice_flush(hw);
-
- /* don't call synchronize_irq() for VF's from the host */
- if (vsi->type == ICE_VSI_VF)
- return;
-
- ice_for_each_q_vector(vsi, i)
- synchronize_irq(vsi->q_vectors[i]->irq.virq);
-}
-
/**
* ice_vsi_release - Delete a VSI and free its resources
* @vsi: the VSI being removed
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h
index f24f5d1e6f9c..dbd0f3409323 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_lib.h
@@ -110,8 +110,6 @@ void
ice_write_qrxflxp_cntxt(struct ice_hw *hw, u16 pf_q, u32 rxdid, u32 prio,
bool ena_ts);
-void ice_vsi_dis_irq(struct ice_vsi *vsi);
-
void ice_vsi_free_irq(struct ice_vsi *vsi);
void ice_vsi_free_rx_rings(struct ice_vsi *vsi);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index c882c218281a..685635a22616 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -7023,6 +7023,50 @@ static void ice_napi_disable_all(struct ice_vsi *vsi)
}
}
+/**
+ * ice_vsi_dis_irq - Mask off queue interrupt generation on the VSI
+ * @vsi: the VSI being un-configured
+ */
+static void ice_vsi_dis_irq(struct ice_vsi *vsi)
+{
+ struct ice_pf *pf = vsi->back;
+ struct ice_hw *hw = &pf->hw;
+ u32 val;
+ int i;
+
+ /* disable interrupt causation from each Rx queue; Tx queues are
+ * handled in ice_vsi_stop_tx_ring()
+ */
+ if (vsi->rx_rings) {
+ ice_for_each_rxq(vsi, i) {
+ if (vsi->rx_rings[i]) {
+ u16 reg;
+
+ reg = vsi->rx_rings[i]->reg_idx;
+ val = rd32(hw, QINT_RQCTL(reg));
+ val &= ~QINT_RQCTL_CAUSE_ENA_M;
+ wr32(hw, QINT_RQCTL(reg), val);
+ }
+ }
+ }
+
+ /* disable each interrupt */
+ ice_for_each_q_vector(vsi, i) {
+ if (!vsi->q_vectors[i])
+ continue;
+ wr32(hw, GLINT_DYN_CTL(vsi->q_vectors[i]->reg_idx), 0);
+ }
+
+ ice_flush(hw);
+
+ /* don't call synchronize_irq() for VF's from the host */
+ if (vsi->type == ICE_VSI_VF)
+ return;
+
+ ice_for_each_q_vector(vsi, i)
+ synchronize_irq(vsi->q_vectors[i]->irq.virq);
+}
+
/**
* ice_down - Shutdown the connection
* @vsi: The VSI being stopped
--
2.43.0

274
kernel-std/debian/patches/ice-VDF/0034-ice-Fix-improper-extts-handling.patch Normal file
View File

@ -0,0 +1,274 @@
From e53280c20bbe58015a91178268244d5e831276f4 Mon Sep 17 00:00:00 2001
From: Milena Olech <milena.olech@intel.com>
Date: Tue, 2 Jul 2024 10:14:54 -0700
Subject: [PATCH 34/36] ice: Fix improper extts handling
Extts events are disabled and enabled by the application ts2phc.
However, in case where the driver is removed when the application is
running, a specific extts event remains enabled and can cause a kernel
crash.
As a side effect, when the driver is reloaded and application is started
again, remaining extts event for the channel from a previous run will
keep firing and the message "extts on unexpected channel" might be
printed to the user.
To avoid that, extts events shall be disabled when PTP is released.
Fixes: 172db5f91d5f ("ice: add support for auxiliary input/output pins")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Co-developed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Milena Olech <milena.olech@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20240702171459.2606611-2-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 00d3b4f54582d4e4a02cda5886bb336eeab268cc)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 105 ++++++++++++++++++-----
drivers/net/ethernet/intel/ice/ice_ptp.h | 8 ++
2 files changed, 91 insertions(+), 22 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 48ec59fc5d87..6e06c5d596b9 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1603,27 +1603,24 @@ void ice_ptp_extts_event(struct ice_pf *pf)
/**
* ice_ptp_cfg_extts - Configure EXTTS pin and channel
* @pf: Board private structure
- * @ena: true to enable; false to disable
* @chan: GPIO channel (0-3)
- * @gpio_pin: GPIO pin
- * @extts_flags: request flags from the ptp_extts_request.flags
+ * @config: desired EXTTS configuration.
+ * @store: If set to true, the values will be stored
+ *
+ * Configure an external timestamp event on the requested channel.
*/
-static int
-ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
- unsigned int extts_flags)
+static void ice_ptp_cfg_extts(struct ice_pf *pf, unsigned int chan,
+ struct ice_extts_channel *config, bool store)
{
u32 func, aux_reg, gpio_reg, irq_reg;
struct ice_hw *hw = &pf->hw;
u8 tmr_idx;
- if (chan > (unsigned int)pf->ptp.info.n_ext_ts)
- return -EINVAL;
-
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
irq_reg = rd32(hw, PFINT_OICR_ENA);
- if (ena) {
+ if (config->ena) {
/* Enable the interrupt */
irq_reg |= PFINT_OICR_TSYN_EVNT_M;
aux_reg = GLTSYN_AUX_IN_0_INT_ENA_M;
@@ -1632,9 +1629,9 @@ ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
#define GLTSYN_AUX_IN_0_EVNTLVL_FALLING_EDGE BIT(1)
/* set event level to requested edge */
- if (extts_flags & PTP_FALLING_EDGE)
+ if (config->flags & PTP_FALLING_EDGE)
aux_reg |= GLTSYN_AUX_IN_0_EVNTLVL_FALLING_EDGE;
- if (extts_flags & PTP_RISING_EDGE)
+ if (config->flags & PTP_RISING_EDGE)
aux_reg |= GLTSYN_AUX_IN_0_EVNTLVL_RISING_EDGE;
/* Write GPIO CTL reg.
@@ -1656,9 +1653,47 @@ ice_ptp_cfg_extts(struct ice_pf *pf, bool ena, unsigned int chan, u32 gpio_pin,
wr32(hw, PFINT_OICR_ENA, irq_reg);
wr32(hw, GLTSYN_AUX_IN(chan, tmr_idx), aux_reg);
- wr32(hw, GLGEN_GPIO_CTL(gpio_pin), gpio_reg);
+ wr32(hw, GLGEN_GPIO_CTL(config->gpio_pin), gpio_reg);
- return 0;
+ if (store)
+ memcpy(&pf->ptp.extts_channels[chan], config, sizeof(*config));
+}
+
+/**
+ * ice_ptp_disable_all_extts - Disable all EXTTS channels
+ * @pf: Board private structure
+ */
+static void ice_ptp_disable_all_extts(struct ice_pf *pf)
+{
+ struct ice_extts_channel extts_cfg = {};
+ int i;
+
+ for (i = 0; i < pf->ptp.info.n_ext_ts; i++) {
+ if (pf->ptp.extts_channels[i].ena) {
+ extts_cfg.gpio_pin = pf->ptp.extts_channels[i].gpio_pin;
+ extts_cfg.ena = false;
+ ice_ptp_cfg_extts(pf, i, &extts_cfg, false);
+ }
+ }
+
+ synchronize_irq(pf->oicr_irq.virq);
+}
+
+/**
+ * ice_ptp_enable_all_extts - Enable all EXTTS channels
+ * @pf: Board private structure
+ *
+ * Called during reset to restore user configuration.
+ */
+static void ice_ptp_enable_all_extts(struct ice_pf *pf)
+{
+ int i;
+
+ for (i = 0; i < pf->ptp.info.n_ext_ts; i++) {
+ if (pf->ptp.extts_channels[i].ena)
+ ice_ptp_cfg_extts(pf, i, &pf->ptp.extts_channels[i],
+ false);
+ }
}
/**
@@ -1815,7 +1850,6 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
struct ptp_clock_request *rq, int on)
{
struct ice_pf *pf = ptp_info_to_pf(info);
- struct ice_perout_channel clk_cfg = {0};
bool sma_pres = false;
unsigned int chan;
u32 gpio_pin;
@@ -1826,6 +1860,9 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
switch (rq->type) {
case PTP_CLK_REQ_PEROUT:
+ {
+ struct ice_perout_channel clk_cfg = {};
+
chan = rq->perout.index;
if (sma_pres) {
if (chan == ice_pin_desc_e810t[SMA1].chan)
@@ -1853,7 +1890,11 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
err = ice_ptp_cfg_clkout(pf, chan, &clk_cfg, true);
break;
+ }
case PTP_CLK_REQ_EXTTS:
+ {
+ struct ice_extts_channel extts_cfg = {};
+
chan = rq->extts.index;
if (sma_pres) {
if (chan < ice_pin_desc_e810t[SMA2].chan)
@@ -1869,9 +1910,13 @@ ice_ptp_gpio_enable_e810(struct ptp_clock_info *info,
gpio_pin = chan;
}
- err = ice_ptp_cfg_extts(pf, !!on, chan, gpio_pin,
- rq->extts.flags);
- break;
+ extts_cfg.flags = rq->extts.flags;
+ extts_cfg.gpio_pin = gpio_pin;
+ extts_cfg.ena = !!on;
+
+ ice_ptp_cfg_extts(pf, chan, &extts_cfg, true);
+ return 0;
+ }
default:
return -EOPNOTSUPP;
}
@@ -1889,21 +1934,31 @@ static int ice_ptp_gpio_enable_e823(struct ptp_clock_info *info,
struct ptp_clock_request *rq, int on)
{
struct ice_pf *pf = ptp_info_to_pf(info);
- struct ice_perout_channel clk_cfg = {0};
int err;
switch (rq->type) {
case PTP_CLK_REQ_PPS:
+ {
+ struct ice_perout_channel clk_cfg = {};
+
clk_cfg.gpio_pin = PPS_PIN_INDEX;
clk_cfg.period = NSEC_PER_SEC;
clk_cfg.ena = !!on;
err = ice_ptp_cfg_clkout(pf, PPS_CLK_GEN_CHAN, &clk_cfg, true);
break;
+ }
case PTP_CLK_REQ_EXTTS:
- err = ice_ptp_cfg_extts(pf, !!on, rq->extts.index,
- TIME_SYNC_PIN_INDEX, rq->extts.flags);
- break;
+ {
+ struct ice_extts_channel extts_cfg = {};
+
+ extts_cfg.flags = rq->extts.flags;
+ extts_cfg.gpio_pin = TIME_SYNC_PIN_INDEX;
+ extts_cfg.ena = !!on;
+
+ ice_ptp_cfg_extts(pf, rq->extts.index, &extts_cfg, true);
+ return 0;
+ }
default:
return -EOPNOTSUPP;
}
@@ -2735,6 +2790,10 @@ static int ice_ptp_rebuild_owner(struct ice_pf *pf)
ice_ptp_restart_all_phy(pf);
}
+ /* Re-enable all periodic outputs and external timestamp events */
+ ice_ptp_enable_all_clkout(pf);
+ ice_ptp_enable_all_extts(pf);
+
return 0;
}
@@ -3286,6 +3345,8 @@ void ice_ptp_release(struct ice_pf *pf)
ice_ptp_release_tx_tracker(pf, &pf->ptp.port.tx);
+ ice_ptp_disable_all_extts(pf);
+
kthread_cancel_delayed_work_sync(&pf->ptp.work);
ice_ptp_port_phy_stop(&pf->ptp.port);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.h b/drivers/net/ethernet/intel/ice/ice_ptp.h
index 352405a2daf2..c6469a5a7afb 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.h
@@ -33,6 +33,12 @@ struct ice_perout_channel {
u64 start_time;
};
+struct ice_extts_channel {
+ bool ena;
+ u32 gpio_pin;
+ u32 flags;
+};
+
/* The ice hardware captures Tx hardware timestamps in the PHY. The timestamp
* is stored in a buffer of registers. Depending on the specific hardware,
* this buffer might be shared across multiple PHY ports.
@@ -226,6 +232,7 @@ enum ice_ptp_state {
* @ext_ts_irq: the external timestamp IRQ in use
* @kworker: kwork thread for handling periodic work
* @perout_channels: periodic output data
+ * @extts_channels: channels for external timestamps
* @info: structure defining PTP hardware capabilities
* @clock: pointer to registered PTP clock device
* @tstamp_config: hardware timestamping configuration
@@ -249,6 +256,7 @@ struct ice_ptp {
u8 ext_ts_irq;
struct kthread_worker *kworker;
struct ice_perout_channel perout_channels[GLTSYN_TGT_H_IDX_MAX];
+ struct ice_extts_channel extts_channels[GLTSYN_TGT_H_IDX_MAX];
struct ptp_clock_info info;
struct ptp_clock *clock;
struct hwtstamp_config tstamp_config;
--
2.43.0

49
kernel-std/debian/patches/ice-VDF/0035-ice-Don-t-process-extts-if-PTP-is-disabled.patch Normal file
View File

@ -0,0 +1,49 @@
From 6c24a32820031f9713d0c0cf7ac6f4ead6b58052 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 2 Jul 2024 10:14:55 -0700
Subject: [PATCH 35/36] ice: Don't process extts if PTP is disabled
The ice_ptp_extts_event() function can race with ice_ptp_release() and
result in a NULL pointer dereference which leads to a kernel panic.
Panic occurs because the ice_ptp_extts_event() function calls
ptp_clock_event() with a NULL pointer. The ice driver has already
released the PTP clock by the time the interrupt for the next external
timestamp event occurs.
To fix this, modify the ice_ptp_extts_event() function to check the
PTP state and bail early if PTP is not ready.
Fixes: 172db5f91d5f ("ice: add support for auxiliary input/output pins")
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://patch.msgid.link/20240702171459.2606611-3-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 996422e3230e41468f652d754fefd1bdbcd4604e)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index 6e06c5d596b9..ceb4ba19c511 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -1578,6 +1578,10 @@ void ice_ptp_extts_event(struct ice_pf *pf)
u8 chan, tmr_idx;
u32 hi, lo;
+ /* Don't process timestamp events if PTP is not ready */
+ if (pf->ptp.state != ICE_PTP_READY)
+ return;
+
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
/* Event time is captured by one of the two matched registers
* GLTSYN_EVNT_L: 32 LSB of sampled time event
--
2.43.0

628
kernel-std/debian/patches/ice-VDF/0036-ice-Introduce-ice_ptp_hw-struct.patch Normal file
View File

@ -0,0 +1,628 @@
From 1ce01cb7cdb0bf4c18a546a62f224c8032d75ebd Mon Sep 17 00:00:00 2001
From: Karol Kolacinski <karol.kolacinski@intel.com>
Date: Tue, 28 May 2024 16:03:51 -0700
Subject: [PATCH 36/36] ice: Introduce ice_ptp_hw struct
Create new ice_ptp_hw struct and use it for all HW and PTP-related
fields from struct ice_hw.
Replace definitions with struct fields, which values are set accordingly
to a specific device.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528-next-2024-05-28-ptp-refactors-v1-1-c082739bb6f6@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit d551d075b043821880b8afc0010ef70d050716d0)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_common.c | 24 ++++
drivers/net/ethernet/intel/ice/ice_common.h | 1 +
drivers/net/ethernet/intel/ice/ice_ptp.c | 22 ++--
drivers/net/ethernet/intel/ice/ice_ptp_hw.c | 134 ++++++++++++--------
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 4 +-
drivers/net/ethernet/intel/ice/ice_type.h | 17 ++-
6 files changed, 126 insertions(+), 76 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c
index 59ede77a1473..147004e0170b 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.c
+++ b/drivers/net/ethernet/intel/ice/ice_common.c
@@ -209,6 +209,30 @@ bool ice_is_e810t(struct ice_hw *hw)
return false;
}
+/**
+ * ice_is_e822 - Check if a device is E822 family device
+ * @hw: pointer to the hardware structure
+ *
+ * Return: true if the device is E822 based, false if not.
+ */
+bool ice_is_e822(struct ice_hw *hw)
+{
+ switch (hw->device_id) {
+ case ICE_DEV_ID_E822C_BACKPLANE:
+ case ICE_DEV_ID_E822C_QSFP:
+ case ICE_DEV_ID_E822C_SFP:
+ case ICE_DEV_ID_E822C_10G_BASE_T:
+ case ICE_DEV_ID_E822C_SGMII:
+ case ICE_DEV_ID_E822L_BACKPLANE:
+ case ICE_DEV_ID_E822L_SFP:
+ case ICE_DEV_ID_E822L_10G_BASE_T:
+ case ICE_DEV_ID_E822L_SGMII:
+ return true;
+ default:
+ return false;
+ }
+}
+
/**
* ice_is_e823
* @hw: pointer to the hardware structure
diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h
index 1c3c29d30815..9d38777310e5 100644
--- a/drivers/net/ethernet/intel/ice/ice_common.h
+++ b/drivers/net/ethernet/intel/ice/ice_common.h
@@ -245,6 +245,7 @@ void
ice_stat_update32(struct ice_hw *hw, u32 reg, bool prev_stat_loaded,
u64 *prev_stat, u64 *cur_stat);
bool ice_is_e810t(struct ice_hw *hw);
+bool ice_is_e822(struct ice_hw *hw);
bool ice_is_e823(struct ice_hw *hw);
int
ice_sched_query_elem(struct ice_hw *hw, u32 node_teid,
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index ceb4ba19c511..bb1572a353d0 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -812,7 +812,7 @@ static enum ice_tx_tstamp_work ice_ptp_tx_tstamp_owner(struct ice_pf *pf)
}
mutex_unlock(&pf->ptp.ports_owner.lock);
- for (i = 0; i < ICE_MAX_QUAD; i++) {
+ for (i = 0; i < ICE_GET_QUAD_NUM(pf->hw.ptp.num_lports); i++) {
u64 tstamp_ready;
int err;
@@ -1026,7 +1026,7 @@ ice_ptp_release_tx_tracker(struct ice_pf *pf, struct ice_ptp_tx *tx)
static int
ice_ptp_init_tx_e82x(struct ice_pf *pf, struct ice_ptp_tx *tx, u8 port)
{
- tx->block = port / ICE_PORTS_PER_QUAD;
+ tx->block = ICE_GET_QUAD_NUM(port);
tx->offset = (port % ICE_PORTS_PER_QUAD) * INDEX_PER_PORT_E82X;
tx->len = INDEX_PER_PORT_E82X;
tx->has_ready_bitmap = 1;
@@ -1248,8 +1248,8 @@ static u64 ice_base_incval(struct ice_pf *pf)
*/
static int ice_ptp_check_tx_fifo(struct ice_ptp_port *port)
{
- int quad = port->port_num / ICE_PORTS_PER_QUAD;
int offs = port->port_num % ICE_PORTS_PER_QUAD;
+ int quad = ICE_GET_QUAD_NUM(port->port_num);
struct ice_pf *pf;
struct ice_hw *hw;
u32 val, phy_sts;
@@ -1448,7 +1448,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
if (pf->ptp.state != ICE_PTP_READY)
return;
- if (WARN_ON_ONCE(port >= ICE_NUM_EXTERNAL_PORTS))
+ if (WARN_ON_ONCE(port >= hw->ptp.num_lports))
return;
ptp_port = &pf->ptp.port;
@@ -1458,7 +1458,7 @@ void ice_ptp_link_change(struct ice_pf *pf, u8 port, bool linkup)
/* Update cached link status for this port immediately */
ptp_port->link_up = linkup;
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
/* Do not reconfigure E810 PHY */
return;
@@ -1487,7 +1487,7 @@ static int ice_ptp_cfg_phy_interrupt(struct ice_pf *pf, bool ena, u32 threshold)
ice_ptp_reset_ts_memory(hw);
- for (quad = 0; quad < ICE_MAX_QUAD; quad++) {
+ for (quad = 0; quad < ICE_GET_QUAD_NUM(hw->ptp.num_lports); quad++) {
err = ice_read_quad_reg_e82x(hw, quad, Q_REG_TX_MEM_GBL_CFG,
&val);
if (err)
@@ -2038,7 +2038,7 @@ ice_ptp_settime64(struct ptp_clock_info *info, const struct timespec64 *ts)
ice_ptp_enable_all_clkout(pf);
/* Recalibrate and re-enable timestamp blocks for E822/E823 */
- if (hw->phy_model == ICE_PHY_E82X)
+ if (hw->ptp.phy_model == ICE_PHY_E82X)
ice_ptp_restart_all_phy(pf);
exit:
if (err) {
@@ -2652,7 +2652,7 @@ static void ice_ptp_maybe_trigger_tx_interrupt(struct ice_pf *pf)
if (!ice_pf_src_tmr_owned(pf))
return;
- for (i = 0; i < ICE_MAX_QUAD; i++) {
+ for (i = 0; i < ICE_GET_QUAD_NUM(hw->ptp.num_lports); i++) {
u64 tstamp_ready;
int err;
@@ -3152,7 +3152,7 @@ static int ice_ptp_init_port(struct ice_pf *pf, struct ice_ptp_port *ptp_port)
mutex_init(&ptp_port->ps_lock);
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_ptp_init_tx_e810(pf, &ptp_port->tx);
case ICE_PHY_E82X:
@@ -3245,7 +3245,7 @@ static void ice_ptp_remove_auxbus_device(struct ice_pf *pf)
*/
static void ice_ptp_init_tx_interrupt_mode(struct ice_pf *pf)
{
- switch (pf->hw.phy_model) {
+ switch (pf->hw.ptp.phy_model) {
case ICE_PHY_E82X:
/* E822 based PHY has the clock owner process the interrupt
* for all ports.
@@ -3281,7 +3281,7 @@ void ice_ptp_init(struct ice_pf *pf)
ptp->state = ICE_PTP_INITIALIZING;
- ice_ptp_init_phy_model(hw);
+ ice_ptp_init_hw(hw);
ice_ptp_init_tx_interrupt_mode(pf);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
index 7337e7e710ed..313a72dad813 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.c
@@ -285,18 +285,21 @@ static void ice_ptp_exec_tmr_cmd(struct ice_hw *hw)
/**
* ice_fill_phy_msg_e82x - Fill message data for a PHY register access
+ * @hw: pointer to the HW struct
* @msg: the PHY message buffer to fill in
* @port: the port to access
* @offset: the register offset
*/
-static void
-ice_fill_phy_msg_e82x(struct ice_sbq_msg_input *msg, u8 port, u16 offset)
+static void ice_fill_phy_msg_e82x(struct ice_hw *hw,
+ struct ice_sbq_msg_input *msg, u8 port,
+ u16 offset)
{
int phy_port, phy, quadtype;
- phy_port = port % ICE_PORTS_PER_PHY_E82X;
- phy = port / ICE_PORTS_PER_PHY_E82X;
- quadtype = (port / ICE_PORTS_PER_QUAD) % ICE_QUADS_PER_PHY_E82X;
+ phy_port = port % hw->ptp.ports_per_phy;
+ phy = port / hw->ptp.ports_per_phy;
+ quadtype = ICE_GET_QUAD_NUM(port) %
+ ICE_GET_QUAD_NUM(hw->ptp.ports_per_phy);
if (quadtype == 0) {
msg->msg_addr_low = P_Q0_L(P_0_BASE + offset, phy_port);
@@ -427,7 +430,7 @@ ice_read_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- ice_fill_phy_msg_e82x(&msg, port, offset);
+ ice_fill_phy_msg_e82x(hw, &msg, port, offset);
msg.opcode = ice_sbq_msg_rd;
err = ice_sbq_rw_reg(hw, &msg);
@@ -504,7 +507,7 @@ ice_write_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- ice_fill_phy_msg_e82x(&msg, port, offset);
+ ice_fill_phy_msg_e82x(hw, &msg, port, offset);
msg.opcode = ice_sbq_msg_wr;
msg.data = val;
@@ -614,24 +617,30 @@ ice_write_64b_phy_reg_e82x(struct ice_hw *hw, u8 port, u16 low_addr, u64 val)
/**
* ice_fill_quad_msg_e82x - Fill message data for quad register access
+ * @hw: pointer to the HW struct
* @msg: the PHY message buffer to fill in
* @quad: the quad to access
* @offset: the register offset
*
* Fill a message buffer for accessing a register in a quad shared between
* multiple PHYs.
+ *
+ * Return:
+ * * %0 - OK
+ * * %-EINVAL - invalid quad number
*/
-static int
-ice_fill_quad_msg_e82x(struct ice_sbq_msg_input *msg, u8 quad, u16 offset)
+static int ice_fill_quad_msg_e82x(struct ice_hw *hw,
+ struct ice_sbq_msg_input *msg, u8 quad,
+ u16 offset)
{
u32 addr;
- if (quad >= ICE_MAX_QUAD)
+ if (quad >= ICE_GET_QUAD_NUM(hw->ptp.num_lports))
return -EINVAL;
msg->dest_dev = rmn_0;
- if ((quad % ICE_QUADS_PER_PHY_E82X) == 0)
+ if (!(quad % ICE_GET_QUAD_NUM(hw->ptp.ports_per_phy)))
addr = Q_0_BASE + offset;
else
addr = Q_1_BASE + offset;
@@ -658,7 +667,7 @@ ice_read_quad_reg_e82x(struct ice_hw *hw, u8 quad, u16 offset, u32 *val)
struct ice_sbq_msg_input msg = {0};
int err;
- err = ice_fill_quad_msg_e82x(&msg, quad, offset);
+ err = ice_fill_quad_msg_e82x(hw, &msg, quad, offset);
if (err)
return err;
@@ -692,7 +701,7 @@ ice_write_quad_reg_e82x(struct ice_hw *hw, u8 quad, u16 offset, u32 val)
struct ice_sbq_msg_input msg = {0};
int err;
- err = ice_fill_quad_msg_e82x(&msg, quad, offset);
+ err = ice_fill_quad_msg_e82x(hw, &msg, quad, offset);
if (err)
return err;
@@ -813,7 +822,7 @@ static void ice_ptp_reset_ts_memory_e82x(struct ice_hw *hw)
{
unsigned int quad;
- for (quad = 0; quad < ICE_MAX_QUAD; quad++)
+ for (quad = 0; quad < ICE_GET_QUAD_NUM(hw->ptp.num_lports); quad++)
ice_ptp_reset_ts_memory_quad_e82x(hw, quad);
}
@@ -1110,7 +1119,7 @@ static int ice_ptp_set_vernier_wl(struct ice_hw *hw)
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_write_phy_reg_e82x(hw, port, P_REG_WL,
@@ -1175,7 +1184,7 @@ ice_ptp_prep_phy_time_e82x(struct ice_hw *hw, u32 time)
*/
phy_time = (u64)time << 32;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
/* Tx case */
err = ice_write_64b_phy_reg_e82x(hw, port,
P_REG_TX_TIMER_INC_PRE_L,
@@ -1278,7 +1287,7 @@ ice_ptp_prep_phy_adj_e82x(struct ice_hw *hw, s32 adj)
else
cycles = -(((s64)-adj) << 32);
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_ptp_prep_port_adj_e82x(hw, port, cycles);
@@ -1304,7 +1313,7 @@ ice_ptp_prep_phy_incval_e82x(struct ice_hw *hw, u64 incval)
int err;
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
err = ice_write_40b_phy_reg_e82x(hw, port, P_REG_TIMETUS_L,
incval);
if (err)
@@ -1460,7 +1469,7 @@ ice_ptp_one_port_cmd(struct ice_hw *hw, u8 configured_port,
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
enum ice_ptp_tmr_cmd cmd;
int err;
@@ -1490,7 +1499,7 @@ ice_ptp_port_cmd_e82x(struct ice_hw *hw, enum ice_ptp_tmr_cmd cmd)
{
u8 port;
- for (port = 0; port < ICE_NUM_EXTERNAL_PORTS; port++) {
+ for (port = 0; port < hw->ptp.num_lports; port++) {
int err;
err = ice_ptp_write_port_cmd_e82x(hw, port, cmd);
@@ -1603,7 +1612,7 @@ static void ice_phy_cfg_lane_e82x(struct ice_hw *hw, u8 port)
return;
}
- quad = port / ICE_PORTS_PER_QUAD;
+ quad = ICE_GET_QUAD_NUM(port);
err = ice_read_quad_reg_e82x(hw, quad, Q_REG_TX_MEM_GBL_CFG, &val);
if (err) {
@@ -2632,6 +2641,17 @@ ice_get_phy_tx_tstamp_ready_e82x(struct ice_hw *hw, u8 quad, u64 *tstamp_ready)
return 0;
}
+/**
+ * ice_ptp_init_phy_e82x - initialize PHY parameters
+ * @ptp: pointer to the PTP HW struct
+ */
+static void ice_ptp_init_phy_e82x(struct ice_ptp_hw *ptp)
+{
+ ptp->phy_model = ICE_PHY_E82X;
+ ptp->num_lports = 8;
+ ptp->ports_per_phy = 8;
+}
+
/* E810 functions
*
* The following functions operate on the E810 series devices which use
@@ -2859,17 +2879,21 @@ static int ice_clear_phy_tstamp_e810(struct ice_hw *hw, u8 lport, u8 idx)
}
/**
- * ice_ptp_init_phy_e810 - Enable PTP function on the external PHY
+ * ice_ptp_init_phc_e810 - Perform E810 specific PHC initialization
* @hw: pointer to HW struct
*
- * Enable the timesync PTP functionality for the external PHY connected to
- * this function.
+ * Perform E810-specific PTP hardware clock initialization steps.
+ *
+ * Return: 0 on success, other error codes when failed to initialize TimeSync
*/
-int ice_ptp_init_phy_e810(struct ice_hw *hw)
+static int ice_ptp_init_phc_e810(struct ice_hw *hw)
{
u8 tmr_idx;
int err;
+ /* Ensure synchronization delay is zero */
+ wr32(hw, GLTSYN_SYNC_DLAY, 0);
+
tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned;
err = ice_write_phy_reg_e810(hw, ETH_GLTSYN_ENA(tmr_idx),
GLTSYN_ENA_TSYN_ENA_M);
@@ -2880,21 +2904,6 @@ int ice_ptp_init_phy_e810(struct ice_hw *hw)
return err;
}
-/**
- * ice_ptp_init_phc_e810 - Perform E810 specific PHC initialization
- * @hw: pointer to HW struct
- *
- * Perform E810-specific PTP hardware clock initialization steps.
- */
-static int ice_ptp_init_phc_e810(struct ice_hw *hw)
-{
- /* Ensure synchronization delay is zero */
- wr32(hw, GLTSYN_SYNC_DLAY, 0);
-
- /* Initialize the PHY */
- return ice_ptp_init_phy_e810(hw);
-}
-
/**
* ice_ptp_prep_phy_time_e810 - Prepare PHY port with initial time
* @hw: Board private structure
@@ -3238,6 +3247,17 @@ int ice_read_pca9575_reg_e810t(struct ice_hw *hw, u8 offset, u8 *data)
return ice_aq_read_i2c(hw, link_topo, 0, addr, 1, data, NULL);
}
+/**
+ * ice_ptp_init_phy_e810 - initialize PHY parameters
+ * @ptp: pointer to the PTP HW struct
+ */
+static void ice_ptp_init_phy_e810(struct ice_ptp_hw *ptp)
+{
+ ptp->phy_model = ICE_PHY_E810;
+ ptp->num_lports = 8;
+ ptp->ports_per_phy = 4;
+}
+
/* Device agnostic functions
*
* The following functions implement shared behavior common to both E822 and
@@ -3295,18 +3315,22 @@ void ice_ptp_unlock(struct ice_hw *hw)
}
/**
- * ice_ptp_init_phy_model - Initialize hw->phy_model based on device type
+ * ice_ptp_init_hw - Initialize hw based on device type
* @hw: pointer to the HW structure
*
- * Determine the PHY model for the device, and initialize hw->phy_model
+ * Determine the PHY model for the device, and initialize hw
* for use by other functions.
*/
-void ice_ptp_init_phy_model(struct ice_hw *hw)
+void ice_ptp_init_hw(struct ice_hw *hw)
{
- if (ice_is_e810(hw))
- hw->phy_model = ICE_PHY_E810;
+ struct ice_ptp_hw *ptp = &hw->ptp;
+
+ if (ice_is_e822(hw) || ice_is_e823(hw))
+ ice_ptp_init_phy_e82x(ptp);
+ else if (ice_is_e810(hw))
+ ice_ptp_init_phy_e810(ptp);
else
- hw->phy_model = ICE_PHY_E82X;
+ ptp->phy_model = ICE_PHY_UNSUP;
}
/**
@@ -3327,7 +3351,7 @@ static int ice_ptp_tmr_cmd(struct ice_hw *hw, enum ice_ptp_tmr_cmd cmd)
ice_ptp_src_cmd(hw, cmd);
/* Next, prepare the ports */
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_port_cmd_e810(hw, cmd);
break;
@@ -3379,7 +3403,7 @@ int ice_ptp_init_time(struct ice_hw *hw, u64 time)
/* PHY timers */
/* Fill Rx and Tx ports and send msg to PHY */
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_time_e810(hw, time & 0xFFFFFFFF);
break;
@@ -3421,7 +3445,7 @@ int ice_ptp_write_incval(struct ice_hw *hw, u64 incval)
wr32(hw, GLTSYN_SHADJ_L(tmr_idx), lower_32_bits(incval));
wr32(hw, GLTSYN_SHADJ_H(tmr_idx), upper_32_bits(incval));
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_incval_e810(hw, incval);
break;
@@ -3487,7 +3511,7 @@ int ice_ptp_adj_clock(struct ice_hw *hw, s32 adj)
wr32(hw, GLTSYN_SHADJ_L(tmr_idx), 0);
wr32(hw, GLTSYN_SHADJ_H(tmr_idx), adj);
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
err = ice_ptp_prep_phy_adj_e810(hw, adj);
break;
@@ -3517,7 +3541,7 @@ int ice_ptp_adj_clock(struct ice_hw *hw, s32 adj)
*/
int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_read_phy_tstamp_e810(hw, block, idx, tstamp);
case ICE_PHY_E82X:
@@ -3545,7 +3569,7 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp)
*/
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_clear_phy_tstamp_e810(hw, block, idx);
case ICE_PHY_E82X:
@@ -3606,7 +3630,7 @@ int ice_get_pf_c827_idx(struct ice_hw *hw, u8 *idx)
*/
void ice_ptp_reset_ts_memory(struct ice_hw *hw)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E82X:
ice_ptp_reset_ts_memory_e82x(hw);
break;
@@ -3632,7 +3656,7 @@ int ice_ptp_init_phc(struct ice_hw *hw)
/* Clear event err indications for auxiliary pins */
(void)rd32(hw, GLTSYN_STAT(src_idx));
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_ptp_init_phc_e810(hw);
case ICE_PHY_E82X:
@@ -3655,7 +3679,7 @@ int ice_ptp_init_phc(struct ice_hw *hw)
*/
int ice_get_phy_tx_tstamp_ready(struct ice_hw *hw, u8 block, u64 *tstamp_ready)
{
- switch (hw->phy_model) {
+ switch (hw->ptp.phy_model) {
case ICE_PHY_E810:
return ice_get_phy_tx_tstamp_ready_e810(hw, block,
tstamp_ready);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index 7e8fd369ef7c..d788221eba57 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -211,6 +211,7 @@ int ice_read_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx, u64 *tstamp);
int ice_clear_phy_tstamp(struct ice_hw *hw, u8 block, u8 idx);
void ice_ptp_reset_ts_memory(struct ice_hw *hw);
int ice_ptp_init_phc(struct ice_hw *hw);
+void ice_ptp_init_hw(struct ice_hw *hw);
int ice_get_phy_tx_tstamp_ready(struct ice_hw *hw, u8 block, u64 *tstamp_ready);
/* E822 family functions */
@@ -265,7 +266,6 @@ int ice_phy_cfg_tx_offset_e82x(struct ice_hw *hw, u8 port);
int ice_phy_cfg_rx_offset_e82x(struct ice_hw *hw, u8 port);
/* E810 family functions */
-int ice_ptp_init_phy_e810(struct ice_hw *hw);
int ice_read_sma_ctrl_e810t(struct ice_hw *hw, u8 *data);
int ice_write_sma_ctrl_e810t(struct ice_hw *hw, u8 data);
int ice_read_pca9575_reg_e810t(struct ice_hw *hw, u8 offset, u8 *data);
@@ -280,8 +280,6 @@ int ice_get_cgu_state(struct ice_hw *hw, u8 dpll_idx,
u8 *ref_state, u8 *eec_mode, s64 *phase_offset,
enum dpll_lock_status *dpll_state);
int ice_get_cgu_rclk_pin_info(struct ice_hw *hw, u8 *base_idx, u8 *pin_num);
-
-void ice_ptp_init_phy_model(struct ice_hw *hw);
int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
unsigned long *caps);
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 28e47bb78eaf..6fc4cd1030d0 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -807,6 +807,9 @@ struct ice_mbx_data {
u16 async_watermark_val;
};
+#define ICE_PORTS_PER_QUAD 4
+#define ICE_GET_QUAD_NUM(port) ((port) / ICE_PORTS_PER_QUAD)
+
/* PHY model */
enum ice_phy_model {
ICE_PHY_UNSUP = -1,
@@ -814,6 +817,12 @@ enum ice_phy_model {
ICE_PHY_E82X,
};
+struct ice_ptp_hw {
+ enum ice_phy_model phy_model;
+ u8 num_lports;
+ u8 ports_per_phy;
+};
+
/* Port hardware description */
struct ice_hw {
u8 __iomem *hw_addr;
@@ -835,7 +844,6 @@ struct ice_hw {
u8 revision_id;
u8 pf_id; /* device profile info */
- enum ice_phy_model phy_model;
u16 max_burst_size; /* driver sets this value */
@@ -896,12 +904,7 @@ struct ice_hw {
/* INTRL granularity in 1 us */
u8 intrl_gran;
-#define ICE_MAX_QUAD 2
-#define ICE_QUADS_PER_PHY_E82X 2
-#define ICE_PORTS_PER_PHY_E82X 8
-#define ICE_PORTS_PER_QUAD 4
-#define ICE_PORTS_PER_PHY_E810 4
-#define ICE_NUM_EXTERNAL_PORTS (ICE_MAX_QUAD * ICE_PORTS_PER_QUAD)
+ struct ice_ptp_hw ptp;
/* Active package version (currently active) */
struct ice_pkg_ver active_pkg_ver;
--
2.43.0

80
kernel-std/debian/patches/ice-VDF/0037-ice-Introduce-ice_get_base_incval-helper.patch Normal file
View File

@ -0,0 +1,80 @@
From cd12b5c8239993e395436ff9a01b524103aa0641 Mon Sep 17 00:00:00 2001
From: Jacob Keller <jacob.e.keller@intel.com>
Date: Tue, 28 May 2024 16:03:56 -0700
Subject: [PATCH] ice: Introduce ice_get_base_incval() helper
Add a new helper for getting base clock increment value for specific HW.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Reviewed-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Karol Kolacinski <karol.kolacinski@intel.com>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20240528-next-2024-05-28-ptp-refactors-v1-6-c082739bb6f6@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
(cherry picked from commit 1f374d57c39386520586539641cafc999d0f3ef5)
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
drivers/net/ethernet/intel/ice/ice_ptp.c | 9 +--------
drivers/net/ethernet/intel/ice/ice_ptp_hw.h | 18 ++++++++++++++++++
2 files changed, 19 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp.c b/drivers/net/ethernet/intel/ice/ice_ptp.c
index bb1572a353d0..44b8fc8021cd 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp.c
+++ b/drivers/net/ethernet/intel/ice/ice_ptp.c
@@ -7,8 +7,6 @@
#define E810_OUT_PROP_DELAY_NS 1
-#define UNKNOWN_INCVAL_E82X 0x100000000ULL
-
static const struct ptp_pin_desc ice_pin_desc_e810t[] = {
/* name idx func chan */
{ "GNSS", GNSS, PTP_PF_EXTTS, 0, { 0, } },
@@ -1229,12 +1227,7 @@ static u64 ice_base_incval(struct ice_pf *pf)
struct ice_hw *hw = &pf->hw;
u64 incval;
- if (ice_is_e810(hw))
- incval = ICE_PTP_NOMINAL_INCVAL_E810;
- else if (ice_e82x_time_ref(hw) < NUM_ICE_TIME_REF_FREQ)
- incval = ice_e82x_nominal_incval(ice_e82x_time_ref(hw));
- else
- incval = UNKNOWN_INCVAL_E82X;
+ incval = ice_get_base_incval(hw);
dev_dbg(ice_pf_to_dev(pf), "PTP: using base increment value of 0x%016llx\n",
incval);
diff --git a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
index d788221eba57..749a3f2d8293 100644
--- a/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
+++ b/drivers/net/ethernet/intel/ice/ice_ptp_hw.h
@@ -283,6 +283,24 @@ int ice_get_cgu_rclk_pin_info(struct ice_hw *hw, u8 *base_idx, u8 *pin_num);
int ice_cgu_get_output_pin_state_caps(struct ice_hw *hw, u8 pin_id,
unsigned long *caps);
+/**
+ * ice_get_base_incval - Get base clock increment value
+ * @hw: pointer to the HW struct
+ *
+ * Return: base clock increment value for supported PHYs, 0 otherwise
+ */
+static inline u64 ice_get_base_incval(struct ice_hw *hw)
+{
+ switch (hw->ptp.phy_model) {
+ case ICE_PHY_E810:
+ return ICE_PTP_NOMINAL_INCVAL_E810;
+ case ICE_PHY_E82X:
+ return ice_e82x_nominal_incval(ice_e82x_time_ref(hw));
+ default:
+ return 0;
+ }
+}
+
#define PFTSYN_SEM_BYTES 4
#define ICE_PTP_CLOCK_INDEX_0 0x00
--
2.43.0

32
kernel-std/debian/patches/ice-VDF/0038-ice-modify-the-ice-driver-version-to-stx.4.patch Normal file
View File

@ -0,0 +1,32 @@
From 437206483113743a4ef40c2f7e14f09705049672 Mon Sep 17 00:00:00 2001
From: Jiping Ma <jiping.ma2@windriver.com>
Date: Mon, 2 Sep 2024 03:18:08 +0000
Subject: [PATCH] ice:modify the ice driver version to stx.4
Change the ice driver min version to stx.4 because we back ported
the upstream 36 commits to our code base to support the customer's
requirement.
The ice driver version should be ice-6.6.40-stx.4.
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Makefile b/Makefile
index 5807b310bdca..7163d25405f8 100644
--- a/Makefile
+++ b/Makefile
@@ -1227,7 +1227,7 @@ uapi-asm-generic:
# KERNELRELEASE can change from a few different places, meaning version.h
# needs to be updated, so this check is forced on all builds
-ICE_STX = "-stx.3"
+ICE_STX = "-stx.4"
I40E_STX = "-stx.0"
IAVF_STX = "-stx.0"
--
2.43.0

38
kernel-std/debian/patches/series
View File

@ -64,3 +64,41 @@ ice-dpll/0046-dpll-fix-dpll_xa_ref_-_del-for-multiple-registration.patch
ice-dpll/0047-ice-modify-the-ice-driver-min-version-to-stx.2.patch
ice-mdd/0001-ice-Add-automatic-VF-reset-on-Tx-MDD-events.patch
ice-mdd/0002-ice-modify-the-ice-driver-min-version-to-stx.3.patch
ice-VDF/0001-ice-Auxbus-devices-driver-for-E822-TS.patch
ice-VDF/0002-ice-introduce-ice_pf_src_tmr_owned.patch
ice-VDF/0003-ice-Re-enable-timestamping-correctly-after-reset.patch
ice-VDF/0004-ice-periodically-kick-Tx-timestamp-interrupt.patch
ice-VDF/0005-ice-PTP-Rename-macros-used-for-PHY-QUAD-port-definit.patch
ice-VDF/0006-ice-PTP-move-quad-value-check-inside-ice_fill_phy_ms.patch
ice-VDF/0007-ice-remove-ptp_tx-ring-parameter-flag.patch
ice-VDF/0008-ice-unify-logic-for-programming-PFINT_TSYN_MSK.patch
ice-VDF/0009-ice-PTP-Clean-up-timestamp-registers-correctly.patch
ice-VDF/0010-ice-Use-PTP-auxbus-for-all-PHYs-restart-in-E822.patch
ice-VDF/0011-ice-Rename-E822-to-E82X.patch
ice-VDF/0012-ice-Schedule-service-task-in-IRQ-top-half.patch
ice-VDF/0013-ice-Enable-SW-interrupt-from-FW-for-LL-TS.patch
ice-VDF/0014-ice-PTP-add-clock-domain-number-to-auxiliary-interfa.patch
ice-VDF/0015-ice-restore-timestamp-configuration-after-device-res.patch
ice-VDF/0016-ice-introduce-PTP-state-machine.patch
ice-VDF/0017-ice-pass-reset-type-to-PTP-reset-functions.patch
ice-VDF/0018-ice-rename-verify_cached-to-has_ready_bitmap.patch
ice-VDF/0019-ice-don-t-check-has_ready_bitmap-in-E810-functions.patch
ice-VDF/0020-ice-rename-ice_ptp_tx_cfg_intr.patch
ice-VDF/0021-ice-factor-out-ice_ptp_rebuild_owner.patch
ice-VDF/0022-ice-stop-destroying-and-reinitalizing-Tx-tracker-dur.patch
ice-VDF/0023-ice-Remove-and-readd-netdev-during-devlink-reload.patch
ice-VDF/0024-ice-remove-FW-logging-code.patch
ice-VDF/0025-ice-configure-FW-logging.patch
ice-VDF/0026-ice-enable-FW-logging.patch
ice-VDF/0027-ice-add-ability-to-read-and-configure-FW-log-data.patch
ice-VDF/0028-ice-Fix-debugfs-with-devlink-reload.patch
ice-VDF/0029-ice-remove-vf-lan_vsi_num-field.patch
ice-VDF/0030-ice-rename-ice_write_-functions-to-ice_pack_ctx_.patch
ice-VDF/0031-ice-use-GENMASK-instead-of-BIT-n-1-in-pack-functions.patch
ice-VDF/0032-ice-cleanup-line-splitting-for-context-set-functions.patch
ice-VDF/0033-ice-do-not-disable-Tx-queues-twice-in-ice_down.patch
ice-VDF/0034-ice-Fix-improper-extts-handling.patch
ice-VDF/0035-ice-Don-t-process-extts-if-PTP-is-disabled.patch
ice-VDF/0036-ice-Introduce-ice_ptp_hw-struct.patch
ice-VDF/0037-ice-Introduce-ice_get_base_incval-helper.patch
ice-VDF/0038-ice-modify-the-ice-driver-version-to-stx.4.patch