starlingx/kernel
Code Issues Proposed changes

Merge "workqueue: Affine rescuer threads and unbound wqs"

Browse Source
This commit is contained in:
Zuul 2021-11-15 16:47:01 +00:00 committed by Gerrit Code Review
commit 058219762f
4 changed files with 112 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
kernel-rt/centos/kernel-rt.spec
View File

@ -787,6 +787,7 @@ Patch7: 0008-Allow-dmar-quirks-for-broken-bioses.patch
Patch8: 0009-tpm-ignore-burstcount-to-improve-tpm_tis-send-perfor.patch
Patch9: 0010-restrict-iSCSI-kthreads-to-CPUs-in-cpu_kthread_mask.patch
Patch10: 0011-scsi-smartpqi-Enable-sas_address-sysfs-for-SATA-dev.patch
Patch11: 0012-workqueue-Affine-rescuer-threads-and-unbound-wqs.patch
# END OF PATCH DEFINITIONS
%endif

55
kernel-rt/centos/patches/0012-workqueue-Affine-rescuer-threads-and-unbound-wqs.patch Normal file
View File

@ -0,0 +1,55 @@
From 9c5c44019a17897c46a2579a50abcd69f1bcb4a8 Mon Sep 17 00:00:00 2001
From: "M. Vefa Bicakci" <vefa.bicakci@windriver.com>
Date: Thu, 9 Sep 2021 04:56:46 -0400
Subject: [PATCH] workqueue: Affine rescuer threads and unbound wqs
This commit ensures that workqueue rescuer threads are affined to the
platform CPUs specified by the "kthread_cpus" kernel argument. Prior to
this commit, rescuer threads could be bound to any CPU. Rescuer threads
are described in "kernel/workqueue.c" as follows:
"Regular work processing on a pool may block trying to create a new
worker which uses GFP_KERNEL allocation which has slight chance of
developing into deadlock if some works currently on the same queue
need to be processed to satisfy the GFP_KERNEL allocation. This is
the problem rescuer solves.
When such condition is possible, the pool summons rescuers of all
workqueues which have works queued on the pool and let them process
those works so that forward progress can be guaranteed."
This commit also affines unbound workqueues to the platform CPUs instead
of the housekeeping CPUs, because the latter can be a superset of the
former.
Signed-off-by: M. Vefa Bicakci <vefa.bicakci@windriver.com>
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
kernel/workqueue.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 1e2ca744dadb..f208096f28f1 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -4235,7 +4235,7 @@ static int init_rescuer(struct workqueue_struct *wq)
}
wq->rescuer = rescuer;
- kthread_bind_mask(rescuer->task, cpu_possible_mask);
+ kthread_bind_mask(rescuer->task, cpu_kthread_mask);
wake_up_process(rescuer->task);
return 0;
@@ -5927,7 +5927,7 @@ void __init workqueue_init_early(void)
BUILD_BUG_ON(__alignof__(struct pool_workqueue) < __alignof__(long long));
BUG_ON(!alloc_cpumask_var(&wq_unbound_cpumask, GFP_KERNEL));
- cpumask_copy(wq_unbound_cpumask, housekeeping_cpumask(hk_flags));
+ cpumask_copy(wq_unbound_cpumask, cpu_kthread_mask);
pwq_cache = KMEM_CACHE(pool_workqueue, SLAB_PANIC);
--
2.29.2

1
kernel-std/centos/kernel.spec
View File

@ -820,6 +820,7 @@ Patch7: 0008-Allow-dmar-quirks-for-broken-bioses.patch
Patch8: 0009-tpm-ignore-burstcount-to-improve-tpm_tis-send-perfor.patch
Patch9: 0010-bpf-cgroups-Fix-cgroup-v2-fallback-on-v1-v2-mixed-mo.patch
Patch10: 0011-scsi-smartpqi-Enable-sas_address-sysfs-for-SATA-dev.patch
Patch11: 0012-workqueue-Affine-rescuer-threads-and-unbound-wqs.patch
# END OF PATCH DEFINITIONS
%endif

55
kernel-std/centos/patches/0012-workqueue-Affine-rescuer-threads-and-unbound-wqs.patch Normal file
View File

@ -0,0 +1,55 @@
From 9c5c44019a17897c46a2579a50abcd69f1bcb4a8 Mon Sep 17 00:00:00 2001
From: "M. Vefa Bicakci" <vefa.bicakci@windriver.com>
Date: Thu, 9 Sep 2021 04:56:46 -0400
Subject: [PATCH] workqueue: Affine rescuer threads and unbound wqs
This commit ensures that workqueue rescuer threads are affined to the
platform CPUs specified by the "kthread_cpus" kernel argument. Prior to
this commit, rescuer threads could be bound to any CPU. Rescuer threads
are described in "kernel/workqueue.c" as follows:
"Regular work processing on a pool may block trying to create a new
worker which uses GFP_KERNEL allocation which has slight chance of
developing into deadlock if some works currently on the same queue
need to be processed to satisfy the GFP_KERNEL allocation. This is
the problem rescuer solves.
When such condition is possible, the pool summons rescuers of all
workqueues which have works queued on the pool and let them process
those works so that forward progress can be guaranteed."
This commit also affines unbound workqueues to the platform CPUs instead
of the housekeeping CPUs, because the latter can be a superset of the
former.
Signed-off-by: M. Vefa Bicakci <vefa.bicakci@windriver.com>
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
---
kernel/workqueue.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 1e2ca744dadb..f208096f28f1 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -4235,7 +4235,7 @@ static int init_rescuer(struct workqueue_struct *wq)
}
wq->rescuer = rescuer;
- kthread_bind_mask(rescuer->task, cpu_possible_mask);
+ kthread_bind_mask(rescuer->task, cpu_kthread_mask);
wake_up_process(rescuer->task);
return 0;
@@ -5927,7 +5927,7 @@ void __init workqueue_init_early(void)
BUILD_BUG_ON(__alignof__(struct pool_workqueue) < __alignof__(long long));
BUG_ON(!alloc_cpumask_var(&wq_unbound_cpumask, GFP_KERNEL));
- cpumask_copy(wq_unbound_cpumask, housekeeping_cpumask(hk_flags));
+ cpumask_copy(wq_unbound_cpumask, cpu_kthread_mask);
pwq_cache = KMEM_CACHE(pool_workqueue, SLAB_PANIC);
--
2.29.2