The patch below does not apply to the 6.1-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y
git checkout FETCH_HEAD
git cherry-pick -x 3e74859ee35edc33a022c3f3971df066ea0ca6b9
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024123045-parka-sublet-a95d@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^..
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 3e74859ee35edc33a022c3f3971df066ea0ca6b9 Mon Sep 17 00:00:00 2001
From: Boris Burkov <boris(a)bur.io>
Date: Fri, 13 Dec 2024 12:22:32 -0800
Subject: [PATCH] btrfs: check folio mapping after unlock in
relocate_one_folio()
When we call btrfs_read_folio() to bring a folio uptodate, we unlock the
folio. The result of that is that a different thread can modify the
mapping (like remove it with invalidate) before we call folio_lock().
This results in an invalid page and we need to try again.
In particular, if we are relocating concurrently with aborting a
transaction, this can result in a crash like the following:
BUG: kernel NULL pointer dereference, address: 0000000000000000
PGD 0 P4D 0
Oops: 0000 [#1] SMP
CPU: 76 PID: 1411631 Comm: kworker/u322:5
Workqueue: events_unbound btrfs_reclaim_bgs_work
RIP: 0010:set_page_extent_mapped+0x20/0xb0
RSP: 0018:ffffc900516a7be8 EFLAGS: 00010246
RAX: ffffea009e851d08 RBX: ffffea009e0b1880 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffc900516a7b90 RDI: ffffea009e0b1880
RBP: 0000000003573000 R08: 0000000000000001 R09: ffff88c07fd2f3f0
R10: 0000000000000000 R11: 0000194754b575be R12: 0000000003572000
R13: 0000000003572fff R14: 0000000000100cca R15: 0000000005582fff
FS: 0000000000000000(0000) GS:ffff88c07fd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 000000407d00f002 CR4: 00000000007706f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
<TASK>
? __die+0x78/0xc0
? page_fault_oops+0x2a8/0x3a0
? __switch_to+0x133/0x530
? wq_worker_running+0xa/0x40
? exc_page_fault+0x63/0x130
? asm_exc_page_fault+0x22/0x30
? set_page_extent_mapped+0x20/0xb0
relocate_file_extent_cluster+0x1a7/0x940
relocate_data_extent+0xaf/0x120
relocate_block_group+0x20f/0x480
btrfs_relocate_block_group+0x152/0x320
btrfs_relocate_chunk+0x3d/0x120
btrfs_reclaim_bgs_work+0x2ae/0x4e0
process_scheduled_works+0x184/0x370
worker_thread+0xc6/0x3e0
? blk_add_timer+0xb0/0xb0
kthread+0xae/0xe0
? flush_tlb_kernel_range+0x90/0x90
ret_from_fork+0x2f/0x40
? flush_tlb_kernel_range+0x90/0x90
ret_from_fork_asm+0x11/0x20
</TASK>
This occurs because cleanup_one_transaction() calls
destroy_delalloc_inodes() which calls invalidate_inode_pages2() which
takes the folio_lock before setting mapping to NULL. We fail to check
this, and subsequently call set_extent_mapping(), which assumes that
mapping != NULL (in fact it asserts that in debug mode)
Note that the "fixes" patch here is not the one that introduced the
race (the very first iteration of this code from 2009) but a more recent
change that made this particular crash happen in practice.
Fixes: e7f1326cc24e ("btrfs: set page extent mapped after read_folio in relocate_one_page")
CC: stable(a)vger.kernel.org # 6.1+
Reviewed-by: Qu Wenruo <wqu(a)suse.com>
Signed-off-by: Boris Burkov <boris(a)bur.io>
Signed-off-by: David Sterba <dsterba(a)suse.com>
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index bf267bdfa8f8..db8b42f674b7 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2902,6 +2902,7 @@ static int relocate_one_folio(struct reloc_control *rc,
const bool use_rst = btrfs_need_stripe_tree_update(fs_info, rc->block_group->flags);
ASSERT(index <= last_index);
+again:
folio = filemap_lock_folio(inode->i_mapping, index);
if (IS_ERR(folio)) {
@@ -2937,6 +2938,11 @@ static int relocate_one_folio(struct reloc_control *rc,
ret = -EIO;
goto release_folio;
}
+ if (folio->mapping != inode->i_mapping) {
+ folio_unlock(folio);
+ folio_put(folio);
+ goto again;
+ }
}
/*
The patch below does not apply to the 6.6-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y
git checkout FETCH_HEAD
git cherry-pick -x 3e74859ee35edc33a022c3f3971df066ea0ca6b9
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024123042-limelight-doily-8703@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^..
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 3e74859ee35edc33a022c3f3971df066ea0ca6b9 Mon Sep 17 00:00:00 2001
From: Boris Burkov <boris(a)bur.io>
Date: Fri, 13 Dec 2024 12:22:32 -0800
Subject: [PATCH] btrfs: check folio mapping after unlock in
relocate_one_folio()
When we call btrfs_read_folio() to bring a folio uptodate, we unlock the
folio. The result of that is that a different thread can modify the
mapping (like remove it with invalidate) before we call folio_lock().
This results in an invalid page and we need to try again.
In particular, if we are relocating concurrently with aborting a
transaction, this can result in a crash like the following:
BUG: kernel NULL pointer dereference, address: 0000000000000000
PGD 0 P4D 0
Oops: 0000 [#1] SMP
CPU: 76 PID: 1411631 Comm: kworker/u322:5
Workqueue: events_unbound btrfs_reclaim_bgs_work
RIP: 0010:set_page_extent_mapped+0x20/0xb0
RSP: 0018:ffffc900516a7be8 EFLAGS: 00010246
RAX: ffffea009e851d08 RBX: ffffea009e0b1880 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffc900516a7b90 RDI: ffffea009e0b1880
RBP: 0000000003573000 R08: 0000000000000001 R09: ffff88c07fd2f3f0
R10: 0000000000000000 R11: 0000194754b575be R12: 0000000003572000
R13: 0000000003572fff R14: 0000000000100cca R15: 0000000005582fff
FS: 0000000000000000(0000) GS:ffff88c07fd00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 000000407d00f002 CR4: 00000000007706f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
<TASK>
? __die+0x78/0xc0
? page_fault_oops+0x2a8/0x3a0
? __switch_to+0x133/0x530
? wq_worker_running+0xa/0x40
? exc_page_fault+0x63/0x130
? asm_exc_page_fault+0x22/0x30
? set_page_extent_mapped+0x20/0xb0
relocate_file_extent_cluster+0x1a7/0x940
relocate_data_extent+0xaf/0x120
relocate_block_group+0x20f/0x480
btrfs_relocate_block_group+0x152/0x320
btrfs_relocate_chunk+0x3d/0x120
btrfs_reclaim_bgs_work+0x2ae/0x4e0
process_scheduled_works+0x184/0x370
worker_thread+0xc6/0x3e0
? blk_add_timer+0xb0/0xb0
kthread+0xae/0xe0
? flush_tlb_kernel_range+0x90/0x90
ret_from_fork+0x2f/0x40
? flush_tlb_kernel_range+0x90/0x90
ret_from_fork_asm+0x11/0x20
</TASK>
This occurs because cleanup_one_transaction() calls
destroy_delalloc_inodes() which calls invalidate_inode_pages2() which
takes the folio_lock before setting mapping to NULL. We fail to check
this, and subsequently call set_extent_mapping(), which assumes that
mapping != NULL (in fact it asserts that in debug mode)
Note that the "fixes" patch here is not the one that introduced the
race (the very first iteration of this code from 2009) but a more recent
change that made this particular crash happen in practice.
Fixes: e7f1326cc24e ("btrfs: set page extent mapped after read_folio in relocate_one_page")
CC: stable(a)vger.kernel.org # 6.1+
Reviewed-by: Qu Wenruo <wqu(a)suse.com>
Signed-off-by: Boris Burkov <boris(a)bur.io>
Signed-off-by: David Sterba <dsterba(a)suse.com>
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index bf267bdfa8f8..db8b42f674b7 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2902,6 +2902,7 @@ static int relocate_one_folio(struct reloc_control *rc,
const bool use_rst = btrfs_need_stripe_tree_update(fs_info, rc->block_group->flags);
ASSERT(index <= last_index);
+again:
folio = filemap_lock_folio(inode->i_mapping, index);
if (IS_ERR(folio)) {
@@ -2937,6 +2938,11 @@ static int relocate_one_folio(struct reloc_control *rc,
ret = -EIO;
goto release_folio;
}
+ if (folio->mapping != inode->i_mapping) {
+ folio_unlock(folio);
+ folio_put(folio);
+ goto again;
+ }
}
/*
The patch below does not apply to the 6.1-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y
git checkout FETCH_HEAD
git cherry-pick -x dc7eb8755797ed41a0d1b5c0c39df3c8f401b3d9
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024012617-overlap-reborn-e124@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^..
Possible dependencies:
dc7eb8755797 ("arm64/sme: Always exit sme_alloc() early with existing storage")
5d0a8d2fba50 ("arm64/ptrace: Ensure that SME is set up for target when writing SSVE state")
f90b529bcbe5 ("arm64/sme: Implement ZT0 ptrace support")
ce514000da4f ("arm64/sme: Rename za_state to sme_state")
1192b93ba352 ("arm64/fp: Use a struct to pass data to fpsimd_bind_state_to_cpu()")
deeb8f9a80fd ("arm64/fpsimd: Have KVM explicitly say which FP registers to save")
baa8515281b3 ("arm64/fpsimd: Track the saved FPSIMD state type separately to TIF_SVE")
93ae6b01bafe ("KVM: arm64: Discard any SVE state when entering KVM guests")
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From dc7eb8755797ed41a0d1b5c0c39df3c8f401b3d9 Mon Sep 17 00:00:00 2001
From: Mark Brown <broonie(a)kernel.org>
Date: Mon, 15 Jan 2024 20:15:46 +0000
Subject: [PATCH] arm64/sme: Always exit sme_alloc() early with existing
storage
When sme_alloc() is called with existing storage and we are not flushing we
will always allocate new storage, both leaking the existing storage and
corrupting the state. Fix this by separating the checks for flushing and
for existing storage as we do for SVE.
Callers that reallocate (eg, due to changing the vector length) should
call sme_free() themselves.
Fixes: 5d0a8d2fba50 ("arm64/ptrace: Ensure that SME is set up for target when writing SSVE state")
Signed-off-by: Mark Brown <broonie(a)kernel.org>
Cc: <stable(a)vger.kernel.org>
Link: https://lore.kernel.org/r/20240115-arm64-sme-flush-v1-1-7472bd3459b7@kernel…
Signed-off-by: Will Deacon <will(a)kernel.org>
diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
index 0983be2b1b61..a5dc6f764195 100644
--- a/arch/arm64/kernel/fpsimd.c
+++ b/arch/arm64/kernel/fpsimd.c
@@ -1217,8 +1217,10 @@ void fpsimd_release_task(struct task_struct *dead_task)
*/
void sme_alloc(struct task_struct *task, bool flush)
{
- if (task->thread.sme_state && flush) {
- memset(task->thread.sme_state, 0, sme_state_size(task));
+ if (task->thread.sme_state) {
+ if (flush)
+ memset(task->thread.sme_state, 0,
+ sme_state_size(task));
return;
}
Commit 60e3318e3e900 ("cifs: use fs_context for automounts") was
released in v6.1.54 and broke the failover when one of the servers
inside DFS becomes unavailable. We reproduced the problem on the EC2
instances of different types. Reverting aforementioned commint on top of
the latest stable verison v6.1.94 helps to resolve the problem.
Earliest working version is v6.2-rc1. There were two big merges of CIFS fixes:
[1] and [2]. We would like to ask for the help to investigate this problem and
if some of those patches need to be backported. Also, is it safe to just revert
problematic commit until proper fixes/backports will be available?
We will help to do testing and confirm if fix works, but let me also list the
steps we used to reproduce the problem if it will help to identify the problem:
1. Create Active Directory domain eg. 'corp.fsxtest.local' in AWS Directory
Service with:
- three AWS FSX file systems filesystem1..filesystem3
- three Windows servers; They have DFS installed as per
https://learn.microsoft.com/en-us/windows-server/storage/dfs-namespaces/dfs…:
- dfs-srv1: EC2AMAZ-2EGTM59
- dfs-srv2: EC2AMAZ-1N36PRD
- dfs-srv3: EC2AMAZ-0PAUH2U
2. Create DFS namespace eg. 'dfs-namespace' in Windows server 2008 mode
and three folders targets in it:
- referral-a mapped to filesystem1.corp.local
- referral-b mapped to filesystem2.corp.local
- referral-c mapped to filesystem3.corp.local
- local folders dfs-srv1..dfs-srv3 in C:\DFSRoots\dfs-namespace of every
Windows server. This helps to quickly define underlying server when
DFS is mounted.
3. Enabled cifs debug logs:
```
echo 'module cifs +p' > /sys/kernel/debug/dynamic_debug/control
echo 'file fs/cifs/* +p' > /sys/kernel/debug/dynamic_debug/control
echo 7 > /proc/fs/cifs/cifsFYI
```
4. Mount DFS namespace on Amazon Linux 2023 instance running any vanilla
kernel v6.1.54+:
```
dmesg -c &>/dev/null
cd /mnt
mount -t cifs -o cred=/mnt/creds,echo_interval=5 \
//corp.fsxtest.local/dfs-namespace \
./dfs-namespace
```
5. List DFS root, it's also required to avoid recursive mounts that happen
during regular 'ls' run:
```
sh -c 'ls dfs-namespace'
dfs-srv2 referral-a referral-b
```
The DFS server is EC2AMAZ-1N36PRD, it's also listed in mount:
```
[root@ip-172-31-2-82 mnt]# mount | grep dfs
//corp.fsxtest.local/dfs-namespace on /mnt/dfs-namespace type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.11.26,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1)
//EC2AMAZ-1N36PRD.corp.fsxtest.local/dfs-namespace/referral-a on /mnt/dfs-namespace/referral-a type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.12.80,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1)
```
List files in first folder:
```
sh -c 'ls dfs-namespace/referral-a'
filea.txt.txt
```
6. Shutdown DFS server-2.
List DFS root again, server changed from dfs-srv2 to dfs-srv1 EC2AMAZ-2EGTM59:
```
sh -c 'ls dfs-namespace'
dfs-srv1 referral-a referral-b
```
7. Try to list files in another folder, this causes ls to fail with error:
```
sh -c 'ls dfs-namespace/referral-b'
ls: cannot access 'dfs-namespace/referral-b': No route to host```
Sometimes it's also 'Operation now in progress' error.
mount shows the same output:
```
//corp.fsxtest.local/dfs-namespace on /mnt/dfs-namespace type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.11.26,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1)
//EC2AMAZ-1N36PRD.corp.fsxtest.local/dfs-namespace/referral-a on /mnt/dfs-namespace/referral-a type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.12.80,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1)
```
I also attached kernel debug logs from this test.
[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?…
[2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?…
Reported-by: Andrei Paniakin <apanyaki(a)amazon.com>
Bisected-by: Simba Bonga <simbarb(a)amazon.com>
---
#regzbot introduced: v6.1.54..v6.2-rc1
The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y
git checkout FETCH_HEAD
git cherry-pick -x 6bda857bcbb86fb9d0e54fbef93a093d51172acc
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024120342-monsoon-wildcat-d0a1@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^..
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 6bda857bcbb86fb9d0e54fbef93a093d51172acc Mon Sep 17 00:00:00 2001
From: Muchun Song <muchun.song(a)linux.dev>
Date: Mon, 14 Oct 2024 17:29:33 +0800
Subject: [PATCH] block: fix ordering between checking QUEUE_FLAG_QUIESCED
request adding
Supposing the following scenario.
CPU0 CPU1
blk_mq_insert_request() 1) store
blk_mq_unquiesce_queue()
blk_queue_flag_clear() 3) store
blk_mq_run_hw_queues()
blk_mq_run_hw_queue()
if (!blk_mq_hctx_has_pending()) 4) load
return
blk_mq_run_hw_queue()
if (blk_queue_quiesced()) 2) load
return
blk_mq_sched_dispatch_requests()
The full memory barrier should be inserted between 1) and 2), as well as
between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED
is cleared or CPU1 sees dispatch list or setting of bitmap of software
queue. Otherwise, either CPU will not rerun the hardware queue causing
starvation.
So the first solution is to 1) add a pair of memory barrier to fix the
problem, another solution is to 2) use hctx->queue->queue_lock to
synchronize QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since
memory barrier is not easy to be maintained.
Fixes: f4560ffe8cec ("blk-mq: use QUEUE_FLAG_QUIESCED to quiesce queue")
Cc: stable(a)vger.kernel.org
Cc: Muchun Song <muchun.song(a)linux.dev>
Signed-off-by: Muchun Song <songmuchun(a)bytedance.com>
Reviewed-by: Ming Lei <ming.lei(a)redhat.com>
Link: https://lore.kernel.org/r/20241014092934.53630-3-songmuchun@bytedance.com
Signed-off-by: Jens Axboe <axboe(a)kernel.dk>
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 5deb9dffca0a..bb4ee2380dce 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2227,6 +2227,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs)
}
EXPORT_SYMBOL(blk_mq_delay_run_hw_queue);
+static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx)
+{
+ bool need_run;
+
+ /*
+ * When queue is quiesced, we may be switching io scheduler, or
+ * updating nr_hw_queues, or other things, and we can't run queue
+ * any more, even blk_mq_hctx_has_pending() can't be called safely.
+ *
+ * And queue will be rerun in blk_mq_unquiesce_queue() if it is
+ * quiesced.
+ */
+ __blk_mq_run_dispatch_ops(hctx->queue, false,
+ need_run = !blk_queue_quiesced(hctx->queue) &&
+ blk_mq_hctx_has_pending(hctx));
+ return need_run;
+}
+
/**
* blk_mq_run_hw_queue - Start to run a hardware queue.
* @hctx: Pointer to the hardware queue to run.
@@ -2247,20 +2265,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async)
might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING);
- /*
- * When queue is quiesced, we may be switching io scheduler, or
- * updating nr_hw_queues, or other things, and we can't run queue
- * any more, even __blk_mq_hctx_has_pending() can't be called safely.
- *
- * And queue will be rerun in blk_mq_unquiesce_queue() if it is
- * quiesced.
- */
- __blk_mq_run_dispatch_ops(hctx->queue, false,
- need_run = !blk_queue_quiesced(hctx->queue) &&
- blk_mq_hctx_has_pending(hctx));
+ need_run = blk_mq_hw_queue_need_run(hctx);
+ if (!need_run) {
+ unsigned long flags;
- if (!need_run)
- return;
+ /*
+ * Synchronize with blk_mq_unquiesce_queue(), because we check
+ * if hw queue is quiesced locklessly above, we need the use
+ * ->queue_lock to make sure we see the up-to-date status to
+ * not miss rerunning the hw queue.
+ */
+ spin_lock_irqsave(&hctx->queue->queue_lock, flags);
+ need_run = blk_mq_hw_queue_need_run(hctx);
+ spin_unlock_irqrestore(&hctx->queue->queue_lock, flags);
+
+ if (!need_run)
+ return;
+ }
if (async || !cpumask_test_cpu(raw_smp_processor_id(), hctx->cpumask)) {
blk_mq_delay_run_hw_queue(hctx, 0);
The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y
git checkout FETCH_HEAD
git cherry-pick -x 2003ee8a9aa14d766b06088156978d53c2e9be3d
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024120323-snowiness-subway-3844@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^..
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 2003ee8a9aa14d766b06088156978d53c2e9be3d Mon Sep 17 00:00:00 2001
From: Muchun Song <muchun.song(a)linux.dev>
Date: Mon, 14 Oct 2024 17:29:32 +0800
Subject: [PATCH] block: fix missing dispatching request when queue is started
or unquiesced
Supposing the following scenario with a virtio_blk driver.
CPU0 CPU1 CPU2
blk_mq_try_issue_directly()
__blk_mq_issue_directly()
q->mq_ops->queue_rq()
virtio_queue_rq()
blk_mq_stop_hw_queue()
virtblk_done()
blk_mq_try_issue_directly()
if (blk_mq_hctx_stopped())
blk_mq_request_bypass_insert() blk_mq_run_hw_queue()
blk_mq_run_hw_queue() blk_mq_run_hw_queue()
blk_mq_insert_request()
return
After CPU0 has marked the queue as stopped, CPU1 will see the queue is
stopped. But before CPU1 puts the request on the dispatch list, CPU2
receives the interrupt of completion of request, so it will run the
hardware queue and marks the queue as non-stopped. Meanwhile, CPU1 also
runs the same hardware queue. After both CPU1 and CPU2 complete
blk_mq_run_hw_queue(), CPU1 just puts the request to the same hardware
queue and returns. It misses dispatching a request. Fix it by running
the hardware queue explicitly. And blk_mq_request_issue_directly()
should handle a similar situation. Fix it as well.
Fixes: d964f04a8fde ("blk-mq: fix direct issue")
Cc: stable(a)vger.kernel.org
Cc: Muchun Song <muchun.song(a)linux.dev>
Signed-off-by: Muchun Song <songmuchun(a)bytedance.com>
Reviewed-by: Ming Lei <ming.lei(a)redhat.com>
Link: https://lore.kernel.org/r/20241014092934.53630-2-songmuchun@bytedance.com
Signed-off-by: Jens Axboe <axboe(a)kernel.dk>
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 7d05a56e3639..5deb9dffca0a 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2647,6 +2647,7 @@ static void blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx,
if (blk_mq_hctx_stopped(hctx) || blk_queue_quiesced(rq->q)) {
blk_mq_insert_request(rq, 0);
+ blk_mq_run_hw_queue(hctx, false);
return;
}
@@ -2677,6 +2678,7 @@ static blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last)
if (blk_mq_hctx_stopped(hctx) || blk_queue_quiesced(rq->q)) {
blk_mq_insert_request(rq, 0);
+ blk_mq_run_hw_queue(hctx, false);
return BLK_STS_OK;
}