This patch converts the ls_root_lock semaphore to a rwlock to not
sleep during dlm message processing.
Signed-off-by: Alexander Aring
---
fs/dlm/dir.c | 14 +++---
fs/dlm/dlm_internal.h | 2 +-
fs/dlm/lock.c | 4 ++--
fs/dlm/lockspace.c| 2 +-
fs/dlm/recover.c
This patch removes the context parameter for message allocations and
replace it by always do GFP_ATOMIC. We are preparing to process dlm
message in softirq context therefore it's necessary to switch to
GFP_ATOMIC allocation as we cannot sleep in this context. To simplify
the code overall we just
This patch converts the per dlm rsb res_lock from a mutex to a spinlock.
Signed-off-by: Alexander Aring
---
fs/dlm/dlm_internal.h | 2 +-
fs/dlm/lock.c | 2 +-
fs/dlm/lock.h | 4 ++--
3 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/fs/dlm/dlm_internal.h
This patch change the dlm message parsing context from a workqueue to
a softirq context. This will hopefully speed up our dlm message
processing by removing a bunch of implicit scheduling points such a
cond_reched() depends on the preemption model setting. A softirq
(except PREEMPT_RT) can only be
This patch prepares to switch some locks to a spinlock. In this case we
need to remove some explicit schedule points when a spinlock is held.
We might have less scheduling points to try to serve others, we need to
see if this still makes problems when we remove them and find other
solutions.
This patch converts all spinlocks involved in message parsing to it's
_bh version. The reason to do that is to convert the message parsing
into softirq context and we need to prevent that those locks can be
interrupted by a softirq if those are held.
Signed-off-by: Alexander Aring
---
This patch converts the ls_recv_active semaphore to a rwlock to not
sleep during dlm message processing.
Signed-off-by: Alexander Aring
---
fs/dlm/dlm_internal.h | 2 +-
fs/dlm/lock.c | 4 ++--
fs/dlm/lockspace.c| 2 +-
fs/dlm/member.c | 4 ++--
fs/dlm/recoverd.c | 4 ++--
This patch removes the ls_requestqueue_wait and convert the
ls_requestqueue_mutex to a rw lock. Instead of calling wait_event() in
dlm processing which waits until all messages are processed and allow
new message processing after recovering is done, this patch is using a
bitflag to signal when a
This patch prepares to process dlm messages in softirq context. To
switch dlm to parse messages in softirq context some code parts either
runs inside the softirq context or need to switch to run while a spinlock
is held. This patch prepares to switch the allocation context to
GFP_ATOMIC for those
This patch converts the per dlm lockspace waiters lock from a mutex to a
spinlock.
Signed-off-by: Alexander Aring
---
fs/dlm/debug_fs.c | 4 ++--
fs/dlm/dlm_internal.h | 2 +-
fs/dlm/lock.c | 33 +
fs/dlm/lockspace.c| 2 +-
4 files changed, 21
Hello,
syzbot found the following issue on:
HEAD commit:3f86ed6ec0b3 Merge tag 'arc-6.6-rc1' of git://git.kernel.o..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=12cda4e7a8
kernel config: https://syzkaller.appspot.com/x/.config?x=fe0cf825f8fbc075
On 9/3/23 23:30, Dave Chinner wrote:
On Wed, Aug 30, 2023 at 02:11:31PM +0800, Hao Xu wrote:
On 8/29/23 19:53, Matthew Wilcox wrote:
On Tue, Aug 29, 2023 at 03:46:13PM +0800, Hao Xu wrote:
On 8/28/23 05:32, Matthew Wilcox wrote:
On Sun, Aug 27, 2023 at 09:28:31PM +0800, Hao Xu wrote:
From:
12 matches
Mail list logo
