[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <3c262735-accf-41a8-aa74-a96a7ae05945@kernel.org>
Date: Wed, 19 Jun 2024 22:14:18 +0800
From: Chao Yu <chao@...nel.org>
To: jaegeuk@...nel.org
Cc: linux-kernel@...r.kernel.org, linux-f2fs-devel@...ts.sourceforge.net
Subject: Re: [f2fs-dev] [PATCH 2/2] f2fs: fix to use sb_{start,
end}_intwrite{_trylock, }() in gc_thread_func()
On 2024/6/6 17:52, Chao Yu wrote:
> Since background GC is f2fs inner operation, so, let's use
> sb_{start,end}_intwrite{_trylock,}() instead of
> sb_{start,end}_write{_trylock,}() in gc_thread_func().
It may cause racing in between gc_thread and freeze_super(), result in
writeback of dirty page after ->freeze, so please ignore this patch.
- sb_wait_write(sb, SB_FREEZE_WRITE)
- sync_filesystem(sb)
- sb_start_intwrite_trylock
- f2fs_gc
: dirty pages
- sb_wait_write(sb, SB_FREEZE_FS)
Thanks,
>
> Signed-off-by: Chao Yu <chao@...nel.org>
> ---
> fs/f2fs/gc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
> index ef667fec9a12..004587ac5530 100644
> --- a/fs/f2fs/gc.c
> +++ b/fs/f2fs/gc.c
> @@ -76,7 +76,7 @@ static int gc_thread_func(void *data)
> f2fs_stop_checkpoint(sbi, false,
> STOP_CP_REASON_FAULT_INJECT);
>
> - if (!sb_start_write_trylock(sbi->sb)) {
> + if (!sb_start_intwrite_trylock(sbi->sb)) {
> stat_other_skip_bggc_count(sbi);
> continue;
> }
> @@ -163,7 +163,7 @@ static int gc_thread_func(void *data)
> }
> spin_unlock(&sbi->gc_remaining_trials_lock);
> }
> - sb_end_write(sbi->sb);
> + sb_end_intwrite(sbi->sb);
>
> } while (!kthread_should_stop());
> return 0;
Powered by blists - more mailing lists