mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "yohan.joung" <yohan.joung@sk.com>
To: linux-f2fs-devel@lists.sourceforge.net
Cc: chao@kernel.org, daeho43@gmail.com, jaegeuk@kernel.org,
	jyh429@gmail.com, linux-kernel@vger.kernel.org,
	pilhyun.kim@sk.com, yohan.joung@sk.com
Subject: Re: [f2fs-dev] [External Mail] Re: [External Mail] Re: [PATCH] f2fs: prevent the current section from being selected as a victim during garbage collection
Date: Fri, 28 Mar 2025 12:40:36 +0900	[thread overview]
Message-ID: <20250328034042.2444-1-yohan.joung@sk.com> (raw)
In-Reply-To: <60e763d5-dff4-44a1-9e80-48d384335027@kernel.org>

> From: Chao Yu <chao@kernel.org>
> Sent: Thursday, March 27, 2025 10:48 PM
> To: 정요한(JOUNG YOHAN) Mobile AE <yohan.joung@sk.com>; Yohan Joung
> <jyh429@gmail.com>; jaegeuk@kernel.org; daeho43@gmail.com
> Cc: chao@kernel.org; linux-f2fs-devel@lists.sourceforge.net; linux-
> kernel@vger.kernel.org; 김필현(KIM PILHYUN) Mobile AE <pilhyun.kim@sk.com>
> Subject: [External Mail] Re: [External Mail] Re: [External Mail] Re:
> [PATCH] f2fs: prevent the current section from being selected as a victim
> during garbage collection
> 
> On 2025/3/27 16:00, yohan.joung@sk.com wrote:
> >> From: Chao Yu <chao@kernel.org>
> >> Sent: Thursday, March 27, 2025 4:30 PM
> >> To: 정요한(JOUNG YOHAN) Mobile AE <yohan.joung@sk.com>; Yohan Joung
> >> <jyh429@gmail.com>; jaegeuk@kernel.org; daeho43@gmail.com
> >> Cc: chao@kernel.org; linux-f2fs-devel@lists.sourceforge.net; linux-
> >> kernel@vger.kernel.org; 김필현(KIM PILHYUN) Mobile AE
> >> <pilhyun.kim@sk.com>
> >> Subject: [External Mail] Re: [External Mail] Re: [PATCH] f2fs:
> >> prevent the current section from being selected as a victim during
> >> garbage collection
> >>
> >> On 3/27/25 14:43, yohan.joung@sk.com wrote:
> >>>> From: Chao Yu <chao@kernel.org>
> >>>> Sent: Thursday, March 27, 2025 3:02 PM
> >>>> To: Yohan Joung <jyh429@gmail.com>; jaegeuk@kernel.org;
> >>>> daeho43@gmail.com
> >>>> Cc: chao@kernel.org; linux-f2fs-devel@lists.sourceforge.net; linux-
> >>>> kernel@vger.kernel.org; 정요한(JOUNG YOHAN) Mobile AE
> >>>> <yohan.joung@sk.com>
> >>>> Subject: [External Mail] Re: [PATCH] f2fs: prevent the current
> >>>> section from being selected as a victim during garbage collection
> >>>>
> >>>> On 3/26/25 22:14, Yohan Joung wrote:
> >>>>> When selecting a victim using next_victim_seg in a large section,
> >>>>> the selected section might already have been cleared and
> >>>>> designated as the new current section, making it actively in use.
> >>>>> This behavior causes inconsistency between the SIT and SSA.
> >>>>
> >>>> Hi, does this fix your issue?
> >>>
> >>> This is an issue that arises when dividing a large section into
> >>> segments for garbage collection.
> >>> caused by the background GC (garbage collection) thread in large
> >>> section
> >>> f2fs_gc(victim_section) ->
> >>> f2fs_clear_prefree_segments(victim_section)->
> >>> cursec(victim_section) -> f2fs_gc(victim_section by next_victim_seg)
> >>
> >> I didn't get it, why f2fs_get_victim() will return section which is
> >> used by curseg? It should be avoided by checking w/ sec_usage_check().
> >>
> >> Or we missed to check gcing section which next_victim_seg points to
> >> during get_new_segment()?
> >>
> >> Can this happen?
> >>
> >> e.g.
> >> - bggc selects sec #0
> >> - next_victim_seg: seg #0
> >> - migrate seg #0 and stop
> >> - next_victim_seg: seg #1
> >> - checkpoint, set sec #0 free if sec #0 has no valid blocks
> >> - allocate seg #0 in sec #0 for curseg
> >> - curseg moves to seg #1 after allocation
> >> - bggc tries to migrate seg #1
> >>
> >> Thanks,
> > That's correct
> > In f2fs_get_victim, we use next_victim_seg to directly jump to
> > got_result, thereby bypassing sec_usage_check What do you think about
> > this change?
> >
> > @@ -850,15 +850,20 @@ int f2fs_get_victim(struct f2fs_sb_info *sbi,
> unsigned int *result,
> >                          p.min_segno = sbi->next_victim_seg[BG_GC];
> >                          *result = p.min_segno;
> >                          sbi->next_victim_seg[BG_GC] = NULL_SEGNO;
> > -                       goto got_result;
> >                  }
> >                  if (gc_type == FG_GC &&
> >                                  sbi->next_victim_seg[FG_GC] != NULL_SEGNO) {
> >                          p.min_segno = sbi->next_victim_seg[FG_GC];
> >                          *result = p.min_segno;
> >                          sbi->next_victim_seg[FG_GC] = NULL_SEGNO;
> > -                       goto got_result;
> >                  }
> > +
> > +               secno = GET_SEC_FROM_SEG(sbi, segno);
> > +
> > +               if (sec_usage_check(sbi, secno))
> > +                       goto next;
> > +
> > +               goto got_result;
> >          }
> 
> But still allocator can assign this segment after sec_usage_check() in
> race condition, right?
Since the BG GC using next_victim  takes place after the SIT update in do_checkpoint, 
it seems unlikely that a race condition with sec_usage_check will occur.
> 
> IMO, we can clear next_victim_seg[] once section is free in
> __set_test_and_free()? something like this:
I will test it according to your suggestion.
If there are no issues, can I submit it again with the patch?
Thanks
> 
> ---
>   fs/f2fs/segment.h | 13 ++++++++++---
>   1 file changed, 10 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index
> 0465dc00b349..826e37999085 100644
> --- a/fs/f2fs/segment.h
> +++ b/fs/f2fs/segment.h
> @@ -473,9 +473,16 @@ static inline void __set_test_and_free(struct
> f2fs_sb_info *sbi,
>   			goto skip_free;
>   		next = find_next_bit(free_i->free_segmap,
>   				start_segno + SEGS_PER_SEC(sbi), start_segno);
> -		if (next >= start_segno + usable_segs) {
> -			if (test_and_clear_bit(secno, free_i->free_secmap))
> -				free_i->free_sections++;
> +		if ((next >= start_segno + usable_segs) &&
> +			test_and_clear_bit(secno, free_i->free_secmap)) {
> +			free_i->free_sections++;
> +
> +			if (GET_SEC_FROM_SEG(sbi->next_victim_seg[BG_GC]) ==
> +									secno)
> +				sbi->next_victim_seg[BG_GC] = NULL_SEGNO;
> +			if (GET_SEC_FROM_SEG(sbi->next_victim_seg[FG_GC]) ==
> +									secno)
> +				sbi->next_victim_seg[FG_GC] = NULL_SEGNO;
>   		}
>   	}
>   skip_free:
> --
> 2.40.1
> 
> >>
> >>>
> >>> Because the call stack is different, I think that in order to handle
> >>> everything at once, we need to address it within do_garbage_collect,
> >>> or otherwise include it on both sides.
> >>> What do you think?
> >>>
> >>> [30146.337471][ T1300] F2FS-fs (dm-54): Inconsistent segment (70961)
> >>> type [0, 1] in SSA and SIT [30146.346151][ T1300] Call trace:
> >>> [30146.346152][ T1300]  dump_backtrace+0xe8/0x10c [30146.346157][
> >>> T1300]  show_stack+0x18/0x28 [30146.346158][ T1300]
> >>> dump_stack_lvl+0x50/0x6c [30146.346161][ T1300]
> >>> dump_stack+0x18/0x28 [30146.346162][ T1300]
> >>> f2fs_stop_checkpoint+0x1c/0x3c [30146.346165][ T1300]
> >>> do_garbage_collect+0x41c/0x271c [30146.346167][ T1300]
> >>> f2fs_gc+0x27c/0x828 [30146.346168][ T1300]
> >>> gc_thread_func+0x290/0x88c [30146.346169][ T1300]
> >>> kthread+0x11c/0x164 [30146.346172][ T1300]
> >>> ret_from_fork+0x10/0x20
> >>>
> >>> struct curseg_info : 0xffffff803f95e800 {
> >>> 	segno        : 0x11531 : 70961
> >>> }
> >>>
> >>> struct f2fs_sb_info : 0xffffff8811d12000 {
> >>> 	next_victim_seg[0] : 0x11531 : 70961 }
> >>>
> >>>>
> >>>> https://lore.kernel.org/linux-f2fs-devel/20250325080646.3291947-2-
> >>>> chao@kernel.org
> >>>>
> >>>> Thanks,
> >>>>
> >>>>>
> >>>>> Signed-off-by: Yohan Joung <yohan.joung@sk.com>
> >>>>> ---
> >>>>>   fs/f2fs/gc.c | 4 ++++
> >>>>>   1 file changed, 4 insertions(+)
> >>>>>
> >>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index
> >>>>> 2b8f9239bede..4b5d18e395eb 100644
> >>>>> --- a/fs/f2fs/gc.c
> >>>>> +++ b/fs/f2fs/gc.c
> >>>>> @@ -1926,6 +1926,10 @@ int f2fs_gc(struct f2fs_sb_info *sbi,
> >>>>> struct
> >>>> f2fs_gc_control *gc_control)
> >>>>>   		goto stop;
> >>>>>   	}
> >>>>>
> >>>>> +	if (__is_large_section(sbi) &&
> >>>>> +			IS_CURSEC(sbi, GET_SEC_FROM_SEG(sbi, segno)))
> >>>>> +		goto stop;
> >>>>> +
> >>>>>   	seg_freed = do_garbage_collect(sbi, segno, &gc_list, gc_type,
> >>>>>   				gc_control->should_migrate_blocks,
> >>>>>   				gc_control->one_time);
> >>>
> >


  reply	other threads:[~2025-03-28  3:40 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-03-26 14:14 Yohan Joung
2025-03-27  6:02 ` Chao Yu
2025-03-27  6:43   ` [External Mail] " 정요한(JOUNG YOHAN) Mobile AE
2025-03-27  7:29     ` Chao Yu
2025-03-27  8:00       ` [External Mail] " 정요한(JOUNG YOHAN) Mobile AE
2025-03-27 13:47         ` Chao Yu
2025-03-28  3:40           ` yohan.joung [this message]
2025-03-28  6:17             ` [f2fs-dev] " Chao Yu
2025-03-28  7:25               ` [External Mail] " yohan.joung
2025-03-28  8:00                 ` Chao Yu
2025-03-31  5:13                   ` yohan.joung
2025-03-31 11:36                     ` Chao Yu
2025-04-01  1:51                   ` yohan.joung
2025-04-01  5:41                     ` Chao Yu
2025-04-01  7:47                       ` yohan.joung
2025-04-01  8:44                         ` Chao Yu
2025-04-02  0:33                           ` yohan.joung

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250328034042.2444-1-yohan.joung@sk.com \
    --to=yohan.joung@sk.com \
    --cc=chao@kernel.org \
    --cc=daeho43@gmail.com \
    --cc=jaegeuk@kernel.org \
    --cc=jyh429@gmail.com \
    --cc=linux-f2fs-devel@lists.sourceforge.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pilhyun.kim@sk.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®