mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Chao Yu <chao@kernel.org>
To: Chunhai Guo <guochunhai@vivo.com>, jaegeuk@kernel.org
Cc: Chao Yu <chao@kernel.org>,
	linux-f2fs-devel@lists.sourceforge.net,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/2] f2fs: fix missing discard candidates in fstrim
Date: Fri, 3 Jan 2025 11:26:23 +0800	[thread overview]
Message-ID: <62b14417-1297-4ed6-9ff3-b24115d433c4@kernel.org> (raw)
In-Reply-To: <20250102101310.580277-1-guochunhai@vivo.com>

On 2025/1/2 18:13, Chunhai Guo wrote:
> fstrim may miss candidates that need to be discarded in fstrim, as shown in
> the examples below.
> The root cause is that when cpc->reason is set with CP_DISCARD,
> add_discard_addrs() expects that ckpt_valid_map and cur_valid_map have been
> synced by seg_info_to_raw_sit() [1] and tries to find the candidates based
> on ckpt_valid_map and discard_map. However, seg_info_to_raw_sit() does not
> actually run before f2fs_exist_trim_candidates(), which results in failure.

Chunhai,

Can you please use nodiscard option due to fstrim stopped to return
trimmed length after below commit:

5a6154920faf ("f2fs: don't issue discard commands in online discard is on")

> 
> root# cp testfile /f2fs_mountpoint
> 
> root# f2fs_io fiemap 0 1 /f2fs_mountpoint/testfile
> Fiemap: offset = 0 len = 1
>          logical addr.    physical addr.   length           flags
> 0       0000000000000000 0000000406a00000 000000003d800000 00001000
> 
> root# rm /f2fs_mountpoint/testfile
> 
> root# fstrim -v -o 0x406a00000 -l 1024M /f2fs_mountpoint -- no candidate is found
> /f2fs_mountpoint: 0 B (0 bytes) trimmed
> 
> [1] Please refer to commit d7bc2484b8d4 ("f2fs: fix small discards not to
> issue redundantly") for the relationship between seg_info_to_raw_sit() and
> add_discard_addrs().
> 
> Fixes: 25290fa5591d ("f2fs: return fs_trim if there is no candidate")
> Signed-off-by: Chunhai Guo <guochunhai@vivo.com>
> ---
>   fs/f2fs/segment.c | 10 +++++-----
>   1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index eade36c5ef13..8fe9f794b581 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -2070,7 +2070,7 @@ static int f2fs_issue_discard(struct f2fs_sb_info *sbi,
>   }
>   
>   static bool add_discard_addrs(struct f2fs_sb_info *sbi, struct cp_control *cpc,
> -							bool check_only)
> +					bool synced, bool check_only)
>   {
>   	int entries = SIT_VBLOCK_MAP_SIZE / sizeof(unsigned long);
>   	struct seg_entry *se = get_seg_entry(sbi, cpc->trim_start);
> @@ -2098,7 +2098,7 @@ static bool add_discard_addrs(struct f2fs_sb_info *sbi, struct cp_control *cpc,
>   
>   	/* SIT_VBLOCK_MAP_SIZE should be multiple of sizeof(unsigned long) */
>   	for (i = 0; i < entries; i++)
> -		dmap[i] = force ? ~ckpt_map[i] & ~discard_map[i] :
> +		dmap[i] = synced ? ~ckpt_map[i] & ~discard_map[i] :

I guess this condition "force ? ~ckpt_map[i] & ~discard_map[i]" didn't cover
all below cases, thoughts?
- ckpt_map[i] == 0 // write data, and then remove data before checkpoint
- ckpt_map[i] != 0 // remove data existed in previous checkpoint

Thanks,

>   				(cur_map[i] ^ ckpt_map[i]) & ckpt_map[i];
>   
>   	while (force || SM_I(sbi)->dcc_info->nr_discards <=
> @@ -3275,7 +3275,7 @@ bool f2fs_exist_trim_candidates(struct f2fs_sb_info *sbi,
>   
>   	down_write(&SIT_I(sbi)->sentry_lock);
>   	for (; cpc->trim_start <= cpc->trim_end; cpc->trim_start++) {
> -		if (add_discard_addrs(sbi, cpc, true)) {
> +		if (add_discard_addrs(sbi, cpc, false, true)) {
>   			has_candidate = true;
>   			break;
>   		}
> @@ -4611,7 +4611,7 @@ void f2fs_flush_sit_entries(struct f2fs_sb_info *sbi, struct cp_control *cpc)
>   			/* add discard candidates */
>   			if (!(cpc->reason & CP_DISCARD)) {
>   				cpc->trim_start = segno;
> -				add_discard_addrs(sbi, cpc, false);
> +				add_discard_addrs(sbi, cpc, false, false);
>   			}
>   
>   			if (to_journal) {
> @@ -4653,7 +4653,7 @@ void f2fs_flush_sit_entries(struct f2fs_sb_info *sbi, struct cp_control *cpc)
>   		__u64 trim_start = cpc->trim_start;
>   
>   		for (; cpc->trim_start <= cpc->trim_end; cpc->trim_start++)
> -			add_discard_addrs(sbi, cpc, false);
> +			add_discard_addrs(sbi, cpc, true, false);
>   
>   		cpc->trim_start = trim_start;
>   	}


  parent reply	other threads:[~2025-01-03  3:26 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-01-02 10:13 Chunhai Guo
2025-01-02 10:13 ` [PATCH 2/2] f2fs: fix missing small discard " Chunhai Guo
2025-01-03  3:36   ` Chao Yu
2025-01-03  7:00     ` Chunhai Guo
2025-05-21 14:05       ` Chunhai Guo
2025-05-22  8:48       ` Chao Yu
2025-05-22 11:11   ` Chao Yu
2025-05-23  8:15     ` Chunhai Guo
2025-01-03  3:26 ` Chao Yu [this message]
2025-01-03  8:07   ` [PATCH 1/2] f2fs: fix missing discard candidates " Chunhai Guo
2025-01-08 12:46     ` Chao Yu
2025-01-09 10:03       ` Chunhai Guo
2025-01-13 13:32         ` Chao Yu
2025-01-19 13:34           ` Chunhai Guo

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=62b14417-1297-4ed6-9ff3-b24115d433c4@kernel.org \
    --to=chao@kernel.org \
    --cc=guochunhai@vivo.com \
    --cc=jaegeuk@kernel.org \
    --cc=linux-f2fs-devel@lists.sourceforge.net \
    --cc=linux-kernel@vger.kernel.org \
    /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

Powered by JetHome