Re: [RFC PATCH] ext4: improve discard efficiency

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



This patch free blocks too earily when do discard, maybe cause data
corrupt, I'll fix this in
next version.

Fengnan Chang <changfengnan@xxxxxxxxxxxxx> 于2023年6月15日周四 19:12写道:
>
> In commit a015434480dc("ext4: send parallel discards on commit
> completions"), issue all discard commands in parallel make all
> bios could merged into one request, so lowlevel drive can issue
> multi segments in one time which is more efficiency, but commit
> 55cdd0af2bc5 ("ext4: get discard out of jbd2 commit kthread contex")
> seems broke this way, let's fix it.
> In my test, the time of fstrim fs with multi big sparse file
> reduce from 6.7s to 1.3s.
>
> Signed-off-by: Fengnan Chang <changfengnan@xxxxxxxxxxxxx>
> ---
>  fs/ext4/mballoc.c | 17 ++++++++++++++---
>  1 file changed, 14 insertions(+), 3 deletions(-)
>
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index a2475b8c9fb5..e5a27fd2e959 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -6790,7 +6790,7 @@ int ext4_group_add_blocks(handle_t *handle, struct super_block *sb,
>   * be called with under the group lock.
>   */
>  static int ext4_trim_extent(struct super_block *sb,
> -               int start, int count, struct ext4_buddy *e4b)
> +               int start, int count, struct ext4_buddy *e4b, struct bio **biop)
>  __releases(bitlock)
>  __acquires(bitlock)
>  {
> @@ -6812,7 +6812,7 @@ __acquires(bitlock)
>          */
>         mb_mark_used(e4b, &ex);
>         ext4_unlock_group(sb, group);
> -       ret = ext4_issue_discard(sb, group, start, count, NULL);
> +       ret = ext4_issue_discard(sb, group, start, count, biop);
>         ext4_lock_group(sb, group);
>         mb_free_blocks(NULL, e4b, start, ex.fe_len);
>         return ret;
> @@ -6826,12 +6826,15 @@ __releases(ext4_group_lock_ptr(sb, e4b->bd_group))
>  {
>         ext4_grpblk_t next, count, free_count;
>         void *bitmap;
> +       struct bio *discard_bio = NULL;
> +       struct blk_plug plug;
>
>         bitmap = e4b->bd_bitmap;
>         start = (e4b->bd_info->bb_first_free > start) ?
>                 e4b->bd_info->bb_first_free : start;
>         count = 0;
>         free_count = 0;
> +       blk_start_plug(&plug);
>
>         while (start <= max) {
>                 start = mb_find_next_zero_bit(bitmap, max + 1, start);
> @@ -6840,7 +6843,7 @@ __releases(ext4_group_lock_ptr(sb, e4b->bd_group))
>                 next = mb_find_next_bit(bitmap, max + 1, start);
>
>                 if ((next - start) >= minblocks) {
> -                       int ret = ext4_trim_extent(sb, start, next - start, e4b);
> +                       int ret = ext4_trim_extent(sb, start, next - start, e4b, &discard_bio);
>
>                         if (ret && ret != -EOPNOTSUPP)
>                                 break;
> @@ -6864,6 +6867,14 @@ __releases(ext4_group_lock_ptr(sb, e4b->bd_group))
>                         break;
>         }
>
> +       if (discard_bio) {
> +               ext4_unlock_group(sb, e4b->bd_group);
> +               submit_bio_wait(discard_bio);
> +               bio_put(discard_bio);
> +               ext4_lock_group(sb, e4b->bd_group);
> +       }
> +       blk_finish_plug(&plug);
> +
>         return count;
>  }
>
> --
> 2.37.1 (Apple Git-137.1)
>




[Index of Archives]     [Reiser Filesystem Development]     [Ceph FS]     [Kernel Newbies]     [Security]     [Netfilter]     [Bugtraq]     [Linux FS]     [Yosemite National Park]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Samba]     [Device Mapper]     [Linux Media]

  Powered by Linux