[Devel] fuse: call proper fiemap_prep as it is required by new kernel api

[Kui Liu]

Found 2 issues:

  1.  It’s better to move fiemap_prep() before inode_lock, as it was the case before when same thing is done in ioctl_fiemap().
  2.  FIEMAP_FLAG_SYNC as the last parameter is unnecessary, because it is always checked in fiemap_prep().

From: Andrey Zaitsev <Andrey.Zaitsev at acronis.com>
Date: Monday, 16 October 2023 at 5:23 PM
To: "Denis V. Lunev" <den at virtuozzo.com>, Alexey Kuznetsov <kuznet at virtuozzo.com>, Devel <devel at openvz.org>, "khorenko at virtuozzo.com" <khorenko at virtuozzo.com>, Kui Liu <Kui.Liu at acronis.com>
Subject: Re: fuse: call proper fiemap_prep as it is required by new kernel api


ok for me

________________________________
From: Denis V. Lunev <den at virtuozzo.com>
Sent: Sunday, October 15, 2023 10:08:40 PM
To: Alexey Kuznetsov; Devel; khorenko at virtuozzo.com; Andrey Zaitsev; Kui Liu
Subject: Re: fuse: call proper fiemap_prep as it is required by new kernel api

On 10/13/23 20:35, Alexey Kuznetsov wrote:
> Otherwise we consider locations with dirty page cache as holes.
>
> Affects: #VSTOR-76073, #PSBM-151414, #PSBM-151424, #PSBM-151464
>
> Signed-off-by: Alexey Kuznetsov <kuznet at acronis.com>
> ---
>   fs/fuse/file.c | 4 ++++
>   1 file changed, 4 insertions(+)
>
> diff --git a/fs/fuse/file.c b/fs/fuse/file.c
> index 5e23bd5..c685e01 100644
> --- a/fs/fuse/file.c
> +++ b/fs/fuse/file.c
> @@ -3711,6 +3711,10 @@ int fuse_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
>
>        inode_lock(inode);
>
> +     err = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
> +     if (err)
> +             goto out;
> +
>        fuse_sync_writes(inode);
>        fuse_read_dio_wait(fi);
>
Can we get acknowledgement for this patch in order to
start merge/build ASAP?

Thanks you in advance,
     Den
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openvz.org/pipermail/devel/attachments/20231016/036e7d12/attachment.html>