All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bruce Fields <bfields@redhat.com>
To: trondmy@kernel.org
Cc: Chuck Lever <chuck.lever@oracle.com>,
	Linux NFS Mailing List <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH 5/9] nfsd: NFSv3 should allow zero length writes
Date: Fri, 17 Dec 2021 17:23:28 -0500	[thread overview]
Message-ID: <CAPL3RVFaWgdWQnWOe5B_6=1pNGSOZXp=SVFOBs24aucXphi6wQ@mail.gmail.com> (raw)
In-Reply-To: <20211217215046.40316-6-trondmy@kernel.org>

On Fri, Dec 17, 2021 at 5:07 PM <trondmy@kernel.org> wrote:
>
> From: Trond Myklebust <trond.myklebust@hammerspace.com>
>
> According to RFC1813: "If count is 0, the WRITE will succeed
> and return a count of 0, barring errors due to permissions checking."

Yes, I'm surprised we're not already doing this right.

I wonder how far back this bug goes.

The svc.c code is from 8154ef2776aa "NFSD: Clean up legacy NFS WRITE
argument XDR decoders", but the behavior might predate that code.  The
nfsd_vfs_write() logic, I'm not sure I understand.

We have a pynfs test for the v4 case (WRT4), but I guess we must have
nothing testing the v3 case.

--b.

>
> Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
> ---
>  fs/nfsd/vfs.c    | 3 +++
>  net/sunrpc/svc.c | 2 +-
>  2 files changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
> index 4d57befdac6e..38fdfcbb079e 100644
> --- a/fs/nfsd/vfs.c
> +++ b/fs/nfsd/vfs.c
> @@ -969,6 +969,9 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfsd_file *nf,
>
>         trace_nfsd_write_opened(rqstp, fhp, offset, *cnt);
>
> +       if (!*cnt)
> +               return nfs_ok;
> +
>         if (sb->s_export_op)
>                 exp_op_flags = sb->s_export_op->flags;
>
> diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
> index a3bbe5ce4570..d1ccf37a4588 100644
> --- a/net/sunrpc/svc.c
> +++ b/net/sunrpc/svc.c
> @@ -1692,7 +1692,7 @@ unsigned int svc_fill_write_vector(struct svc_rqst *rqstp, struct page **pages,
>          * entirely in rq_arg.pages. In this case, @first is empty.
>          */
>         i = 0;
> -       if (first->iov_len) {
> +       if (first->iov_len || !total) {
>                 vec[i].iov_base = first->iov_base;
>                 vec[i].iov_len = min_t(size_t, total, first->iov_len);
>                 total -= vec[i].iov_len;
> --
> 2.33.1
>


  parent reply	other threads:[~2021-12-17 22:23 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-17 21:50 [PATCH 0/9] Assorted patches for knfsd trondmy
2021-12-17 21:50 ` [PATCH 1/9] nfsd: map EBADF trondmy
2021-12-17 21:50   ` [PATCH 2/9] nfsd: Add errno mapping for EREMOTEIO trondmy
2021-12-17 21:50     ` [PATCH 3/9] nfsd: Retry once in nfsd_open on an -EOPENSTALE return trondmy
2021-12-17 21:50       ` [PATCH 4/9] nfs: Add export support for weak cache consistency attributes trondmy
2021-12-17 21:50         ` [PATCH 5/9] nfsd: NFSv3 should allow zero length writes trondmy
2021-12-17 21:50           ` [PATCH 6/9] nfsd: Add a tracepoint for errors in nfsd4_clone_file_range() trondmy
2021-12-17 21:50             ` [PATCH 7/9] nfsd: Replace use of rwsem with errseq_t trondmy
2021-12-17 21:50               ` [PATCH 8/9] nfsd: allow lockd to be forcibly disabled trondmy
2021-12-17 21:50                 ` [PATCH 9/9] nfsd: Ignore rpcbind errors on nfsd startup trondmy
2021-12-18 18:07                 ` [PATCH 8/9] nfsd: allow lockd to be forcibly disabled Chuck Lever III
2021-12-19 22:21                 ` Bruce Fields
2021-12-17 22:23           ` Bruce Fields [this message]
2021-12-18 18:41             ` [PATCH 5/9] nfsd: NFSv3 should allow zero length writes Chuck Lever III
2021-12-19 22:25               ` Bruce Fields
2021-12-18 21:16         ` [PATCH 4/9] nfs: Add export support for weak cache consistency attributes Chuck Lever III

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='CAPL3RVFaWgdWQnWOe5B_6=1pNGSOZXp=SVFOBs24aucXphi6wQ@mail.gmail.com' \
    --to=bfields@redhat.com \
    --cc=chuck.lever@oracle.com \
    --cc=linux-nfs@vger.kernel.org \
    --cc=trondmy@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.