From: bfields@fieldses.org (J. Bruce Fields)
To: Olga Kornievskaia <olga.kornievskaia@gmail.com>
Cc: bfields@redhat.com, chuck.lever@oracle.com, linux-nfs@vger.kernel.org
Subject: Re: [PATCH v2 1/1] NFSD add vfs_fsync after async copy is done
Date: Wed, 19 May 2021 17:32:36 -0400 [thread overview]
Message-ID: <20210519213236.GD19450@fieldses.org> (raw)
In-Reply-To: <20210519184827.5460-1-olga.kornievskaia@gmail.com>
On Wed, May 19, 2021 at 02:48:27PM -0400, Olga Kornievskaia wrote:
> From: Olga Kornievskaia <kolga@netapp.com>
>
> Currently, the server does all copies as NFS_UNSTABLE. For synchronous
> copies linux client will append a COMMIT to the COPY compound but for
> async copies it does not (because COMMIT needs to be done after all
> bytes are copied and not as a reply to the COPY operation).
>
> However, in order to save the client doing a COMMIT as a separate
> rpc, the server can reply back with NFS_FILE_SYNC copy. This patch
> proposed to add vfs_fsync() call at the end of the async copy.
>
> --- v2: moved the committed argument into the nfsd4_copy structure
Makes sense to me, thanks. Applying if nobody sees an objection.--b.
>
> Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
> ---
> fs/nfsd/nfs4proc.c | 14 +++++++++++++-
> fs/nfsd/xdr4.h | 1 +
> 2 files changed, 14 insertions(+), 1 deletion(-)
>
> diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
> index f4ce93d7f26e..dc7ac4b39eff 100644
> --- a/fs/nfsd/nfs4proc.c
> +++ b/fs/nfsd/nfs4proc.c
> @@ -1375,7 +1375,8 @@ static const struct nfsd4_callback_ops nfsd4_cb_offload_ops = {
>
> static void nfsd4_init_copy_res(struct nfsd4_copy *copy, bool sync)
> {
> - copy->cp_res.wr_stable_how = NFS_UNSTABLE;
> + copy->cp_res.wr_stable_how =
> + copy->committed ? NFS_FILE_SYNC : NFS_UNSTABLE;
> copy->cp_synchronous = sync;
> gen_boot_verifier(©->cp_res.wr_verifier, copy->cp_clp->net);
> }
> @@ -1386,6 +1387,7 @@ static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy)
> u64 bytes_total = copy->cp_count;
> u64 src_pos = copy->cp_src_pos;
> u64 dst_pos = copy->cp_dst_pos;
> + __be32 status;
>
> /* See RFC 7862 p.67: */
> if (bytes_total == 0)
> @@ -1403,6 +1405,16 @@ static ssize_t _nfsd_copy_file_range(struct nfsd4_copy *copy)
> src_pos += bytes_copied;
> dst_pos += bytes_copied;
> } while (bytes_total > 0 && !copy->cp_synchronous);
> + /* for a non-zero asynchronous copy do a commit of data */
> + if (!copy->cp_synchronous && copy->cp_res.wr_bytes_written > 0) {
> + down_write(©->nf_dst->nf_rwsem);
> + status = vfs_fsync_range(copy->nf_dst->nf_file,
> + copy->cp_dst_pos,
> + copy->cp_res.wr_bytes_written, 0);
> + up_write(©->nf_dst->nf_rwsem);
> + if (!status)
> + copy->committed = true;
> + }
> return bytes_copied;
> }
>
> diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h
> index a7c425254fee..3e4052e3bd50 100644
> --- a/fs/nfsd/xdr4.h
> +++ b/fs/nfsd/xdr4.h
> @@ -567,6 +567,7 @@ struct nfsd4_copy {
> struct vfsmount *ss_mnt;
> struct nfs_fh c_fh;
> nfs4_stateid stateid;
> + bool committed;
> };
>
> struct nfsd4_seek {
> --
> 2.27.0
prev parent reply other threads:[~2021-05-19 21:32 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-19 18:48 [PATCH v2 1/1] NFSD add vfs_fsync after async copy is done Olga Kornievskaia
2021-05-19 21:32 ` J. Bruce Fields [this message]
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=20210519213236.GD19450@fieldses.org \
--to=bfields@fieldses.org \
--cc=bfields@redhat.com \
--cc=chuck.lever@oracle.com \
--cc=linux-nfs@vger.kernel.org \
--cc=olga.kornievskaia@gmail.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