From: "J. Bruce Fields" <bfields@fieldses.org>
To: Chuck Lever <chuck.lever@oracle.com>
Cc: linux-rdma@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: Re: [PATCH v1 3/4] NFSD: Refactor the generic write vector fill helper
Date: Thu, 15 Nov 2018 11:32:46 -0500 [thread overview]
Message-ID: <20181115163246.GB9022@fieldses.org> (raw)
In-Reply-To: <20180727151905.21878.62693.stgit@klimt.1015granger.net>
Um, I somehow managed to overlook a pynfs regression till just now:
On Fri, Jul 27, 2018 at 11:19:05AM -0400, Chuck Lever wrote:
> fill_in_write_vector() is nearly the same logic as
> svc_fill_write_vector(), but there are a few differences so that
> the former can handle multiple WRITE payloads in a single COMPOUND.
>
> svc_fill_write_vector() can be adjusted so that it can be used in
> the NFSv4 WRITE code path too. Instead of assuming the pages are
> coming from rq_args.pages, have the caller pass in the page list.
>
> The immediate benefit is a reduction of code duplication. It also
> prevents the NFSv4 WRITE decoder from passing an empty vector
> element when the transport has provided the payload in the xdr_buf's
> page array.
>
...
> @@ -1027,7 +1009,10 @@ static int fill_in_write_vector(struct kvec *vec, struct nfsd4_write *write)
> write->wr_how_written = write->wr_stable_how;
> gen_boot_verifier(&write->wr_verifier, SVC_NET(rqstp));
>
> - nvecs = fill_in_write_vector(rqstp->rq_vec, write);
> + nvecs = svc_fill_write_vector(rqstp, write->wr_pagelist,
> + &write->wr_head, write->wr_buflen);
> + if (!nvecs)
> + return nfserr_io;
Do you remember why you added this check?
It's causing zero-length writes to fail, in violation of the spec:
"If the count is zero, the WRITE will succeed and return a count
of zero subject to permissions checking."
I'm not seeing a reason why it wouldn't be safe just to remove that
check.
--b.
> WARN_ON_ONCE(nvecs > ARRAY_SIZE(rqstp->rq_vec));
>
> status = nfsd_vfs_write(rqstp, &cstate->current_fh, filp,
> diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c
> index f107f9f..a6faee5 100644
> --- a/fs/nfsd/nfsproc.c
> +++ b/fs/nfsd/nfsproc.c
> @@ -218,7 +218,8 @@
> SVCFH_fmt(&argp->fh),
> argp->len, argp->offset);
>
> - nvecs = svc_fill_write_vector(rqstp, &argp->first, cnt);
> + nvecs = svc_fill_write_vector(rqstp, rqstp->rq_arg.pages,
> + &argp->first, cnt);
> if (!nvecs)
> return nfserr_io;
> nfserr = nfsd_write(rqstp, fh_copy(&resp->fh, &argp->fh),
> diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h
> index 574368e..43f88bd 100644
> --- a/include/linux/sunrpc/svc.h
> +++ b/include/linux/sunrpc/svc.h
> @@ -496,6 +496,7 @@ int svc_register(const struct svc_serv *, struct net *, const int,
> struct svc_pool * svc_pool_for_cpu(struct svc_serv *serv, int cpu);
> char * svc_print_addr(struct svc_rqst *, char *, size_t);
> unsigned int svc_fill_write_vector(struct svc_rqst *rqstp,
> + struct page **pages,
> struct kvec *first, size_t total);
> char *svc_fill_symlink_pathname(struct svc_rqst *rqstp,
> struct kvec *first, size_t total);
> diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
> index 30a4226..2194ed5 100644
> --- a/net/sunrpc/svc.c
> +++ b/net/sunrpc/svc.c
> @@ -1537,16 +1537,16 @@ u32 svc_max_payload(const struct svc_rqst *rqstp)
> /**
> * svc_fill_write_vector - Construct data argument for VFS write call
> * @rqstp: svc_rqst to operate on
> + * @pages: list of pages containing data payload
> * @first: buffer containing first section of write payload
> * @total: total number of bytes of write payload
> *
> - * Returns the number of elements populated in the data argument array.
> + * Fills in rqstp::rq_vec, and returns the number of elements.
> */
> -unsigned int svc_fill_write_vector(struct svc_rqst *rqstp, struct kvec *first,
> - size_t total)
> +unsigned int svc_fill_write_vector(struct svc_rqst *rqstp, struct page **pages,
> + struct kvec *first, size_t total)
> {
> struct kvec *vec = rqstp->rq_vec;
> - struct page **pages;
> unsigned int i;
>
> /* Some types of transport can present the write payload
> @@ -1560,14 +1560,11 @@ unsigned int svc_fill_write_vector(struct svc_rqst *rqstp, struct kvec *first,
> ++i;
> }
>
> - WARN_ON_ONCE(rqstp->rq_arg.page_base != 0);
> - pages = rqstp->rq_arg.pages;
> while (total) {
> vec[i].iov_base = page_address(*pages);
> vec[i].iov_len = min_t(size_t, total, PAGE_SIZE);
> total -= vec[i].iov_len;
> ++i;
> -
> ++pages;
> }
>
next prev parent reply other threads:[~2018-11-15 16:32 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-27 15:18 [PATCH v1 0/4] NFS/RDMA server patches for v4.19 Chuck Lever
2018-07-27 15:18 ` [PATCH v1 1/4] svcrdma: Avoid releasing a page in svc_xprt_release() Chuck Lever
2018-07-27 15:18 ` [PATCH v1 2/4] svcrdma: Clean up Read chunk path Chuck Lever
2018-07-27 15:19 ` [PATCH v1 3/4] NFSD: Refactor the generic write vector fill helper Chuck Lever
2018-11-15 16:32 ` J. Bruce Fields [this message]
2018-11-16 0:19 ` Chuck Lever
2018-11-19 19:54 ` Bruce Fields
2018-11-19 19:58 ` Chuck Lever
2018-11-19 19:59 ` Bruce Fields
2018-07-27 15:19 ` [PATCH v1 4/4] NFSD: Handle full-length symlinks Chuck Lever
2018-08-01 14:14 ` J. Bruce Fields
2018-08-01 14:16 ` Chuck Lever
2018-08-01 14:36 ` [PATCH v1 0/4] NFS/RDMA server patches for v4.19 J. Bruce Fields
2018-08-01 15:14 ` Chuck Lever
2018-08-01 15:20 ` Bruce Fields
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=20181115163246.GB9022@fieldses.org \
--to=bfields@fieldses.org \
--cc=chuck.lever@oracle.com \
--cc=linux-nfs@vger.kernel.org \
--cc=linux-rdma@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;
as well as URLs for NNTP newsgroup(s).