From: "J. Bruce Fields" <bfields@redhat.com>
To: linux-nfs@vger.kernel.org
Cc: Christoph Hellwig <hch@infradead.org>,
"J. Bruce Fields" <bfields@redhat.com>
Subject: [PATCH 35/43] nfsd4: turn off zero-copy-read in exotic cases
Date: Sun, 11 May 2014 16:52:40 -0400 [thread overview]
Message-ID: <1399841568-19716-36-git-send-email-bfields@redhat.com> (raw)
In-Reply-To: <1399841568-19716-1-git-send-email-bfields@redhat.com>
From: "J. Bruce Fields" <bfields@redhat.com>
We currently allow only one read per compound, with operations before
and after whose responses will require no more than about a page to
encode.
While we don't expect clients to violate those limits any time soon,
this limitation isn't really condoned by the spec, so to future proof
the server we should lift the limitation.
At the same time we'd like to continue to support zero-copy reads.
Supporting multiple zero-copy-reads per compound would require a new
data structure to replace struct xdr_buf, which can represent only one
set of included pages.
So for now we plan to modify encode_read() to support either zero-copy
or non-zero-copy reads, and use some heuristics at the start of the
compound processing to decide whether a zero-copy read will work.
This will allow us to support more exotic compounds without introducing
a performance regression in the normal case.
Later patches handle those "exotic compounds", this one just makes sure
zero-copy is turned off in those cases.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
---
fs/nfsd/nfs4xdr.c | 13 +++++++++++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 4dba311..f69906d 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -1629,6 +1629,8 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
struct nfsd4_op *op;
bool cachethis = false;
int max_reply = 2 * RPC_MAX_AUTH_SIZE; /* uh, kind of a guess */
+ int readcount = 0;
+ int readbytes = 0;
int i;
READ_BUF(4);
@@ -1679,14 +1681,21 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp)
*/
cachethis |= nfsd4_cache_this_op(op);
- max_reply += nfsd4_max_reply(argp->rqstp, op);
+ if (op->opnum == OP_READ) {
+ readcount++;
+ readbytes += nfsd4_max_reply(argp->rqstp, op);
+ } else
+ max_reply += nfsd4_max_reply(argp->rqstp, op);
}
/* Sessions make the DRC unnecessary: */
if (argp->minorversion)
cachethis = false;
- svc_reserve(argp->rqstp, max_reply);
+ svc_reserve(argp->rqstp, max_reply + readbytes);
argp->rqstp->rq_cachetype = cachethis ? RC_REPLBUFF : RC_NOCACHE;
+ if (readcount > 1 || max_reply > PAGE_SIZE - 2*RPC_MAX_AUTH_SIZE)
+ argp->rqstp->rq_splice_ok = false;
+
DECODE_TAIL;
}
--
1.7.9.5
next prev parent reply other threads:[~2014-05-11 20:53 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-11 20:52 nfsd4 xdr encoding fixes v2 J. Bruce Fields
2014-05-11 20:52 ` [PATCH 01/43] nfsd4: embed xdr_stream in nfsd4_compoundres J. Bruce Fields
2014-05-12 5:34 ` Christoph Hellwig
2014-05-16 9:58 ` Kinglong Mee
2014-05-22 15:13 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 02/43] nfsd4: tweak nfsd4_encode_getattr to take xdr_stream J. Bruce Fields
2014-05-12 5:35 ` Christoph Hellwig
2014-05-12 16:06 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 03/43] nfsd4: move proc_compound xdr encode init to helper J. Bruce Fields
2014-05-12 5:36 ` Christoph Hellwig
2014-05-11 20:52 ` [PATCH 04/43] nfsd4: reserve head space for krb5 integ/priv info J. Bruce Fields
2014-05-12 5:37 ` Christoph Hellwig
2014-05-12 21:45 ` J. Bruce Fields
2014-05-13 5:05 ` Christoph Hellwig
2014-05-13 14:47 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 05/43] nfsd4: move nfsd4_operation to xdr4.h J. Bruce Fields
2014-05-12 5:41 ` Christoph Hellwig
2014-05-22 15:56 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 06/43] nfsd4: fix encoding of out-of-space replies J. Bruce Fields
2014-05-12 8:18 ` Christoph Hellwig
2014-05-12 21:47 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 07/43] nfsd4: allow space for final error return J. Bruce Fields
2014-05-12 8:18 ` Christoph Hellwig
2014-05-12 14:06 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 08/43] nfsd4: use xdr_reserve_space in attribute encoding J. Bruce Fields
2014-05-11 20:52 ` [PATCH 09/43] nfsd4: use xdr_stream throughout compound encoding J. Bruce Fields
2014-05-11 20:52 ` [PATCH 10/43] nfsd4: remove ADJUST_ARGS J. Bruce Fields
2014-05-11 20:52 ` [PATCH 11/43] nfsd4: no need for encode_compoundres to adjust lengths J. Bruce Fields
2014-05-11 20:52 ` [PATCH 12/43] nfsd4: keep xdr buf length updated J. Bruce Fields
2014-05-11 20:52 ` [PATCH 13/43] rpc: xdr_truncate_encode J. Bruce Fields
2014-05-11 20:52 ` [PATCH 14/43] nfsd4: use xdr_truncate_encode J. Bruce Fields
2014-05-11 20:52 ` [PATCH 15/43] nfsd4: "backfill" using write_bytes_to_xdr_buf J. Bruce Fields
2014-05-11 20:52 ` [PATCH 16/43] nfsd4: teach encoders to handle reserve_space failures J. Bruce Fields
2014-05-11 20:52 ` [PATCH 17/43] nfsd4: reserve space before inlining 0-copy pages J. Bruce Fields
2014-05-11 20:52 ` [PATCH 18/43] nfsd4: nfsd4_check_resp_size needn't recalculate length J. Bruce Fields
2014-05-11 20:52 ` [PATCH 19/43] nfsd4: remove redundant encode buffer size checking J. Bruce Fields
2014-05-11 20:52 ` [PATCH 20/43] nfsd4: size-checking cleanup J. Bruce Fields
2014-05-11 20:52 ` [PATCH 21/43] nfsd4: allow encoding across page boundaries J. Bruce Fields
2014-05-11 20:52 ` [PATCH 22/43] nfsd4: convert 4.1 replay encoding J. Bruce Fields
2014-05-11 20:52 ` [PATCH 23/43] nfsd4: don't try to encode conflicting owner if low on space J. Bruce Fields
2014-05-11 20:52 ` [PATCH 24/43] nfsd4: more precise nfsd4_max_reply J. Bruce Fields
2014-05-11 20:52 ` [PATCH 25/43] nfsd4: minor encode_read cleanup J. Bruce Fields
2014-05-11 20:52 ` [PATCH 26/43] nfsd4: nfsd4_check_resp_size should check against whole buffer J. Bruce Fields
2014-05-11 20:52 ` [PATCH 27/43] rpc: define xdr_restrict_buflen J. Bruce Fields
2014-05-11 20:52 ` [PATCH 28/43] nfsd4: adjust buflen to session channel limit J. Bruce Fields
2014-05-11 20:52 ` [PATCH 29/43] nfsd4: use session limits to release send buffer reservation J. Bruce Fields
2014-05-11 20:52 ` [PATCH 30/43] nfsd4: allow large readdirs J. Bruce Fields
2014-05-11 20:52 ` [PATCH 31/43] nfsd4: enforce rd_dircount J. Bruce Fields
2014-05-11 20:52 ` [PATCH 32/43] nfsd4: don't treat readlink like a zero-copy operation J. Bruce Fields
2014-05-11 20:52 ` [PATCH 33/43] nfsd4: better estimate of getattr response size J. Bruce Fields
2014-05-11 20:52 ` [PATCH 34/43] nfsd4: estimate sequence " J. Bruce Fields
2014-05-11 20:52 ` J. Bruce Fields [this message]
2014-05-11 20:52 ` [PATCH 36/43] nfsd4: nfsd_vfs_read doesn't use file handle parameter J. Bruce Fields
2014-05-11 20:52 ` [PATCH 37/43] nfsd4: separate splice and readv cases J. Bruce Fields
2014-05-11 20:52 ` [PATCH 38/43] nfsd4: allow exotic read compounds J. Bruce Fields
2014-05-11 20:52 ` [PATCH 39/43] nfsd4: really fix nfs4err_resource in 4.1 case J. Bruce Fields
2014-05-12 5:33 ` Christoph Hellwig
2014-05-12 14:18 ` J. Bruce Fields
2014-05-11 20:52 ` [PATCH 40/43] nfsd4: kill WRITE32 J. Bruce Fields
2014-05-11 20:52 ` [PATCH 41/43] nfsd4: kill WRITE64 J. Bruce Fields
2014-05-11 20:52 ` [PATCH 42/43] nfsd4: kill WRITEMEM J. Bruce Fields
2014-05-11 20:52 ` [PATCH 43/43] nfsd4: kill write32, write64 J. Bruce Fields
2014-05-12 8:20 ` nfsd4 xdr encoding fixes v2 Christoph Hellwig
2014-05-12 16:07 ` J. Bruce Fields
2014-05-12 16:11 ` Christoph Hellwig
2014-05-13 11:09 ` Christoph Hellwig
2014-05-13 14:48 ` J. Bruce Fields
2014-05-13 21:18 ` J. Bruce Fields
2014-05-13 21:33 ` J. Bruce Fields
2014-05-22 19:17 ` J. 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=1399841568-19716-36-git-send-email-bfields@redhat.com \
--to=bfields@redhat.com \
--cc=hch@infradead.org \
--cc=linux-nfs@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).