From: Chuck Lever <cel@kernel.org>
To: NeilBrown <neilb@ownmail.net>, Jeff Layton <jlayton@kernel.org>,
Olga Kornievskaia <okorniev@redhat.com>,
Dai Ngo <dai.ngo@oracle.com>, Tom Talpey <tom@talpey.com>
Cc: <linux-nfs@vger.kernel.org>, Chuck Lever <chuck.lever@oracle.com>
Subject: [PATCH v2 5/6] SUNRPC: Track consumed rq_pages entries
Date: Thu, 26 Feb 2026 09:47:38 -0500 [thread overview]
Message-ID: <20260226144739.193129-6-cel@kernel.org> (raw)
In-Reply-To: <20260226144739.193129-1-cel@kernel.org>
From: Chuck Lever <chuck.lever@oracle.com>
The rq_pages array holds pages allocated for incoming RPC requests.
Two transport receive paths NULL entries in rq_pages to prevent
svc_rqst_release_pages() from freeing pages that the transport has
taken ownership of:
- svc_tcp_save_pages() moves partial request data pages to
svsk->sk_pages during multi-fragment TCP reassembly.
- svc_rdma_clear_rqst_pages() moves request data pages to
head->rc_pages because they are targets of active RDMA Read WRs.
A new rq_pages_nfree field in struct svc_rqst records how many
entries were NULLed. svc_alloc_arg() uses it to refill only those
entries rather than scanning the full rq_pages array. In steady
state, the transport NULLs a handful of entries per RPC, so the
allocator visits only those entries instead of the full ~259 slots
(for 1MB messages).
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
include/linux/sunrpc/svc.h | 10 ++++++++++
net/sunrpc/svc.c | 1 +
net/sunrpc/svc_xprt.c | 11 ++++++++---
net/sunrpc/svcsock.c | 1 +
net/sunrpc/xprtrdma/svc_rdma_rw.c | 1 +
5 files changed, 21 insertions(+), 3 deletions(-)
diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h
index 3559de664f64..b5a842dd97a4 100644
--- a/include/linux/sunrpc/svc.h
+++ b/include/linux/sunrpc/svc.h
@@ -143,6 +143,15 @@ extern u32 svc_max_payload(const struct svc_rqst *rqstp);
* server thread needs to allocate more to replace those used in
* sending.
*
+ * rq_pages request page contract:
+ *
+ * Transport receive paths that move request data pages out of
+ * rq_pages -- TCP multi-fragment reassembly (svc_tcp_save_pages)
+ * and RDMA Read I/O (svc_rdma_clear_rqst_pages) -- NULL those
+ * entries to prevent svc_rqst_release_pages() from freeing pages
+ * still in transport use, and set rq_pages_nfree to the count.
+ * svc_alloc_arg() refills only that many rq_pages entries.
+ *
* xdr_buf holds responses; the structure fits NFS read responses
* (header, data pages, optional tail) and enables sharing of
* client-side routines.
@@ -201,6 +210,7 @@ struct svc_rqst {
struct folio *rq_scratch_folio;
struct xdr_buf rq_res;
unsigned long rq_maxpages; /* entries per page array */
+ unsigned long rq_pages_nfree; /* rq_pages entries NULLed by transport */
struct page * *rq_pages; /* Call buffer pages */
struct page * *rq_respages; /* Reply buffer pages */
struct page * *rq_next_page; /* next reply page to use */
diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
index 0ce16e9abdf6..6e57e35fa6d6 100644
--- a/net/sunrpc/svc.c
+++ b/net/sunrpc/svc.c
@@ -655,6 +655,7 @@ svc_init_buffer(struct svc_rqst *rqstp, const struct svc_serv *serv, int node)
return false;
}
+ rqstp->rq_pages_nfree = rqstp->rq_maxpages;
return true;
}
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index e027765f4307..795b5729525f 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -675,12 +675,17 @@ static bool svc_fill_pages(struct svc_rqst *rqstp, struct page **pages,
static bool svc_alloc_arg(struct svc_rqst *rqstp)
{
struct xdr_buf *arg = &rqstp->rq_arg;
- unsigned long pages;
+ unsigned long pages, nfree;
pages = rqstp->rq_maxpages;
- if (!svc_fill_pages(rqstp, rqstp->rq_pages, pages))
- return false;
+ nfree = rqstp->rq_pages_nfree;
+ if (nfree) {
+ if (!svc_fill_pages(rqstp, rqstp->rq_pages, nfree))
+ return false;
+ rqstp->rq_pages_nfree = 0;
+ }
+
if (!svc_fill_pages(rqstp, rqstp->rq_respages, pages))
return false;
rqstp->rq_next_page = rqstp->rq_respages;
diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c
index c86f28f720f7..2ce43f9995f1 100644
--- a/net/sunrpc/svcsock.c
+++ b/net/sunrpc/svcsock.c
@@ -1009,6 +1009,7 @@ static void svc_tcp_save_pages(struct svc_sock *svsk, struct svc_rqst *rqstp)
svsk->sk_pages[i] = rqstp->rq_pages[i];
rqstp->rq_pages[i] = NULL;
}
+ rqstp->rq_pages_nfree = npages;
}
static void svc_tcp_clear_pages(struct svc_sock *svsk)
diff --git a/net/sunrpc/xprtrdma/svc_rdma_rw.c b/net/sunrpc/xprtrdma/svc_rdma_rw.c
index 4ec2f9ae06aa..cf4a1762b629 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_rw.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_rw.c
@@ -1107,6 +1107,7 @@ static void svc_rdma_clear_rqst_pages(struct svc_rqst *rqstp,
head->rc_pages[i] = rqstp->rq_pages[i];
rqstp->rq_pages[i] = NULL;
}
+ rqstp->rq_pages_nfree = head->rc_page_count;
}
/**
--
2.53.0
next prev parent reply other threads:[~2026-02-26 14:47 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-02-26 14:47 [PATCH v2 0/6] Optimize NFSD buffer page management Chuck Lever
2026-02-26 14:47 ` [PATCH v2 1/6] SUNRPC: Tighten bounds checking in svc_rqst_replace_page Chuck Lever
2026-03-10 18:01 ` Jeff Layton
2026-02-26 14:47 ` [PATCH v2 2/6] SUNRPC: Allocate a separate Reply page array Chuck Lever
2026-03-10 18:10 ` Jeff Layton
2026-02-26 14:47 ` [PATCH v2 3/6] SUNRPC: Handle NULL entries in svc_rqst_release_pages Chuck Lever
2026-03-10 18:11 ` Jeff Layton
2026-02-26 14:47 ` [PATCH v2 4/6] svcrdma: preserve rq_next_page in svc_rdma_save_io_pages Chuck Lever
2026-03-10 18:13 ` Jeff Layton
2026-02-26 14:47 ` Chuck Lever [this message]
2026-03-10 18:16 ` [PATCH v2 5/6] SUNRPC: Track consumed rq_pages entries Jeff Layton
2026-02-26 14:47 ` [PATCH v2 6/6] SUNRPC: Optimize rq_respages allocation in svc_alloc_arg Chuck Lever
2026-03-10 18:18 ` Jeff Layton
2026-03-10 18:19 ` [PATCH v2 0/6] Optimize NFSD buffer page management Jeff Layton
2026-03-10 18:24 ` Chuck Lever
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=20260226144739.193129-6-cel@kernel.org \
--to=cel@kernel.org \
--cc=chuck.lever@oracle.com \
--cc=dai.ngo@oracle.com \
--cc=jlayton@kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=neilb@ownmail.net \
--cc=okorniev@redhat.com \
--cc=tom@talpey.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