From: Oleg Drokin <green@linuxhacker.ru>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
devel@driverdev.osuosl.org,
Andreas Dilger <andreas.dilger@intel.com>
Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Lustre Development List <lustre-devel@lists.lustre.org>,
Al Viro <viro@zeniv.linux.org.uk>
Subject: [PATCH 06/15] ksocknal_lib_send_kiov(): sendmsg doesn't bugger iovec...
Date: Sat, 23 Jul 2016 02:37:03 -0400 [thread overview]
Message-ID: <1469255832-746785-7-git-send-email-green@linuxhacker.ru> (raw)
In-Reply-To: <1469255832-746785-1-git-send-email-green@linuxhacker.ru>
From: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
.../lustre/lnet/klnds/socklnd/socklnd_lib.c | 29 ++++------------------
1 file changed, 5 insertions(+), 24 deletions(-)
diff --git a/drivers/staging/lustre/lnet/klnds/socklnd/socklnd_lib.c b/drivers/staging/lustre/lnet/klnds/socklnd/socklnd_lib.c
index 8eb4a68..6c95e98 100644
--- a/drivers/staging/lustre/lnet/klnds/socklnd/socklnd_lib.c
+++ b/drivers/staging/lustre/lnet/klnds/socklnd/socklnd_lib.c
@@ -106,10 +106,6 @@ ksocknal_lib_send_kiov(struct ksock_conn *conn, struct ksock_tx *tx)
/* Not NOOP message */
LASSERT(tx->tx_lnetmsg);
- /*
- * NB we can't trust socket ops to either consume our iovs
- * or leave them alone.
- */
if (tx->tx_msg.ksm_zc_cookies[0]) {
/* Zero copy is enabled */
struct sock *sk = sock->sk;
@@ -132,34 +128,19 @@ ksocknal_lib_send_kiov(struct ksock_conn *conn, struct ksock_tx *tx)
rc = tcp_sendpage(sk, page, offset, fragsize, msgflg);
}
} else {
-#if SOCKNAL_SINGLE_FRAG_TX || !SOCKNAL_RISK_KMAP_DEADLOCK
- struct kvec scratch;
- struct kvec *scratchiov = &scratch;
- unsigned int niov = 1;
-#else
-#ifdef CONFIG_HIGHMEM
-#warning "XXX risk of kmap deadlock on multiple frags..."
-#endif
- struct kvec *scratchiov = conn->ksnc_scheduler->kss_scratch_iov;
- unsigned int niov = tx->tx_nkiov;
-#endif
struct msghdr msg = {.msg_flags = MSG_DONTWAIT};
int i;
- for (nob = i = 0; i < niov; i++) {
- scratchiov[i].iov_base = kmap(kiov[i].bv_page) +
- kiov[i].bv_offset;
- nob += scratchiov[i].iov_len = kiov[i].bv_len;
- }
+ for (nob = i = 0; i < tx->tx_nkiov; i++)
+ nob += kiov[i].bv_len;
if (!list_empty(&conn->ksnc_tx_queue) ||
nob < tx->tx_resid)
msg.msg_flags |= MSG_MORE;
- rc = kernel_sendmsg(sock, &msg, (struct kvec *)scratchiov, niov, nob);
-
- for (i = 0; i < niov; i++)
- kunmap(kiov[i].bv_page);
+ iov_iter_bvec(&msg.msg_iter, WRITE | ITER_BVEC,
+ kiov, tx->tx_nkiov, nob);
+ rc = sock_sendmsg(sock, &msg);
}
return rc;
}
--
2.7.4
next prev parent reply other threads:[~2016-07-23 6:37 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-07-23 6:36 [PATCH 00/15] Lustre cleanups Oleg Drokin
2016-07-23 6:36 ` [PATCH 01/15] lustre: switch lnet_sock_write() to sock_sendmsg() Oleg Drokin
2016-07-23 6:36 ` [PATCH 02/15] lustre: simplify the living hell out of ksocknal_lib_recv_kiov() Oleg Drokin
2016-07-23 6:37 ` [PATCH 03/15] lustre: don't reinvent struct bio_vec Oleg Drokin
2016-07-23 6:37 ` [PATCH 04/15] ksocknal_lib_recv_iov(): recvmsg doesn't bugger iovec anymore Oleg Drokin
2016-07-23 6:37 ` [PATCH 05/15] ksocknal_lib_send_iov(): sendmsg doesn't bugger iovec Oleg Drokin
2016-07-23 6:37 ` Oleg Drokin [this message]
2016-07-23 6:37 ` [PATCH 07/15] lustre: ->kss_scratch... are unused now Oleg Drokin
2016-08-15 16:55 ` Greg Kroah-Hartman
2016-07-23 6:37 ` [PATCH 08/15] lustre: constify lib-move.c stuff Oleg Drokin
2016-07-23 6:37 ` [PATCH 09/15] lustre: pass iov_iter to ->lnd_recv() Oleg Drokin
2016-07-23 6:37 ` [PATCH 10/15] lustre: introduce lnet_copy_{k,}iov2iter(), kill lnet_copy_{k,}iov2{k,}iov() Oleg Drokin
2016-07-23 6:37 ` [PATCH 11/15] staging/lustre: Always return EEXIST on mkdir for existing names Oleg Drokin
2016-07-23 6:37 ` [PATCH 12/15] staging/lustre: Add spaces preferred around that '{+,-,*,/,|,<<,>>,&}' Oleg Drokin
2016-07-23 17:31 ` Joe Perches
2016-07-23 18:01 ` Oleg Drokin
2016-07-26 1:47 ` Joe Perches
2016-07-23 6:37 ` [PATCH 13/15] staging/lustre: Fix unnecessary parentheses around variables Oleg Drokin
2016-07-23 6:37 ` [PATCH 14/15] staging/lustre: Make alignment match open parenthesis Oleg Drokin
2016-07-23 6:37 ` [PATCH 15/15] staging/lustre: Remove unused cp_error from struct cl_page Oleg Drokin
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=1469255832-746785-7-git-send-email-green@linuxhacker.ru \
--to=green@linuxhacker.ru \
--cc=andreas.dilger@intel.com \
--cc=devel@driverdev.osuosl.org \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lustre-devel@lists.lustre.org \
--cc=viro@zeniv.linux.org.uk \
/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