From: Josef Bacik Date: Thu, 19 Jan 2017 21:08:49 +0000 (-0500) Subject: nbd: only set MSG_MORE when we have more to send X-Git-Tag: v3.2.87~84 X-Git-Url: https://git.openpandora.org/cgi-bin/gitweb.cgi?p=pandora-kernel.git;a=commitdiff_plain;h=5486d50a9d86b9ca21c5d1be97cc00b4bd2179a5 nbd: only set MSG_MORE when we have more to send commit d61b7f972dab2a7d187c38254845546dfc8eed85 upstream. A user noticed that write performance was horrible over loopback and we traced it to an inversion of when we need to set MSG_MORE. It should be set when we have more bvec's to send, not when we are on the last bvec. This patch made the test go from 20 iops to 78k iops. Signed-off-by: Josef Bacik Fixes: 429a787be679 ("nbd: fix use-after-free of rq/bio in the xmit path") Signed-off-by: Jens Axboe [bwh: Backported to 3.2: adjust context] Signed-off-by: Ben Hutchings --- diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 3d3f54a4c14c..2a80552633dd 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -232,7 +232,7 @@ static inline int sock_send_bvec(struct nbd_device *lo, struct bio_vec *bvec, /* always call with the tx_lock held */ static int nbd_send_req(struct nbd_device *lo, struct request *req) { - int result, flags; + int result; struct nbd_request request; unsigned long size = blk_rq_bytes(req); struct bio *bio; @@ -259,7 +259,6 @@ static int nbd_send_req(struct nbd_device *lo, struct request *req) if (nbd_cmd(req) != NBD_CMD_WRITE) return 0; - flags = 0; bio = req->bio; while (bio) { struct bio *next = bio->bi_next; @@ -268,9 +267,8 @@ static int nbd_send_req(struct nbd_device *lo, struct request *req) bio_for_each_segment(bvec, bio, i) { bool is_last = !next && i == bio->bi_vcnt - 1; + int flags = is_last ? 0 : MSG_MORE; - if (is_last) - flags = MSG_MORE; dprintk(DBG_TX, "%s: request %p: sending %d bytes data\n", lo->disk->disk_name, req, bvec->bv_len); result = sock_send_bvec(lo, bvec, flags);