From: Eric Dumazet Date: Sun, 6 Jan 2013 18:21:49 +0000 (+0000) Subject: tcp: fix MSG_SENDPAGE_NOTLAST logic X-Git-Tag: v2.6.34.15~173 X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=fb883a89a99481dca7cd6b9ef942c35165ff2d54;p=thirdparty%2Fkernel%2Fstable.git tcp: fix MSG_SENDPAGE_NOTLAST logic commit ae62ca7b03217be5e74759dc6d7698c95df498b3 upstream. commit 35f9c09fe9c72e (tcp: tcp_sendpages() should call tcp_push() once) added an internal flag : MSG_SENDPAGE_NOTLAST meant to be set on all frags but the last one for a splice() call. The condition used to set the flag in pipe_to_sendpage() relied on splice() user passing the exact number of bytes present in the pipe, or a smaller one. But some programs pass an arbitrary high value, and the test fails. The effect of this bug is a lack of tcp_push() at the end of a splice(pipe -> socket) call, and possibly very slow or erratic TCP sessions. We should both test sd->total_len and fact that another fragment is in the pipe (pipe->nrbufs > 1) Many thanks to Willy for providing very clear bug report, bisection and test programs. Reported-by: Willy Tarreau Bisected-by: Willy Tarreau Tested-by: Willy Tarreau Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Paul Gortmaker --- diff --git a/fs/splice.c b/fs/splice.c index 3bec7c63be642..1c991d6a64b46 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -640,8 +640,10 @@ static int pipe_to_sendpage(struct pipe_inode_info *pipe, ret = buf->ops->confirm(pipe, buf); if (!ret) { more = (sd->flags & SPLICE_F_MORE) ? MSG_MORE : 0; - if (sd->len < sd->total_len) + + if (sd->len < sd->total_len && pipe->nrbufs > 1) more |= MSG_SENDPAGE_NOTLAST; + if (file->f_op && file->f_op->sendpage) ret = file->f_op->sendpage(file, buf->page, buf->offset, sd->len, &pos, more);