mirrors/passt
1
0
Fork 0
mirror of https://passt.top/passt synced 2025-05-05 18:28:52 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

tcp_splice: Don't clobber errno before checking for EAGAIN

Browse source 
Like many places, tcp_splice_sock_handler() needs to handle EAGAIN
specially, in this case for both of its splice() calls.  Unfortunately it
tests for EAGAIN some time after those calls.  In between there has been
at least a flow_trace() which could have clobbered errno.  Move the test on
errno closer to the relevant system calls to avoid this problem.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This commit is contained in:
David Gibson 2025-04-09 16:35:41 +10:00 • committed by Stefano Brivio
parent d3f33f3b8e
commit 6693fa1158
1 changed files with 10 additions and 8 deletions

18
tcp_splice.c
View file

@ -526,13 +526,15 @@ retry:
c->tcp.pipe_size,
SPLICE_F_MOVE | SPLICE_F_NONBLOCK);
while (readlen < 0 && errno == EINTR);
if (readlen < 0 && errno != EAGAIN)
goto close;
flow_trace(conn, "%zi from read-side call", readlen);
if (readlen < 0) {
if (errno != EAGAIN)
goto close;
} else if (!readlen) {
if (!readlen) {
eof = 1;
} else {
} else if (readlen > 0) {
never_read = 0;
if (readlen >= (long)c->tcp.pipe_size * 90 / 100)
@ -549,6 +551,9 @@ retry:
SPLICE_F_MOVE | more | SPLICE_F_NONBLOCK);
while (written < 0 && errno == EINTR);
if (written < 0 && errno != EAGAIN)
goto close;
flow_trace(conn, "%zi from write-side call (passed %zi)",
written, c->tcp.pipe_size);
@ -580,9 +585,6 @@ retry:
conn->written[fromsidei] += written > 0 ? written : 0;
if (written < 0) {
if (errno != EAGAIN)
goto close;
if (conn->read[fromsidei] == conn->written[fromsidei])
break;