[BUG] check: we must not check for error before reading a response
We can receive data with a notification of socket error. But we
must not check for the error before reading the data, because it
may be an asynchronous error notification that we check too early
while the response we're waiting for is available. If there is an
error, recv() will get it.
This should help with servers that close very fast after the response
and should also slightly lower the CPU usage during very fast checks
on massive amounts of servers since we eliminate one system call.
This should probably be backported to 1.3.
diff --git a/src/checks.c b/src/checks.c
index ed5ca84..7eeced2 100644
--- a/src/checks.c
+++ b/src/checks.c
@@ -768,19 +768,14 @@
int len;
struct task *t = fdtab[fd].owner;
struct server *s = t->context;
- int skerr;
char *desc;
- socklen_t lskerr = sizeof(skerr);
len = -1;
if (unlikely((s->result & SRV_CHK_ERROR) ||
(fdtab[fd].state == FD_STERROR) ||
- (fdtab[fd].ev & FD_POLL_ERR) ||
- (getsockopt(fd, SOL_SOCKET, SO_ERROR, &skerr, &lskerr) == -1) ||
- (skerr != 0))) {
+ (fdtab[fd].ev & FD_POLL_ERR))) {
/* in case of TCP only, this tells us if the connection failed */
-
if (!(s->result & SRV_CHK_ERROR))
set_server_check_status(s, HCHK_STATUS_SOCKERR, NULL);
@@ -792,10 +787,16 @@
* works correctly and we don't need to do the getsockopt() on linux.
*/
len = recv(fd, trash, sizeof(trash), 0);
- if (unlikely(len < 0 && errno == EAGAIN)) {
- /* we want some polling to happen first */
- fdtab[fd].ev &= ~FD_POLL_IN;
- return 0;
+ if (unlikely(len < 0)) {
+ if (errno == EAGAIN) {
+ /* not ready, we want to poll first */
+ fdtab[fd].ev &= ~FD_POLL_IN;
+ return 0;
+ }
+ /* network error, report it */
+ if (!(s->result & SRV_CHK_ERROR))
+ set_server_check_status(s, HCHK_STATUS_SOCKERR, NULL);
+ goto out_wakeup;
}
if (len < sizeof(trash))