On at least one occassion, I saw the repair queue fail with an EAGAIN.
libsoccr doesn't do this, so I'm not sure whether this is really the right
thing.
Signed-off-by: David Gibson <david(a)gibson.dropbear.id.au>
---
tcp.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tcp.c b/tcp.c
index f18b2913..39035e14 100644
--- a/tcp.c
+++ b/tcp.c
@@ -2957,6 +2957,8 @@ static int tcp_flow_repair_queue(int s, size_t len, uint8_t *buf)
ssize_t rc = send(s, p, MIN(len, chunk), 0);
if (rc < 0) {
+ if (errno == EAGAIN || errno == EWOULDBLOCK)
+ continue;
if ((errno == ENOBUFS || errno == ENOMEM) &&
chunk >= TCP_MIGRATE_RESTORE_CHUNK_MIN) {
chunk /= 2;
--
2.48.1