Serious bug, acks for partial writes were missing. Unfortunately this was not seen while normal work. In vstorage context it shows itself when a CS is restarted. Then client may hang for 5 seconds closing each socket, which triggers shaman timeout.
https://pmc.acronis.work/browse/VSTOR-84977 Signed-off-by: Alexey Kuznetsov <kuz...@acronis.com> --- net/unix/af_unix.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 2f4d192..5ca5b32 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -2171,7 +2171,12 @@ static int unix_stream_sendmsg(struct socket *sock, struct msghdr *msg, send_sig(SIGPIPE, current, 0); err = -EPIPE; out_err: - net_zcopy_put_abort(uarg, true); + if (uarg) { + if (!sent) + net_zcopy_put_abort(uarg, true); + else + net_zcopy_put(uarg); + } scm_destroy(&scm); return sent ? : err; } -- 1.8.3.1 _______________________________________________ Devel mailing list Devel@openvz.org https://lists.openvz.org/mailman/listinfo/devel
