From 5a79ba6272814dda72b387ea13b97d1f0e79dbcf Mon Sep 17 00:00:00 2001 From: David Gibson Date: Tue, 7 Nov 2023 13:42:42 +1100 Subject: [PATCH] tcp_splice: Don't handle EPOLL_CTL_DEL as part of tcp_splice_epoll_ctl() tcp_splice_epoll_ctl() removes both sockets from the epoll set if called when conn->flags & CLOSING. This will always happen immediately after setting that flag, since conn_flag_do() makes the call itself. That's also the _only_ time it can happen: we perform the EPOLL_CTL_DEL without clearing the conn->in_epoll flag, meaning that any further calls to tcp_splice_epoll_ctl() would attempt EPOLL_CTL_MOD, which would necessarily fail since the fds are no longer in the epoll. The EPOLL_CTL_DEL path in tcp_splice_epoll_ctl() has essentially zero overlap with anything else the function does, so just move them to be open coded in conn_flag_do(). This does require kernel 2.6.9 or later, in order to pass NULL as the event structure for epoll_ctl(). However, we already require at least 3.13 to allow unprivileged user namespaces. Given that, simply directly perform the EPOLL_CTL_DEL operations from conn_flag_do() rather than unnecessarily multiplexini Signed-off-by: David Gibson Signed-off-by: Stefano Brivio --- tcp_splice.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-) diff --git a/tcp_splice.c b/tcp_splice.c index 6c4c664..570a8c6 100644 --- a/tcp_splice.c +++ b/tcp_splice.c @@ -152,8 +152,10 @@ static void conn_flag_do(const struct ctx *c, struct tcp_splice_conn *conn, } } - if (flag == CLOSING) - tcp_splice_epoll_ctl(c, conn); + if (flag == CLOSING) { + epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->a, NULL); + epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->b, NULL); + } } #define conn_flag(c, conn, flag) \ @@ -182,12 +184,6 @@ static int tcp_splice_epoll_ctl(const struct ctx *c, struct epoll_event ev_b = { .data.u64 = ref_b.u64 }; uint32_t events_a, events_b; - if (conn->flags & CLOSING) { - epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->a, &ev_a); - epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->b, &ev_b); - return 0; - } - tcp_splice_conn_epoll_events(conn->events, &events_a, &events_b); ev_a.events = events_a; ev_b.events = events_b;