The branch main has been updated by rmacklem:

URL: 
https://cgit.FreeBSD.org/src/commit/?id=2420e64c463a4873a7f0cb86837148bc66a6474a

commit 2420e64c463a4873a7f0cb86837148bc66a6474a
Author:     Rick Macklem <rmack...@freebsd.org>
AuthorDate: 2025-07-18 18:27:50 +0000
Commit:     Rick Macklem <rmack...@freebsd.org>
CommitDate: 2025-07-18 18:27:50 +0000

    rpctls_impl.c: Fix handling of hung rpc tls daemon
    
    Commit 26ee05939209 replaced soshutdown()
    with soclose() for the case where the RPC TLS
    daemon has not removed the socket from the RB
    tree during an upcall.
    
    It turns out that soshutdown() is needed for the
    "else" case, which is the case that occurs when the
    daemon is stuck in SSL_accept() or SSL_connect().
    
    This patch adds a soshutdown() call to this else case,
    which unsticks the daemon when the other end does
    not do the actual TLS handshake. This can occur if
    the daemon is not running at the other end.
    
    Reviewed by:    glebius
    Differential Revision:  https://reviews.freebsd.org/D51083
    Fixes:  26ee05939209 ("rpctls_impl.c: Fix handling of socket for daemon 
failure")
---
 sys/rpc/rpcsec_tls/rpctls_impl.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/sys/rpc/rpcsec_tls/rpctls_impl.c b/sys/rpc/rpcsec_tls/rpctls_impl.c
index 93fe283e65fd..51fe270b13d9 100644
--- a/sys/rpc/rpcsec_tls/rpctls_impl.c
+++ b/sys/rpc/rpcsec_tls/rpctls_impl.c
@@ -240,6 +240,14 @@ rpctls_rpc_failed(struct upsock *ups, struct socket *so)
                 * failed to do the handshake.
                 */
                mtx_unlock(&rpctls_lock);
+               /*
+                * Do a shutdown on the socket, since the daemon is
+                * probably stuck in SSL_accept() or SSL_connect() trying to
+                * read the socket.  Do not soclose() the socket, since the
+                * daemon will close() the socket after SSL_accept()
+                * returns an error.
+                */
+               soshutdown(so, SHUT_RD);
        }
 }
 

Reply via email to