5.15-stable review patch. If anyone has any objections, please let me know. ------------------ From: Johan Hovold [ Upstream commit 56964f6a31238fb41d5830bb42520c5b10ff7d8f ] Make sure to restore the TTY_IO_ERROR flag if activation fails to prevent I/O attempts on a port that is gone or powered down. This specifically avoids issues like kernel panic due to unclocked accesses or NULL-pointer dereferences when accessing powered down or deregistered ports when a TTY ioctl races with hangup and open. Note that the flag is set again in tty_port_close() but not restoring it on activation failure leaves a window where drivers fail to handle the ioctl-hangup race. Fixes: a9a37ec33a1b ("tty: tty_port: Move the IO_ERROR clear") Cc: stable # 2.6.33 Signed-off-by: Johan Hovold Link: https://patch.msgid.link/20260908084000.218528-1-johan@kernel.org Signed-off-by: Greg Kroah-Hartman [ sashal: Adapted to the existing activation-failure block with explicit mutex unlocking. ] Signed-off-by: Sasha Levin --- drivers/tty/tty_port.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/tty/tty_port.c b/drivers/tty/tty_port.c index 2f1061a9d9263..3afdbe5ea156d 100644 --- a/drivers/tty/tty_port.c +++ b/drivers/tty/tty_port.c @@ -694,6 +694,7 @@ int tty_port_open(struct tty_port *port, struct tty_struct *tty, int retval = port->ops->activate(port, tty); if (retval) { + set_bit(TTY_IO_ERROR, &tty->flags); mutex_unlock(&port->mutex); return retval; } -- 2.53.0