serial: protect uart_port_dtr_rts() in uart_shutdown() too
PATCH 1/1 serial: protect uart_port_dtr_rts() in uart_shutdown() too
From: "Jiri Slaby (SUSE)" <[email protected]>
[ Upstream commit 602babaa84d627923713acaf5f7e9a4369e77473 ]
Commit af224ca2df29 (serial: core: Prevent unsafe uart port access, part
3) added few uport == NULL checks. It added one to uart_shutdown(), so the commit assumes, uport can be NULL in there. But right after that protection, there is an unprotected "uart_port_dtr_rts(uport, false);" call. That is invoked only if HUPCL is set, so I assume that is the reason why we do not see lots of these reports.
Or it cannot be NULL at this point at all for some reason :P.
Until the above is investigated, stay on the safe side and move this dereference to the if too.
I got this inconsistency from Coverity under CID 1585130. Thanks.
[Adapted over commit 5701cb8bf50e ("tty: Call ->dtr_rts() parameter active consistently") not in the tree]
--- drivers/tty/serial/serial_core.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-)
drivers/tty/serial/serial_core.c +9 -7
@@ -339,14 +339,16 @@ static void uart_shutdown(struct tty_struct *tty, struct uart_state *state)/* * Turn off DTR and RTS early. */ -if (uport &&uart_console (uport) && tty) { - uport->cons->cflag = tty->termios.c_cflag; - uport->cons->ispeed = tty->termios.c_ispeed; - uport->cons->ospeed = tty->termios.c_ospeed; - } +if (uport) { +if (uart_console (uport) && tty) { + uport->cons->cflag = tty->termios.c_cflag; + uport->cons->ispeed = tty->termios.c_ispeed; + uport->cons->ospeed = tty->termios.c_ospeed; + } -if (!tty ||C_HUPCL (tty)) -uart_port_dtr_rts (uport,0 ); +if (!tty ||C_HUPCL (tty)) +uart_port_dtr_rts (uport,0 ); + }uart_port_shutdown (port); }
Signature · 1 line
2.40.1