netdev - [PATCH] hso_serial_open hso_serial

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Date:	Thu, 09 Oct 2008 13:22:59 +0200
From:	Denis Joseph Barrow <D.Barow@...ion.com>
To:	Linux USB kernel mailing list <linux-usb@...r.kernel.org>,
	Linux netdev Mailing list <netdev@...r.kernel.org>,
	Greg Kroah-Hartman <gregkh@...e.de>
Subject: [PATCH] hso_serial_open hso_serial_close fixups

Hi all,
After my discussion with Alan Cox this should make hot disconnection of
the hso device more stable.

Disconnection still sometimes crashes the kernel when a modem port is connected
over ppp because the large patch for the hso mutex stuff hasn't yet been accepted.
i.e. unplugging a modem while using a ppp connection is in use is much more stable for our
internally maintained driver.



Moved serial_open_count in hso_serial_open to
prevent crashes owing to the serial structure being made NULL
when hso_serial_close is called even though hso_serial_open
returned -ENODEV, Alan Cox pointed out this happens,
also put in sanity check in hso_serial_close
to check for a valid serial structure which should prevent
the most reproducable crash in the driver when the hso device
is disconnected while in use.
Signed-off-by: Denis Joseph Barrow <D.Barow@...ion.com>
---
Index: linux-2.6.27-rc9/drivers/net/usb/hso.c
===================================================================
--- linux-2.6.27-rc9.orig/drivers/net/usb/hso.c	2008-10-09 12:40:41.000000000 +0200
+++ linux-2.6.27-rc9/drivers/net/usb/hso.c	2008-10-09 12:46:50.000000000 +0200
@@ -1053,6 +1053,11 @@
 	}
 
 	mutex_lock(&serial->parent->mutex);
+	/* check for port already opened, if not set the termios */
+	/* The serial->open count needs to be here as hso_serial_close
+	 *  will be called even if hso_serial_open returns -ENODEV.
+	 */
+	serial->open_count++;
 	result = usb_autopm_get_interface(serial->parent->interface);
 	if (result < 0)
 		goto err_out;
@@ -1064,8 +1069,6 @@
 	tty->driver_data = serial;
 	serial->tty = tty;
 
-	/* check for port allready opened, if not set the termios */
-	serial->open_count++;
 	if (serial->open_count == 1) {
 		tty->low_latency = 1;
 		serial->flags = 0;
@@ -1098,6 +1101,10 @@
 	u8 usb_gone;
 
 	D1("Closing serial port");
+	if (serial == NULL || serial->magic != HSO_SERIAL_MAGIC) {
+		D1("invalid serial structure bailing out. ");
+		return;
+	}
 
 	mutex_lock(&serial->parent->mutex);
 	usb_gone = serial->parent->usb_gone;

-- 
best regards,
D.J. Barrow

View attachment "hso_close_fix.patch" of type "text/x-diff" (1654 bytes)