[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20090902140201.GA10854@elte.hu>
Date: Wed, 2 Sep 2009 16:02:01 +0200
From: Ingo Molnar <mingo@...e.hu>
To: Arjan van de Ven <arjan@...radead.org>
Cc: linux-kernel@...r.kernel.org, Karsten Keil <isdn@...ux-pingi.de>,
isdn4linux@...tserv.isdn4linux.de,
Andrew Morton <akpm@...ux-foundation.org>, tj@...e.hu
Subject: [PATCH, v3] isdn: Fix stack corruption in isdnloop_init()
* Arjan van de Ven <arjan@...radead.org> wrote:
> > - char rev[10];
> > + char rev[strlen(revision)+1];
> >
> > if ((p = strchr(revision, ':'))) {
> > strcpy(rev, p + 1);
>
> now it;s a runtime variable sized array.
> NotNice(tm)
ah, indeed - i thought GCC would figure out its constness but it
doesnt. v3 should fix this:
-------------->
>From aa9aff8c2633cf24ad6465d8bf5aa93aa5d91f83 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@...e.hu>
Date: Tue, 26 May 2009 21:18:22 +0200
Subject: [PATCH] isdn: Fix stack corruption in isdnloop_init()
-tip testing found this stack corruption and bootup crash
in the ISDN subsystem, reported by stackprotector:
[ 25.656688] calling isdn_init+0x0/0x2c2 @ 1
[ 25.660388] ISDN subsystem Rev: 1.1.2.3/1.1.2.3/1.1.2.2/1.1.2.3/1.1.2.2/1.1.2.2
[ 25.668179] initcall isdn_init+0x0/0x2c2 returned 0 after 6510 usecs
[ 25.670005] calling isdn_bsdcomp_init+0x0/0x45 @ 1
[ 25.673336] PPP BSD Compression module registered
[ 25.676674] initcall isdn_bsdcomp_init+0x0/0x45 returned 0 after 3255 usecs
[ 25.680005] calling isdnloop_init+0x0/0x88 @ 1
[ 25.683337] isdnloop-ISDN-driver Rev 1.11.6.7
[ 25.686705] isdnloop: (loop0) virtual card added
[ 25.690004] Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: c1de2d8b
[ 25.690006]
[ 25.693338] Pid: 1, comm: swapper Not tainted 2.6.31-rc8-tip-01250-geed031c-dirty #9565
[ 25.696672] Call Trace:
[ 25.700008] [<c190f517>] ? printk+0x1d/0x30
[ 25.703339] [<c190f45d>] panic+0x50/0xed
[ 25.706677] [<c1059194>] __stack_chk_fail+0x1e/0x42
[ 25.710005] [<c1de2d8b>] ? isdnloop_init+0x83/0x88
[ 25.713338] [<c1de2d8b>] isdnloop_init+0x83/0x88
[ 25.716674] [<c1001056>] _stext+0x56/0x15a
[ 25.720007] [<c1da8368>] kernel_init+0x8f/0xf1
[ 25.723338] [<c1da82d9>] ? kernel_init+0x0/0xf1
[ 25.726675] [<c1025c67>] kernel_thread_helper+0x7/0x58
[ 25.730005] Rebooting in 1 seconds..Press any key to enter the menu
The bug is that the temporary array:
char rev[10];
Is sized one byte too small to store strings based on
the 'revision' string.
This is a truly ancient bug: it has been introduced in
the v2.4.2.1 kernel, ~8.5 years ago, which extended
the length of 'revision' by 1 byte.
Instead of using a fixed size temporary array, size
it based on the 'revision' string.
Cc: <stable@...nel.org>
Signed-off-by: Ingo Molnar <mingo@...e.hu>
---
drivers/isdn/isdnloop/isdnloop.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/isdn/isdnloop/isdnloop.c b/drivers/isdn/isdnloop/isdnloop.c
index a335c85..ea07fdd 100644
--- a/drivers/isdn/isdnloop/isdnloop.c
+++ b/drivers/isdn/isdnloop/isdnloop.c
@@ -15,7 +15,7 @@
#include <linux/sched.h>
#include "isdnloop.h"
-static char *revision = "$Revision: 1.11.6.7 $";
+static char revision[] = "$Revision: 1.11.6.7 $";
static char *isdnloop_id = "loop0";
MODULE_DESCRIPTION("ISDN4Linux: Pseudo Driver that simulates an ISDN card");
@@ -1494,7 +1494,7 @@ static int __init
isdnloop_init(void)
{
char *p;
- char rev[10];
+ char rev[sizeof(revision)];
if ((p = strchr(revision, ':'))) {
strcpy(rev, p + 1);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists