MIPS: stack protector: Fix per-task canary switch
diff mbox

Message ID -
State Accepted
Delegated to: Ralf Baechle
Headers show

Commit Message

James Hogan Oct. 7, 2013, 11:14 a.m. UTC
Commit 1400eb6 (MIPS: r4k,octeon,r2300: stack protector: change canary
per task) was merged in v3.11 and introduced assembly in the MIPS resume
functions to update the value of the current canary in
__stack_chk_guard. However it used PTR_L resulting in a load of the
canary value, instead of PTR_LA to construct its address. The value is
intended to be random but is then treated as an address in the
subsequent LONG_S (store).

This was observed to cause a fault and panic:

CPU 0 Unable to handle kernel paging request at virtual address 139fea20, epc == 8000cc0c, ra == 8034f2a4
Oops[#1]:
...
$24   : 139fea20 1e1f7cb6
...
Call Trace:
[<8000cc0c>] resume+0xac/0x118
[<8034f2a4>] __schedule+0x5f8/0x78c
[<8034f4e0>] schedule_preempt_disabled+0x20/0x2c
[<80348eec>] rest_init+0x74/0x84
[<804dc990>] start_kernel+0x43c/0x454
Code: 3c18804b  8f184030  8cb901f8 <af190000> 00c0e021  8cb002f0 8cb102f4  8cb202f8  8cb302fc

This can also be forced by modifying
arch/mips/include/asm/stackprotector.h so that the default
__stack_chk_guard value is more likely to be a bad (or unaligned)
pointer.

Fix it to use PTR_LA instead, to load the address of the canary value,
which the LONG_S can then use to write into it.

Reported-by: bobjones (via #mipslinux on IRC)
Signed-off-by: James Hogan <>
Cc: Ralf Baechle <>
Cc: Gregory Fong <>
Cc: 
Cc: <> #3.11
---
Ralf: This is a regression in v3.11, so please consider for v3.12.

(Sorry, resent to Cc stable correctly)

 arch/mips/kernel/octeon_switch.S | 2 +-
 arch/mips/kernel/r2300_switch.S  | 2 +-
 arch/mips/kernel/r4k_switch.S    | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

Comments

root Oct. 7, 2013, 12:48 p.m. UTC | #1
On Mon, Oct 07, 2013 at 12:14:26PM +0100, James Hogan wrote:

> Ralf: This is a regression in v3.11, so please consider for v3.12.

Applied, will send to Linus with the next pull request.

stable folks - please apply to 3.12-stable.

Thanks!

  Ralf
Greg Kroah-Hartman Oct. 10, 2013, 11:16 p.m. UTC | #2
On Mon, Oct 07, 2013 at 02:48:59PM +0200, Ralf Baechle wrote:
> On Mon, Oct 07, 2013 at 12:14:26PM +0100, James Hogan wrote:
> 
> > Ralf: This is a regression in v3.11, so please consider for v3.12.
> 
> Applied, will send to Linus with the next pull request.

Which would be in time for 3.12-final, right?

> stable folks - please apply to 3.12-stable.

There is no 3.12-stable yet, as 3.12-final isn't out yet.

Confused,

greg k-h
James Hogan Oct. 14, 2013, 8:40 a.m. UTC | #3
Hi Greg,

On 11/10/13 00:16, Greg KH wrote:
> On Mon, Oct 07, 2013 at 02:48:59PM +0200, Ralf Baechle wrote:
>> On Mon, Oct 07, 2013 at 12:14:26PM +0100, James Hogan wrote:
>>
>>> Ralf: This is a regression in v3.11, so please consider for v3.12.
>>
>> Applied, will send to Linus with the next pull request.
> 
> Which would be in time for 3.12-final, right?

Yes, it's included in v3.12-rc5 as:
8b3c569a3999a8fd5a819f892525ab5520777c92

>> stable folks - please apply to 3.12-stable.
> 
> There is no 3.12-stable yet, as 3.12-final isn't out yet.
> 
> Confused,

I think Ralf meant v3.11-stable. It's only needed there as the
regression was introduced in v3.11-rc1.

Thanks
James
root Oct. 14, 2013, 10:27 a.m. UTC | #4
On Mon, Oct 14, 2013 at 09:40:27AM +0100, James Hogan wrote:

> On 11/10/13 00:16, Greg KH wrote:
> > On Mon, Oct 07, 2013 at 02:48:59PM +0200, Ralf Baechle wrote:
> >> On Mon, Oct 07, 2013 at 12:14:26PM +0100, James Hogan wrote:
> >>
> >>> Ralf: This is a regression in v3.11, so please consider for v3.12.
> >>
> >> Applied, will send to Linus with the next pull request.
> > 
> > Which would be in time for 3.12-final, right?
> 
> Yes, it's included in v3.12-rc5 as:
> 8b3c569a3999a8fd5a819f892525ab5520777c92
> 
> >> stable folks - please apply to 3.12-stable.
> > 
> > There is no 3.12-stable yet, as 3.12-final isn't out yet.
> > 
> > Confused,
> 
> I think Ralf meant v3.11-stable. It's only needed there as the
> regression was introduced in v3.11-rc1.

Yes, indeed.  Greg, sorry for the confusion.

Thanks,

  Ralf

Patch
diff mbox

diff --git a/arch/mips/kernel/octeon_switch.S b/arch/mips/kernel/octeon_switch.S
index 4204d76..029e002 100644
--- a/arch/mips/kernel/octeon_switch.S
+++ b/arch/mips/kernel/octeon_switch.S
@@ -73,7 +73,7 @@ 
 3:
 
 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
-	PTR_L	t8, __stack_chk_guard
+	PTR_LA	t8, __stack_chk_guard
 	LONG_L	t9, TASK_STACK_CANARY(a1)
 	LONG_S	t9, 0(t8)
 #endif
diff --git a/arch/mips/kernel/r2300_switch.S b/arch/mips/kernel/r2300_switch.S
index 38af83f..20b7b04 100644
--- a/arch/mips/kernel/r2300_switch.S
+++ b/arch/mips/kernel/r2300_switch.S
@@ -67,7 +67,7 @@  LEAF(resume)
 1:
 
 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
-	PTR_L	t8, __stack_chk_guard
+	PTR_LA	t8, __stack_chk_guard
 	LONG_L	t9, TASK_STACK_CANARY(a1)
 	LONG_S	t9, 0(t8)
 #endif
diff --git a/arch/mips/kernel/r4k_switch.S b/arch/mips/kernel/r4k_switch.S
index 921238a..078de5e 100644
--- a/arch/mips/kernel/r4k_switch.S
+++ b/arch/mips/kernel/r4k_switch.S
@@ -69,7 +69,7 @@ 
 1:
 
 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
-	PTR_L	t8, __stack_chk_guard
+	PTR_LA	t8, __stack_chk_guard
 	LONG_L	t9, TASK_STACK_CANARY(a1)
 	LONG_S	t9, 0(t8)
 #endif