[PATCH] ARM: kprobes: Change testcase with unpredictable STRD instruction

[PATCH] ARM: kprobes: Change testcase with unpredictable STRD instruction

[Tixy]

There is a kprobes testcase for the instruction "strd r2, [r3], r4".
This has unpredictable behaviour as it uses r3 for register writeback
addressing and also stores it to memory.

On a cortex A9, this testcase would fail because the instruction writes
the updated value of r3 to memory, whereas the kprobes emulation code
writes the original value.

Fix this by changing testcase to used r5 instead of r3.

Signed-off-by: Jon Medhurst <tixy@yxit.co.uk>
---
 arch/arm/kernel/kprobes-test-arm.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/arm/kernel/kprobes-test-arm.c b/arch/arm/kernel/kprobes-test-arm.c
index fc82de8..edf9ad8 100644
--- a/arch/arm/kernel/kprobes-test-arm.c
+++ b/arch/arm/kernel/kprobes-test-arm.c
@@ -550,7 +550,7 @@ void kprobe_arm_test_cases(void)
 	TEST_RPR(  "strccd	r",8, VAL2,", [r",13,0, ", r",12,48,"]")
 	TEST_RPR(  "strd	r",4, VAL1,", [r",2, 24,", r",3, 48,"]!")
 	TEST_RPR(  "strcsd	r",12,VAL2,", [r",11,48,", -r",10,24,"]!")
-	TEST_RPR(  "strd	r",2, VAL1,", [r",3, 24,"], r",4,48,"")
+	TEST_RPR(  "strd	r",2, VAL1,", [r",5, 24,"], r",4,48,"")
 	TEST_RPR(  "strd	r",10,VAL2,", [r",9, 48,"], -r",7,24,"")
 	TEST_UNSUPPORTED(".word 0xe1afc0fa	@ strd r12, [pc, r10]!")
 
-- 
1.7.2.5

[PATCH] ARM: kprobes: Change testcase with unpredictable STRD instruction

[Nicolas Pitre]

On Sun, 27 Nov 2011, Tixy wrote:

> There is a kprobes testcase for the instruction "strd r2, [r3], r4".
> This has unpredictable behaviour as it uses r3 for register writeback
> addressing and also stores it to memory.
> 
> On a cortex A9, this testcase would fail because the instruction writes
> the updated value of r3 to memory, whereas the kprobes emulation code
> writes the original value.
> 
> Fix this by changing testcase to used r5 instead of r3.
> 
> Signed-off-by: Jon Medhurst <tixy@yxit.co.uk>

Acked-by: Nicolas Pitre <nico@linaro.org>


> ---
>  arch/arm/kernel/kprobes-test-arm.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/arch/arm/kernel/kprobes-test-arm.c b/arch/arm/kernel/kprobes-test-arm.c
> index fc82de8..edf9ad8 100644
> --- a/arch/arm/kernel/kprobes-test-arm.c
> +++ b/arch/arm/kernel/kprobes-test-arm.c
> @@ -550,7 +550,7 @@ void kprobe_arm_test_cases(void)
>  	TEST_RPR(  "strccd	r",8, VAL2,", [r",13,0, ", r",12,48,"]")
>  	TEST_RPR(  "strd	r",4, VAL1,", [r",2, 24,", r",3, 48,"]!")
>  	TEST_RPR(  "strcsd	r",12,VAL2,", [r",11,48,", -r",10,24,"]!")
> -	TEST_RPR(  "strd	r",2, VAL1,", [r",3, 24,"], r",4,48,"")
> +	TEST_RPR(  "strd	r",2, VAL1,", [r",5, 24,"], r",4,48,"")
>  	TEST_RPR(  "strd	r",10,VAL2,", [r",9, 48,"], -r",7,24,"")
>  	TEST_UNSUPPORTED(".word 0xe1afc0fa	@ strd r12, [pc, r10]!")
>  
> -- 
> 1.7.2.5
> 
> 
> 
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 

[PATCH] ARM: kprobes: Change testcase with unpredictable STRD instruction

[Russell King - ARM Linux]

On Sun, Nov 27, 2011 at 12:11:25PM +0000, Tixy wrote:
> There is a kprobes testcase for the instruction "strd r2, [r3], r4".
> This has unpredictable behaviour as it uses r3 for register writeback
> addressing and also stores it to memory.
> 
> On a cortex A9, this testcase would fail because the instruction writes
> the updated value of r3 to memory, whereas the kprobes emulation code
> writes the original value.
> 
> Fix this by changing testcase to used r5 instead of r3.

Is this a -stable candidate?

[PATCH] ARM: kprobes: Change testcase with unpredictable STRD instruction

[Tixy]

On Wed, 2011-11-30 at 23:56 +0000, Russell King - ARM Linux wrote:
> On Sun, Nov 27, 2011 at 12:11:25PM +0000, Tixy wrote:
> > There is a kprobes testcase for the instruction "strd r2, [r3], r4".
> > This has unpredictable behaviour as it uses r3 for register writeback
> > addressing and also stores it to memory.
> > 
> > On a cortex A9, this testcase would fail because the instruction writes
> > the updated value of r3 to memory, whereas the kprobes emulation code
> > writes the original value.
> > 
> > Fix this by changing testcase to used r5 instead of r3.
> 
> Is this a -stable candidate?

No, the test code is new in 3.2

-- 
Tixy