arm64: Make TSK_STACK_CANARY more accurate defined
Subsystems (2)
[email protected]
maintainerPATCH 1/1 arm64: Make TSK_STACK_CANARY more accurate defined
From: Guo Ren <[email protected]>
TSK_STACK_CANARY only used in arm64/Makefile with CONFIG_STACKPROTECTOR_PER_TASK wrap. So use the same policy in asm-offset.c.
--- arch/arm64/kernel/asm-offsets.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/kernel/asm-offsets.c +1 -1
@@ -39,7 +39,7 @@ int main(void) DEFINE(TSK_TI_SCS_SP, offsetof(struct task_struct, thread_info.scs_sp));#endif DEFINE (TSK_STACK , offsetof(struct task_struct , stack)); -#ifdef CONFIG_STACKPROTECTOR +#ifdef CONFIG_STACKPROTECTOR_PER_TASK DEFINE (TSK_STACK_CANARY , offsetof(struct task_struct , stack_canary));#endif BLANK ();
Signature · 1 line
2.7.4
Will Deacon
maintainerQuoted 19 lines · On Mon, Jul 13, 2020 at 04:03:33AM +0000, [email protected] wrote:
From: Guo Ren <[email protected]>
TSK_STACK_CANARYonly used in arm64/Makefile withCONFIG_STACKPROTECTOR_PER_TASKwrap. So use the same policy inasm-offset.c.Signed-off-by Guo Ren <[email protected]>Co-developed-by Kees Cook <[email protected]>Cc Catalin Marinas <[email protected]>Cc Will Deacon <[email protected]>--- arch/arm64/kernel/asm-offsets.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/arm64/kernel/asm-offsets.c +0
-0
arch/arm64/kernel/asm-offsets.c int main(void) · line 44
Quoted 7 lines
arch/arm64/kernel/asm-offsets.c +1
-1@@ -39,7 +39,7 @@ int main(void) DEFINE(TSK_TI_SCS_SP, offsetof(struct task_struct, thread_info.scs_sp));#endif DEFINE (TSK_STACK , offsetof(struct task_struct , stack)); -#ifdef CONFIG_STACKPROTECTOR +#ifdef CONFIG_STACKPROTECTOR_PER_TASK DEFINE (TSK_STACK_CANARY , offsetof(struct task_struct , stack_canary));#endif
I don't think this really makese much sense. The 'stack_canary' field in 'struct task_struct' is defined as:
#ifdef CONFIG_STACKPROTECTOR
/* Canary value for the -fstack-protector GCC feature: */
unsigned long stack_canary;#endif
so I think it makes sense to follow that in asm-offsets.c
Does the current code actually cause a problem?
Will
Quoted 30 lines · Will Deacon
Quoted 19 lines · On Mon, Jul 13, 2020 at 04:03:33AM +0000, [email protected] wrote:
From: Guo Ren <[email protected]>
TSK_STACK_CANARYonly used in arm64/Makefile withCONFIG_STACKPROTECTOR_PER_TASKwrap. So use the same policy inasm-offset.c.Signed-off-by Guo Ren <[email protected]>Co-developed-by Kees Cook <[email protected]>Cc Catalin Marinas <[email protected]>Cc Will Deacon <[email protected]>--- arch/arm64/kernel/asm-offsets.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/arm64/kernel/asm-offsets.c +0
-0arch/arm64/kernel/asm-offsets.c int main(void) · line 44
Quoted 7 lines
arch/arm64/kernel/asm-offsets.c +1
-1@@ -39,7 +39,7 @@ int main(void) DEFINE(TSK_TI_SCS_SP, offsetof(struct task_struct, thread_info.scs_sp));#endif DEFINE (TSK_STACK , offsetof(struct task_struct , stack)); -#ifdef CONFIG_STACKPROTECTOR +#ifdef CONFIG_STACKPROTECTOR_PER_TASK DEFINE (TSK_STACK_CANARY , offsetof(struct task_struct , stack_canary));#endif I don't think this really makese much sense. The '
stack_canary' field in 'structtask_struct' is defined as:
Quoted 8 lines
#ifdef CONFIG_STACKPROTECTOR /* Canary value for the -fstack-protector GCC feature: */ unsigned long stack_canary;#endif
so I think it makes sense to follow that in
asm-offsets.cDoes the current code actually cause a problem?
No, I just want to know how arm64 reply, ref:
https://lore.kernel.org/linux-riscv/[email protected]/T/#t
Best Regards
Guo Ren
Guo Ren
maintainerBTW, Jim found a GCC security leak in arm64, and would you want to have a look at it?
-------
I notice in the epilogue I get
ld a4, 8(sp)
ld a5, 100(t6)
xor a5, a4, a5
bne a5,zero,.L4This looks like a security leak that the canary value is left in a4. The i386 implementation operates directly on memory without loading into registers. The rs6000 implementation is careful to load 0 into the other register in the stack_protector_test code after the xor. I think this is a bug in the aarch64 code that it isn't clearing the other register. And I think it is a bug in your code too. If we don't need to clear the canary from the two registers, then you should eliminate the xor and just use "bne a5,a4,.L4". But I think the way you have it is right, you just need to clear the a4 register after the xor.
--------
[1] https://gcc.gnu.org/pipermail/gcc-patches/2020-July/549910.html
Quoted 35 lines · Will Deacon
Quoted 19 lines · On Mon, Jul 13, 2020 at 04:03:33AM +0000, [email protected] wrote:
From: Guo Ren <[email protected]>
TSK_STACK_CANARYonly used in arm64/Makefile withCONFIG_STACKPROTECTOR_PER_TASKwrap. So use the same policy inasm-offset.c.Signed-off-by Guo Ren <[email protected]>Co-developed-by Kees Cook <[email protected]>Cc Catalin Marinas <[email protected]>Cc Will Deacon <[email protected]>--- arch/arm64/kernel/asm-offsets.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/arm64/kernel/asm-offsets.c +0
-0arch/arm64/kernel/asm-offsets.c int main(void) · line 44
Quoted 7 lines
arch/arm64/kernel/asm-offsets.c +1
-1@@ -39,7 +39,7 @@ int main(void) DEFINE(TSK_TI_SCS_SP, offsetof(struct task_struct, thread_info.scs_sp));#endif DEFINE (TSK_STACK , offsetof(struct task_struct , stack)); -#ifdef CONFIG_STACKPROTECTOR +#ifdef CONFIG_STACKPROTECTOR_PER_TASK DEFINE (TSK_STACK_CANARY , offsetof(struct task_struct , stack_canary));#endif I don't think this really makese much sense. The '
stack_canary' field in 'structtask_struct' is defined as:#ifdef CONFIG_STACKPROTECTOR /* Canary value for the -fstack-protector GCC feature: */ unsigned long stack_canary;
Quoted 7 lines
#endif
so I think it makes sense to follow that in
asm-offsets.cDoes the current code actually cause a problem?
Will
--
Best Regards
Guo Ren
Will Deacon
maintainerQuoted 2 lines · Guo Ren
BTW, Jim found a GCC security leak in arm64, and would you want to have a look at it?
Thanks. This seems to be tracked in their bugzilla here:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96191
I agree with Jim that this should be fixed.
Will