[PATCH 1/2] Linux/x86: Update cancel_jmp_buf to match __jmp_buf_tag [BZ #22563]

[Florian Weimer]

On 12/07/2017 08:19 PM, H.J. Lu wrote:
> On Thu, Dec 7, 2017 at 11:14 AM, Florian Weimer <fweimer@redhat.com> wrote:
>> On 12/07/2017 08:12 PM, H.J. Lu wrote:
>>>>
>>>> Sorry, what exactly is stored on the shadow stack?  I assumed it was for
>>>> verification of the targets of ret instructions.
>>>>
>>>> In this case, don't need to unwind the shadow stack (or preserve its
>>>> contents) because there are no returns from existing stack frames once
>>>> cancellation has started.
>>>>
>>> Shadow stack is the similar to normal call stack without local variables.
>>> SHSTK checks  the return address of EACH  "RET"  instruction against
>>> shadow stack.
>>
>>
>> Then the shadow stack contents at the time of cancellation does not matter
>> because all future RET instructions on this thread will match CALLs which
>> happened *after* cancellation.  (In other words, I still think I'm right
>> about this.)
>>
> 
> We are updating setjmp/lonjmp to save and restore shadow stack pointer:
> 
> https://sourceware.org/git/?p=glibc.git;a=commit;h=ac195a2d554e3fb577e44474faf3ed7f4521de9f

Please try to understand what I wrote.  You don't need the restore 
during cancellation handling:

     if (__glibc_unlikely (__not_first_call))	      \
       {						      \
	__cancel_routine (__cancel_arg);	      \
	__pthread_unwind_next (&__cancel_buf);	      \
	/* NOTREACHED */			      \
       }						      \

Florian