Re: [Qemu-devel] [PATCH] linux-user: Enable sigaltstack syscall for sh4

[Laurent Vivier]

Hi,

On 19/11/2015 10:28, John Paul Adrian Glaubitz wrote:
> On 11/19/2015 10:17 AM, Peter Maydell wrote:
>> Unfortunately this isn't sufficient. You also need to add
>> the code to the sh4-specific functions in linux-user/signal.c
>> which honours the requested sigaltstack when taking and returning
>> from signal handlers.

it seems all needed functions for sh4 signal handling are already
written in linux-user/signal.c, I thing about setup_frame(),
setup_rt_frame(), do_sigreturn() and do_rt_sigreturn().

Do we need more ?

> My supplied test case shows that sigaltstack works unless I am
> overseeing anything? Laurent Vivier (CC'ed) who has done some
> extensive qemu development thinks that my change should be enough.
> 
> Here's the output of my test case (CC'ing Michael Karcher who
> suggested the test case):
> 
> (sid-sh4-sbuild)address@hidden:/tmp# cat stackoverflow.c
> 
> #include <setjmp.h>
> #include <signal.h>
> #include <stdlib.h>
> #include <stdio.h>
> 
> jmp_buf exit_jmp;
> 
> void handler(int x)
> {
>   longjmp(exit_jmp, 1);
> }
> 
> int f(void)
> {
>   return f();
> }
> 
> int main(void)
> {
>   stack_t sigstack;
>   sigstack.ss_sp = malloc(1024*1024);
>   sigstack.ss_size = 1024*1024;
>   sigstack.ss_flags = 0;
>   sigaltstack(&sigstack, NULL);
>   struct sigaction sa;
>   sa.sa_handler = handler;
>   sigemptyset(&sa.sa_mask);
>   sa.sa_flags = SA_ONSTACK;
>   sigaction(SIGSEGV, &sa, NULL);
>   if (setjmp(exit_jmp) == 0)
>     {
>       return f();
>     }
>   puts("recovered");
>   return 0;
> }
> (sid-sh4-sbuild)address@hidden:/tmp# gcc stackoverflow.c -o stackoverflow
> (sid-sh4-sbuild)address@hidden:/tmp# ./stackoverflow
> recovered
> (sid-sh4-sbuild)address@hidden:/tmp#
> 
> Now commenting "sigaltstack" out:
> 
> (sid-sh4-sbuild)address@hidden:/tmp# cat stackoverflow.c
> 
> #include <setjmp.h>
> #include <signal.h>
> #include <stdlib.h>
> #include <stdio.h>
> 
> jmp_buf exit_jmp;
> 
> void handler(int x)
> {
>   longjmp(exit_jmp, 1);
> }
> 
> int f(void)
> {
>   return f();
> }
> 
> int main(void)
> {
>   stack_t sigstack;
>   sigstack.ss_sp = malloc(1024*1024);
>   sigstack.ss_size = 1024*1024;
>   sigstack.ss_flags = 0;
>   // sigaltstack(&sigstack, NULL);
>   struct sigaction sa;
>   sa.sa_handler = handler;
>   sigemptyset(&sa.sa_mask);
>   sa.sa_flags = SA_ONSTACK;
>   sigaction(SIGSEGV, &sa, NULL);
>   if (setjmp(exit_jmp) == 0)
>     {
>       return f();
>     }
>   puts("recovered");
>   return 0;
> }
> (sid-sh4-sbuild)address@hidden:/tmp# gcc stackoverflow.c -o stackoverflow
> (sid-sh4-sbuild)address@hidden:/tmp# ./stackoverflow
> qemu: uncaught target signal 11 (Segmentation fault) - core dumped
> Segmentation fault
> (sid-sh4-sbuild)address@hidden:/tmp#
> 
> Thus, for me it seems sigaltstack behaves as expected with the patch
> applied.
> 
> Am I missing something obvious?
> 
> Cheers,
> Adrian
>