[LTP] tst_strstatus.c fails on Alpine

[Alexey Kodanev]

On 13.07.2021 09:08, Jan Stancek wrote:
> 
> 
> On Mon, Jul 12, 2021 at 7:02 PM Petr Vorel <pvorel@suse.cz <mailto:pvorel@suse.cz>> wrote:
> 
>     Hi all,
> 
>     I see failures of lib/newlib_tests/tst_strstatus on Alpine:
> 
>     tst_strstatus.c:31: TPASS: exited with 1
>     tst_strstatus.c:31: TPASS: killed by SIGHUP
>     tst_strstatus.c:31: TPASS: is stopped
>     tst_strstatus.c:31: TPASS: is resumed
>     tst_strstatus.c:29: TFAIL: killed by ??? != invalid status 0xff
> 
>     Any idea what could be wrong?
> 
> 
> I'd start with definition of WIFSIGNALED on that system.
> 
> printf("%d\n", WIFSIGNALED(0xff));
> should give you 0, but it does appear to return 1 in output above.
> 

musl defines it as:
#define WIFSIGNALED(s) (((s)&0xffff)-1U < 0xffu)
so passing 0xff trigger this.

This difference from the glibc have appeared since the commit:
41c632824c08 ("fix definitions of WIFSTOPPED and WIFSIGNALED to support up to signal 127")

May be changed 0xff to 0x1ff for invalid status...