On Tue, Jul 28, 2015 at 3:00 PM, Paul Hargrove <phhargr...@lbl.gov> wrote:

> I noticed an early-clobber applied to "1" and "3" in master.
> I am retesting that now (still as applied to 1.11.0)
>

Ok, with the 1.11.0 tarball patched to use the same inline asm as master we
have:

#elif defined(HWLOC_X86_32_ARCH)
  __asm__(
  "mov %%ebx,%1\n\t"
  "cpuid\n\t"
  "xchg %%ebx,%1\n\t"
  : "+a" (*eax), "=&SD" (*ebx), "+c" (*ecx), "=&d" (*edx));
#else


This too works on all 28 platform/compiler combinations for which I have
test scripts setup.

Interestingly, this error:

> All looked fine *except* the following with icc-15 with -m32 on a x86-64
> Ubuntu 12.04.5 system:
>    FAIL: 256ia64-64n2s2c.output
> I don't see how that could possibly be related to the inline asm for
> cpuid, and have confirmed that the same failure occurs without the patch
> (but it is absent w/o -m32).

Occurred with both icc-14 and icc-15 on different hosts this time.

-Paul

-- 
Paul H. Hargrove                          phhargr...@lbl.gov
Computer Languages & Systems Software (CLaSS) Group
Computer Science Department               Tel: +1-510-495-2352
Lawrence Berkeley National Laboratory     Fax: +1-510-486-6900

Reply via email to