On Thu, Sep 15, 2011 at 1:05 PM, Luca Barbato <[email protected]> wrote:
> On 15/09/2011 18:23, Sean McGovern wrote:
>> Not to sound buzzword-compliant or anything but how should I be moving
>> forward on this patch. If it's not welcome then I should just drop both it
>> and the Solaris Studio FATE runs and stop pretending that we build with
>> anything that isn't gcc?
>
> I guess we dug a bit too deep the issue, but at least now everybody know
> what's missing to get this feature is in the assembler.
>
> That said, I guess we could hopefully land a proper patch for it soon.
>
> Thanks for the patience.
>
Here's the result of file(1) on avconv built with gcc:
sean@tsukimi:/BUILD/libav.gcc$ file ./avconv
./avconv: ELF 64-bit LSB executable AMD64 Version 1, dynamically
linked, not stripped
Here's the same on avconv built with suncc (without the patch, but
with several other patches I would like to submit soon that help at
least complete the build):
sean@tsukimi:/BUILD/libav.sol10$ file ./avconv
./avconv: ELF 64-bit LSB executable AMD64 Version 1 [SSSE3 SSE3 SSE2
SSE FXSR AMD_3DNowx AMD_3DNow MMX CMOV TSC FPU], dynamically linked,
not stripped
Here's an attempt to run the suncc version:
sean@tsukimi:/BUILD/libav.sol10$ ./avconv
ld.so.1: avconv: fatal: hardware capability unsupported: 0x300 [
AMD_3DNowx AMD_3DNow ]
Killed
Here's the processor I have:
sean@tsukimi:/BUILD/libav.sol10$ /usr/sbin/psrinfo -pv
The physical processor has 1 virtual processor (0)
x86 (chipid 0x0 GenuineIntel family 6 model 23 step 10 clock 2388 MHz)
Intel(r) Core(tm)2 Duo CPU P8600 @ 2.40GHz
(repeated twice)
See the problem?
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
