Thanks Ted,
> I don't have it converted to an asm format which works with gcc (it's in the
> works as I have time), but I have CPUID source which properly Idents all
> Pentium-and-later chip variants and properly checks for AMD, Cyrix, etc.,
> specific stuff.
>
I took a first look on it, it's w
> If test_all_transform_functions() is used in a few different files,
> maybe it would be better to move it to a .c file instead of having a
> separate instance for each use?
>
you are right. Have a look on my new version with included benchmark
macros. Note that they will only work on 586 CPU's
Holger Waechtler <[EMAIL PROTECTED]> writes:
> All the related code is extracted now into the file debug_xform.h, which
> can be included by any file which calls a init_XXX_transforms (). Have a
> look on it.
> We should change the criteria, on which the test fails, since the 3Dnow!
> functions
>Hope, the test_3dnow program will work now on a PII CPU, it checks
I don't have it converted to an asm format which works with gcc (it's in the
works as I have time), but I have CPUID source which properly Idents all
Pentium-and-later chip variants and properly checks for AMD, Cyrix, etc.,
spec
Hallo,
I worked at the 3Dnow! - extensions last days, the results are in the
attachment.
Hope, the test_3dnow program will work now on a PII CPU, it checks
the CPU vendor string and will only test for extended features, if this
string is 'AuthenticAMD'. If other CPU vendors implement 3Dnow!
func