1 |
On Sun, Jan 18, 2015 at 3:23 PM, Michał Górny <mgorny@g.o> wrote: |
2 |
> Dnia 2015-01-18, o godz. 15:15:22 |
3 |
> Matt Turner <mattst88@g.o> napisał(a): |
4 |
> |
5 |
>> On Sun, Jan 18, 2015 at 12:44 PM, Michał Górny <mgorny@g.o> wrote: |
6 |
>> > Hello, |
7 |
>> > |
8 |
>> > I would like to commit the following flags as cpu_flags_x86_desc. |
9 |
>> > The list combines global USE flags with some local USE flags I've been |
10 |
>> > able to find. |
11 |
>> > |
12 |
>> > |
13 |
>> > 3dnow - Use the 3DNow! instruction set |
14 |
>> > 3dnowext - Use the Enhanced 3DNow! instruction set |
15 |
>> > aes-ni - Enable support for Intel's AES instruction set (aes in cpuinfo) |
16 |
>> > avx - Adds support for Advanced Vector Extensions instructions |
17 |
>> > avx2 - Adds support for Advanced Vector Extensions 2 instructions |
18 |
>> > fma - Use the Fused Multiply Add instruction set |
19 |
>> |
20 |
>> There are fma3 and fma4 instruction sets. We need to determine which |
21 |
>> one the packages that have this flag are actually using and name the |
22 |
>> flag appropriately. |
23 |
> |
24 |
> Ok then: |
25 |
> |
26 |
> fma3 - Use the Fused Multiply Add v3 instruction set |
27 |
> fma4 - Use the Fused Multiply Add v4 instruction set |
28 |
|
29 |
They're not actually versions 3 and 4, like SSE. The number is the |
30 |
number of arguments the instructions take. |
31 |
|
32 |
I'd just remove the 'v' from the description and call it good. |
33 |
|
34 |
See http://en.wikipedia.org/wiki/FMA_instruction_set for more information. |