1 |
Am 02.05.2015 um 14:06 schrieb Nikos Chantziaras: |
2 |
> On 02/05/15 14:37, Volker Armin Hemmann wrote: |
3 |
>> Am 02.05.2015 um 13:25 schrieb Nikos Chantziaras: |
4 |
>>>>> |
5 |
>>>>> The kernel uses -O2 and several -march variants (e.g. -march=core2). |
6 |
>>>>> Several other options are used to prevent GCC from generating |
7 |
>>>>> unsuitable code. |
8 |
>>>>> |
9 |
>>>>> Specifying another -march variant does not affect the optimizer |
10 |
>>>>> though. It only affects the code generator. If you don't modify the |
11 |
>>>>> other CFLAGS and only change -march, you will not get FP instructions |
12 |
>>>>> unless you use FP in the code. |
13 |
>>>> |
14 |
>>>> http://www.agner.org/optimize/calling_conventions.pdf |
15 |
>>> |
16 |
>>> Not sure what you're trying to say. |
17 |
>>> |
18 |
>> |
19 |
>> that simd is not save in kernel if not carefully guarded. |
20 |
>> |
21 |
>> Really people, just don't fuck around with the cflags. |
22 |
> |
23 |
> I still fail to see the relevance. Unless you mean using a different |
24 |
> -O level. In that case, yes. You shouldn't. But I was talking about |
25 |
> -march. |
26 |
> |
27 |
|
28 |
you said this |
29 |
|
30 |
> |
31 |
> (note that SIMD is not FP and is perfectly fine in the kernel.) |
32 |
|
33 |
and I have shown you that you are wrong. |