| 1 |
Ciaran McCreesh wrote: |
| 2 |
> On Fri, 30 May 2008 15:07:43 -0700 |
| 3 |
> Donnie Berkholz <dberkholz@g.o> wrote: |
| 4 |
>> On 22:53 Fri 30 May , Ciaran McCreesh wrote: |
| 5 |
>>> On Sat, 31 May 2008 00:47:44 +0300 |
| 6 |
>>> Mart Raudsepp <leio@g.o> wrote: |
| 7 |
>>>> The story that matters here is, that a C++ corner case that does |
| 8 |
>>>> not work on 0.01% of packages with --as-needed and breaks on |
| 9 |
>>>> non-ELF platforms, should not cause good things for our users to |
| 10 |
>>>> be shot down. |
| 11 |
>>> You could say the same thing for -ffast-math... |
| 12 |
>> When there's a feature that only breaks one package that we know of, |
| 13 |
>> wouldn't it make more sense to enable it globally and add an |
| 14 |
>> exception than to do it the other way around? |
| 15 |
> |
| 16 |
> Both -ffast-math and --as-needed make the compiler / linker violate |
| 17 |
> various standards in ways that can't be used safely unless a package |
| 18 |
> has been explicitly designed to work with it. |
| 19 |
|
| 20 |
I know exactly which standard -ffast-math violates (IEEE/ISO floating |
| 21 |
point spec) and how (the man page is quite complete about this), |
| 22 |
--as-needed doesn't have any warning about this, there isn't any |
| 23 |
standard that it violates since it's the default behavior at least for 2 |
| 24 |
platform (one from those who wrote most of the ELF spec...). |
| 25 |
Point the spec, and the paragraph violated. |
| 26 |
|
| 27 |
lu |
| 28 |
|
| 29 |
-- |
| 30 |
|
| 31 |
Luca Barbato |
| 32 |
Gentoo Council Member |
| 33 |
Gentoo/linux Gentoo/PPC |
| 34 |
http://dev.gentoo.org/~lu_zero |
| 35 |
|
| 36 |
-- |
| 37 |
gentoo-dev@l.g.o mailing list |
Archives