This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Regressions -march=geode/athlon regressions (Was: 4.3 release plan)

On Wed, May 23, 2007 at 10:40:46AM +0200, Rask Ingemann Lambertsen wrote:
> On Tue, May 22, 2007 at 08:45:38PM -0400, Bernardo Innocenti wrote:
>  
> > FAIL: gcc.c-torture/execute/20050316-2.c execution,  -O0
> [...]
> > FAIL: gcc.c-torture/execute/20050604-1.c execution,  -O0
> [...]
> > 
> > I analyzed only the first failing test.  It seems we fail
> > with all MMX enabled processors. But maybe the test has
> > invalid expectations, I'm not sure what it is trying to do.
> 
>    The last time I looked into it, the reason was use of MMX instructions
> without (f)emms afterwards.

Yeah:

        movq    -48(%ebp), %mm0 #, x
        call    test1   #
        movl    %eax, -40(%ebp) # D.1650, u.l
        movl    %edx, -36(%ebp) # D.1650, u.l
        flds    -40(%ebp)       # u.f
        flds    .LC0    #

(and test1 also using movq %mm0, something, also no (f)emms).

	Jakub
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]