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: RFC - COST of const_double for x86 prevents constant copy propagation in cse



On Aug 25, 2005, at 3:59 PM, Fariborz Jahanian wrote:


I think the problem may be somewhere else. I got the same xmm0 code
sequence on Linux/ia32 with -msse3 -mfpmath=sse. However, I got

        xorl    %eax, %eax
        movq    %rax, 16(%rdi)
        movq    %rax, 8(%rdi)
        movq    %rax, (%rdi)

Can you try this with -march=pentium4

pentium4 is not the arch which supports 64bit, it is Nocoa which is a different
core than the orginal pentium4 which is what is meant by pentium4 here.


Also for x86/64, we have 64bit registers so using the integer registers is nice.

-- Pinski


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