[LLVMbugs] [Bug 13850] opportunity for small improvement to zero-initialization codegen [AVX]

[bugzilla-daemon at llvm.org]

http://llvm.org/bugs/show_bug.cgi?id=13850

Benjamin Kramer <benny.kra at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #3 from Benjamin Kramer <benny.kra at gmail.com> 2012-11-22 10:47:01 CST ---
Bug 2962 was finally fixed. We now turn this into:

_orthoBasisUnaligned:                   ## @orthoBasisUnaligned
## BB#0:                                ## %allocas
    pushq    %rbp
    movq    %rsp, %rbp
    vxorps    %ymm0, %ymm0, %ymm0
    vmovups    %ymm0, 64(%rdi)
    vmovups    %ymm0, 32(%rdi)
    vmovups    %ymm0, (%rdi)
    popq    %rbp
    vzeroupper
    ret

    .globl    _orthoBasisAligned
    .align    4, 0x90
_orthoBasisAligned:                     ## @orthoBasisAligned
## BB#0:                                ## %allocas
    pushq    %rbp
    movq    %rsp, %rbp
    vxorps    %ymm0, %ymm0, %ymm0
    vmovups    %ymm0, 64(%rdi)
    vmovups    %ymm0, 32(%rdi)
    vmovups    %ymm0, (%rdi)
    popq    %rbp
    vzeroupper
    ret

not sure if this is actually faster on sandy bridge, but this bug is fixed.

-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.