[llvm-bugs] [Bug 33013] New: [x86] default codegen should be more branchy

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=33013

            Bug ID: 33013
           Summary: [x86] default codegen should be more branchy
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: X86
          Assignee: unassignedbugs at nondot.org
          Reporter: spatel+llvm at rotateright.com
                CC: llvm-bugs at lists.llvm.org

The default x86 target is something like an Intel big core (ie, it's good at
absorbing the cost of correctly predicted branches, good at predicting those
branches, and good at speculating execution past those branches).

Therefore, we shouldn't favor cmov codegen for IR select as much as we
currently do. Example:

int foo(float x) {
  if (x < 42.0f)
    return x;
  return 12;
}

define i32 @foo(float %x) {
  %cmp = fcmp olt float %x, 4.200000e+01
  %conv = fptosi float %x to i32
  %ret = select i1 %cmp, i32 %conv, i32 12
  ret i32 %ret
}

$ clang -O2 cmovfp.c -S -o -
    movss    LCPI0_0(%rip), %xmm1    ## xmm1 = mem[0],zero,zero,zero
    ucomiss    %xmm0, %xmm1
    cvttss2si    %xmm0, %ecx
    movl    $12, %eax
    cmoval    %ecx, %eax
    retq

Note that gcc and icc will use compare and branch on this example.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20170511/476fede9/attachment.html>