[LLVMbugs] [Bug 7500] New: x86-64 missed optimization: pcmpeqd not commuted

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Fri Jun 25 15:30:15 CDT 2010


http://llvm.org/bugs/show_bug.cgi?id=7500

           Summary: x86-64 missed optimization: pcmpeqd not commuted
           Product: new-bugs
           Version: unspecified
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: new bugs
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: sharparrow1 at yahoo.com
                CC: llvmbugs at cs.uiuc.edu


Testcase:
target datalayout =
"e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
target triple = "x86_64-unknown-linux-gnu"

define <2 x double> @a(<2 x double>, <2 x double>) nounwind readnone {
entry:
  %tmp6 = bitcast <2 x double> %0 to <4 x i32>    ; <<4 x i32>> [#uses=2]
  %tmp4 = bitcast <2 x double> %1 to <4 x i32>    ; <<4 x i32>> [#uses=1]
  %cmp = icmp eq <4 x i32> %tmp6, %tmp4           ; <<4 x i1>> [#uses=1]
  %sext = sext <4 x i1> %cmp to <4 x i32>         ; <<4 x i32>> [#uses=1]
  %and = and <4 x i32> %tmp6, %sext               ; <<4 x i32>> [#uses=1]
  %tmp8 = bitcast <4 x i32> %and to <2 x double>  ; <<2 x double>> [#uses=1]
  ret <2 x double> %tmp8
}

Output (relevant part):
    movdqa    %xmm0, %xmm2
    pcmpeqd    %xmm1, %xmm2
    pand    %xmm2, %xmm0
    ret

Expected output:
    pcmpeqd    %xmm0, %xmm1
    pand    %xmm1, %xmm0
    ret

Filing instead of putting into README because I'm not sure why this
optimization isn't happening.

-- 
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.


More information about the LLVMbugs mailing list