[LLVMbugs] [Bug 4340] New: Missed optimisations with vectors.

[bugzilla-daemon at cs.uiuc.edu]

http://llvm.org/bugs/show_bug.cgi?id=4340

           Summary: Missed optimisations with vectors.
           Product: libraries
           Version: trunk
          Platform: Macintosh
        OS/Version: MacOS X
            Status: NEW
          Severity: minor
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: arplynn at gmail.com
                CC: llvmbugs at cs.uiuc.edu


Here is a very short test-case:

define void @vz(<4 x float>* nocapture %val) nounwind {
entry:
        %tmp1 = load <4 x float>* %val          ; <<4 x float>> [#uses=1]
        %vecins = insertelement <4 x float> %tmp1, float 0.000000e+00, i32 0   
        ; <<4 x float>> [#uses=1]
        %vecins4 = insertelement <4 x float> %vecins, float 0.000000e+00, i32 1
        ; <<4 x float>> [#uses=1]
        %vecins7 = insertelement <4 x float> %vecins4, float 0.000000e+00, i32
2                ; <<4 x float>> [#uses=1]
        %vecins10 = insertelement <4 x float> %vecins7, float 0.000000e+00, i32
3               ; <<4 x float>> [#uses=1]
        store <4 x float> %vecins10, <4 x float>* %val
        ret void
}

Two missed things here: the redundant load and the fact that the whole thing
could be reduced to store <4 x float> <float 0.000000e+00, float 0.000000e+00,
float 0.000000e+00, float 0.000000e+00>, <4 x float>* %val


-- 
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.