[Bug rtl-optimization/24082] New: Not that good register allocating on powerpc for vectors

[pinskia at gcc dot gnu dot org]

Take the following code:
vector int bar3(long long x)
{
  return (vector int)(vector long long){x,x};
}
---
Currently we get:
_bar3:
        mfspr r0,256
        stw r0,-8(r1)
        oris r0,r0,0x8000
        mtspr 256,r0
        addi r2,r1,-32
        stw r3,0(r2)
        stw r4,4(r2)
        stw r3,8(r2)
        stw r4,12(r2)
        lvx v0,0,r2
        lwz r12,-8(r1)
        vor v2,v0,v0
        mtspr 256,r12
        blr
But we should really allocate v0 in v2, removing the save/restore/setting of 
VRSAVE.
Like:
_bar3:
        addi r2,r1,-32
        stw r3,0(r2)
        stw r4,4(r2)
        stw r3,8(r2)
        stw r4,12(r2)
        lvx v2,0,r2
        blr

-- 
           Summary: Not that good register allocating on powerpc for vectors
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Keywords: missed-optimization, ra
          Severity: enhancement
          Priority: P2
         Component: rtl-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: pinskia at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org
GCC target triplet: powerpc-darwin


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24082