Nadav Rotem 50e64cfe6e Modify the code that lowers shuffles to blends from using blendvXX to vblendXX. ...

blendv uses a register for the selection while vblend uses an immediate.
On sandybridge they still have the same latency and execute on the same execution ports.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@154396 91177308-0d34-0410-b5e6-96231b3b80d8

2012-04-10 14:33:13 +00:00

..

ARM

Transform div to mul with reciprocal only when fp imm is legal.

2012-04-10 13:22:49 +00:00

CellSPU

1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new

2012-04-07 21:19:08 +00:00

CPP

…

Generic

Don't try to zExt just to check if an integer constant is zero, it might

2012-04-10 00:16:22 +00:00

Hexagon

…

MBlaze

…

Mips

Have TargetLowering::getPICJumpTableRelocBase return a node that points to the

2012-04-09 20:32:12 +00:00

MSP430

…

PowerPC

…

PTX

…

SPARC

…

Thumb

…

Thumb2

Don't break the IV update in TLI::SimplifySetCC().

2012-04-05 20:30:20 +00:00

X86

Modify the code that lowers shuffles to blends from using blendvXX to vblendXX.

2012-04-10 14:33:13 +00:00

XCore

…