mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-24 03:33:20 +01:00
ee0d5cd952
This adds support for the new 32-bit vector float instructions of z14. This includes: - Enabling the instructions for the assembler/disassembler. - CodeGen for the instructions, including new LLVM intrinsics. - Scheduler description support for the instructions. - Update to the vector cost function calculations. In general, CodeGen support for the new v4f32 instructions closely matches support for the existing v2f64 instructions. llvm-svn: 308195
25 lines
721 B
LLVM
25 lines
721 B
LLVM
; Test vector multiplication on z14.
|
|
;
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu -mcpu=z14 | FileCheck %s
|
|
|
|
; Test a v4f32 multiplication.
|
|
define <4 x float> @f1(<4 x float> %dummy, <4 x float> %val1,
|
|
<4 x float> %val2) {
|
|
; CHECK-LABEL: f1:
|
|
; CHECK: vfmsb %v24, %v26, %v28
|
|
; CHECK: br %r14
|
|
%ret = fmul <4 x float> %val1, %val2
|
|
ret <4 x float> %ret
|
|
}
|
|
|
|
; Test an f32 multiplication that uses vector registers.
|
|
define float @f2(<4 x float> %val1, <4 x float> %val2) {
|
|
; CHECK-LABEL: f2:
|
|
; CHECK: wfmsb %f0, %v24, %v26
|
|
; CHECK: br %r14
|
|
%scalar1 = extractelement <4 x float> %val1, i32 0
|
|
%scalar2 = extractelement <4 x float> %val2, i32 0
|
|
%ret = fmul float %scalar1, %scalar2
|
|
ret float %ret
|
|
}
|