6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-12-19 01:34:32 +00:00
Code Issues Projects Releases Wiki Activity
ca5e4f8e19
llvm-6502/test/CodeGen/X86/break-sse-dep.ll

22 lines
491 B
LLVM
Raw Normal View History

Remove target attribute break-sse-dep. Instead, do not fold load into sse partial update instructions unless optimizing for size. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91910 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-22 17:47:23 +00:00
; RUN: llc < %s -march=x86-64 -mattr=+sse2 | FileCheck %s
On recent Intel u-arch's, folding loads into some unary SSE instructions can be non-optimal. To be precise, we should avoid folding loads if the instructions only update part of the destination register, and the non-updated part is not needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks the partial register dependency and it can improve performance. e.g. movss (%rdi), %xmm0 cvtss2sd %xmm0, %xmm0 instead of cvtss2sd (%rdi), %xmm0 An alternative method to break dependency is to clear the register first. e.g. xorps %xmm0, %xmm0 cvtss2sd (%rdi), %xmm0 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-18 07:40:29 +00:00
define double @t1(float* nocapture %x) nounwind readonly ssp {
entry:
Remove target attribute break-sse-dep. Instead, do not fold load into sse partial update instructions unless optimizing for size. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91910 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-22 17:47:23 +00:00
; CHECK: t1:
; CHECK: movss (%rdi), %xmm0
; CHECK; cvtss2sd %xmm0, %xmm0
On recent Intel u-arch's, folding loads into some unary SSE instructions can be non-optimal. To be precise, we should avoid folding loads if the instructions only update part of the destination register, and the non-updated part is not needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks the partial register dependency and it can improve performance. e.g. movss (%rdi), %xmm0 cvtss2sd %xmm0, %xmm0 instead of cvtss2sd (%rdi), %xmm0 An alternative method to break dependency is to clear the register first. e.g. xorps %xmm0, %xmm0 cvtss2sd (%rdi), %xmm0 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-18 07:40:29 +00:00
%0 = load float* %x, align 4
%1 = fpext float %0 to double
ret double %1
}
Remove target attribute break-sse-dep. Instead, do not fold load into sse partial update instructions unless optimizing for size. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91910 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-22 17:47:23 +00:00
define float @t2(double* nocapture %x) nounwind readonly ssp optsize {
On recent Intel u-arch's, folding loads into some unary SSE instructions can be non-optimal. To be precise, we should avoid folding loads if the instructions only update part of the destination register, and the non-updated part is not needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks the partial register dependency and it can improve performance. e.g. movss (%rdi), %xmm0 cvtss2sd %xmm0, %xmm0 instead of cvtss2sd (%rdi), %xmm0 An alternative method to break dependency is to clear the register first. e.g. xorps %xmm0, %xmm0 cvtss2sd (%rdi), %xmm0 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-18 07:40:29 +00:00
entry:
Remove target attribute break-sse-dep. Instead, do not fold load into sse partial update instructions unless optimizing for size. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91910 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-22 17:47:23 +00:00
; CHECK: t2:
; CHECK; cvtsd2ss (%rdi), %xmm0
On recent Intel u-arch's, folding loads into some unary SSE instructions can be non-optimal. To be precise, we should avoid folding loads if the instructions only update part of the destination register, and the non-updated part is not needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks the partial register dependency and it can improve performance. e.g. movss (%rdi), %xmm0 cvtss2sd %xmm0, %xmm0 instead of cvtss2sd (%rdi), %xmm0 An alternative method to break dependency is to clear the register first. e.g. xorps %xmm0, %xmm0 cvtss2sd (%rdi), %xmm0 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
2009-12-18 07:40:29 +00:00
%0 = load double* %x, align 8
%1 = fptrunc double %0 to float
ret float %1
}
Reference in New Issue Copy Permalink