maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/X86/alloca-align-rounding.ll

21 lines
451 B
LLVM
Raw Normal View History

Enable MI Sched for x86. This changes the SelectionDAG scheduling preference to source order. Soon, the SelectionDAG scheduler can be bypassed saving a nice chunk of compile time. Performance differences that result from this change are often a consequence of register coalescing. The register coalescer is far from perfect. Bugs can be filed for deficiencies. On x86 SandyBridge/Haswell, the source order schedule is often preserved, particularly for small blocks. Register pressure is generally improved over the SD scheduler's ILP mode. However, we are still able to handle large blocks that require latency hiding, unlike the SD scheduler's BURR mode. MI scheduler also attempts to discover the critical path in single-block loops and adjust heuristics accordingly. The MI scheduler relies on the new machine model. This is currently unimplemented for AVX, so we may not be generating the best code yet. Unit tests are updated so they don't depend on SD scheduling heuristics. llvm-svn: 192750
2013-10-16 07:33:07 +08:00
; RUN: llc < %s -march=x86-64 -mtriple=i686-pc-linux -enable-misched=false | FileCheck %s
It's not necessary to do rounding for alloca operations when the requested alignment is equal to the stack alignment. llvm-svn: 40004
2007-07-19 00:29:46 +08:00
declare void @bar(<2 x i64>* %n)
Eliminate the restriction that the array size in an alloca must be i32. This will help reduce the amount of casting required on 64-bit targets. llvm-svn: 104911
2010-05-28 09:14:11 +08:00
define void @foo(i64 %h) {
%p = alloca <2 x i64>, i64 %h
It's not necessary to do rounding for alloca operations when the requested alignment is equal to the stack alignment. llvm-svn: 40004
2007-07-19 00:29:46 +08:00
call void @bar(<2 x i64>* %p)
ret void
FileCheckize tests. llvm-svn: 159044
2012-06-23 07:04:02 +08:00
; CHECK: foo
; CHECK-NOT: andq $-32, %rax
It's not necessary to do rounding for alloca operations when the requested alignment is equal to the stack alignment. llvm-svn: 40004
2007-07-19 00:29:46 +08:00
}
Update test: dynamic_stackalloc size *must* be rounded to ensure stack ptr be left in a valid state. llvm-svn: 41134
2007-08-17 07:51:28 +08:00
Eliminate the restriction that the array size in an alloca must be i32. This will help reduce the amount of casting required on 64-bit targets. llvm-svn: 104911
2010-05-28 09:14:11 +08:00
define void @foo2(i64 %h) {
%p = alloca <2 x i64>, i64 %h, align 32
Update test: dynamic_stackalloc size *must* be rounded to ensure stack ptr be left in a valid state. llvm-svn: 41134
2007-08-17 07:51:28 +08:00
call void @bar(<2 x i64>* %p)
ret void
FileCheckize tests. llvm-svn: 159044
2012-06-23 07:04:02 +08:00
; CHECK: foo2
Add support for dynamic stack realignment in the presence of dynamic allocas on X86. Basically, this is a reapplication of r158087 with a few fixes. Specifically, (1) the stack pointer is restored from the base pointer before popping callee-saved registers and (2) in obscure cases (see comments in patch) we must cache the value of the original stack adjustment in the prologue and apply it in the epilogue. rdar://11496434 llvm-svn: 160002
2012-07-11 01:45:53 +08:00
; CHECK: andq $-32, %rsp
FileCheckize tests. llvm-svn: 159044
2012-06-23 07:04:02 +08:00
; CHECK: andq $-32, %rax
Update test: dynamic_stackalloc size *must* be rounded to ensure stack ptr be left in a valid state. llvm-svn: 41134
2007-08-17 07:51:28 +08:00
}