maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[BreakFalseDeps] ignore function with minsize attribute 

This came up in the x86-specific:
https://bugs.llvm.org/show_bug.cgi?id=43239
...but it is a general problem for the BreakFalseDeps pass.
Dependencies may be broken by adding some other instruction,
so that should be avoided if the overall goal is to minimize size.

Differential Revision: https://reviews.llvm.org/D67363

llvm-svn: 372628
This commit is contained in:
Sanjay Patel 2019-09-23 17:01:01 +00:00
parent 7e56dd0229
commit 7414151929
5 changed files with 16 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
llvm/lib/CodeGen/BreakFalseDeps.cpp
View File

@ -177,6 +177,7 @@ void BreakFalseDeps::processDefs(MachineInstr *MI) {
assert(!MI->isDebugInstr() && "Won't process debug values");
// Break dependence on undef uses. Do this before updating LiveRegs below.
// This can remove a false dependence with no additional instructions.
unsigned OpNum;
unsigned Pref = TII->getUndefRegClearance(*MI, OpNum, TRI);
if (Pref) {
@ -188,6 +189,11 @@ void BreakFalseDeps::processDefs(MachineInstr *MI) {
UndefReads.push_back(std::make_pair(MI, OpNum));
}
// The code below allows the target to create a new instruction to break the
// dependence. That opposes the goal of minimizing size, so bail out now.
if (MF->getFunction().hasMinSize())
return;
const MCInstrDesc &MCID = MI->getDesc();
for (unsigned i = 0,
e = MI->isVariadic() ? MI->getNumOperands() : MCID.getNumDefs();
@ -208,6 +214,11 @@ void BreakFalseDeps::processUndefReads(MachineBasicBlock *MBB) {
if (UndefReads.empty())
return;
// The code below allows the target to create a new instruction to break the
// dependence. That opposes the goal of minimizing size, so bail out now.
if (MF->getFunction().hasMinSize())
return;
// Collect this block's live out register units.
LiveRegSet.init(*TRI);
// We do not need to care about pristine registers as they are just preserved

6
llvm/test/CodeGen/ARM/a15-partial-update.ll
View File

@ -56,6 +56,11 @@ ret:
ret void
}
; If minimizing size, that overrides perf, so no extra vmov.f64 here.
; TODO: This (and above) could use a splat load to remove the false
; dependence with no extra instruction.
define void @t2_minsize(<4 x i8> *%in, <4 x i8> *%out, i32 %n) minsize {
; CHECK-LABEL: t2_minsize:
; CHECK: @ %bb.0: @ %entry
@ -63,7 +68,6 @@ define void @t2_minsize(<4 x i8> *%in, <4 x i8> *%out, i32 %n) minsize {
; CHECK-NEXT: add r1, r1, #4
; CHECK-NEXT: .LBB2_1: @ %loop
; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vmov.f64 d16, #5.000000e-01
; CHECK-NEXT: vld1.32 {d16[0]}, [r0:32]
; CHECK-NEXT: vmovl.u8 q8, d16
; CHECK-NEXT: vuzp.8 d16, d18

1
llvm/test/CodeGen/X86/sqrt-partial.ll
View File

@ -67,7 +67,6 @@ define double @minsize(double %x, double %y) minsize {
; SSE-NEXT: mulsd %xmm0, %xmm0
; SSE-NEXT: mulsd %xmm1, %xmm1
; SSE-NEXT: addsd %xmm0, %xmm1
; SSE-NEXT: xorps %xmm0, %xmm0
; SSE-NEXT: sqrtsd %xmm1, %xmm0
; SSE-NEXT: retq
;

1
llvm/test/CodeGen/X86/stack-folding-fp-avx1.ll
View File

@ -2876,7 +2876,6 @@ define double @stack_fold_roundsd_minsize(double %a0) minsize {
; CHECK-NEXT: #APP
; CHECK-NEXT: nop
; CHECK-NEXT: #NO_APP
; CHECK-NEXT: vxorps %xmm0, %xmm0, %xmm0
; CHECK-NEXT: vroundsd $9, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0, %xmm0 # 8-byte Folded Reload
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()

4
llvm/test/CodeGen/X86/stack-folding-fp-sse42.ll
View File

@ -583,7 +583,6 @@ define float @stack_fold_cvtsd2ss(double %a0) minsize {
; CHECK-NEXT: #APP
; CHECK-NEXT: nop
; CHECK-NEXT: #NO_APP
; CHECK-NEXT: xorps %xmm0, %xmm0
; CHECK-NEXT: cvtsd2ss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 8-byte Folded Reload
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
@ -975,7 +974,6 @@ define double @stack_fold_cvtss2sd(float %a0) minsize {
; CHECK-NEXT: #APP
; CHECK-NEXT: nop
; CHECK-NEXT: #NO_APP
; CHECK-NEXT: xorps %xmm0, %xmm0
; CHECK-NEXT: cvtss2sd {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 4-byte Folded Reload
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
@ -2012,7 +2010,6 @@ define float @stack_fold_roundss(float %a0) minsize {
; CHECK-NEXT: #APP
; CHECK-NEXT: nop
; CHECK-NEXT: #NO_APP
; CHECK-NEXT: xorps %xmm0, %xmm0
; CHECK-NEXT: roundss $9, {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 4-byte Folded Reload
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()
@ -2181,7 +2178,6 @@ define float @stack_fold_sqrtss(float %a0) minsize {
; CHECK-NEXT: #APP
; CHECK-NEXT: nop
; CHECK-NEXT: #NO_APP
; CHECK-NEXT: xorps %xmm0, %xmm0
; CHECK-NEXT: sqrtss {{[-0-9]+}}(%r{{[sb]}}p), %xmm0 # 4-byte Folded Reload
; CHECK-NEXT: retq
%1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm1},~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"()