[PATCH] D67363: [BreakFalseDeps] ignore function with minsize attribute

[Sanjay Patel via Phabricator via llvm-commits]

spatel created this revision.
spatel added reviewers: craig.topper, RKSimon.
Herald added subscribers: hiraditya, mcrosier.
Herald added a project: LLVM.

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


https://reviews.llvm.org/D67363

Files:
  llvm/lib/CodeGen/BreakFalseDeps.cpp
  llvm/test/CodeGen/X86/sqrt-partial.ll
  llvm/test/CodeGen/X86/stack-folding-fp-sse42.ll


Index: llvm/test/CodeGen/X86/stack-folding-fp-sse42.ll
===================================================================
--- llvm/test/CodeGen/X86/stack-folding-fp-sse42.ll
+++ llvm/test/CodeGen/X86/stack-folding-fp-sse42.ll
@@ -583,7 +583,6 @@
 ; 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 @@
 ; 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 @@
 ; 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 @@
 ; 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}"()
Index: llvm/test/CodeGen/X86/sqrt-partial.ll
===================================================================
--- llvm/test/CodeGen/X86/sqrt-partial.ll
+++ llvm/test/CodeGen/X86/sqrt-partial.ll
@@ -44,7 +44,6 @@
 ; CHECK-NEXT:    mulsd %xmm0, %xmm0
 ; CHECK-NEXT:    mulsd %xmm1, %xmm1
 ; CHECK-NEXT:    addsd %xmm0, %xmm1
-; CHECK-NEXT:    xorps %xmm0, %xmm0
 ; CHECK-NEXT:    sqrtsd %xmm1, %xmm0
 ; CHECK-NEXT:    retq
   %t3 = fmul fast double %x, %x
Index: llvm/lib/CodeGen/BreakFalseDeps.cpp
===================================================================
--- llvm/lib/CodeGen/BreakFalseDeps.cpp
+++ llvm/lib/CodeGen/BreakFalseDeps.cpp
@@ -9,11 +9,11 @@
 /// \file Break False Dependency pass.
 ///
 /// Some instructions have false dependencies which cause unnecessary stalls.
-/// For exmaple, instructions that only write part of a register, and implicitly
+/// For example, instructions that only write part of a register, and implicitly
 /// need to read the other parts of the register.  This may cause unwanted
 /// stalls preventing otherwise unrelated instructions from executing in
 /// parallel in an out-of-order CPU.
-/// This pass is aimed at identifying and avoiding these depepndencies when
+/// This pass is aimed at identifying and avoiding these dependencies when
 /// possible.
 //
 //===----------------------------------------------------------------------===//
@@ -252,6 +252,12 @@
 bool BreakFalseDeps::runOnMachineFunction(MachineFunction &mf) {
   if (skipFunction(mf.getFunction()))
     return false;
+
+  // This pass adds instructions to remove dependencies. That opposes the goal
+  // of minimizing size.
+  if (mf.getFunction().hasMinSize())
+    return false;
+
   MF = &mf;
   TII = MF->getSubtarget().getInstrInfo();
   TRI = MF->getSubtarget().getRegisterInfo();


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D67363.219403.patch
Type: text/x-patch
Size: 3744 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190909/3cc36028/attachment.bin>