From f5497fb1b474028709f0a6d8556251dc21e34c26 Mon Sep 17 00:00:00 2001 From: Jakob Stoklund Olesen Date: Mon, 16 May 2011 23:50:05 +0000 Subject: [PATCH] Teach LiveInterval::isZeroLength about null SlotIndexes. When instructions are deleted, they leave tombstone SlotIndex entries. The isZeroLength method should ignore these null indexes. This causes RABasic to sometimes spill a callee-saved register in the abi-isel.ll test, so don't run that test with -regalloc=basic. Prioritizing register allocation according to spill weight can cause more registers to be used. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131436 91177308-0d34-0410-b5e6-96231b3b80d8 --- include/llvm/CodeGen/LiveInterval.h | 5 +++-- lib/CodeGen/CalcSpillWeights.cpp | 2 +- test/CodeGen/ARM/vldlane.ll | 8 ++++---- test/CodeGen/X86/abi-isel.ll | 11 ----------- 4 files changed, 8 insertions(+), 18 deletions(-) diff --git a/include/llvm/CodeGen/LiveInterval.h b/include/llvm/CodeGen/LiveInterval.h index c5285cec1e2..5fd4d3df266 100644 --- a/include/llvm/CodeGen/LiveInterval.h +++ b/include/llvm/CodeGen/LiveInterval.h @@ -492,9 +492,10 @@ namespace llvm { /// Returns true if the live interval is zero length, i.e. no live ranges /// span instructions. It doesn't pay to spill such an interval. - bool isZeroLength() const { + bool isZeroLength(SlotIndexes *Indexes) const { for (const_iterator i = begin(), e = end(); i != e; ++i) - if (i->end.getPrevIndex() > i->start) + if (Indexes->getNextNonNullIndex(i->start).getBaseIndex() < + i->end.getBaseIndex()) return false; return true; } diff --git a/lib/CodeGen/CalcSpillWeights.cpp b/lib/CodeGen/CalcSpillWeights.cpp index e5894b8cca9..5d722ee34f7 100644 --- a/lib/CodeGen/CalcSpillWeights.cpp +++ b/lib/CodeGen/CalcSpillWeights.cpp @@ -165,7 +165,7 @@ void VirtRegAuxInfo::CalculateWeightAndHint(LiveInterval &li) { return; // Mark li as unspillable if all live ranges are tiny. - if (li.isZeroLength()) { + if (li.isZeroLength(LIS.getSlotIndexes())) { li.markNotSpillable(); return; } diff --git a/test/CodeGen/ARM/vldlane.ll b/test/CodeGen/ARM/vldlane.ll index 2b2c0d78a2f..0d7d4ec2949 100644 --- a/test/CodeGen/ARM/vldlane.ll +++ b/test/CodeGen/ARM/vldlane.ll @@ -222,7 +222,7 @@ define <8 x i8> @vld3lanei8(i8* %A, <8 x i8>* %B) nounwind { define <4 x i16> @vld3lanei16(i16* %A, <4 x i16>* %B) nounwind { ;CHECK: vld3lanei16: ;Check the (default) alignment value. VLD3 does not support alignment. -;CHECK: vld3.16 {d16[1], d17[1], d18[1]}, [{{r[0-9]+}}] +;CHECK: vld3.16 {d{{.*}}[1], d{{.*}}[1], d{{.*}}[1]}, [{{r[0-9]+}}] %tmp0 = bitcast i16* %A to i8* %tmp1 = load <4 x i16>* %B %tmp2 = call %struct.__neon_int16x4x3_t @llvm.arm.neon.vld3lane.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 8) @@ -265,7 +265,7 @@ define <2 x float> @vld3lanef(float* %A, <2 x float>* %B) nounwind { define <8 x i16> @vld3laneQi16(i16* %A, <8 x i16>* %B) nounwind { ;CHECK: vld3laneQi16: ;Check the (default) alignment value. VLD3 does not support alignment. -;CHECK: vld3.16 {d16[1], d18[1], d20[1]}, [{{r[0-9]+}}] +;CHECK: vld3.16 {d{{.*}}[1], d{{.*}}[1], d{{.*}}[1]}, [{{r[0-9]+}}] %tmp0 = bitcast i16* %A to i8* %tmp1 = load <8 x i16>* %B %tmp2 = call %struct.__neon_int16x8x3_t @llvm.arm.neon.vld3lane.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 1, i32 8) @@ -280,7 +280,7 @@ define <8 x i16> @vld3laneQi16(i16* %A, <8 x i16>* %B) nounwind { ;Check for a post-increment updating load with register increment. define <8 x i16> @vld3laneQi16_update(i16** %ptr, <8 x i16>* %B, i32 %inc) nounwind { ;CHECK: vld3laneQi16_update: -;CHECK: vld3.16 {d16[1], d18[1], d20[1]}, [{{r[0-9]+}}], {{r[0-9]+}} +;CHECK: vld3.16 {d{{.*}}[1], d{{.*}}[1], d{{.*}}[1]}, [{{r[0-9]+}}], {{r[0-9]+}} %A = load i16** %ptr %tmp0 = bitcast i16* %A to i8* %tmp1 = load <8 x i16>* %B @@ -344,7 +344,7 @@ declare %struct.__neon_float32x4x3_t @llvm.arm.neon.vld3lane.v4f32(i8*, <4 x flo define <8 x i8> @vld4lanei8(i8* %A, <8 x i8>* %B) nounwind { ;CHECK: vld4lanei8: ;Check the alignment value. Max for this instruction is 32 bits: -;CHECK: vld4.8 {d16[1], d17[1], d18[1], d19[1]}, [{{r[0-9]+}}, :32] +;CHECK: vld4.8 {d{{.*}}[1], d{{.*}}[1], d{{.*}}[1], d{{.*}}[1]}, [{{r[0-9]+}}, :32] %tmp1 = load <8 x i8>* %B %tmp2 = call %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4lane.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 8) %tmp3 = extractvalue %struct.__neon_int8x8x4_t %tmp2, 0 diff --git a/test/CodeGen/X86/abi-isel.ll b/test/CodeGen/X86/abi-isel.ll index fe5e3726524..5068d2929db 100644 --- a/test/CodeGen/X86/abi-isel.ll +++ b/test/CodeGen/X86/abi-isel.ll @@ -12,17 +12,6 @@ ; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=dynamic-no-pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-DYNAMIC ; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-PIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-32-STATIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-32-PIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-64-STATIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=LINUX-64-PIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=i686-apple-darwin -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=DARWIN-32-STATIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=i686-apple-darwin -march=x86 -relocation-model=dynamic-no-pic -code-model=small | FileCheck %s -check-prefix=DARWIN-32-DYNAMIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=i686-apple-darwin -march=x86 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=DARWIN-32-PIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=DARWIN-64-STATIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=dynamic-no-pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-DYNAMIC -; RUN: llc < %s -asm-verbose=0 -regalloc=basic -join-physregs -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-PIC - @src = external global [131072 x i32] @dst = external global [131072 x i32] @xsrc = external global [32 x i32]