From ec19b35179bcbe4ac4e3d6dfa62d374553099aed Mon Sep 17 00:00:00 2001 From: Wei Mi Date: Thu, 11 Aug 2016 18:42:56 +0000 Subject: [PATCH] Remove the restriction that MachineSinking is now stopped by "insert_subreg, subreg_to_reg, and reg_sequence" instructions. This is to solve PR28852. The restriction was added at 2010 to make better register coalescing. We assumed that it was not necessary any more. Testing results on x86 supported the assumption. We will look closely to any performance impact it will bring and will be prepared to help analyzing performance problem found on other architectures. Differential Revision: https://reviews.llvm.org/D23210 llvm-svn: 278384 --- llvm/include/llvm/Target/TargetInstrInfo.h | 5 +--- llvm/test/CodeGen/X86/MachineSink-SubReg.ll | 37 +++++++++++++++++++++++++++++ 2 files changed, 38 insertions(+), 4 deletions(-) create mode 100644 llvm/test/CodeGen/X86/MachineSink-SubReg.ll diff --git a/llvm/include/llvm/Target/TargetInstrInfo.h b/llvm/include/llvm/Target/TargetInstrInfo.h index 2de98ac..af0fc80 100644 --- a/llvm/include/llvm/Target/TargetInstrInfo.h +++ b/llvm/include/llvm/Target/TargetInstrInfo.h @@ -270,11 +270,8 @@ public: /// MachineSink determines on its own whether the instruction is safe to sink; /// this gives the target a hook to override the default behavior with regards /// to which instructions should be sunk. - /// The default behavior is to not sink insert_subreg, subreg_to_reg, and - /// reg_sequence. These are meant to be close to the source to make it easier - /// to coalesce. virtual bool shouldSink(const MachineInstr &MI) const { - return !MI.isInsertSubreg() && !MI.isSubregToReg() && !MI.isRegSequence(); + return true; } /// Re-issue the specified 'original' instruction at the diff --git a/llvm/test/CodeGen/X86/MachineSink-SubReg.ll b/llvm/test/CodeGen/X86/MachineSink-SubReg.ll new file mode 100644 index 0000000..f78d4d9 --- /dev/null +++ b/llvm/test/CodeGen/X86/MachineSink-SubReg.ll @@ -0,0 +1,37 @@ +; PR28852: Check machine code sinking is not stopped by SUBREG_TO_REG. +; RUN: llc < %s | FileCheck %s + +target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128" +target triple = "x86_64-unknown-linux-gnu" + +; CHECK: foo +; CHECK-NOT: imull +; CHECK: retq +; CHECK: imull + +define void @foo(i64 %value, i32 %kLengthBits, i32* nocapture %bits, i64* nocapture %bit_buffer_64, i32 %x) local_unnamed_addr { +entry: + %mul = mul i32 %x, %kLengthBits + %add = add i32 %mul, 3 + %conv = zext i32 %add to i64 + %mul2 = mul nuw nsw i64 %conv, 5 + %sub = sub i64 64, %value + %conv4 = trunc i64 %sub to i32 + %tmp0 = load i32, i32* %bits, align 4 + %cmp = icmp ult i32 %tmp0, %conv4 + br i1 %cmp, label %if.then, label %if.end, !prof !0 + +if.then: ; preds = %entry + %add7 = add i64 %mul2, %value + %tmp1 = load i64, i64* %bit_buffer_64, align 8 + %add8 = add i64 %add7, %tmp1 + store i64 %add8, i64* %bit_buffer_64, align 8 + %conv9 = trunc i64 %mul2 to i32 + store i32 %conv9, i32* %bits, align 4 + br label %if.end + +if.end: ; preds = %if.then, %entry + ret void +} + +!0 = !{!"branch_weights", i32 1, i32 2000} -- 2.7.4