diff options
| -rw-r--r-- | llvm/include/llvm/Analysis/TargetTransformInfoImpl.h | 3 | ||||
| -rw-r--r-- | llvm/include/llvm/CodeGen/BasicTTIImpl.h | 19 | ||||
| -rw-r--r-- | llvm/test/Transforms/SimplifyCFG/X86/speculate-cttz-ctlz.ll | 189 |
3 files changed, 209 insertions, 2 deletions
diff --git a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h index b90f6888238..364d458b5d0 100644 --- a/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h +++ b/llvm/include/llvm/Analysis/TargetTransformInfoImpl.h @@ -421,8 +421,7 @@ public: return TTI::TCC_Free; } - // Otherwise delegate to the fully generic implementations. - return getOperationCost( + return static_cast<T *>(this)->getOperationCost( Operator::getOpcode(U), U->getType(), U->getNumOperands() == 1 ? U->getOperand(0)->getType() : nullptr); } diff --git a/llvm/include/llvm/CodeGen/BasicTTIImpl.h b/llvm/include/llvm/CodeGen/BasicTTIImpl.h index f3f6ba21eea..25a74b331de 100644 --- a/llvm/include/llvm/CodeGen/BasicTTIImpl.h +++ b/llvm/include/llvm/CodeGen/BasicTTIImpl.h @@ -195,6 +195,25 @@ public: return TargetTransformInfo::TCC_Basic; } + unsigned getOperationCost(unsigned Opcode, Type *Ty, Type *OpTy) { + const TargetLoweringBase *TLI = getTLI(); + switch (Opcode) { + default: break; + case Instruction::Trunc: { + if (TLI->isTruncateFree(OpTy, Ty)) + return TargetTransformInfo::TCC_Free; + return TargetTransformInfo::TCC_Basic; + } + case Instruction::ZExt: { + if (TLI->isZExtFree(OpTy, Ty)) + return TargetTransformInfo::TCC_Free; + return TargetTransformInfo::TCC_Basic; + } + } + + return BaseT::getOperationCost(Opcode, Ty, OpTy); + } + void getUnrollingPreferences(Loop *L, TTI::UnrollingPreferences &UP) { // This unrolling functionality is target independent, but to provide some // motivation for its intended use, for x86: diff --git a/llvm/test/Transforms/SimplifyCFG/X86/speculate-cttz-ctlz.ll b/llvm/test/Transforms/SimplifyCFG/X86/speculate-cttz-ctlz.ll index 6873d7da788..69f6c69059d 100644 --- a/llvm/test/Transforms/SimplifyCFG/X86/speculate-cttz-ctlz.ll +++ b/llvm/test/Transforms/SimplifyCFG/X86/speculate-cttz-ctlz.ll @@ -133,6 +133,195 @@ cond.end: ; preds = %entry, %cond.true ret i16 %cond } +; The following tests verify that calls to cttz/ctlz are speculated even if +; basic block %cond.true has an extra zero extend/truncate which is "free" +; for the target. + +define i64 @test1e(i32 %x) { +; ALL-LABEL: @test1e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i32 %x, 0 +; ALL: [[CTTZ:%[A-Za-z0-9]+]] = tail call i32 @llvm.cttz.i32(i32 %x, i1 true) +; ALL: [[ZEXT:%[A-Za-z0-9]+]] = zext i32 [[CTTZ]] to i64 +; BMI-NEXT: select i1 [[COND]], i64 32, i64 [[ZEXT]] +; LZCNT-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i32 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i32 @llvm.cttz.i32(i32 %x, i1 true) + %phitmp2 = zext i32 %0 to i64 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i64 [ %phitmp2, %cond.true ], [ 32, %entry ] + ret i64 %cond +} + +define i32 @test2e(i64 %x) { +; ALL-LABEL: @test2e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i64 %x, 0 +; ALL: [[CTTZ:%[A-Za-z0-9]+]] = tail call i64 @llvm.cttz.i64(i64 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i64 [[CTTZ]] to i32 +; BMI-NEXT: select i1 [[COND]], i32 64, i32 [[TRUNC]] +; LZCNT-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i64 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i64 @llvm.cttz.i64(i64 %x, i1 true) + %cast = trunc i64 %0 to i32 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i32 [ %cast, %cond.true ], [ 64, %entry ] + ret i32 %cond +} + +define i64 @test3e(i32 %x) { +; ALL-LABEL: @test3e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i32 %x, 0 +; ALL: [[CTLZ:%[A-Za-z0-9]+]] = tail call i32 @llvm.ctlz.i32(i32 %x, i1 true) +; ALL: [[ZEXT:%[A-Za-z0-9]+]] = zext i32 [[CTLZ]] to i64 +; LZCNT-NEXT: select i1 [[COND]], i64 32, i64 [[ZEXT]] +; BMI-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i32 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i32 @llvm.ctlz.i32(i32 %x, i1 true) + %phitmp2 = zext i32 %0 to i64 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i64 [ %phitmp2, %cond.true ], [ 32, %entry ] + ret i64 %cond +} + +define i32 @test4e(i64 %x) { +; ALL-LABEL: @test4e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i64 %x, 0 +; ALL: [[CTLZ:%[A-Za-z0-9]+]] = tail call i64 @llvm.ctlz.i64(i64 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i64 [[CTLZ]] to i32 +; LZCNT-NEXT: select i1 [[COND]], i32 64, i32 [[TRUNC]] +; BMI-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i64 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i64 @llvm.ctlz.i64(i64 %x, i1 true) + %cast = trunc i64 %0 to i32 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i32 [ %cast, %cond.true ], [ 64, %entry ] + ret i32 %cond +} + +define i16 @test5e(i64 %x) { +; ALL-LABEL: @test5e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i64 %x, 0 +; ALL: [[CTLZ:%[A-Za-z0-9]+]] = tail call i64 @llvm.ctlz.i64(i64 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i64 [[CTLZ]] to i16 +; LZCNT-NEXT: select i1 [[COND]], i16 64, i16 [[TRUNC]] +; BMI-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i64 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i64 @llvm.ctlz.i64(i64 %x, i1 true) + %cast = trunc i64 %0 to i16 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i16 [ %cast, %cond.true ], [ 64, %entry ] + ret i16 %cond +} + +define i16 @test6e(i32 %x) { +; ALL-LABEL: @test6e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i32 %x, 0 +; ALL: [[CTLZ:%[A-Za-z0-9]+]] = tail call i32 @llvm.ctlz.i32(i32 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i32 [[CTLZ]] to i16 +; LZCNT-NEXT: select i1 [[COND]], i16 32, i16 [[TRUNC]] +; BMI-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i32 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i32 @llvm.ctlz.i32(i32 %x, i1 true) + %cast = trunc i32 %0 to i16 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i16 [ %cast, %cond.true ], [ 32, %entry ] + ret i16 %cond +} + +define i16 @test7e(i64 %x) { +; ALL-LABEL: @test7e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i64 %x, 0 +; ALL: [[CTTZ:%[A-Za-z0-9]+]] = tail call i64 @llvm.cttz.i64(i64 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i64 [[CTTZ]] to i16 +; BMI-NEXT: select i1 [[COND]], i16 64, i16 [[TRUNC]] +; LZCNT-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i64 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i64 @llvm.cttz.i64(i64 %x, i1 true) + %cast = trunc i64 %0 to i16 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i16 [ %cast, %cond.true ], [ 64, %entry ] + ret i16 %cond +} + +define i16 @test8e(i32 %x) { +; ALL-LABEL: @test8e( +; ALL: [[COND:%[A-Za-z0-9]+]] = icmp eq i32 %x, 0 +; ALL: [[CTTZ:%[A-Za-z0-9]+]] = tail call i32 @llvm.cttz.i32(i32 %x, i1 true) +; ALL: [[TRUNC:%[A-Za-z0-9]+]] = trunc i32 [[CTTZ]] to i16 +; BMI-NEXT: select i1 [[COND]], i16 32, i16 [[TRUNC]] +; LZCNT-NOT: select +; GENERIC-NOT: select +; ALL: ret +entry: + %tobool = icmp eq i32 %x, 0 + br i1 %tobool, label %cond.end, label %cond.true + +cond.true: ; preds = %entry + %0 = tail call i32 @llvm.cttz.i32(i32 %x, i1 true) + %cast = trunc i32 %0 to i16 + br label %cond.end + +cond.end: ; preds = %entry, %cond.true + %cond = phi i16 [ %cast, %cond.true ], [ 32, %entry ] + ret i16 %cond +} + + declare i64 @llvm.ctlz.i64(i64, i1) declare i32 @llvm.ctlz.i32(i32, i1) declare i16 @llvm.ctlz.i16(i16, i1) |
