[ARM CGP] Fix ConvertTruncs
ConvertTruncs is used to replace a trunc for an AND mask, however
this function wasn't working as expected. By performing the change
later, we can create a wide type integer mask instead of a narrow -1
value, which could then be simply removed (incorrectly). Because we
now perform this action later, it's necessary to cache the trunc type
before we perform the promotion.
Differential Revision: https://reviews.llvm.org/D57686
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@354108 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Target/ARM/ARMCodeGenPrepare.cpp b/lib/Target/ARM/ARMCodeGenPrepare.cpp
index bfa836c..e9f5e6a 100644
--- a/lib/Target/ARM/ARMCodeGenPrepare.cpp
+++ b/lib/Target/ARM/ARMCodeGenPrepare.cpp
@@ -686,6 +686,7 @@
}
void IRPromoter::Cleanup() {
+ LLVM_DEBUG(dbgs() << "ARM CGP: Cleanup..\n");
// Some zexts will now have become redundant, along with their trunc
// operands, so remove them
for (auto V : *Visited) {
@@ -729,6 +730,7 @@
}
void IRPromoter::ConvertTruncs() {
+ LLVM_DEBUG(dbgs() << "ARM CGP: Converting truncs..\n");
IRBuilder<> Builder{Ctx};
for (auto *V : *Visited) {
@@ -736,12 +738,13 @@
continue;
auto *Trunc = cast<TruncInst>(V);
- assert(LessThanTypeSize(Trunc) && "expected narrow trunc");
-
Builder.SetInsertPoint(Trunc);
- unsigned NumBits =
- cast<IntegerType>(Trunc->getType())->getScalarSizeInBits();
- ConstantInt *Mask = ConstantInt::get(Ctx, APInt::getMaxValue(NumBits));
+ IntegerType *SrcTy = cast<IntegerType>(Trunc->getOperand(0)->getType());
+ IntegerType *DestTy = cast<IntegerType>(TruncTysMap[Trunc][0]);
+
+ unsigned NumBits = DestTy->getScalarSizeInBits();
+ ConstantInt *Mask =
+ ConstantInt::get(SrcTy, APInt::getMaxValue(NumBits).getZExtValue());
Value *Masked = Builder.CreateAnd(Trunc->getOperand(0), Mask);
if (auto *I = dyn_cast<Instruction>(Masked))
@@ -783,6 +786,12 @@
TruncTysMap[I].push_back(I->getOperand(i)->getType());
}
}
+ for (auto *V : Visited) {
+ if (!isa<TruncInst>(V) || Sources.count(V))
+ continue;
+ auto *Trunc = cast<TruncInst>(V);
+ TruncTysMap[Trunc].push_back(Trunc->getDestTy());
+ }
// Convert adds and subs using negative immediates to equivalent instructions
// that use positive constants.
@@ -791,14 +800,14 @@
// Insert zext instructions between sources and their users.
ExtendSources();
- // Convert any truncs, that aren't sources, into AND masks.
- ConvertTruncs();
-
// Promote visited instructions, mutating their types in place. Also insert
// DSP intrinsics, if enabled, for adds and subs which would be unsafe to
// promote.
PromoteTree();
+ // Convert any truncs, that aren't sources, into AND masks.
+ ConvertTruncs();
+
// Insert trunc instructions for use by calls, stores etc...
TruncateSinks();
diff --git a/test/CodeGen/ARM/CGP/arm-cgp-calls.ll b/test/CodeGen/ARM/CGP/arm-cgp-calls.ll
index 4c95ade..8d58c8e 100644
--- a/test/CodeGen/ARM/CGP/arm-cgp-calls.ll
+++ b/test/CodeGen/ARM/CGP/arm-cgp-calls.ll
@@ -201,7 +201,7 @@
}
; CHECK-LABEL: promote_arg_pass_to_call
-; CHECK-NOT: uxt
+; CHECK: uxtb
define i16 @promote_arg_pass_to_call(i16 zeroext %arg1, i16 zeroext %arg2) {
%conv = add nuw i16 %arg1, 15
%mul = mul nuw nsw i16 %conv, 3
diff --git a/test/CodeGen/ARM/CGP/arm-cgp-casts.ll b/test/CodeGen/ARM/CGP/arm-cgp-casts.ll
index 3e5e2fa..f0f444a 100644
--- a/test/CodeGen/ARM/CGP/arm-cgp-casts.ll
+++ b/test/CodeGen/ARM/CGP/arm-cgp-casts.ll
@@ -1,11 +1,11 @@
-; RUN: llc -mtriple=thumbv8.main -mcpu=cortex-m33 %s -arm-disable-cgp=false -o - | FileCheck %s --check-prefix=CHECK-COMMON --check-prefix=CHECK-NODSP
-; RUN: llc -mtriple=thumbv7-linux-android %s -arm-disable-cgp=false -o - | FileCheck %s --check-prefix=CHECK-COMMON --check-prefix=CHECK-NODSP
-; RUN: llc -mtriple=thumbv7em %s -arm-disable-cgp=false -arm-enable-scalar-dsp=true -o - | FileCheck %s --check-prefix=CHECK-COMMON --check-prefix=CHECK-DSP
-; RUN: llc -mtriple=thumbv8 %s -arm-disable-cgp=false -arm-enable-scalar-dsp=true -arm-enable-scalar-dsp-imms=true -o - | FileCheck %s --check-prefix=CHECK-COMMON --check-prefix=CHECK-DSP-IMM
+; RUN: llc -mtriple=thumbv8.main -mcpu=cortex-m33 %s -arm-disable-cgp=false -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-NODSP
+; RUN: llc -mtriple=thumbv7-linux-android %s -arm-disable-cgp=false -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-NODSP
+; RUN: llc -mtriple=thumbv7em %s -arm-disable-cgp=false -arm-enable-scalar-dsp=true -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-DSP
+; RUN: llc -mtriple=thumbv8 %s -arm-disable-cgp=false -arm-enable-scalar-dsp=true -arm-enable-scalar-dsp-imms=true -o - | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-DSP-IMM
; Transform will fail because the trunc is not a sink.
-; CHECK-COMMON-LABEL: dsp_trunc
-; CHECK-COMMON: add [[ADD:[^ ]+]],
+; CHECK-LABEL: dsp_trunc
+; CHECK: add [[ADD:[^ ]+]],
; CHECK-DSP-NEXT: ldrh r1, [r3]
; CHECK-DSP-NEXT: ldrh r2, [r2]
; CHECK-DSP-NEXT: subs r1, r1, [[ADD]]
@@ -14,15 +14,10 @@
; CHECK-DSP-NEXT: uxth r2, r0
; CHECK-DSP-NEXT: cmp r2, r3
-; With DSP-IMM, we could have:
-; movs r1, #0
-; uxth r0, r0
-; usub16 r1, r1, r0
-; ldrh r0, [r2]
-; ldrh r3, [r3]
-; usub16 r0, r0, r1
-; uadd16 r1, r3, r1
-; cmp r0, r1
+; CHECK-DSP-IMM: usub16
+; CHECK-DSP-IMM: usub16
+; CHECK-DSP-IMM: uadd16
+; CHECK-DSP-IMM: cmp
define i16 @dsp_trunc(i32 %arg0, i32 %arg1, i16* %gep0, i16* %gep1) {
entry:
%add0 = add i32 %arg0, %arg1
@@ -37,10 +32,10 @@
ret i16 %res
}
-; CHECK-COMMON-LABEL: trunc_i16_i8
-; CHECK-COMMON: ldrh
-; CHECK-COMMON: uxtb
-; CHECK-COMMON: cmp
+; CHECK-LABEL: trunc_i16_i8
+; CHECK: ldrh
+; CHECK: uxtb
+; CHECK: cmp
define i8 @trunc_i16_i8(i16* %ptr, i16 zeroext %arg0, i8 zeroext %arg1) {
entry:
%0 = load i16, i16* %ptr
@@ -53,10 +48,10 @@
; The pass perform the transform, but a uxtb will still be inserted to handle
; the zext to the icmp.
-; CHECK-COMMON-LABEL: icmp_i32_zext:
-; CHECK-COMMON: sub
-; CHECK-COMMON: uxtb
-; CHECK-COMMON: cmp
+; CHECK-LABEL: icmp_i32_zext:
+; CHECK: sub
+; CHECK: uxtb
+; CHECK: cmp
define i8 @icmp_i32_zext(i8* %ptr) {
entry:
%gep = getelementptr inbounds i8, i8* %ptr, i32 0
@@ -86,9 +81,9 @@
}
; Won't don't handle sext
-; CHECK-COMMON-LABEL: icmp_sext_zext_store_i8_i16
-; CHECK-COMMON: ldrb
-; CHECK-COMMON: ldrsh
+; CHECK-LABEL: icmp_sext_zext_store_i8_i16
+; CHECK: ldrb
+; CHECK: ldrsh
define i32 @icmp_sext_zext_store_i8_i16() {
entry:
%0 = load i8, i8* getelementptr inbounds ([16 x i8], [16 x i8]* @d_uch, i32 0, i32 2), align 1
@@ -102,11 +97,11 @@
ret i32 %conv3
}
-; CHECK-COMMON-LABEL: or_icmp_ugt:
-; CHECK-COMMON: ldrb
-; CHECK-COMMON: subs.w
-; CHECK-COMMON-NOT: uxt
-; CHECK-COMMON: cmp
+; CHECK-LABEL: or_icmp_ugt:
+; CHECK: ldrb
+; CHECK: subs.w
+; CHECK-NOT: uxt
+; CHECK: cmp
define i1 @or_icmp_ugt(i32 %arg, i8* %ptr) {
entry:
%0 = load i8, i8* %ptr
@@ -122,10 +117,10 @@
; We currently only handle truncs as sinks, so a uxt will still be needed for
; the icmp ugt instruction.
-; CHECK-COMMON-LABEL: urem_trunc_icmps
-; CHECK-COMMON: cmp
-; CHECK-COMMON: uxt
-; CHECK-COMMON: cmp
+; CHECK-LABEL: urem_trunc_icmps
+; CHECK: cmp
+; CHECK: uxt
+; CHECK: cmp
define void @urem_trunc_icmps(i16** %in, i32* %g, i32* %k) {
entry:
%ptr = load i16*, i16** %in, align 4
@@ -157,10 +152,10 @@
; Check that %exp requires uxth in all cases, and will also be required to
; promote %1 for the call - unless we can generate a uadd16.
-; CHECK-COMMON-LABEL: zext_load_sink_call:
-; CHECK-COMMON: uxt
+; CHECK-LABEL: zext_load_sink_call:
+; CHECK: uxt
; CHECK-DSP-IMM: uadd16
-; CHECK-COMMON: cmp
+; CHECK: cmp
; CHECK-NODSP: uxt
; CHECK-DSP-IMM-NOT: uxt
define i32 @zext_load_sink_call(i16* %ptr, i16 %exp) {
@@ -181,8 +176,8 @@
ret i32 %exitval
}
-; CHECK-COMMON-LABEL: bitcast_i16
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: bitcast_i16
+; CHECK-NOT: uxt
define i16 @bitcast_i16(i16 zeroext %arg0, i16 zeroext %arg1) {
entry:
%cast = bitcast i16 12345 to i16
@@ -192,8 +187,8 @@
ret i16 %res
}
-; CHECK-COMMON-LABEL: bitcast_i8
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: bitcast_i8
+; CHECK-NOT: uxt
define i8 @bitcast_i8(i8 zeroext %arg0, i8 zeroext %arg1) {
entry:
%cast = bitcast i8 127 to i8
@@ -203,8 +198,8 @@
ret i8 %res
}
-; CHECK-COMMON-LABEL: bitcast_i16_minus
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: bitcast_i16_minus
+; CHECK-NOT: uxt
define i16 @bitcast_i16_minus(i16 zeroext %arg0, i16 zeroext %arg1) {
entry:
%cast = bitcast i16 -12345 to i16
@@ -214,8 +209,8 @@
ret i16 %res
}
-; CHECK-COMMON-LABEL: bitcast_i8_minus
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: bitcast_i8_minus
+; CHECK-NOT: uxt
define i8 @bitcast_i8_minus(i8 zeroext %arg0, i8 zeroext %arg1) {
entry:
%cast = bitcast i8 -127 to i8
@@ -231,8 +226,8 @@
@sh1 = hidden local_unnamed_addr global i16 0, align 2
@d_sh = hidden local_unnamed_addr global [16 x i16] zeroinitializer, align 2
-; CHECK-COMMON-LABEL: two_stage_zext_trunc_mix
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: two_stage_zext_trunc_mix
+; CHECK-NOT: uxt
define i8* @two_stage_zext_trunc_mix(i32* %this, i32 %__pos1, i32 %__n1, i32** %__str, i32 %__pos2, i32 %__n2) {
entry:
%__size_.i.i.i.i = bitcast i32** %__str to i8*
@@ -255,8 +250,8 @@
ret i8* %res
}
-; CHECK-COMMON-LABEL: search_through_zext_1
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: search_through_zext_1
+; CHECK-NOT: uxt
define i8 @search_through_zext_1(i8 zeroext %a, i8 zeroext %b, i16 zeroext %c) {
entry:
%add = add nuw i8 %a, %b
@@ -279,9 +274,9 @@
; TODO: We should be able to remove the uxtb here. The transform fails because
; the icmp ugt uses an i32, which is too large... but this doesn't matter
; because it won't be writing a large value to a register as a result.
-; CHECK-COMMON-LABEL: search_through_zext_2
-; CHECK-COMMON: uxtb
-; CHECK-COMMON: uxtb
+; CHECK-LABEL: search_through_zext_2
+; CHECK: uxtb
+; CHECK: uxtb
define i8 @search_through_zext_2(i8 zeroext %a, i8 zeroext %b, i16 zeroext %c, i32 %d) {
entry:
%add = add nuw i8 %a, %b
@@ -304,9 +299,9 @@
; TODO: We should be able to remove the uxtb here as all the calculations are
; performed on i8s. The promotion of i8 to i16 and then the later truncation
; results in the uxtb.
-; CHECK-COMMON-LABEL: search_through_zext_3
-; CHECK-COMMON: uxtb
-; CHECK-COMMON: uxtb
+; CHECK-LABEL: search_through_zext_3
+; CHECK: uxtb
+; CHECK: uxtb
define i8 @search_through_zext_3(i8 zeroext %a, i8 zeroext %b, i16 zeroext %c, i32 %d) {
entry:
%add = add nuw i8 %a, %b
@@ -328,8 +323,8 @@
}
; TODO: We should be able to remove the uxt that gets introduced for %conv2
-; CHECK-COMMON-LABEL: search_through_zext_cmp
-; CHECK-COMMON: uxt
+; CHECK-LABEL: search_through_zext_cmp
+; CHECK: uxt
define i8 @search_through_zext_cmp(i8 zeroext %a, i8 zeroext %b, i16 zeroext %c) {
entry:
%cmp = icmp ne i8 %a, %b
@@ -349,8 +344,8 @@
ret i8 %retval
}
-; CHECK-COMMON-LABEL: search_through_zext_load
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: search_through_zext_load
+; CHECK-NOT: uxt
define i8 @search_through_zext_load(i8* %a, i8 zeroext %b, i16 zeroext %c) {
entry:
%load = load i8, i8* %a
@@ -370,10 +365,10 @@
ret i8 %retval
}
-; CHECK-COMMON-LABEL: trunc_sink_less_than
-; CHECK-COMMON-NOT: uxth
-; CHECK-COMMON: cmp
-; CHECK-COMMON: uxtb
+; CHECK-LABEL: trunc_sink_less_than
+; CHECK-NOT: uxth
+; CHECK: cmp
+; CHECK: uxtb
define i16 @trunc_sink_less_than_cmp(i16 zeroext %a, i16 zeroext %b, i16 zeroext %c, i8 zeroext %d) {
entry:
%sub = sub nuw i16 %b, %a
@@ -393,10 +388,10 @@
}
; TODO: We should be able to remove the uxth introduced to handle %sub
-; CHECK-COMMON-LABEL: trunc_sink_less_than_arith
-; CHECK-COMMON: uxth
-; CHECK-COMMON: cmp
-; CHECK-COMMON: uxtb
+; CHECK-LABEL: trunc_sink_less_than_arith
+; CHECK: uxth
+; CHECK: cmp
+; CHECK: uxtb
define i16 @trunc_sink_less_than_arith(i16 zeroext %a, i16 zeroext %b, i16 zeroext %c, i8 zeroext %d, i8 zeroext %e) {
entry:
%sub = sub nuw i16 %b, %a
@@ -415,10 +410,10 @@
ret i16 %retval
}
-; CHECK-COMMON-LABEL: trunc_sink_less_than_store
-; CHECK-COMMON-NOT: uxt
-; CHECK-COMMON: cmp
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: trunc_sink_less_than_store
+; CHECK-NOT: uxt
+; CHECK: cmp
+; CHECK-NOT: uxt
define i16 @trunc_sink_less_than_store(i16 zeroext %a, i16 zeroext %b, i16 zeroext %c, i8 zeroext %d, i8* %e) {
entry:
%sub = sub nuw i16 %b, %a
@@ -436,8 +431,8 @@
ret i16 %retval
}
-; CHECK-COMMON-LABEL: trunc_sink_less_than_ret
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: trunc_sink_less_than_ret
+; CHECK: uxt{{.*}}b
define i8 @trunc_sink_less_than_ret(i16 zeroext %a, i16 zeroext %b, i16 zeroext %c, i8 zeroext %d, i8 zeroext %e) {
entry:
%sub = sub nuw i16 %b, %a
@@ -454,10 +449,10 @@
ret i8 %retval
}
-; CHECK-COMMON-LABEL: trunc_sink_less_than_zext_ret
-; CHECK-COMMON-NOT: uxth
-; CHECK-COMMON: sub
-; CHECK-COMMON: uxtb
+; CHECK-LABEL: trunc_sink_less_than_zext_ret
+; CHECK-NOT: uxth
+; CHECK: sub
+; CHECK: uxtb
define zeroext i8 @trunc_sink_less_than_zext_ret(i16 zeroext %a, i16 zeroext %b, i16 zeroext %c, i8 zeroext %d, i8 zeroext %e) {
entry:
%sub = sub nuw i16 %b, %a
@@ -474,8 +469,8 @@
ret i8 %retval
}
-; CHECK-COMMON-LABEL: bitcast_i1
-; CHECK-COMMON-NOT: uxt
+; CHECK-LABEL: bitcast_i1
+; CHECK-NOT: uxt
define i32 @bitcast_i1(i16 zeroext %a, i32 %b, i32 %c) {
entry:
%0 = bitcast i1 1 to i1
@@ -495,11 +490,11 @@
ret i32 %retval
}
-; CHECK-COMMON-LABEL: search_back_through_trunc
-; CHECK-COMMON-NOT: uxt
-; CHECK-COMMON: cmp
-; CHECK-COMMON: strb
-; CHECK-COMMON: strb
+; CHECK-LABEL: search_back_through_trunc
+; CHECK-NOT: uxt
+; CHECK: cmp
+; CHECK: strb
+; CHECK: strb
define void @search_back_through_trunc(i8* %a, i8* %b, i8* %c, i8* %d, i16* %e) {
entry:
%0 = load i8, i8* %a, align 1
@@ -590,9 +585,10 @@
}
; CHECK-LABEL: dont_replace_trunc_1
-; CHECK: cmp
+; CHECK: sxth
; CHECK: uxtb
-define void @dont_replace_trunc_1(i8* %a, i16* %b, i16* %c, i32* %d, i8* %e, i32* %f) {
+; CHECK: uxth
+define i1 @dont_replace_trunc_1(i8* %a, i16* %b, i16* %c, i32* %d, i8* %e, i32* %f) {
entry:
%0 = load i16, i16* %c, align 2
%1 = load i16, i16* %b, align 2
@@ -608,13 +604,7 @@
%4 = zext i8 %narrow to i16
%conv5 = or i16 %0, %4
%tobool = icmp eq i16 %conv5, 0
- br i1 %tobool, label %if.end, label %for.cond
-
-for.cond: ; preds = %entry, %for.cond
- br label %for.cond
-
-if.end: ; preds = %entry
- ret void
+ ret i1 %tobool
}
; CHECK-LABEL: dont_replace_trunc_2
@@ -632,3 +622,24 @@
%conv5 = zext i8 %or to i32
ret i32 %conv5
}
+
+; CHECK-LABEL: replace_trunk_with_mask
+; CHECK: div
+; CHECK: uxtb
+define i32 @replace_trunk_with_mask(i16* %a) {
+entry:
+ %0 = load i16, i16* %a
+ %cmp = icmp eq i16 %0, 0
+ br i1 %cmp, label %cond.end, label %cond.false
+
+cond.false:
+ %1 = urem i16 535, %0
+ %.lhs.trunc = trunc i16 %1 to i8
+ %2 = udiv i8 %.lhs.trunc, 3
+ %phitmp = zext i8 %2 to i32
+ br label %cond.end
+
+cond.end:
+ %cond = phi i32 [ %phitmp, %cond.false ], [ 0, %entry ]
+ ret i32 %cond
+}
