[AMDGPULowerBufferFatPointers] Simplify and fix GEP offset emission #95115
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Use emitGEPOffset() to emit the GEP offset, which already has all the necessary logic. This also fixes the nuw flag incorrectly being set on the offset calculation, while only nsw is correct.
|
@llvm/pr-subscribers-backend-amdgpu Author: Nikita Popov (nikic) ChangesUse emitGEPOffset() to emit the GEP offset, which already has all the necessary logic. This also fixes the nuw flag incorrectly being set on the offset calculation, while only nsw is correct. Full diff: https://github.com/llvm/llvm-project/pull/95115.diff 3 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp b/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
index f878bd9465d38..77238d184adb0 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
@@ -200,6 +200,7 @@
#include "llvm/ADT/SetOperations.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/Analysis/ConstantFolding.h"
+#include "llvm/Analysis/Utils/Local.h"
#include "llvm/CodeGen/TargetPassConfig.h"
#include "llvm/IR/AttributeMask.h"
#include "llvm/IR/Constants.h"
@@ -1393,50 +1394,17 @@ PtrParts SplitPtrStructs::visitGetElementPtrInst(GetElementPtrInst &GEP) {
IRB.SetInsertPoint(&GEP);
auto [Rsrc, Off] = getPtrParts(Ptr);
- Type *OffTy = Off->getType();
const DataLayout &DL = GEP.getModule()->getDataLayout();
bool InBounds = GEP.isInBounds();
- // In order to call collectOffset() and thus not have to reimplement it,
- // we need the GEP's pointer operand to have ptr addrspace(7) type
- GEP.setOperand(GEP.getPointerOperandIndex(),
- PoisonValue::get(IRB.getPtrTy(AMDGPUAS::BUFFER_FAT_POINTER)));
- MapVector<Value *, APInt> VariableOffs;
- APInt ConstOffVal = APInt::getZero(BufferOffsetWidth);
- if (!GEP.collectOffset(DL, BufferOffsetWidth, VariableOffs, ConstOffVal))
- report_fatal_error("Scalable vector or unsized struct in fat pointer GEP");
- GEP.setOperand(GEP.getPointerOperandIndex(), Ptr);
- Value *OffAccum = nullptr;
- // Accumulate offsets together before adding to the base in order to preserve
- // as many of the inbounds properties as possible.
- for (auto [Arg, Multiple] : VariableOffs) {
- if (auto *OffVecTy = dyn_cast<VectorType>(OffTy))
- if (!Arg->getType()->isVectorTy())
- Arg = IRB.CreateVectorSplat(OffVecTy->getElementCount(), Arg);
- Arg = IRB.CreateIntCast(Arg, OffTy, /*isSigned=*/true);
- if (!Multiple.isOne()) {
- if (Multiple.isPowerOf2())
- Arg = IRB.CreateShl(Arg, Multiple.logBase2(), "", /*hasNUW=*/InBounds,
- /*HasNSW=*/InBounds);
- else
- Arg = IRB.CreateMul(Arg, ConstantExpr::getIntegerValue(OffTy, Multiple),
- "", /*hasNUW=*/InBounds, /*hasNSW=*/InBounds);
- }
- if (OffAccum)
- OffAccum = IRB.CreateAdd(OffAccum, Arg, "", /*hasNUW=*/InBounds,
- /*hasNSW=*/InBounds);
- else
- OffAccum = Arg;
- }
- if (!ConstOffVal.isZero()) {
- Constant *ConstOff = ConstantExpr::getIntegerValue(OffTy, ConstOffVal);
- if (OffAccum)
- OffAccum = IRB.CreateAdd(OffAccum, ConstOff, "", /*hasNUW=*/InBounds,
- /*hasNSW=*/InBounds);
- else
- OffAccum = ConstOff;
- }
-
+ // In order to call emitGEPOffset() and thus not have to reimplement it,
+ // we need the GEP result to have ptr addrspace(7) type.
+ Type *FatPtrTy = IRB.getPtrTy(AMDGPUAS::BUFFER_FAT_POINTER);
+ if (auto *VT = dyn_cast<VectorType>(Off->getType()))
+ FatPtrTy = VectorType::get(FatPtrTy, VT->getElementCount());
+ GEP.mutateType(FatPtrTy);
+ Value *OffAccum = emitGEPOffset(&IRB, DL, &GEP);
+ GEP.mutateType(Ptr->getType());
if (!OffAccum) { // Constant-zero offset
SplitUsers.insert(&GEP);
return {Rsrc, Off};
diff --git a/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-pointer-ops.ll b/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-pointer-ops.ll
index 85cc810988a71..60c6890689ccb 100644
--- a/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-pointer-ops.ll
+++ b/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-pointer-ops.ll
@@ -10,12 +10,13 @@ define ptr addrspace(7) @gep(ptr addrspace(7) %in, i32 %idx) {
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[IN:%.*]], i32 [[IDX:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: [[IN_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[IN]], 0
; CHECK-NEXT: [[IN_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[IN]], 1
-; CHECK-NEXT: [[TMP1:%.*]] = mul nuw nsw i32 [[IDX]], 40
-; CHECK-NEXT: [[TMP2:%.*]] = add nuw nsw i32 [[TMP1]], 32
-; CHECK-NEXT: [[RET:%.*]] = add i32 [[IN_OFF]], [[TMP2]]
-; CHECK-NEXT: [[TMP3:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[IN_RSRC]], 0
-; CHECK-NEXT: [[TMP4:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP3]], i32 [[RET]], 1
-; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP4]]
+; CHECK-NEXT: [[RET_IDX:%.*]] = mul nsw i32 [[IDX]], 40
+; CHECK-NEXT: [[RET_OFFS:%.*]] = add nsw i32 [[RET_IDX]], 8
+; CHECK-NEXT: [[RET_OFFS1:%.*]] = add nsw i32 [[RET_OFFS]], 24
+; CHECK-NEXT: [[RET:%.*]] = add i32 [[IN_OFF]], [[RET_OFFS1]]
+; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[IN_RSRC]], 0
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[RET]], 1
+; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
;
%ret = getelementptr inbounds {i32, [4 x ptr]}, ptr addrspace(7) %in, i32 %idx, i32 1, i32 3
ret ptr addrspace(7) %ret
@@ -26,12 +27,13 @@ define <2 x ptr addrspace(7)> @gep_vectors(<2 x ptr addrspace(7)> %in, <2 x i32>
; CHECK-SAME: ({ <2 x ptr addrspace(8)>, <2 x i32> } [[IN:%.*]], <2 x i32> [[IDX:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: [[IN_RSRC:%.*]] = extractvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[IN]], 0
; CHECK-NEXT: [[IN_OFF:%.*]] = extractvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[IN]], 1
-; CHECK-NEXT: [[TMP1:%.*]] = mul nuw nsw <2 x i32> [[IDX]], <i32 40, i32 40>
-; CHECK-NEXT: [[TMP2:%.*]] = add nuw nsw <2 x i32> [[TMP1]], <i32 32, i32 32>
-; CHECK-NEXT: [[RET:%.*]] = add <2 x i32> [[IN_OFF]], [[TMP2]]
-; CHECK-NEXT: [[TMP3:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } poison, <2 x ptr addrspace(8)> [[IN_RSRC]], 0
-; CHECK-NEXT: [[TMP4:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP3]], <2 x i32> [[RET]], 1
-; CHECK-NEXT: ret { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP4]]
+; CHECK-NEXT: [[RET_IDX:%.*]] = mul nsw <2 x i32> [[IDX]], <i32 40, i32 40>
+; CHECK-NEXT: [[RET_OFFS:%.*]] = add nsw <2 x i32> [[RET_IDX]], <i32 8, i32 8>
+; CHECK-NEXT: [[RET_OFFS1:%.*]] = add nsw <2 x i32> [[RET_OFFS]], <i32 24, i32 24>
+; CHECK-NEXT: [[RET:%.*]] = add <2 x i32> [[IN_OFF]], [[RET_OFFS1]]
+; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } poison, <2 x ptr addrspace(8)> [[IN_RSRC]], 0
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP1]], <2 x i32> [[RET]], 1
+; CHECK-NEXT: ret { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP2]]
;
%ret = getelementptr inbounds {i32, [4 x ptr]}, <2 x ptr addrspace(7)> %in, <2 x i32> %idx, i32 1, i32 3
ret <2 x ptr addrspace(7)> %ret
@@ -44,13 +46,14 @@ define <2 x ptr addrspace(7)> @gep_vector_scalar(<2 x ptr addrspace(7)> %in, i64
; CHECK-NEXT: [[IN_OFF:%.*]] = extractvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[IN]], 1
; CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <2 x i64> poison, i64 [[IDX]], i64 0
; CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <2 x i64> [[DOTSPLATINSERT]], <2 x i64> poison, <2 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP1:%.*]] = trunc <2 x i64> [[DOTSPLAT]] to <2 x i32>
-; CHECK-NEXT: [[TMP2:%.*]] = mul nuw nsw <2 x i32> [[TMP1]], <i32 40, i32 40>
-; CHECK-NEXT: [[TMP3:%.*]] = add nuw nsw <2 x i32> [[TMP2]], <i32 32, i32 32>
-; CHECK-NEXT: [[RET:%.*]] = add <2 x i32> [[IN_OFF]], [[TMP3]]
-; CHECK-NEXT: [[TMP4:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } poison, <2 x ptr addrspace(8)> [[IN_RSRC]], 0
-; CHECK-NEXT: [[TMP5:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP4]], <2 x i32> [[RET]], 1
-; CHECK-NEXT: ret { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP5]]
+; CHECK-NEXT: [[DOTSPLAT_C:%.*]] = trunc <2 x i64> [[DOTSPLAT]] to <2 x i32>
+; CHECK-NEXT: [[RET_IDX:%.*]] = mul nsw <2 x i32> [[DOTSPLAT_C]], <i32 40, i32 40>
+; CHECK-NEXT: [[RET_OFFS:%.*]] = add nsw <2 x i32> [[RET_IDX]], <i32 8, i32 8>
+; CHECK-NEXT: [[RET_OFFS1:%.*]] = add nsw <2 x i32> [[RET_OFFS]], <i32 24, i32 24>
+; CHECK-NEXT: [[RET:%.*]] = add <2 x i32> [[IN_OFF]], [[RET_OFFS1]]
+; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } poison, <2 x ptr addrspace(8)> [[IN_RSRC]], 0
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP1]], <2 x i32> [[RET]], 1
+; CHECK-NEXT: ret { <2 x ptr addrspace(8)>, <2 x i32> } [[TMP2]]
;
%ret = getelementptr inbounds {i32, [4 x ptr]}, <2 x ptr addrspace(7)> %in, i64 %idx, i32 1, i32 3
ret <2 x ptr addrspace(7)> %ret
@@ -61,11 +64,11 @@ define ptr addrspace(7) @simple_gep(ptr addrspace(7) %ptr, i32 %off) {
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[PTR:%.*]], i32 [[OFF:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: [[PTR_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 0
; CHECK-NEXT: [[PTR_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 1
-; CHECK-NEXT: [[TMP1:%.*]] = shl i32 [[OFF]], 2
-; CHECK-NEXT: [[RET:%.*]] = add i32 [[PTR_OFF]], [[TMP1]]
-; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[PTR_RSRC]], 0
-; CHECK-NEXT: [[TMP3:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP2]], i32 [[RET]], 1
-; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP3]]
+; CHECK-NEXT: [[RET_IDX:%.*]] = mul i32 [[OFF]], 4
+; CHECK-NEXT: [[RET:%.*]] = add i32 [[PTR_OFF]], [[RET_IDX]]
+; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[PTR_RSRC]], 0
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[RET]], 1
+; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
;
%ret = getelementptr i32, ptr addrspace(7) %ptr, i32 %off
ret ptr addrspace(7) %ret
@@ -76,11 +79,11 @@ define ptr addrspace(7) @simple_inbounds_gep(ptr addrspace(7) %ptr, i32 %off) {
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[PTR:%.*]], i32 [[OFF:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: [[PTR_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 0
; CHECK-NEXT: [[PTR_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 1
-; CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i32 [[OFF]], 2
-; CHECK-NEXT: [[RET:%.*]] = add i32 [[PTR_OFF]], [[TMP1]]
-; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[PTR_RSRC]], 0
-; CHECK-NEXT: [[TMP3:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP2]], i32 [[RET]], 1
-; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP3]]
+; CHECK-NEXT: [[RET_IDX:%.*]] = mul nsw i32 [[OFF]], 4
+; CHECK-NEXT: [[RET:%.*]] = add i32 [[PTR_OFF]], [[RET_IDX]]
+; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[PTR_RSRC]], 0
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[RET]], 1
+; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
;
%ret = getelementptr inbounds i32, ptr addrspace(7) %ptr, i32 %off
ret ptr addrspace(7) %ret
@@ -91,9 +94,10 @@ define ptr addrspace(7) @zero_gep(ptr addrspace(7) %ptr) {
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[PTR:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: [[PTR_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 0
; CHECK-NEXT: [[PTR_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[PTR]], 1
+; CHECK-NEXT: [[RET:%.*]] = add i32 [[PTR_OFF]], 0
; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[PTR_RSRC]], 0
-; CHECK-NEXT: [[RET:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[PTR_OFF]], 1
-; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[RET]]
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[RET]], 1
+; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
;
%ret = getelementptr i8, ptr addrspace(7) %ptr, i32 0
ret ptr addrspace(7) %ret
@@ -105,9 +109,10 @@ define ptr addrspace(7) @zero_gep_goes_second(ptr addrspace(7) %v0, i32 %arg) {
; CHECK-NEXT: [[V0_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 0
; CHECK-NEXT: [[V0_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 1
; CHECK-NEXT: [[V1:%.*]] = add i32 [[V0_OFF]], [[ARG]]
+; CHECK-NEXT: [[V2:%.*]] = add i32 [[V1]], 0
; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[V0_RSRC]], 0
-; CHECK-NEXT: [[V2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[V1]], 1
-; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[V2]]
+; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[V2]], 1
+; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
;
%v1 = getelementptr i8, ptr addrspace(7) %v0, i32 %arg
%v2 = getelementptr i8, ptr addrspace(7) %v1, i32 0
@@ -119,7 +124,8 @@ define ptr addrspace(7) @zero_gep_goes_first(ptr addrspace(7) %v0, i32 %arg) {
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[V0:%.*]], i32 [[ARG:%.*]]) #[[ATTR0]] {
; CHECK-NEXT: [[V0_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 0
; CHECK-NEXT: [[V0_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 1
-; CHECK-NEXT: [[V2:%.*]] = add i32 [[V0_OFF]], [[ARG]]
+; CHECK-NEXT: [[V1:%.*]] = add i32 [[V0_OFF]], 0
+; CHECK-NEXT: [[V2:%.*]] = add i32 [[V1]], [[ARG]]
; CHECK-NEXT: [[TMP1:%.*]] = insertvalue { ptr addrspace(8), i32 } poison, ptr addrspace(8) [[V0_RSRC]], 0
; CHECK-NEXT: [[TMP2:%.*]] = insertvalue { ptr addrspace(8), i32 } [[TMP1]], i32 [[V2]], 1
; CHECK-NEXT: ret { ptr addrspace(8), i32 } [[TMP2]]
diff --git a/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-unoptimized-debug-data.ll b/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-unoptimized-debug-data.ll
index 69387e67c1c7e..5828d5bf6f11c 100644
--- a/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-unoptimized-debug-data.ll
+++ b/llvm/test/CodeGen/AMDGPU/lower-buffer-fat-pointers-unoptimized-debug-data.ll
@@ -28,32 +28,32 @@ define float @debug_stash_pointer(ptr addrspace(8) %buf, i32 %idx, ptr addrspace
; CHECK-NEXT: [[BUF_PTR_2_PTR_RSRC:%.*]] = inttoptr i128 [[TMP4]] to ptr addrspace(8), !dbg [[DBG27]]
; CHECK-NEXT: [[BUF_PTR_2_PTR_OFF:%.*]] = trunc i160 [[BUF_PTR_2]] to i32, !dbg [[DBG27]]
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata { ptr addrspace(8), i32 } undef, metadata [[META16:![0-9]+]], metadata !DIExpression()), !dbg [[DBG27]]
-; CHECK-NEXT: [[TMP5:%.*]] = shl i32 [[IDX]], 2, !dbg [[DBG28:![0-9]+]]
-; CHECK-NEXT: [[BUF_PTR_3:%.*]] = add i32 [[BUF_PTR_2_PTR_OFF]], [[TMP5]], !dbg [[DBG28]]
+; CHECK-NEXT: [[BUF_PTR_3_IDX:%.*]] = mul i32 [[IDX]], 4, !dbg [[DBG28:![0-9]+]]
+; CHECK-NEXT: [[BUF_PTR_3:%.*]] = add i32 [[BUF_PTR_2_PTR_OFF]], [[BUF_PTR_3_IDX]], !dbg [[DBG28]]
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata { ptr addrspace(8), i32 } undef, metadata [[META17:![0-9]+]], metadata !DIExpression()), !dbg [[DBG28]]
; CHECK-NEXT: [[BUF_PTR_3_INT_RSRC:%.*]] = ptrtoint ptr addrspace(8) [[BUF_PTR_2_PTR_RSRC]] to i160, !dbg [[DBG29:![0-9]+]]
-; CHECK-NEXT: [[TMP6:%.*]] = shl nuw i160 [[BUF_PTR_3_INT_RSRC]], 32, !dbg [[DBG29]]
+; CHECK-NEXT: [[TMP5:%.*]] = shl nuw i160 [[BUF_PTR_3_INT_RSRC]], 32, !dbg [[DBG29]]
; CHECK-NEXT: [[BUF_PTR_3_INT_OFF:%.*]] = zext i32 [[BUF_PTR_3]] to i160, !dbg [[DBG29]]
-; CHECK-NEXT: [[BUF_PTR_3_INT:%.*]] = or i160 [[TMP6]], [[BUF_PTR_3_INT_OFF]], !dbg [[DBG29]]
+; CHECK-NEXT: [[BUF_PTR_3_INT:%.*]] = or i160 [[TMP5]], [[BUF_PTR_3_INT_OFF]], !dbg [[DBG29]]
; CHECK-NEXT: store i160 [[BUF_PTR_3_INT]], ptr addrspace(5) [[BUF_PTR_VAR]], align 32, !dbg [[DBG29]]
; CHECK-NEXT: [[BUF_PTR_4:%.*]] = load i160, ptr addrspace(5) [[BUF_PTR_VAR]], align 32, !dbg [[DBG30:![0-9]+]]
-; CHECK-NEXT: [[TMP7:%.*]] = lshr i160 [[BUF_PTR_4]], 32, !dbg [[DBG30]]
-; CHECK-NEXT: [[TMP8:%.*]] = trunc i160 [[TMP7]] to i128, !dbg [[DBG30]]
-; CHECK-NEXT: [[BUF_PTR_4_PTR_RSRC:%.*]] = inttoptr i128 [[TMP8]] to ptr addrspace(8), !dbg [[DBG30]]
+; CHECK-NEXT: [[TMP6:%.*]] = lshr i160 [[BUF_PTR_4]], 32, !dbg [[DBG30]]
+; CHECK-NEXT: [[TMP7:%.*]] = trunc i160 [[TMP6]] to i128, !dbg [[DBG30]]
+; CHECK-NEXT: [[BUF_PTR_4_PTR_RSRC:%.*]] = inttoptr i128 [[TMP7]] to ptr addrspace(8), !dbg [[DBG30]]
; CHECK-NEXT: [[BUF_PTR_4_PTR_OFF:%.*]] = trunc i160 [[BUF_PTR_4]] to i32, !dbg [[DBG30]]
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata { ptr addrspace(8), i32 } undef, metadata [[META18:![0-9]+]], metadata !DIExpression()), !dbg [[DBG30]]
; CHECK-NEXT: [[RET:%.*]] = call float @llvm.amdgcn.raw.ptr.buffer.load.f32(ptr addrspace(8) align 4 [[BUF_PTR_4_PTR_RSRC]], i32 [[BUF_PTR_4_PTR_OFF]], i32 0, i32 0), !dbg [[DBG31:![0-9]+]]
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata float [[RET]], metadata [[META19:![0-9]+]], metadata !DIExpression()), !dbg [[DBG31]]
; CHECK-NEXT: [[AUX_PTR_2:%.*]] = load i160, ptr addrspace(5) [[AUX_PTR_VAR]], align 32, !dbg [[DBG32:![0-9]+]]
-; CHECK-NEXT: [[TMP9:%.*]] = lshr i160 [[AUX_PTR_2]], 32, !dbg [[DBG32]]
-; CHECK-NEXT: [[TMP10:%.*]] = trunc i160 [[TMP9]] to i128, !dbg [[DBG32]]
-; CHECK-NEXT: [[AUX_PTR_2_PTR_RSRC:%.*]] = inttoptr i128 [[TMP10]] to ptr addrspace(8), !dbg [[DBG32]]
+; CHECK-NEXT: [[TMP8:%.*]] = lshr i160 [[AUX_PTR_2]], 32, !dbg [[DBG32]]
+; CHECK-NEXT: [[TMP9:%.*]] = trunc i160 [[TMP8]] to i128, !dbg [[DBG32]]
+; CHECK-NEXT: [[AUX_PTR_2_PTR_RSRC:%.*]] = inttoptr i128 [[TMP9]] to ptr addrspace(8), !dbg [[DBG32]]
; CHECK-NEXT: [[AUX_PTR_2_PTR_OFF:%.*]] = trunc i160 [[AUX_PTR_2]] to i32, !dbg [[DBG32]]
; CHECK-NEXT: tail call void @llvm.dbg.value(metadata { ptr addrspace(8), i32 } undef, metadata [[META20:![0-9]+]], metadata !DIExpression()), !dbg [[DBG32]]
; CHECK-NEXT: [[BUF_PTR_4_PTR_INT_RSRC:%.*]] = ptrtoint ptr addrspace(8) [[BUF_PTR_4_PTR_RSRC]] to i160, !dbg [[DBG33:![0-9]+]]
-; CHECK-NEXT: [[TMP11:%.*]] = shl nuw i160 [[BUF_PTR_4_PTR_INT_RSRC]], 32, !dbg [[DBG33]]
+; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i160 [[BUF_PTR_4_PTR_INT_RSRC]], 32, !dbg [[DBG33]]
; CHECK-NEXT: [[BUF_PTR_4_PTR_INT_OFF:%.*]] = zext i32 [[BUF_PTR_4_PTR_OFF]] to i160, !dbg [[DBG33]]
-; CHECK-NEXT: [[BUF_PTR_4_PTR_INT:%.*]] = or i160 [[TMP11]], [[BUF_PTR_4_PTR_INT_OFF]], !dbg [[DBG33]]
+; CHECK-NEXT: [[BUF_PTR_4_PTR_INT:%.*]] = or i160 [[TMP10]], [[BUF_PTR_4_PTR_INT_OFF]], !dbg [[DBG33]]
; CHECK-NEXT: call void @llvm.amdgcn.raw.ptr.buffer.store.i160(i160 [[BUF_PTR_4_PTR_INT]], ptr addrspace(8) align 32 [[AUX_PTR_2_PTR_RSRC]], i32 [[AUX_PTR_2_PTR_OFF]], i32 0, i32 0), !dbg [[DBG33]]
; CHECK-NEXT: ret float [[RET]], !dbg [[DBG34:![0-9]+]]
;
|
krzysz00
approved these changes
Jun 11, 2024
| @@ -119,7 +124,8 @@ define ptr addrspace(7) @zero_gep_goes_first(ptr addrspace(7) %v0, i32 %arg) { | |||
| ; CHECK-SAME: ({ ptr addrspace(8), i32 } [[V0:%.*]], i32 [[ARG:%.*]]) #[[ATTR0]] { | |||
| ; CHECK-NEXT: [[V0_RSRC:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 0 | |||
| ; CHECK-NEXT: [[V0_OFF:%.*]] = extractvalue { ptr addrspace(8), i32 } [[V0]], 1 | |||
| ; CHECK-NEXT: [[V2:%.*]] = add i32 [[V0_OFF]], [[ARG]] | |||
| ; CHECK-NEXT: [[V1:%.*]] = add i32 [[V0_OFF]], 0 | |||
There was a problem hiding this comment.
Silly question: is there a good way to fold away these sort of trivial computations? Because I think I've seen them make their way down into ISel and beyond
(Maybe there's a way to do a localized instcombine?)
There was a problem hiding this comment.
InstSimplify. Can also use the IRBuilder that tries to simplify everything
There was a problem hiding this comment.
Yes, we could use InstSimplifyFolder. We could also avoid emitting these in emitGEPOffset() directly -- it's currently used primarily in InstCombine where this doesn't matter, but for other usages it's probably nice to not generate "add 0", especially as zero-indices are somewhat common in "old-style" GEPs.
There was a problem hiding this comment.
Turns out that this particular zero case was actually handled by the pass itself, and I broke it. Fixed by cb3a6bd.
|
I will note that addrspace(7) has the offset part being unsigned |
Does that mean that doing a |
|
@nikic Right, no, it's the final offset that's unsigned |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Use emitGEPOffset() to emit the GEP offset, which already has all the necessary logic.
This also fixes the nuw flag incorrectly being set on the offset calculation, while only nsw is correct.