[AMDGPU] Support nuw and nusw in buffer fat pointer lowering
#115039
Merged
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
This commit usis the `nuw` flag on `getelemnetptr` to set the `nuw` flag on buffer offset additions, and also moves from `inbounds` to the looser `nusw` for the existing case.
Member
|
@llvm/pr-subscribers-backend-amdgpu Author: Krzysztof Drewniak (krzysz00) ChangesThis commit usis the Full diff: https://github.com/llvm/llvm-project/pull/115039.diff 2 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp b/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
index 787747e6055805..c7cdd7a37282c7 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPULowerBufferFatPointers.cpp
@@ -1256,7 +1256,8 @@ PtrParts SplitPtrStructs::visitGetElementPtrInst(GetElementPtrInst &GEP) {
auto [Rsrc, Off] = getPtrParts(Ptr);
const DataLayout &DL = GEP.getDataLayout();
- bool InBounds = GEP.isInBounds();
+ bool IsNUW = GEP.hasNoUnsignedWrap();
+ bool IsNUSW = GEP.hasNoUnsignedSignedWrap();
// In order to call emitGEPOffset() and thus not have to reimplement it,
// we need the GEP result to have ptr addrspace(7) type.
@@ -1280,7 +1281,7 @@ PtrParts SplitPtrStructs::visitGetElementPtrInst(GetElementPtrInst &GEP) {
NewOff = OffAccum;
} else {
NewOff = IRB.CreateAdd(Off, OffAccum, "",
- /*hasNUW=*/InBounds && HasNonNegativeOff,
+ /*hasNUW=*/IsNUW || (IsNUSW && HasNonNegativeOff),
/*hasNSW=*/false);
}
copyMetadata(NewOff, &GEP);
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 cc98b5333c5bb8..9cff68303a5390 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
@@ -89,6 +89,53 @@ define ptr addrspace(7) @simple_inbounds_gep(ptr addrspace(7) %ptr, i32 %off) {
ret ptr addrspace(7) %ret
}
+define ptr addrspace(7) @simple_nuw_gep(ptr addrspace(7) %ptr, i32 %off) {
+; CHECK-LABEL: define { ptr addrspace(8), i32 } @simple_nuw_gep
+; 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: [[RET_IDX:%.*]] = mul nuw i32 [[OFF]], 4
+; CHECK-NEXT: [[RET:%.*]] = add nuw 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 nuw i32, ptr addrspace(7) %ptr, i32 %off
+ ret ptr addrspace(7) %ret
+}
+
+define ptr addrspace(7) @simple_nusw_gep(ptr addrspace(7) %ptr, i32 %off) {
+; CHECK-LABEL: define { ptr addrspace(8), i32 } @simple_nusw_gep
+; 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: [[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 nusw i32, ptr addrspace(7) %ptr, i32 %off
+ ret ptr addrspace(7) %ret
+}
+
+define ptr addrspace(7) @nusw_gep_pair(ptr addrspace(7) %ptr, i32 %off) {
+; CHECK-LABEL: define { ptr addrspace(8), i32 } @nusw_gep_pair
+; 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: [[P1_IDX:%.*]] = mul nsw i32 [[OFF]], 4
+; CHECK-NEXT: [[P1:%.*]] = add i32 [[PTR_OFF]], [[P1_IDX]]
+; CHECK-NEXT: [[RET:%.*]] = add nuw i32 [[P1]], 16
+; 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]]
+;
+ %p1 = getelementptr nusw i32, ptr addrspace(7) %ptr, i32 %off
+ %ret = getelementptr nusw i32, ptr addrspace(7) %p1, i32 4
+ ret ptr addrspace(7) %ret
+}
+
define ptr addrspace(7) @zero_gep(ptr addrspace(7) %ptr) {
; CHECK-LABEL: define { ptr addrspace(8), i32 } @zero_gep
; CHECK-SAME: ({ ptr addrspace(8), i32 } [[PTR:%.*]]) #[[ATTR0]] {
|
arsenm
approved these changes
Nov 5, 2024
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/134/builds/8161 Here is the relevant piece of the build log for the reference |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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.
This commit usis the
nuwflag ongetelemnetptrto set thenuwflag on buffer offset additions, and also moves frominboundsto the loosernuswfor the existing case.