[LoongArch] Don't crash on instruction prefetch intrinsics #135760
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
Instead of failing to select during isel, drop the intrinsic in lowering. Similar as the X86's PR. Seeing: https://reviews.llvm.org/D151050. Fixes llvm#134624
Member
|
@llvm/pr-subscribers-backend-loongarch Author: None (leecheechen) ChangesInstead of failing to select during isel, drop the intrinsic in lowering. Similar as the X86's PR. Seeing: https://reviews.llvm.org/D151050. Fixes #134624 Full diff: https://github.com/llvm/llvm-project/pull/135760.diff 3 Files Affected:
diff --git a/llvm/lib/Target/LoongArch/LoongArchISelLowering.cpp b/llvm/lib/Target/LoongArch/LoongArchISelLowering.cpp
index 002d88cbeeba3..44194b5829182 100644
--- a/llvm/lib/Target/LoongArch/LoongArchISelLowering.cpp
+++ b/llvm/lib/Target/LoongArch/LoongArchISelLowering.cpp
@@ -99,7 +99,7 @@ LoongArchTargetLowering::LoongArchTargetLowering(const TargetMachine &TM,
setOperationAction(ISD::INTRINSIC_W_CHAIN, MVT::Other, Custom);
setOperationAction(ISD::INTRINSIC_WO_CHAIN, MVT::Other, Custom);
- setOperationAction(ISD::PREFETCH, MVT::Other, Legal);
+ setOperationAction(ISD::PREFETCH, MVT::Other, Custom);
// Expand bitreverse.i16 with native-width bitrev and shift for now, before
// we get to know which of sll and revb.2h is faster.
@@ -465,10 +465,24 @@ SDValue LoongArchTargetLowering::LowerOperation(SDValue Op,
return lowerBITREVERSE(Op, DAG);
case ISD::SCALAR_TO_VECTOR:
return lowerSCALAR_TO_VECTOR(Op, DAG);
+ case ISD::PREFETCH:
+ return lowerPREFETCH(Op, DAG);
}
return SDValue();
}
+SDValue LoongArchTargetLowering::lowerPREFETCH(SDValue Op,
+ SelectionDAG &DAG) const {
+ unsigned IsData = Op.getConstantOperandVal(4);
+
+ // We don't support non-data prefetch.
+ // Just preserve the chain.
+ if (!IsData)
+ return Op.getOperand(0);
+
+ return Op;
+}
+
SDValue
LoongArchTargetLowering::lowerSCALAR_TO_VECTOR(SDValue Op,
SelectionDAG &DAG) const {
diff --git a/llvm/lib/Target/LoongArch/LoongArchISelLowering.h b/llvm/lib/Target/LoongArch/LoongArchISelLowering.h
index 52d88b9b24a6b..8c2d2597a26ec 100644
--- a/llvm/lib/Target/LoongArch/LoongArchISelLowering.h
+++ b/llvm/lib/Target/LoongArch/LoongArchISelLowering.h
@@ -353,6 +353,7 @@ class LoongArchTargetLowering : public TargetLowering {
SDValue lowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) const;
SDValue lowerBITREVERSE(SDValue Op, SelectionDAG &DAG) const;
SDValue lowerSCALAR_TO_VECTOR(SDValue Op, SelectionDAG &DAG) const;
+ SDValue lowerPREFETCH(SDValue Op, SelectionDAG &DAG) const;
bool isFPImmLegal(const APFloat &Imm, EVT VT,
bool ForCodeSize) const override;
diff --git a/llvm/test/CodeGen/LoongArch/prefetchi.ll b/llvm/test/CodeGen/LoongArch/prefetchi.ll
new file mode 100644
index 0000000000000..a00f6f8161862
--- /dev/null
+++ b/llvm/test/CodeGen/LoongArch/prefetchi.ll
@@ -0,0 +1,33 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc --mtriple=loongarch32 < %s | FileCheck %s --check-prefix=LA32
+; RUN: llc --mtriple=loongarch64 < %s | FileCheck %s --check-prefix=LA64
+
+declare void @llvm.prefetch(ptr, i32, i32, i32) nounwind
+
+define dso_local void @prefetch_no_offset(ptr %ptr) nounwind {
+; LA32-LABEL: prefetch_no_offset:
+; LA32: # %bb.0: # %entry
+; LA32-NEXT: ret
+;
+; LA64-LABEL: prefetch_no_offset:
+; LA64: # %bb.0: # %entry
+; LA64-NEXT: ret
+entry:
+ tail call void @llvm.prefetch(ptr %ptr, i32 0, i32 3, i32 0)
+ ret void
+}
+
+
+define dso_local void @prefetch_with_offset(ptr %ptr) nounwind {
+; LA32-LABEL: prefetch_with_offset:
+; LA32: # %bb.0: # %entry
+; LA32-NEXT: ret
+;
+; LA64-LABEL: prefetch_with_offset:
+; LA64: # %bb.0: # %entry
+; LA64-NEXT: ret
+entry:
+ %addr = getelementptr i8, ptr %ptr, i64 200
+ tail call void @llvm.prefetch(ptr %addr, i32 0, i32 3, i32 0)
+ ret void
+}
|
heiher
approved these changes
Apr 16, 2025
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/59/builds/16107 Here is the relevant piece of the build log for the reference |
swift-ci
pushed a commit
to swiftlang/llvm-project
that referenced
this pull request
Apr 25, 2025
Instead of failing to select during isel, drop the intrinsic in lowering. Similar as the X86's PR. Seeing: https://reviews.llvm.org/D151050. Fixes llvm#134624 (cherry picked from commit dfb5b6e)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
Instead of failing to select during isel, drop the intrinsic in lowering.
Similar as the X86's PR. Seeing: https://reviews.llvm.org/D151050.
Fixes #134624