[clang] [HLSL] Implement the `lit` intrinsic (PR #134171)
https://github.com/farzonl edited https://github.com/llvm/llvm-project/pull/134171 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Improve integer comparison by matching valid expressions outside implicitCastExpr (PR #134188)
llvmbot wrote: @llvm/pr-subscribers-clang-tools-extra Author: David Rivera (RiverDave) Changes Covered the edge case where an int expression is not necessarily directly wrapped around an ImplicitCastExpr which seemed to be a requirement in this check to trigger. **For instance**: ```cpp #includebool f() { std::vector v; unsigned int i = 0; return i >= v.size(); } ``` **See AST:**  --- Full diff: https://github.com/llvm/llvm-project/pull/134188.diff 3 Files Affected: - (modified) clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp (+13-3) - (modified) clang-tools-extra/docs/ReleaseNotes.rst (+4) - (modified) clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp (+26) ``diff diff --git a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp index eeba5cce80da5..089df7ece3f82 100644 --- a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp +++ b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp @@ -39,9 +39,11 @@ intCastExpression(bool IsSigned, // std::cmp_{} functions trigger a compile-time error if either LHS or RHS // is a non-integer type, char, enum or bool // (unsigned char/ signed char are Ok and can be used). - auto IntTypeExpr = expr(hasType(hasCanonicalType(qualType( + const auto HasIntegerType = hasType(hasCanonicalType(qualType( isInteger(), IsSigned ? isSignedInteger() : isUnsignedInteger(), - unless(isActualChar()), unless(booleanType()), unless(enumType()); + unless(isActualChar()), unless(booleanType()), unless(enumType(); + + auto IntTypeExpr = expr(HasIntegerType); const auto ImplicitCastExpr = CastBindName.empty() ? implicitCastExpr(hasSourceExpression(IntTypeExpr)) @@ -52,8 +54,16 @@ intCastExpression(bool IsSigned, const auto StaticCastExpr = cxxStaticCastExpr(has(ImplicitCastExpr)); const auto FunctionalCastExpr = cxxFunctionalCastExpr(has(ImplicitCastExpr)); + // Match function calls or variable references not wrapped by an implicit cast + const auto CallIntExpr = CastBindName.empty() + ? callExpr(HasIntegerType) + : callExpr(HasIntegerType).bind(CastBindName); + const auto DeclRefIntExpr = + CastBindName.empty() ? declRefExpr(HasIntegerType) + : declRefExpr(HasIntegerType).bind(CastBindName); + return expr(anyOf(ImplicitCastExpr, CStyleCastExpr, StaticCastExpr, -FunctionalCastExpr)); +FunctionalCastExpr, CallIntExpr)); } static StringRef parseOpCode(BinaryOperator::Opcode Code) { diff --git a/clang-tools-extra/docs/ReleaseNotes.rst b/clang-tools-extra/docs/ReleaseNotes.rst index 6cb8d572d3a78..b5f2d8e8fcbd7 100644 --- a/clang-tools-extra/docs/ReleaseNotes.rst +++ b/clang-tools-extra/docs/ReleaseNotes.rst @@ -177,6 +177,10 @@ Changes in existing checks matched scenarios of ``find`` and ``rfind`` methods and fixing false positives when those methods were called with 3 arguments. +- Improved :doc:`modernize-use-integer-sign-comparison + ` check by matching + valid integer expressions not wrapped around an Implicit Cast. + - Improved :doc:`modernize-use-std-numbers ` check to support math functions of different precisions. diff --git a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp index 99f00444c2d3f..1d2f64a359a2c 100644 --- a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp +++ b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp @@ -120,3 +120,29 @@ int AllComparisons() { return 0; } + +namespace PR127471 { +int getSignedValue(); +unsigned int getUnsignedValue(); + +void callExprTest() { + +if (getSignedValue() < getUnsignedValue()) +return; +// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and 'unsigned' integers [modernize-use-integer-sign-comparison] +// CHECK-FIXES: if (std::cmp_less(getSignedValue() , getUnsignedValue())) + +int sVar = 0; +if (getUnsignedValue() > sVar) +return; +// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and 'unsigned' integers [modernize-use-integer-sign-comparison] +// CHECK-FIXES: if (std::cmp_greater(getUnsignedValue() , sVar)) + +unsigned int uVar = 0; +if (getSignedValue() > uVar) +return; +// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and 'unsigned' integers [modernize-use-integer-sign-comparison] +// CHECK-FIXES: if (std::cmp_greater(ge
[clang-tools-extra] [clang-tidy] Improve integer comparison by matching valid expressions outside implicitCastExpr (PR #134188)
https://github.com/RiverDave created
https://github.com/llvm/llvm-project/pull/134188
Covered the edge case where an int expression is not necessarily directly
wrapped around an ImplicitCastExpr which seemed to be a requirement in this
check to trigger.
**For instance**:
```cpp
#include
bool f() {
std::vector v;
unsigned int i = 0;
return i >= v.size();
}
```
**See AST:**
>From ebce879a943f5817aca4b0d3d637a17a626c9683 Mon Sep 17 00:00:00 2001
From: David Rivera
Date: Wed, 2 Apr 2025 21:02:00 -0400
Subject: [PATCH] [clang-tidy] Improve integer comparison by matching valid
expressions outside implicitCastExpr
---
.../UseIntegerSignComparisonCheck.cpp | 16 +---
clang-tools-extra/docs/ReleaseNotes.rst | 4 +++
.../modernize/use-integer-sign-comparison.cpp | 26 +++
3 files changed, 43 insertions(+), 3 deletions(-)
diff --git
a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
index eeba5cce80da5..089df7ece3f82 100644
--- a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
+++ b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
@@ -39,9 +39,11 @@ intCastExpression(bool IsSigned,
// std::cmp_{} functions trigger a compile-time error if either LHS or RHS
// is a non-integer type, char, enum or bool
// (unsigned char/ signed char are Ok and can be used).
- auto IntTypeExpr = expr(hasType(hasCanonicalType(qualType(
+ const auto HasIntegerType = hasType(hasCanonicalType(qualType(
isInteger(), IsSigned ? isSignedInteger() : isUnsignedInteger(),
- unless(isActualChar()), unless(booleanType()), unless(enumType());
+ unless(isActualChar()), unless(booleanType()), unless(enumType();
+
+ auto IntTypeExpr = expr(HasIntegerType);
const auto ImplicitCastExpr =
CastBindName.empty() ? implicitCastExpr(hasSourceExpression(IntTypeExpr))
@@ -52,8 +54,16 @@ intCastExpression(bool IsSigned,
const auto StaticCastExpr = cxxStaticCastExpr(has(ImplicitCastExpr));
const auto FunctionalCastExpr = cxxFunctionalCastExpr(has(ImplicitCastExpr));
+ // Match function calls or variable references not wrapped by an implicit
cast
+ const auto CallIntExpr = CastBindName.empty()
+ ? callExpr(HasIntegerType)
+ : callExpr(HasIntegerType).bind(CastBindName);
+ const auto DeclRefIntExpr =
+ CastBindName.empty() ? declRefExpr(HasIntegerType)
+ : declRefExpr(HasIntegerType).bind(CastBindName);
+
return expr(anyOf(ImplicitCastExpr, CStyleCastExpr, StaticCastExpr,
-FunctionalCastExpr));
+FunctionalCastExpr, CallIntExpr));
}
static StringRef parseOpCode(BinaryOperator::Opcode Code) {
diff --git a/clang-tools-extra/docs/ReleaseNotes.rst
b/clang-tools-extra/docs/ReleaseNotes.rst
index 6cb8d572d3a78..b5f2d8e8fcbd7 100644
--- a/clang-tools-extra/docs/ReleaseNotes.rst
+++ b/clang-tools-extra/docs/ReleaseNotes.rst
@@ -177,6 +177,10 @@ Changes in existing checks
matched scenarios of ``find`` and ``rfind`` methods and fixing false
positives when those methods were called with 3 arguments.
+- Improved :doc:`modernize-use-integer-sign-comparison
+ ` check by matching
+ valid integer expressions not wrapped around an Implicit Cast.
+
- Improved :doc:`modernize-use-std-numbers
` check to support math
functions of different precisions.
diff --git
a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
index 99f00444c2d3f..1d2f64a359a2c 100644
---
a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
+++
b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
@@ -120,3 +120,29 @@ int AllComparisons() {
return 0;
}
+
+namespace PR127471 {
+int getSignedValue();
+unsigned int getUnsignedValue();
+
+void callExprTest() {
+
+if (getSignedValue() < getUnsignedValue())
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsigned' integers [modernize-use-integer-sign-comparison]
+// CHECK-FIXES: if (std::cmp_less(getSignedValue() , getUnsignedValue()))
+
+int sVar = 0;
+if (getUnsignedValue() > sVar)
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsigned' integers [modernize-use-integer-sign-comparison]
+// CHECK-FIXES: if (std::cmp_greater(getUnsignedValue() , sVar))
+
+unsigned int uVar = 0;
+if (getSignedValue() > uVar)
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsig
[clang] [CIR] Upstream support for break and continue statements (PR #134181)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Andy Kaylor (andykaylor)
Changes
This adds ClangIR support for break and continue statements in loops.
Because only loops are currently implemented upstream in CIR, only breaks in
loops are supported here, but this same code will work (with minor changes to
the verification and cfg flattening) when switch statements are upstreamed.
---
Full diff: https://github.com/llvm/llvm-project/pull/134181.diff
8 Files Affected:
- (modified) clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h (+10)
- (modified) clang/include/clang/CIR/Dialect/IR/CIROps.td (+29)
- (modified) clang/include/clang/CIR/MissingFeatures.h (-2)
- (modified) clang/lib/CIR/CodeGen/CIRGenFunction.h (+2)
- (modified) clang/lib/CIR/CodeGen/CIRGenStmt.cpp (+23)
- (modified) clang/lib/CIR/Dialect/IR/CIRDialect.cpp (+21)
- (modified) clang/lib/CIR/Dialect/Transforms/FlattenCFG.cpp (+11-10)
- (modified) clang/test/CIR/CodeGen/loop.cpp (+122)
``diff
diff --git a/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h
b/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h
index e666be0b25d75..900d78c401ebf 100644
--- a/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h
+++ b/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h
@@ -136,6 +136,16 @@ class CIRBaseBuilderTy : public mlir::OpBuilder {
return create(loc, condBuilder, bodyBuilder, stepBuilder);
}
+ /// Create a break operation.
+ cir::BreakOp createBreak(mlir::Location loc) {
+return create(loc);
+ }
+
+ /// Create a continue operation.
+ cir::ContinueOp createContinue(mlir::Location loc) {
+return create(loc);
+ }
+
mlir::TypedAttr getConstPtrAttr(mlir::Type type, int64_t value) {
auto valueAttr = mlir::IntegerAttr::get(
mlir::IntegerType::get(type.getContext(), 64), value);
diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td
b/clang/include/clang/CIR/Dialect/IR/CIROps.td
index 3965372755685..7ddf5e843bde2 100644
--- a/clang/include/clang/CIR/Dialect/IR/CIROps.td
+++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td
@@ -569,6 +569,35 @@ def YieldOp : CIR_Op<"yield", [ReturnLike, Terminator,
];
}
+//===--===//
+// BreakOp
+//===--===//
+
+def BreakOp : CIR_Op<"break", [Terminator]> {
+ let summary = "C/C++ `break` statement equivalent";
+ let description = [{
+The `cir.break` operation is used to cease the control flow to the parent
+operation, exiting its region's control flow. It is only allowed if it is
+within a breakable operation (loops and `switch`).
+ }];
+ let assemblyFormat = "attr-dict";
+ let hasVerifier = 1;
+}
+
+//===--===//
+// ContinueOp
+//===--===//
+
+def ContinueOp : CIR_Op<"continue", [Terminator]> {
+ let summary = "C/C++ `continue` statement equivalent";
+ let description = [{
+The `cir.continue` operation is used to continue execution to the next
+iteration of a loop. It is only allowed within `cir.loop` regions.
+ }];
+ let assemblyFormat = "attr-dict";
+ let hasVerifier = 1;
+}
+
//===--===//
// ScopeOp
//===--===//
diff --git a/clang/include/clang/CIR/MissingFeatures.h
b/clang/include/clang/CIR/MissingFeatures.h
index 3a102d90aba8f..be16b5c063f3b 100644
--- a/clang/include/clang/CIR/MissingFeatures.h
+++ b/clang/include/clang/CIR/MissingFeatures.h
@@ -122,12 +122,10 @@ struct MissingFeatures {
// Future CIR operations
static bool awaitOp() { return false; }
- static bool breakOp() { return false; }
static bool callOp() { return false; }
static bool complexCreateOp() { return false; }
static bool complexImagOp() { return false; }
static bool complexRealOp() { return false; }
- static bool continueOp() { return false; }
static bool ifOp() { return false; }
static bool labelOp() { return false; }
static bool ptrDiffOp() { return false; }
diff --git a/clang/lib/CIR/CodeGen/CIRGenFunction.h
b/clang/lib/CIR/CodeGen/CIRGenFunction.h
index 5cae4d5da9516..c244e3752d7fe 100644
--- a/clang/lib/CIR/CodeGen/CIRGenFunction.h
+++ b/clang/lib/CIR/CodeGen/CIRGenFunction.h
@@ -395,6 +395,8 @@ class CIRGenFunction : public CIRGenTypeCache {
LValue emitBinaryOperatorLValue(const BinaryOperator *e);
+ mlir::LogicalResult emitBreakStmt(const clang::BreakStmt &s);
+ mlir::LogicalResult emitContinueStmt(const clang::ContinueStmt &s);
mlir::LogicalResult emitDoStmt(const clang::DoStmt &s);
/// Emit an expression as an initializer for an object (variable, field,
etc.)
diff --git a/clang/lib/CIR/CodeGen/CIRGenStmt.cpp
b/clang/lib/CIR/CodeGen/CIRGenStmt.c
[clang] [Clang] fixed clang frontend crash with friend class declaration and overload == (PR #133878)
https://github.com/zyn0217 edited https://github.com/llvm/llvm-project/pull/133878 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Sema] Use llvm::erase_if (NFC) (PR #134017)
https://github.com/kazutakahirata closed https://github.com/llvm/llvm-project/pull/134017 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Reland [Clang][Cmake] fix libtool duplicate member name warnings (PR #133850)
https://github.com/farzonl updated
https://github.com/llvm/llvm-project/pull/133850
>From cd2f6658a1a609edf7471f629b766eb4ee495122 Mon Sep 17 00:00:00 2001
From: Farzon Lotfi
Date: Sun, 30 Mar 2025 00:59:48 -0400
Subject: [PATCH 1/3] [Clang][Cmake] fix libtool duplicate member name warnings
fixes #133199
PR #132252 Created a second file that shared .cpp in
`clang/lib/CodeGen/CMakeLists.txt`
For example There were two AMDGPU.cpp's one in TargetBuiltins and the
other in Targets. Even though these were in different directories
libtool warns that it might not distinguish them because they share the same
base name.
There are two fixes. The easy fix is to rename one of them and keep one
cmake file. That solution though doesn't future proof this problem in
the event of a third .cpp and it seems teams want to
just use the target name
https://github.com/llvm/llvm-project/pull/132252#issuecomment-2758178483.
The alternative fix is to seperate the cmake files into their own sub
directories. I chose to create static libraries. It might of been
possible to build an OBJECT, but I only saw examples of $
in compiler-rt and test directories so assumed there was a reason it wasn't
used.
---
clang/lib/CodeGen/CMakeLists.txt | 49 +--
clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp | 2 +-
.../lib/CodeGen/TargetBuiltins/CMakeLists.txt | 14 ++
clang/lib/CodeGen/Targets/CMakeLists.txt | 30
4 files changed, 57 insertions(+), 38 deletions(-)
create mode 100644 clang/lib/CodeGen/TargetBuiltins/CMakeLists.txt
create mode 100644 clang/lib/CodeGen/Targets/CMakeLists.txt
diff --git a/clang/lib/CodeGen/CMakeLists.txt b/clang/lib/CodeGen/CMakeLists.txt
index ebe2fbd7db295..cdf9f909a3675 100644
--- a/clang/lib/CodeGen/CMakeLists.txt
+++ b/clang/lib/CodeGen/CMakeLists.txt
@@ -116,45 +116,8 @@ add_clang_library(clangCodeGen
PatternInit.cpp
SanitizerMetadata.cpp
SwiftCallingConv.cpp
- TargetBuiltins/ARM.cpp
- TargetBuiltins/AMDGPU.cpp
- TargetBuiltins/Hexagon.cpp
- TargetBuiltins/NVPTX.cpp
- TargetBuiltins/PPC.cpp
- TargetBuiltins/RISCV.cpp
- TargetBuiltins/SPIR.cpp
- TargetBuiltins/SystemZ.cpp
- TargetBuiltins/WebAssembly.cpp
- TargetBuiltins/X86.cpp
TargetInfo.cpp
- Targets/AArch64.cpp
- Targets/AMDGPU.cpp
- Targets/ARC.cpp
- Targets/ARM.cpp
- Targets/AVR.cpp
- Targets/BPF.cpp
- Targets/CSKY.cpp
- Targets/DirectX.cpp
- Targets/Hexagon.cpp
- Targets/Lanai.cpp
- Targets/LoongArch.cpp
- Targets/M68k.cpp
- Targets/MSP430.cpp
- Targets/Mips.cpp
- Targets/NVPTX.cpp
- Targets/PNaCl.cpp
- Targets/PPC.cpp
- Targets/RISCV.cpp
- Targets/SPIR.cpp
- Targets/Sparc.cpp
- Targets/SystemZ.cpp
- Targets/TCE.cpp
- Targets/VE.cpp
- Targets/WebAssembly.cpp
- Targets/X86.cpp
- Targets/XCore.cpp
VarBypassDetector.cpp
-
DEPENDS
vt_gen
intrinsics_gen
@@ -170,4 +133,16 @@ add_clang_library(clangCodeGen
clangFrontend
clangLex
clangSerialization
+ clangCodeGenTargetBuiltins
+ clangCodeGenTargets
+ )
+
+ target_include_directories(clangCodeGen
+PUBLIC
+${CMAKE_CURRENT_SOURCE_DIR}
+${CMAKE_CURRENT_SOURCE_DIR}/TargetBuiltins
+${CMAKE_CURRENT_SOURCE_DIR}/Targets
)
+
+ add_subdirectory(TargetBuiltins)
+ add_subdirectory(Targets)
diff --git a/clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp
b/clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp
index b56b739094ff3..577fee05d4af6 100644
--- a/clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp
+++ b/clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp
@@ -1,4 +1,4 @@
-//===--- AMDCPU.cpp - Emit LLVM Code for builtins
-===//
+//===--- AMDGPU.cpp - Emit LLVM Code for builtins
-===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
diff --git a/clang/lib/CodeGen/TargetBuiltins/CMakeLists.txt
b/clang/lib/CodeGen/TargetBuiltins/CMakeLists.txt
new file mode 100644
index 0..76be68a11d02a
--- /dev/null
+++ b/clang/lib/CodeGen/TargetBuiltins/CMakeLists.txt
@@ -0,0 +1,14 @@
+include_directories(${CMAKE_CURRENT_SOURCE_DIR}/..)
+
+add_clang_library(clangCodeGenTargetBuiltins STATIC
+ ARM.cpp
+ AMDGPU.cpp
+ Hexagon.cpp
+ NVPTX.cpp
+ PPC.cpp
+ RISCV.cpp
+ SPIR.cpp
+ SystemZ.cpp
+ WebAssembly.cpp
+ X86.cpp
+)
diff --git a/clang/lib/CodeGen/Targets/CMakeLists.txt
b/clang/lib/CodeGen/Targets/CMakeLists.txt
new file mode 100644
index 0..6b6e7ce5b0b29
--- /dev/null
+++ b/clang/lib/CodeGen/Targets/CMakeLists.txt
@@ -0,0 +1,30 @@
+include_directories(${CMAKE_CURRENT_SOURCE_DIR}/..)
+
+add_clang_library(clangCodeGenTargets STATIC
+ AArch64.cpp
+ AMDGPU.cpp
+ ARC.cpp
+ ARM.cpp
+ AVR.cpp
+ BPF.cpp
+ CSKY.cpp
+ DirectX.cpp
+ Hexagon.cpp
+ Lanai.cpp
+ LoongArch.cpp
+ M68k.cpp
+ MSP430.cpp
+ Mips.cpp
+ NVPTX.cpp
+ PNaCl.cpp
+ PPC.cpp
+ RISCV.cpp
+ SPIR.cpp
+ Sparc.cpp
+ SystemZ.cpp
+ TCE.cpp
+ VE.cpp
+ WebAssem
[clang] [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (PR #133863)
https://github.com/zyn0217 closed https://github.com/llvm/llvm-project/pull/133863 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] b384d6d - [CodeGen] Don't include CGDebugInfo.h in CodeGenFunction.h (NFC) (#134100)
Author: Nikita Popov
Date: 2025-04-03T08:04:19+02:00
New Revision: b384d6d6ccc8f4452cd7086061c657ce76b41224
URL:
https://github.com/llvm/llvm-project/commit/b384d6d6ccc8f4452cd7086061c657ce76b41224
DIFF:
https://github.com/llvm/llvm-project/commit/b384d6d6ccc8f4452cd7086061c657ce76b41224.diff
LOG: [CodeGen] Don't include CGDebugInfo.h in CodeGenFunction.h (NFC) (#134100)
This is an expensive header, only include it where needed. Move some
functions out of line to achieve that.
This reduces time to build clang by ~0.5% in terms of instructions
retired.
Added:
Modified:
clang/lib/CodeGen/CGBuiltin.cpp
clang/lib/CodeGen/CGCall.cpp
clang/lib/CodeGen/CGCoroutine.cpp
clang/lib/CodeGen/CGDebugInfo.cpp
clang/lib/CodeGen/CGDeclCXX.cpp
clang/lib/CodeGen/CGException.cpp
clang/lib/CodeGen/CGExprAgg.cpp
clang/lib/CodeGen/CGExprComplex.cpp
clang/lib/CodeGen/CGNonTrivialStruct.cpp
clang/lib/CodeGen/CGOpenMPRuntime.cpp
clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp
clang/lib/CodeGen/CGStmtOpenMP.cpp
clang/lib/CodeGen/CGVTables.cpp
clang/lib/CodeGen/CodeGenFunction.h
clang/lib/CodeGen/CodeGenPGO.cpp
clang/lib/CodeGen/CodeGenTypes.cpp
clang/lib/CodeGen/ItaniumCXXABI.cpp
clang/lib/CodeGen/MicrosoftCXXABI.cpp
clang/lib/CodeGen/TargetBuiltins/ARM.cpp
Removed:
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 91ac7c5847b02..310addebd50e9 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -14,6 +14,7 @@
#include "ABIInfo.h"
#include "CGCUDARuntime.h"
#include "CGCXXABI.h"
+#include "CGDebugInfo.h"
#include "CGObjCRuntime.h"
#include "CGOpenCLRuntime.h"
#include "CGRecordLayout.h"
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index 3cefa3b0c585c..b202255c3a15b 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -17,6 +17,7 @@
#include "CGBlocks.h"
#include "CGCXXABI.h"
#include "CGCleanup.h"
+#include "CGDebugInfo.h"
#include "CGRecordLayout.h"
#include "CodeGenFunction.h"
#include "CodeGenModule.h"
diff --git a/clang/lib/CodeGen/CGCoroutine.cpp
b/clang/lib/CodeGen/CGCoroutine.cpp
index a9795c2c0dc8f..0fc488e98aaf0 100644
--- a/clang/lib/CodeGen/CGCoroutine.cpp
+++ b/clang/lib/CodeGen/CGCoroutine.cpp
@@ -11,10 +11,11 @@
//===--===//
#include "CGCleanup.h"
+#include "CGDebugInfo.h"
#include "CodeGenFunction.h"
-#include "llvm/ADT/ScopeExit.h"
#include "clang/AST/StmtCXX.h"
#include "clang/AST/StmtVisitor.h"
+#include "llvm/ADT/ScopeExit.h"
using namespace clang;
using namespace CodeGen;
diff --git a/clang/lib/CodeGen/CGDebugInfo.cpp
b/clang/lib/CodeGen/CGDebugInfo.cpp
index 52aa956121d73..d659243d38d5f 100644
--- a/clang/lib/CodeGen/CGDebugInfo.cpp
+++ b/clang/lib/CodeGen/CGDebugInfo.cpp
@@ -6223,3 +6223,23 @@ CGDebugInfo::createConstantValueExpression(const
clang::ValueDecl *VD,
return nullptr;
}
+
+CodeGenFunction::LexicalScope::LexicalScope(CodeGenFunction &CGF,
+SourceRange Range)
+: RunCleanupsScope(CGF), Range(Range), ParentScope(CGF.CurLexicalScope) {
+ CGF.CurLexicalScope = this;
+ if (CGDebugInfo *DI = CGF.getDebugInfo())
+DI->EmitLexicalBlockStart(CGF.Builder, Range.getBegin());
+}
+
+CodeGenFunction::LexicalScope::~LexicalScope() {
+ if (CGDebugInfo *DI = CGF.getDebugInfo())
+DI->EmitLexicalBlockEnd(CGF.Builder, Range.getEnd());
+
+ // If we should perform a cleanup, force them now. Note that
+ // this ends the cleanup scope before rescoping any labels.
+ if (PerformCleanup) {
+ApplyDebugLocation DL(CGF, Range.getEnd());
+ForceCleanup();
+ }
+}
diff --git a/clang/lib/CodeGen/CGDeclCXX.cpp b/clang/lib/CodeGen/CGDeclCXX.cpp
index 33c048b48795c..e0921993bd14e 100644
--- a/clang/lib/CodeGen/CGDeclCXX.cpp
+++ b/clang/lib/CodeGen/CGDeclCXX.cpp
@@ -11,6 +11,7 @@
//===--===//
#include "CGCXXABI.h"
+#include "CGDebugInfo.h"
#include "CGHLSLRuntime.h"
#include "CGObjCRuntime.h"
#include "CGOpenMPRuntime.h"
diff --git a/clang/lib/CodeGen/CGException.cpp
b/clang/lib/CodeGen/CGException.cpp
index b4b8c2952b02b..ebecb3aa5241d 100644
--- a/clang/lib/CodeGen/CGException.cpp
+++ b/clang/lib/CodeGen/CGException.cpp
@@ -12,6 +12,7 @@
#include "CGCXXABI.h"
#include "CGCleanup.h"
+#include "CGDebugInfo.h"
#include "CGObjCRuntime.h"
#include "CodeGenFunction.h"
#include "ConstantEmitter.h"
diff --git a/clang/lib/CodeGen/CGExprAgg.cpp b/clang/lib/CodeGen/CGExprAgg.cpp
index c8bdda375d1b1..87b2a73fb0c03 100644
--- a/clang/lib/CodeGen/CGExprAgg.cpp
+++ b/clang/lib/CodeGen/CGExprAgg.cpp
@@ -11,6 +11,7 @@
//===--===//
[clang] [llvm] [HLSL] Implement dot2add intrinsic (PR #131237)
@@ -45,6 +45,14 @@ distance_vec_impl(vector X, vector Y) {
return length_vec_impl(X - Y);
}
+constexpr float dot2add_impl(half2 a, half2 b, float c) {
+#if defined(__DIRECTX__)
farzonl wrote:
Also to answer your question we only need to do one or the other for spirv
because the HLSL header is intended to be used with clang and clang will set
both defines when some targets spriv.
https://github.com/llvm/llvm-project/pull/131237
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Hlsl dst function (PR #133828)
@@ -0,0 +1,51 @@
+// RUN: %clang_cc1 -finclude-default-header -triple
dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only
-disable-llvm-passes -verify
+
+float4 test_too_many_arg(float4 p0)
+{
+dst(p0, p0, p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+}
+
+float4 test_no_second_arg(float4 p0)
+{
+return dst(p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+}
+
+float4 test_no_args()
+{
+return dst();
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+}
+
+float4 test_3_components(float3 p0, float3 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector<[...], 3>' to 'vector<[...], 4>' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+}
+
+float4 test_float2(double4 p0, float4 p1)
farzonl wrote:
the test name is not correct. this test does nothing with float2
https://github.com/llvm/llvm-project/pull/133828
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] c57b9c2 - [CIR] Generate the nsw flag correctly for unary ops (#133815)
Author: Andy Kaylor
Date: 2025-04-02T15:48:55-07:00
New Revision: c57b9c233a87f37e034445596ed09260cc6b23f5
URL:
https://github.com/llvm/llvm-project/commit/c57b9c233a87f37e034445596ed09260cc6b23f5
DIFF:
https://github.com/llvm/llvm-project/commit/c57b9c233a87f37e034445596ed09260cc6b23f5.diff
LOG: [CIR] Generate the nsw flag correctly for unary ops (#133815)
A previous checkin used a workaround to generate the nsw flag where
needed for unary ops. This change upstreams a subsequent change that was
made in the incubator to generate the flag correctly.
Added:
clang/test/CIR/IR/unary.cir
Modified:
clang/include/clang/CIR/Dialect/IR/CIROps.td
clang/include/clang/CIR/MissingFeatures.h
clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp
clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp
clang/test/CIR/CodeGen/unary.cpp
Removed:
diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td
b/clang/include/clang/CIR/Dialect/IR/CIROps.td
index 3965372755685..c17abfd752a1a 100644
--- a/clang/include/clang/CIR/Dialect/IR/CIROps.td
+++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td
@@ -697,17 +697,24 @@ def UnaryOp : CIR_Op<"unary", [Pure,
SameOperandsAndResultType]> {
It requires one input operand and has one result, both types
should be the same.
+If the `nsw` (no signed wrap) attribute is present, the result is poison if
+signed overflow occurs.
+
```mlir
%7 = cir.unary(inc, %1) : i32 -> i32
-%8 = cir.unary(dec, %2) : i32 -> i32
+%8 = cir.unary(dec, %2) nsw : i32 -> i32
```
}];
let results = (outs CIR_AnyType:$result);
- let arguments = (ins Arg:$kind,
Arg:$input);
+ let arguments = (ins Arg:$kind,
+ Arg:$input,
+ UnitAttr:$no_signed_wrap);
let assemblyFormat = [{
- `(` $kind `,` $input `)` `:` type($input) `,` type($result) attr-dict
+ `(` $kind `,` $input `)`
+ (`nsw` $no_signed_wrap^)?
+ `:` type($input) `,` type($result) attr-dict
}];
let hasVerifier = 1;
@@ -961,9 +968,21 @@ def BinOp : CIR_Op<"binop", [Pure,
It requires two input operands and has one result, all types
should be the same.
+If the `nsw` (no signed wrap) or `nuw` (no unsigned wrap) attributes are
+present, the result is poison if signed or unsigned overflow occurs
+(respectively).
+
+If the `sat` (saturated) attribute is present, the result is clamped to
+the maximum value representatable by the type if it would otherwise
+exceed that value and is clamped to the minimum representable value if
+it would otherwise be below that value.
+
```mlir
-%7 = cir.binop(add, %1, %2) : !s32i
-%7 = cir.binop(mul, %1, %2) : !u8i
+%5 = cir.binop(add, %1, %2) : !s32i
+%6 = cir.binop(mul, %1, %2) : !u8i
+%7 = cir.binop(add, %1, %2) nsw : !s32i
+%8 = cir.binop(add, %3, %4) nuw : !u32i
+%9 = cir.binop(add, %1, %2) sat : !s32i
```
}];
diff --git a/clang/include/clang/CIR/MissingFeatures.h
b/clang/include/clang/CIR/MissingFeatures.h
index 3a102d90aba8f..23bf826d19a69 100644
--- a/clang/include/clang/CIR/MissingFeatures.h
+++ b/clang/include/clang/CIR/MissingFeatures.h
@@ -76,7 +76,6 @@ struct MissingFeatures {
static bool opScopeCleanupRegion() { return false; }
// Unary operator handling
- static bool opUnarySignedOverflow() { return false; }
static bool opUnaryPromotionType() { return false; }
// Clang early optimizations or things defered to LLVM lowering.
diff --git a/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp
b/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp
index 2cf92dfbf3a5b..5ac1dc1052c2e 100644
--- a/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp
+++ b/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp
@@ -374,7 +374,7 @@ class ScalarExprEmitter : public
StmtVisitor {
cir::UnaryOpKind kind =
e->isIncrementOp() ? cir::UnaryOpKind::Inc : cir::UnaryOpKind::Dec;
// NOTE(CIR): clang calls CreateAdd but folds this to a unary op
-value = emitUnaryOp(e, kind, input);
+value = emitUnaryOp(e, kind, input, /*nsw=*/false);
}
} else if (isa(type)) {
cgf.cgm.errorNYI(e->getSourceRange(), "Unary inc/dec pointer");
@@ -429,19 +429,17 @@ class ScalarExprEmitter : public
StmtVisitor {
mlir::Value emitIncDecConsiderOverflowBehavior(const UnaryOperator *e,
mlir::Value inVal,
bool isInc) {
-assert(!cir::MissingFeatures::opUnarySignedOverflow());
cir::UnaryOpKind kind =
e->isIncrementOp() ? cir::UnaryOpKind::Inc : cir::UnaryOpKind::Dec;
switch (cgf.getLangOpts().getSignedOverflowBehavior()) {
case LangOptions::SOB_Defined:
- return emitUnaryOp(e, kind, inVal);
+ return emitUnaryOp(e, kind, inVal, /*nsw=*/false);
case LangOptions::SOB_Undefined:
[clang] [clang-tools-extra] [clang] support pack expansions for trailing requires clauses (PR #133190)
@@ -7379,8 +7378,11 @@ bool ASTContext::isSameEntity(const NamedDecl *X, const
NamedDecl *Y) const {
return false;
}
-if (!isSameConstraintExpr(FuncX->getTrailingRequiresClause(),
cor3ntin wrote:
Can we add an overload, at least + comment?
https://github.com/llvm/llvm-project/pull/133190
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][NFC] Refactor CodeGen's hasBooleanRepresentation (PR #134159)
https://github.com/andykaylor created
https://github.com/llvm/llvm-project/pull/134159
The ClangIR upstreaming project needs the same logic for
hasBooleanRepresentation() that is currently implemented in the standard clang
codegen. In order to share this code, this change moves the implementation of
this function into the AST Type class.
No functional change is intended by this change. The ClangIR use of this
function will be added separately in a later change.
>From 3f0fe2316040c114c6637545ae8fd0463dbe2b94 Mon Sep 17 00:00:00 2001
From: Andy Kaylor
Date: Wed, 2 Apr 2025 14:32:27 -0700
Subject: [PATCH] [clang][NFC] Refactor CodeGen's hasBooleanRepresentation
The ClangIR upstreaming project needs the same logic for
hasBooleanRepresentation() that is currently implemented in
the standard clang codegen. In order to share this code, this
change moves the implementation of this function into the AST
Type class.
No functional change is intended by this change. The ClangIR
use of this function will be added separately in a later change.
---
clang/include/clang/AST/Type.h | 4
clang/lib/AST/Type.cpp | 13 +
clang/lib/CodeGen/CGExpr.cpp | 23 +--
3 files changed, 22 insertions(+), 18 deletions(-)
diff --git a/clang/include/clang/AST/Type.h b/clang/include/clang/AST/Type.h
index a809102c069a8..c38250a13a6e7 100644
--- a/clang/include/clang/AST/Type.h
+++ b/clang/include/clang/AST/Type.h
@@ -2769,6 +2769,10 @@ class alignas(TypeAlignment) Type : public
ExtQualsTypeCommonBase {
/// of some sort, e.g., it is a floating-point type or a vector thereof.
bool hasFloatingRepresentation() const;
+ /// Determine whether this type has a boolean representation
+ /// of some sort.
+ bool hasBooleanRepresentation() const;
+
// Type Checking Functions: Check to see if this type is structurally the
// specified type, ignoring typedefs and qualifiers, and return a pointer to
// the best type we can.
diff --git a/clang/lib/AST/Type.cpp b/clang/lib/AST/Type.cpp
index 08798219c0b83..42fec3a13cf2f 100644
--- a/clang/lib/AST/Type.cpp
+++ b/clang/lib/AST/Type.cpp
@@ -2334,6 +2334,19 @@ bool Type::isArithmeticType() const {
return isa(CanonicalType) || isBitIntType();
}
+bool Type::hasBooleanRepresentation() const {
+ if (isBooleanType())
+return true;
+
+ if (const EnumType *ET = getAs())
+return ET->getDecl()->getIntegerType()->isBooleanType();
+
+ if (const AtomicType *AT = getAs())
+return AT->getValueType()->hasBooleanRepresentation();
+
+ return false;
+}
+
Type::ScalarTypeKind Type::getScalarTypeKind() const {
assert(isScalarType());
diff --git a/clang/lib/CodeGen/CGExpr.cpp b/clang/lib/CodeGen/CGExpr.cpp
index 3d3a111f0514a..1bde01490a428 100644
--- a/clang/lib/CodeGen/CGExpr.cpp
+++ b/clang/lib/CodeGen/CGExpr.cpp
@@ -1893,19 +1893,6 @@ llvm::Value *CodeGenFunction::EmitLoadOfScalar(LValue
lvalue,
lvalue.getTBAAInfo(), lvalue.isNontemporal());
}
-static bool hasBooleanRepresentation(QualType Ty) {
- if (Ty->isBooleanType())
-return true;
-
- if (const EnumType *ET = Ty->getAs())
-return ET->getDecl()->getIntegerType()->isBooleanType();
-
- if (const AtomicType *AT = Ty->getAs())
-return hasBooleanRepresentation(AT->getValueType());
-
- return false;
-}
-
static bool getRangeForType(CodeGenFunction &CGF, QualType Ty,
llvm::APInt &Min, llvm::APInt &End,
bool StrictEnums, bool IsBool) {
@@ -1928,7 +1915,7 @@ static bool getRangeForType(CodeGenFunction &CGF,
QualType Ty,
llvm::MDNode *CodeGenFunction::getRangeForLoadFromType(QualType Ty) {
llvm::APInt Min, End;
if (!getRangeForType(*this, Ty, Min, End, CGM.getCodeGenOpts().StrictEnums,
- hasBooleanRepresentation(Ty)))
+ Ty->hasBooleanRepresentation()))
return nullptr;
llvm::MDBuilder MDHelper(getLLVMContext());
@@ -1942,7 +1929,7 @@ bool CodeGenFunction::EmitScalarRangeCheck(llvm::Value
*Value, QualType Ty,
if (!HasBoolCheck && !HasEnumCheck)
return false;
- bool IsBool = hasBooleanRepresentation(Ty) ||
+ bool IsBool = Ty->hasBooleanRepresentation() ||
NSAPI(CGM.getContext()).isObjCBOOLType(Ty);
bool NeedsBoolCheck = HasBoolCheck && IsBool;
bool NeedsEnumCheck = HasEnumCheck && Ty->getAs();
@@ -2070,7 +2057,7 @@ llvm::Value *CodeGenFunction::EmitLoadOfScalar(Address
Addr, bool Volatile,
/// by ConvertType) to its load/store type (as returned by
/// convertTypeForLoadStore).
llvm::Value *CodeGenFunction::EmitToMemory(llvm::Value *Value, QualType Ty) {
- if (hasBooleanRepresentation(Ty) || Ty->isBitIntType()) {
+ if (Ty->hasBooleanRepresentation() || Ty->isBitIntType()) {
llvm::Type *StoreTy = convertTypeForLoadStore(Ty, Value->getType());
bool Signed = Ty->isSignedIntegerOrEnumerationType();
return Builder.CreateIntCast(Value, StoreTy, Signed, "storedv");
@
[clang] Hlsl dst function (PR #133828)
@@ -0,0 +1,37 @@
+// RUN: %clang_cc1 -finclude-default-header -triple
dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only
-disable-llvm-passes -verify
+
+float4 test_too_many_arg(float4 p0)
+{
+dst(p0, p0, p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+}
+
+float4 test_no_second_arg(float4 p0)
+{
+return dst(p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+}
+
+float4 test_no_args()
+{
+return dst();
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+}
+
+float4 test_3_components(float3 p0, float3 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector<[...], 3>' to 'vector<[...], 4>' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+}
farzonl wrote:
@metkarpoonam this is my bad the exports are only needed in godbolt. I didn't
intend for you to add them in these test cases. I appolgize.
https://github.com/llvm/llvm-project/pull/133828
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Hlsl dst function (PR #133828)
https://github.com/metkarpoonam updated
https://github.com/llvm/llvm-project/pull/133828
>From 3a45246453d120da108e597d23da0fb8d9df0b1b Mon Sep 17 00:00:00 2001
From: Poonam Vilas Metkar
Date: Mon, 31 Mar 2025 16:49:18 -0700
Subject: [PATCH 1/7] Implement a dst function with test cases for HLSL codegen
and sema
---
.../lib/Headers/hlsl/hlsl_intrinsic_helpers.h | 6 +++
clang/lib/Headers/hlsl/hlsl_intrinsics.h | 25 ++
clang/test/CodeGenHLSL/builtins/dst.hlsl | 48 +++
clang/test/SemaHLSL/BuiltIns/dst-error.hlsl | 37 ++
4 files changed, 116 insertions(+)
create mode 100644 clang/test/CodeGenHLSL/builtins/dst.hlsl
create mode 100644 clang/test/SemaHLSL/BuiltIns/dst-error.hlsl
diff --git a/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
b/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
index 8cdd63d7e07bb..5ea8faf169380 100644
--- a/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
+++ b/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
@@ -35,6 +35,12 @@ length_vec_impl(vector X) {
#endif
}
+template
+constexpr vector dst_impl(vector src0, vector src1) {
+ vector dest = {1, src0[1] * src1[1], src0[2], src1[3]};
+ return dest;
+}
+
template constexpr T distance_impl(T X, T Y) {
return length_impl(X - Y);
}
diff --git a/clang/lib/Headers/hlsl/hlsl_intrinsics.h
b/clang/lib/Headers/hlsl/hlsl_intrinsics.h
index fd799b8d874ae..7ae94731234f9 100644
--- a/clang/lib/Headers/hlsl/hlsl_intrinsics.h
+++ b/clang/lib/Headers/hlsl/hlsl_intrinsics.h
@@ -174,6 +174,31 @@ const inline float
distance(__detail::HLSL_FIXED_VECTOR X,
__detail::HLSL_FIXED_VECTOR Y) {
return __detail::distance_vec_impl(X, Y);
}
+//===--===//
+// dst builtins
+//===--===//
+
+/// \fn fvector dst( fvector, fvector)
+/// \brief Returns the length of a vector
+/// \param src0 [in] The first vector contain {_, d*d, d*d, _}
+/// \param src1 [in] The second vector contain {_, 1/d, _, 1/d}
+///
+/// Return the computed distance vector contain {1, d, d*d, 1/d}
+
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+const inline vector dst(vector src0, vector src1) {
+ return __detail::dst_impl(src0, src1);
+}
+
+const inline vector dst(vector src0,
+ vector src1) {
+ return __detail::dst_impl(src0, src1);
+}
+
+const inline vector dst(vector src0,
+ vector src1) {
+ return __detail::dst_impl(src0, src1);
+}
//===--===//
// fmod builtins
diff --git a/clang/test/CodeGenHLSL/builtins/dst.hlsl
b/clang/test/CodeGenHLSL/builtins/dst.hlsl
new file mode 100644
index 0..c62c9be5b0c1d
--- /dev/null
+++ b/clang/test/CodeGenHLSL/builtins/dst.hlsl
@@ -0,0 +1,48 @@
+// RUN: %clang_cc1 -finclude-default-header -x hlsl -triple
dxil-pc-shadermodel6.2-library %s -fnative-half-type -emit-llvm -O1 -o - |
FileCheck %s
+
+
+// CHECK-LABEL: define noundef nofpclass(nan inf) <4 x float>
@_Z12dstWithFloatDv4_fS_(
+// CHECK-SAME: <4 x float> noundef nofpclass(nan inf) [[P:%.*]], <4 x float>
noundef nofpclass(nan inf) [[Q:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[VECEXT:%.*]] = extractelement <4 x float> [[P]], i64 1
+// CHECK-NEXT: [[VECEXT1:%.*]] = extractelement <4 x float> [[Q]], i64 1
+// CHECK-NEXT: [[MULRES:%.*]] = fmul reassoc nnan ninf nsz arcp afn float
[[VECEXT1]], [[VECEXT]]
+// CHECK-NEXT: [[VECINIT:%.*]] = insertelement <4 x float> , float [[MULRES]], i64 1
+// CHECK-NEXT: [[VECINIT3:%.*]] = shufflevector <4 x float> [[VECINIT]], <4 x
float> [[P]], <4 x i32>
+// CHECK-NEXT: [[VECINIT5:%.*]] = shufflevector <4 x float> [[VECINIT3]], <4 x
float> [[Q]], <4 x i32>
+// CHECK-NEXT: ret <4 x float> [[VECINIT5]]
+
+float4 dstWithFloat(float4 p1, float4 p2)
+{
+return dst(p1, p2);
+}
+
+// CHECK-LABEL: define noundef nofpclass(nan inf) <4 x half>
@_Z11dstwithHalfDv4_DhS_(
+// CHECK-SAME: <4 x half> noundef nofpclass(nan inf) [[P:%.*]], <4 x half>
noundef nofpclass(nan inf) [[Q:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[VECEXT:%.*]] = extractelement <4 x half> [[P]], i64 1
+// CHECK-NEXT: [[VECEXT1:%.*]] = extractelement <4 x half> [[Q]], i64 1
+// CHECK-NEXT: [[MULRES:%.*]] = fmul reassoc nnan ninf nsz arcp afn half
[[VECEXT1]], [[VECEXT]]
+// CHECK-NEXT: [[VECINIT:%.*]] = insertelement <4 x half> , half [[MULRES]], i64 1
+// CHECK-NEXT: [[VECINIT3:%.*]] = shufflevector <4 x half> [[VECINIT]], <4 x
half> [[P]], <4 x i32>
+// CHECK-NEXT: [[VECINIT5:%.*]] = shufflevector <4 x half> [[VECINIT3]], <4 x
half> [[Q]], <4 x i32>
+// CHECK-NEXT: ret <4 x half> [[VECINIT5]]
+half4 dstwithHalf(half4 p1, half4 p2)
+{
+return dst(p1, p2);
+}
+
+// CHECK-LABEL: define noundef nofpclass(nan inf) <4 x double>
[clang] Hlsl dst function (PR #133828)
@@ -0,0 +1,51 @@
+// RUN: %clang_cc1 -finclude-default-header -triple
dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only
-disable-llvm-passes -verify
+
+float4 test_too_many_arg(float4 p0)
+{
+dst(p0, p0, p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+}
+
+float4 test_no_second_arg(float4 p0)
+{
+return dst(p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+}
+
+float4 test_no_args()
+{
+return dst();
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+}
+
+float4 test_3_components(float3 p0, float3 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector<[...], 3>' to 'vector<[...], 4>' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+}
+
+float4 test_float2(double4 p0, float4 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{call to 'dst' is ambiguous}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function}}
+}
+
+float2 test_float2(float4 p0, float4 p1)
farzonl wrote:
this test name is not descriptive.
https://github.com/llvm/llvm-project/pull/133828
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Fix crash on invalid `std::initializer_list` template-id (PR #132284)
https://github.com/offsetof updated
https://github.com/llvm/llvm-project/pull/132284
>From 553ced1e367c0ec6399e71e7a3a3685a550f3431 Mon Sep 17 00:00:00 2001
From: offsetof
Date: Thu, 20 Mar 2025 20:54:45 +
Subject: [PATCH 1/2] [clang] Fix crash on invalid `std::initializer_list`
template-id
In `Sema::BuildStdInitializerList`, check that the synthesized
template-id `std::initializer_list` is valid (which might not be the
case if the template has associated constraints or subsequent parameters
with default arguments) before forming the type.
---
clang/lib/Sema/SemaDeclCXX.cpp | 8 ++--
.../test/SemaCXX/invalid-std-initializer-list.cpp | 14 ++
2 files changed, 20 insertions(+), 2 deletions(-)
create mode 100644 clang/test/SemaCXX/invalid-std-initializer-list.cpp
diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp
index 7b2e0df8cb55d..54f7bc9a3b021 100644
--- a/clang/lib/Sema/SemaDeclCXX.cpp
+++ b/clang/lib/Sema/SemaDeclCXX.cpp
@@ -12182,10 +12182,14 @@ QualType Sema::BuildStdInitializerList(QualType
Element, SourceLocation Loc) {
Args.addArgument(TemplateArgumentLoc(TemplateArgument(Element),
Context.getTrivialTypeSourceInfo(Element,
Loc)));
+
+ QualType T = CheckTemplateIdType(TemplateName(StdInitializerList), Loc,
Args);
+ if (T.isNull())
+return QualType();
+
return Context.getElaboratedType(
ElaboratedTypeKeyword::None,
- NestedNameSpecifier::Create(Context, nullptr, getStdNamespace()),
- CheckTemplateIdType(TemplateName(StdInitializerList), Loc, Args));
+ NestedNameSpecifier::Create(Context, nullptr, getStdNamespace()), T);
}
bool Sema::isInitListConstructor(const FunctionDecl *Ctor) {
diff --git a/clang/test/SemaCXX/invalid-std-initializer-list.cpp
b/clang/test/SemaCXX/invalid-std-initializer-list.cpp
new file mode 100644
index 0..080a712759c45
--- /dev/null
+++ b/clang/test/SemaCXX/invalid-std-initializer-list.cpp
@@ -0,0 +1,14 @@
+// RUN: %clang_cc1 %s -verify -std=c++20
+
+namespace std {
+
+template // expected-error 2 {{type 'int' cannot be
used prior to '::' because it has no members}}
+class initializer_list;
+
+}
+
+auto x = {1}; // expected-note {{in instantiation of default argument for
'initializer_list' required here}}
+
+void f() {
+ for(int x : {1, 2}); // expected-note {{in instantiation of default
argument for 'initializer_list' required here}}
+}
>From bd5c32e5716d32a0134d99a6b16bb5cd3039f655 Mon Sep 17 00:00:00 2001
From: offsetof
Date: Mon, 24 Mar 2025 15:58:36 +
Subject: [PATCH 2/2] fixup! [clang] Fix crash on invalid
`std::initializer_list` template-id
---
clang/docs/ReleaseNotes.rst | 1 +
clang/test/SemaCXX/invalid-std-initializer-list.cpp | 4
2 files changed, 5 insertions(+)
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 159991e8db981..f1b4310134702 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -357,6 +357,7 @@ Bug Fixes to C++ Support
- Fixed a Clang regression in C++20 mode where unresolved dependent call
expressions were created inside non-dependent contexts (#GH122892)
- Clang now emits the ``-Wunused-variable`` warning when some structured
bindings are unused
and the ``[[maybe_unused]]`` attribute is not applied. (#GH125810)
+- Fixed a crash caused by invalid declarations of ``std::initializer_list``.
(#GH132256)
Bug Fixes to AST Handling
^
diff --git a/clang/test/SemaCXX/invalid-std-initializer-list.cpp
b/clang/test/SemaCXX/invalid-std-initializer-list.cpp
index 080a712759c45..93246b5f03fd4 100644
--- a/clang/test/SemaCXX/invalid-std-initializer-list.cpp
+++ b/clang/test/SemaCXX/invalid-std-initializer-list.cpp
@@ -7,8 +7,12 @@ class initializer_list;
}
+namespace gh132256 {
+
auto x = {1}; // expected-note {{in instantiation of default argument for
'initializer_list' required here}}
void f() {
for(int x : {1, 2}); // expected-note {{in instantiation of default
argument for 'initializer_list' required here}}
}
+
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] a0e1e68 - [clang][bytecode] Return Invalid() on non-constexpr builtins (#133700)
Author: Timm Baeder
Date: 2025-03-31T18:53:12+02:00
New Revision: a0e1e680d28c4ef5f87be948d1d223fbfda2950c
URL:
https://github.com/llvm/llvm-project/commit/a0e1e680d28c4ef5f87be948d1d223fbfda2950c
DIFF:
https://github.com/llvm/llvm-project/commit/a0e1e680d28c4ef5f87be948d1d223fbfda2950c.diff
LOG: [clang][bytecode] Return Invalid() on non-constexpr builtins (#133700)
So the diagnostic output matches with the current interpreter
Added:
Modified:
clang/lib/AST/ByteCode/InterpBuiltin.cpp
clang/test/AST/ByteCode/builtin-functions.cpp
Removed:
diff --git a/clang/lib/AST/ByteCode/InterpBuiltin.cpp
b/clang/lib/AST/ByteCode/InterpBuiltin.cpp
index 4d125e4c202d2..3029314ddbad8 100644
--- a/clang/lib/AST/ByteCode/InterpBuiltin.cpp
+++ b/clang/lib/AST/ByteCode/InterpBuiltin.cpp
@@ -2125,7 +2125,7 @@ static bool interp__builtin_memchr(InterpState &S,
CodePtr OpPC,
bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const Function *F,
const CallExpr *Call, uint32_t BuiltinID) {
if (!S.getASTContext().BuiltinInfo.isConstantEvaluated(BuiltinID))
-return false;
+return Invalid(S, OpPC);
const InterpFrame *Frame = S.Current;
diff --git a/clang/test/AST/ByteCode/builtin-functions.cpp
b/clang/test/AST/ByteCode/builtin-functions.cpp
index 8408286314bb8..40f7a18119751 100644
--- a/clang/test/AST/ByteCode/builtin-functions.cpp
+++ b/clang/test/AST/ByteCode/builtin-functions.cpp
@@ -1591,3 +1591,12 @@ namespace WMemChr {
constexpr bool c = !wcschr(L"hello", L'h'); // both-error {{constant
expression}} \
// both-note {{non-constexpr
function 'wcschr' cannot be used in a constant expression}}
}
+
+namespace Invalid {
+ constexpr int test() { // both-error {{never produces a constant expression}}
+__builtin_abort(); // both-note 2{{subexpression not valid in a constant
expression}}
+return 0;
+ }
+ static_assert(test() == 0); // both-error {{not an integral constant
expression}} \
+ // both-note {{in call to}}
+}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] remove unused frontend flag -fretain_subst_template_type_parm_type_ast_nodes (PR #134177)
https://github.com/mizvekov created https://github.com/llvm/llvm-project/pull/134177 This is a follow-up to #132748, where we deferred the flag removal in order to ease transition for external users. The plan is to merge this in the nearish future, in two weeks or so is my best guess. >From 4526115a6afbd5bb78b7ee11b9ff8c9f137fcdab Mon Sep 17 00:00:00 2001 From: Matheus Izvekov Date: Wed, 2 Apr 2025 17:47:17 -0300 Subject: [PATCH] [clang] remove unused frontend flag -fretain_subst_template_type_parm_type_ast_nodes This is a folow-up to #132748, where we deferred the flag removal in order to ease transition for external users. --- clang/include/clang/Basic/LangOptions.def | 1 - clang/include/clang/Driver/Options.td | 6 -- 2 files changed, 7 deletions(-) diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def index 3879cc7942877..930c1c06d1a76 100644 --- a/clang/include/clang/Basic/LangOptions.def +++ b/clang/include/clang/Basic/LangOptions.def @@ -161,7 +161,6 @@ LANGOPT(Coroutines, 1, 0, "C++20 coroutines") LANGOPT(CoroAlignedAllocation, 1, 0, "prefer Aligned Allocation according to P2014 Option 2") LANGOPT(DllExportInlines , 1, 1, "dllexported classes dllexport inline methods") LANGOPT(ExperimentalLibrary, 1, 0, "enable unstable and experimental library features") -LANGOPT(RetainSubstTemplateTypeParmTypeAstNodes, 1, 0, "retain SubstTemplateTypeParmType nodes in the AST's representation of alias template specializations") LANGOPT(PointerAuthIntrinsics, 1, 0, "pointer authentication intrinsics") LANGOPT(PointerAuthCalls , 1, 0, "function pointer authentication") diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td index e69b804de63b5..f8b59d5c6e4aa 100644 --- a/clang/include/clang/Driver/Options.td +++ b/clang/include/clang/Driver/Options.td @@ -3508,12 +3508,6 @@ defm application_extension : BoolFOption<"application-extension", PosFlag, NegFlag>; -defm retain_subst_template_type_parm_type_ast_nodes : BoolFOption<"retain-subst-template-type-parm-type-ast-nodes", - LangOpts<"RetainSubstTemplateTypeParmTypeAstNodes">, DefaultFalse, - PosFlag, - NegFlag, - BothFlags<[], [], " retain SubstTemplateTypeParmType nodes in the AST's representation" - " of alias template specializations">>; defm sized_deallocation : BoolFOption<"sized-deallocation", LangOpts<"SizedDeallocation">, Default, PosFlag, ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
llvm-ci wrote: LLVM Buildbot has detected a new failure on builder `clang-ppc64-aix` running on `aix-ppc64` while building `clang` at step 6 "test-build-unified-tree-check-all". Full details are available at: https://lab.llvm.org/buildbot/#/builders/64/builds/2751 Here is the relevant piece of the build log for the reference ``` Step 6 (test-build-unified-tree-check-all) failure: test (failure) TEST 'Clang :: ClangScanDeps/verbose.test' FAILED Exit Code: 1 Command Output (stderr): -- rm -rf /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp # RUN: at line 1 + rm -rf /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp split-file /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp # RUN: at line 2 + split-file /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp sed -e "s|DIR|/home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp|g" /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/cdb.json.in > /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/cdb.json # RUN: at line 3 + sed -e 's|DIR|/home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp|g' /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/cdb.json.in /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/bin/clang-scan-deps -compilation-database /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/cdb.json -v -o /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/result.json 2>&1 | /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/bin/FileCheck /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test # RUN: at line 5 + /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/bin/clang-scan-deps -compilation-database /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/cdb.json -v -o /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/tools/clang/test/ClangScanDeps/Output/verbose.test.tmp/result.json + /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/build/bin/FileCheck /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test:6:11: error: CHECK: expected string not found in input // CHECK: *** Virtual File System Stats: ^ :1:1: note: scanning from here PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. ^ :1:8: note: possible intended match here PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. ^ Input file: Check file: /home/llvm/llvm-external-buildbots/workers/aix-ppc64/clang-ppc64-aix/llvm-project/clang/test/ClangScanDeps/verbose.test -dump-input=help explains the following input dump. Input was: << 1: PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace. check:6'0 X~~~ error: no match found check:6'1? possible intended match >> -- ``` https://github.com/llvm/llvm-project/pull/133967 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang][OpenCL][AMDGPU] Allow a kernel to call another kernel (PR #115821)
lalaniket8 wrote: pinging for review @arsenm @yxsamliu @rjmccall https://github.com/llvm/llvm-project/pull/115821 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [HLSL] Implement dot2add intrinsic (PR #131237)
https://github.com/farzonl edited https://github.com/llvm/llvm-project/pull/131237 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang][CodeGen][UBSan] Add more precise attributes to recoverable ubsan handlers (PR #130990)
https://github.com/efriedma-quic approved this pull request. LGTM https://github.com/llvm/llvm-project/pull/130990 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] f302f35 - [clang] Track final substitution for Subst* AST nodes (#132748)
Author: Matheus Izvekov
Date: 2025-04-02T19:27:29-03:00
New Revision: f302f35526553abcb46dab278c4494c3d01deb45
URL:
https://github.com/llvm/llvm-project/commit/f302f35526553abcb46dab278c4494c3d01deb45
DIFF:
https://github.com/llvm/llvm-project/commit/f302f35526553abcb46dab278c4494c3d01deb45.diff
LOG: [clang] Track final substitution for Subst* AST nodes (#132748)
Added:
Modified:
clang/include/clang/AST/ASTContext.h
clang/include/clang/AST/ExprCXX.h
clang/include/clang/AST/PropertiesBase.td
clang/include/clang/AST/TemplateName.h
clang/include/clang/AST/Type.h
clang/include/clang/AST/TypeProperties.td
clang/lib/AST/ASTContext.cpp
clang/lib/AST/ASTImporter.cpp
clang/lib/AST/ExprCXX.cpp
clang/lib/AST/TemplateName.cpp
clang/lib/AST/TextNodeDumper.cpp
clang/lib/AST/Type.cpp
clang/lib/Sema/SemaTemplate.cpp
clang/lib/Sema/SemaTemplateInstantiate.cpp
clang/lib/Sema/TreeTransform.h
clang/test/AST/ast-dump-template-decls.cpp
clang/test/Misc/diag-template-diffing-cxx11.cpp
clang/test/SemaTemplate/make_integer_seq.cpp
Removed:
clang/test/AST/ast-dump-retain-subst-template-type-parm-type-ast-nodes.cpp
diff --git a/clang/include/clang/AST/ASTContext.h
b/clang/include/clang/AST/ASTContext.h
index a24f30815e6b9..b3010fa888fa4 100644
--- a/clang/include/clang/AST/ASTContext.h
+++ b/clang/include/clang/AST/ASTContext.h
@@ -1795,10 +1795,10 @@ class ASTContext : public RefCountedBase {
QualType Wrapped, QualType Contained,
const HLSLAttributedResourceType::Attributes &Attrs);
- QualType
- getSubstTemplateTypeParmType(QualType Replacement, Decl *AssociatedDecl,
- unsigned Index,
- std::optional PackIndex) const;
+ QualType getSubstTemplateTypeParmType(QualType Replacement,
+Decl *AssociatedDecl, unsigned Index,
+std::optional PackIndex,
+bool Final) const;
QualType getSubstTemplateTypeParmPackType(Decl *AssociatedDecl,
unsigned Index, bool Final,
const TemplateArgument &ArgPack);
@@ -2393,10 +2393,11 @@ class ASTContext : public RefCountedBase {
TemplateName
getDependentTemplateName(const DependentTemplateStorage &Name) const;
- TemplateName
- getSubstTemplateTemplateParm(TemplateName replacement, Decl *AssociatedDecl,
- unsigned Index,
- std::optional PackIndex) const;
+ TemplateName getSubstTemplateTemplateParm(TemplateName replacement,
+Decl *AssociatedDecl,
+unsigned Index,
+std::optional PackIndex,
+bool Final) const;
TemplateName getSubstTemplateTemplateParmPack(const TemplateArgument
&ArgPack,
Decl *AssociatedDecl,
unsigned Index,
diff --git a/clang/include/clang/AST/ExprCXX.h
b/clang/include/clang/AST/ExprCXX.h
index 223d74993e9e6..028ee82718d50 100644
--- a/clang/include/clang/AST/ExprCXX.h
+++ b/clang/include/clang/AST/ExprCXX.h
@@ -4514,7 +4514,9 @@ class SubstNonTypeTemplateParmExpr : public Expr {
llvm::PointerIntPair AssociatedDeclAndRef;
unsigned Index : 15;
- unsigned PackIndex : 16;
+ unsigned PackIndex : 15;
+ LLVM_PREFERRED_TYPE(bool)
+ unsigned Final : 1;
explicit SubstNonTypeTemplateParmExpr(EmptyShell Empty)
: Expr(SubstNonTypeTemplateParmExprClass, Empty) {}
@@ -4523,11 +4525,12 @@ class SubstNonTypeTemplateParmExpr : public Expr {
SubstNonTypeTemplateParmExpr(QualType Ty, ExprValueKind ValueKind,
SourceLocation Loc, Expr *Replacement,
Decl *AssociatedDecl, unsigned Index,
- std::optional PackIndex, bool
RefParam)
+ std::optional PackIndex, bool
RefParam,
+ bool Final)
: Expr(SubstNonTypeTemplateParmExprClass, Ty, ValueKind, OK_Ordinary),
Replacement(Replacement),
AssociatedDeclAndRef(AssociatedDecl, RefParam), Index(Index),
-PackIndex(PackIndex ? *PackIndex + 1 : 0) {
+PackIndex(PackIndex ? *PackIndex + 1 : 0), Final(Final) {
assert(AssociatedDecl != nullptr);
SubstNonTypeTemplateParmExprBits.NameLoc = Loc;
setDependence(computeDependence(this));
@@ -4555,6 +4558,10 @@ class SubstNonTypeTemplateParmExpr : public Expr {
return PackIndex - 1;
}
+ // This substitution is Final, which means the substitution is fully
+ // sugared: it doesn't need to be res
[clang] [clang] Track final substitution for Subst* AST nodes (PR #132748)
https://github.com/mizvekov closed https://github.com/llvm/llvm-project/pull/132748 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [HLSL] Make it possible to assign an array from a cbuffer (PR #134174)
https://github.com/spall created
https://github.com/llvm/llvm-project/pull/134174
Update Sema Checking to always do an HLSL Array RValue cast in the case we are
dealing with hlsl constant array types
Instead of comparing canonical types, compare canonical unqualified types
Add a test to show it is possible to assign an array from a cbuffer.
Closes #133767
>From 11ecff90d8d51bce41bc8063afb31bd27e3981cc Mon Sep 17 00:00:00 2001
From: Sarah Spall
Date: Wed, 2 Apr 2025 12:04:18 -0700
Subject: [PATCH] compare unqualified canonical types and add an hlsl array
rvalue cast
---
clang/lib/Sema/SemaExprCXX.cpp | 13 +++--
clang/lib/Sema/SemaOverload.cpp | 10 --
clang/test/CodeGenHLSL/ArrayAssignable.hlsl | 20
3 files changed, 27 insertions(+), 16 deletions(-)
diff --git a/clang/lib/Sema/SemaExprCXX.cpp b/clang/lib/Sema/SemaExprCXX.cpp
index fa492bc124abd..17e9311a631f7 100644
--- a/clang/lib/Sema/SemaExprCXX.cpp
+++ b/clang/lib/Sema/SemaExprCXX.cpp
@@ -4418,19 +4418,12 @@ Sema::PerformImplicitConversion(Expr *From, QualType
ToType,
case ICK_HLSL_Array_RValue:
if (ToType->isArrayParameterType()) {
FromType = Context.getArrayParameterType(FromType);
- From = ImpCastExprToType(From, FromType, CK_HLSLArrayRValue, VK_PRValue,
- /*BasePath=*/nullptr, CCK)
- .get();
-} else { // FromType must be ArrayParameterType
- assert(FromType->isArrayParameterType() &&
- "FromType must be ArrayParameterType in ICK_HLSL_Array_RValue \
- if it is not ToType");
+} else if (FromType->isArrayParameterType()) {
const ArrayParameterType *APT = cast(FromType);
FromType = APT->getConstantArrayType(Context);
- From = ImpCastExprToType(From, FromType, CK_HLSLArrayRValue, VK_PRValue,
- /*BasePath=*/nullptr, CCK)
- .get();
}
+ From = ImpCastExprToType(From, FromType, CK_HLSLArrayRValue, VK_PRValue,
+ /*BasePath=*/nullptr, CCK).get();
break;
case ICK_Function_To_Pointer:
diff --git a/clang/lib/Sema/SemaOverload.cpp b/clang/lib/Sema/SemaOverload.cpp
index 1802f8f4e1f91..950114046fe5e 100644
--- a/clang/lib/Sema/SemaOverload.cpp
+++ b/clang/lib/Sema/SemaOverload.cpp
@@ -2268,17 +2268,15 @@ static bool IsStandardConversion(Sema &S, Expr* From,
QualType ToType,
// handling here.
if (ToType->isArrayParameterType()) {
FromType = S.Context.getArrayParameterType(FromType);
- SCS.First = ICK_HLSL_Array_RValue;
} else if (FromType->isArrayParameterType()) {
const ArrayParameterType *APT = cast(FromType);
FromType = APT->getConstantArrayType(S.Context);
- SCS.First = ICK_HLSL_Array_RValue;
-} else {
- SCS.First = ICK_Identity;
}
-if (S.Context.getCanonicalType(FromType) !=
-S.Context.getCanonicalType(ToType))
+SCS.First = ICK_HLSL_Array_RValue;
+
+if (FromType.getCanonicalType().getUnqualifiedType() !=
+ ToType.getCanonicalType().getUnqualifiedType())
return false;
SCS.setAllToTypes(ToType);
diff --git a/clang/test/CodeGenHLSL/ArrayAssignable.hlsl
b/clang/test/CodeGenHLSL/ArrayAssignable.hlsl
index e2ff2de68ed99..ce806e1492d80 100644
--- a/clang/test/CodeGenHLSL/ArrayAssignable.hlsl
+++ b/clang/test/CodeGenHLSL/ArrayAssignable.hlsl
@@ -1,5 +1,13 @@
// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.3-library -emit-llvm
-disable-llvm-passes -o - %s | FileCheck %s --enable-var-scope
+
+//CHECK-DAG: [[CBLayout:%.*]] = type <{ [2 x float] }>
+//CHECK-DAG: @CBArrays.cb = global target("dx.CBuffer", target("dx.Layout",
[[CBLayout]], 20, 0)) poison
+//CHECK-DAG: @c1 = external addrspace(2) global [2 x float], align 4
+cbuffer CBArrays {
+ float c1[2];
+}
+
// CHECK-LABEL: define void {{.*}}arr_assign1
// CHECK: [[Arr:%.*]] = alloca [2 x i32], align 4
// CHECK-NEXT: [[Arr2:%.*]] = alloca [2 x i32], align 4
@@ -116,3 +124,15 @@ void arr_assign7() {
int Arr2[2][2] = {{0, 0}, {1, 1}};
(Arr = Arr2)[0] = {6, 6};
}
+
+// Verify you can assign from a cbuffer array
+
+// CHECK-LABEL: define void {{.*}}arr_assign8
+// CHECK: [[C:%.*]] = alloca [2 x float], align 4
+// CHECK-NEXT: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[C]], ptr align 4
{{.*}}, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0.p2.i32(ptr align 4 [[C]], ptr
addrspace(2) align 4 @c1, i32 8, i1 false)
+// CHECK-NEXT: ret void
+void arr_assign8() {
+ float C[2] = {1.0, 2.0};
+ C = c1;
+}
\ No newline at end of file
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [HLSL] Make it possible to assign an array from a cbuffer (PR #134174)
github-actions[bot] wrote: :warning: C/C++ code formatter, clang-format found issues in your code. :warning: You can test this locally with the following command: ``bash git-clang-format --diff HEAD~1 HEAD --extensions cpp -- clang/lib/Sema/SemaExprCXX.cpp clang/lib/Sema/SemaOverload.cpp `` View the diff from clang-format here. ``diff diff --git a/clang/lib/Sema/SemaExprCXX.cpp b/clang/lib/Sema/SemaExprCXX.cpp index 9e9229ff5..9ab39a824 100644 --- a/clang/lib/Sema/SemaExprCXX.cpp +++ b/clang/lib/Sema/SemaExprCXX.cpp @@ -4420,8 +4420,9 @@ Sema::PerformImplicitConversion(Expr *From, QualType ToType, const ArrayParameterType *APT = cast(FromType); FromType = APT->getConstantArrayType(Context); } - From = ImpCastExprToType(From, FromType, CK_HLSLArrayRValue, VK_PRValue, - /*BasePath=*/nullptr, CCK).get(); +From = ImpCastExprToType(From, FromType, CK_HLSLArrayRValue, VK_PRValue, + /*BasePath=*/nullptr, CCK) + .get(); break; case ICK_Function_To_Pointer: diff --git a/clang/lib/Sema/SemaOverload.cpp b/clang/lib/Sema/SemaOverload.cpp index 950114046..d282fe50e 100644 --- a/clang/lib/Sema/SemaOverload.cpp +++ b/clang/lib/Sema/SemaOverload.cpp @@ -2276,7 +2276,7 @@ static bool IsStandardConversion(Sema &S, Expr* From, QualType ToType, SCS.First = ICK_HLSL_Array_RValue; if (FromType.getCanonicalType().getUnqualifiedType() != - ToType.getCanonicalType().getUnqualifiedType()) +ToType.getCanonicalType().getUnqualifiedType()) return false; SCS.setAllToTypes(ToType); `` https://github.com/llvm/llvm-project/pull/134174 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-tidy] Improve integer comparison by matching valid expressions outside implicitCastExpr (PR #134188)
https://github.com/RiverDave updated
https://github.com/llvm/llvm-project/pull/134188
>From ac9833cd5abdaf4a91c785d5fa9d51190219851a Mon Sep 17 00:00:00 2001
From: David Rivera
Date: Wed, 2 Apr 2025 21:02:00 -0400
Subject: [PATCH] [clang-tidy] Improve integer comparison by matching valid
expressions outside implicitCastExpr
---
.../UseIntegerSignComparisonCheck.cpp | 16 +---
clang-tools-extra/docs/ReleaseNotes.rst | 4 +++
.../modernize/use-integer-sign-comparison.cpp | 26 +++
3 files changed, 43 insertions(+), 3 deletions(-)
diff --git
a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
index eeba5cce80da5..63e044f8ca993 100644
--- a/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
+++ b/clang-tools-extra/clang-tidy/modernize/UseIntegerSignComparisonCheck.cpp
@@ -39,9 +39,11 @@ intCastExpression(bool IsSigned,
// std::cmp_{} functions trigger a compile-time error if either LHS or RHS
// is a non-integer type, char, enum or bool
// (unsigned char/ signed char are Ok and can be used).
- auto IntTypeExpr = expr(hasType(hasCanonicalType(qualType(
+ const auto HasIntegerType = hasType(hasCanonicalType(qualType(
isInteger(), IsSigned ? isSignedInteger() : isUnsignedInteger(),
- unless(isActualChar()), unless(booleanType()), unless(enumType());
+ unless(isActualChar()), unless(booleanType()), unless(enumType();
+
+ auto IntTypeExpr = expr(HasIntegerType);
const auto ImplicitCastExpr =
CastBindName.empty() ? implicitCastExpr(hasSourceExpression(IntTypeExpr))
@@ -52,8 +54,16 @@ intCastExpression(bool IsSigned,
const auto StaticCastExpr = cxxStaticCastExpr(has(ImplicitCastExpr));
const auto FunctionalCastExpr = cxxFunctionalCastExpr(has(ImplicitCastExpr));
+ // Match function calls or variable references not directly wrapped by an
implicit cast
+ const auto CallIntExpr = CastBindName.empty()
+ ? callExpr(HasIntegerType)
+ : callExpr(HasIntegerType).bind(CastBindName);
+ const auto DeclRefIntExpr =
+ CastBindName.empty() ? declRefExpr(HasIntegerType)
+ : declRefExpr(HasIntegerType).bind(CastBindName);
+
return expr(anyOf(ImplicitCastExpr, CStyleCastExpr, StaticCastExpr,
-FunctionalCastExpr));
+FunctionalCastExpr, CallIntExpr));
}
static StringRef parseOpCode(BinaryOperator::Opcode Code) {
diff --git a/clang-tools-extra/docs/ReleaseNotes.rst
b/clang-tools-extra/docs/ReleaseNotes.rst
index 6cb8d572d3a78..d90f277045003 100644
--- a/clang-tools-extra/docs/ReleaseNotes.rst
+++ b/clang-tools-extra/docs/ReleaseNotes.rst
@@ -177,6 +177,10 @@ Changes in existing checks
matched scenarios of ``find`` and ``rfind`` methods and fixing false
positives when those methods were called with 3 arguments.
+- Improved :doc:`modernize-use-integer-sign-comparison
+ ` check by matching
+ valid integer expressions not directly wrapped around an Implicit Cast.
+
- Improved :doc:`modernize-use-std-numbers
` check to support math
functions of different precisions.
diff --git
a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
index 99f00444c2d3f..1d2f64a359a2c 100644
---
a/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
+++
b/clang-tools-extra/test/clang-tidy/checkers/modernize/use-integer-sign-comparison.cpp
@@ -120,3 +120,29 @@ int AllComparisons() {
return 0;
}
+
+namespace PR127471 {
+int getSignedValue();
+unsigned int getUnsignedValue();
+
+void callExprTest() {
+
+if (getSignedValue() < getUnsignedValue())
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsigned' integers [modernize-use-integer-sign-comparison]
+// CHECK-FIXES: if (std::cmp_less(getSignedValue() , getUnsignedValue()))
+
+int sVar = 0;
+if (getUnsignedValue() > sVar)
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsigned' integers [modernize-use-integer-sign-comparison]
+// CHECK-FIXES: if (std::cmp_greater(getUnsignedValue() , sVar))
+
+unsigned int uVar = 0;
+if (getSignedValue() > uVar)
+return;
+// CHECK-MESSAGES: :[[@LINE-2]]:13: warning: comparison between 'signed' and
'unsigned' integers [modernize-use-integer-sign-comparison]
+// CHECK-FIXES: if (std::cmp_greater(getSignedValue() , uVar))
+
+}
+} // namespace PR127471
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (PR #133863)
zyn0217 wrote: /cherry-pick dcc2182bce https://github.com/llvm/llvm-project/pull/133863 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [HLSL] Implement the `lit` intrinsic (PR #134171)
@@ -280,6 +280,22 @@ constexpr bool4 isinf(double4 V) { return
isinf((float4)V); }
_DXC_COMPAT_TERNARY_DOUBLE_OVERLOADS(lerp)
_DXC_COMPAT_TERNARY_INTEGER_OVERLOADS(lerp)
+//===--===//
+// lit builtins overloads
+//===--===//
+
+template
+constexpr __detail::enable_if_t<__detail::is_arithmetic::Value &&
+(__detail::is_same::value ||
farzonl wrote:
I'm curious do we have to check all these types or can we just check
`__detail::is_arithmetic` and not half or float so as to not impact the
other templates?
https://github.com/llvm/llvm-project/pull/134171
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] dcc2182 - [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (#133863)
Author: Younan Zhang
Date: 2025-04-03T11:15:42+08:00
New Revision: dcc2182bce3d2ef0e0a991664c51b4b3bfcf7197
URL:
https://github.com/llvm/llvm-project/commit/dcc2182bce3d2ef0e0a991664c51b4b3bfcf7197
DIFF:
https://github.com/llvm/llvm-project/commit/dcc2182bce3d2ef0e0a991664c51b4b3bfcf7197.diff
LOG: [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (#133863)
In ecc7e6ce4, we tried to inspect the `LambdaScopeInfo` on stack to
recover the instantiating lambda captures. However, there was a mismatch
in how we compared the pattern declarations of lambdas: the constraint
instantiation used a tailored `getPatternFunctionDecl()` which is
localized in SemaLambda that finds the very primal template declaration
of a lambda, while `FunctionDecl::getTemplateInstantiationPattern` finds
the latest template pattern of a lambda. This difference causes issues
when lambdas are nested, as we always want the primary template
declaration.
This corrects that by moving `Sema::addInstantiatedCapturesToScope` from
SemaConcept to SemaLambda, allowing it to use the localized version of
`getPatternFunctionDecl`.
It is also worth exploring to coalesce the implementation of
`getPatternFunctionDecl` with
`FunctionDecl::getTemplateInstantiationPattern`. But I’m leaving that
for the future, as I’d like to backport this fix (ecc7e6ce4 made the
issue more visible in clang 20, sorry!), and changing Sema’s ABI would
not be suitable in that regards. Hence, no release note.
Fixes https://github.com/llvm/llvm-project/issues/133719
Added:
Modified:
clang/lib/Sema/SemaConcept.cpp
clang/lib/Sema/SemaLambda.cpp
clang/test/SemaTemplate/concepts-lambda.cpp
Removed:
diff --git a/clang/lib/Sema/SemaConcept.cpp b/clang/lib/Sema/SemaConcept.cpp
index 5e1cd62530c5b..16f9e3d60560e 100644
--- a/clang/lib/Sema/SemaConcept.cpp
+++ b/clang/lib/Sema/SemaConcept.cpp
@@ -697,75 +697,6 @@ bool Sema::CheckConstraintSatisfaction(
.isInvalid();
}
-bool Sema::addInstantiatedCapturesToScope(
-FunctionDecl *Function, const FunctionDecl *PatternDecl,
-LocalInstantiationScope &Scope,
-const MultiLevelTemplateArgumentList &TemplateArgs) {
- const auto *LambdaClass = cast(Function)->getParent();
- const auto *LambdaPattern = cast(PatternDecl)->getParent();
-
- unsigned Instantiated = 0;
-
- // FIXME: This is a workaround for not having deferred lambda body
- // instantiation.
- // When transforming a lambda's body, if we encounter another call to a
- // nested lambda that contains a constraint expression, we add all of the
- // outer lambda's instantiated captures to the current instantiation scope to
- // facilitate constraint evaluation. However, these captures don't appear in
- // the CXXRecordDecl until after the lambda expression is rebuilt, so we
- // pull them out from the corresponding LSI.
- LambdaScopeInfo *InstantiatingScope = nullptr;
- if (LambdaPattern->capture_size() && !LambdaClass->capture_size()) {
-for (FunctionScopeInfo *Scope : llvm::reverse(FunctionScopes)) {
- auto *LSI = dyn_cast(Scope);
- if (!LSI ||
- LSI->CallOperator->getTemplateInstantiationPattern() != PatternDecl)
-continue;
- InstantiatingScope = LSI;
- break;
-}
-assert(InstantiatingScope);
- }
-
- auto AddSingleCapture = [&](const ValueDecl *CapturedPattern,
- unsigned Index) {
-ValueDecl *CapturedVar =
-InstantiatingScope ? InstantiatingScope->Captures[Index].getVariable()
- : LambdaClass->getCapture(Index)->getCapturedVar();
-assert(CapturedVar->isInitCapture());
-Scope.InstantiatedLocal(CapturedPattern, CapturedVar);
- };
-
- for (const LambdaCapture &CapturePattern : LambdaPattern->captures()) {
-if (!CapturePattern.capturesVariable()) {
- Instantiated++;
- continue;
-}
-ValueDecl *CapturedPattern = CapturePattern.getCapturedVar();
-
-if (!CapturedPattern->isInitCapture()) {
- Instantiated++;
- continue;
-}
-
-if (!CapturedPattern->isParameterPack()) {
- AddSingleCapture(CapturedPattern, Instantiated++);
-} else {
- Scope.MakeInstantiatedLocalArgPack(CapturedPattern);
- SmallVector Unexpanded;
- SemaRef.collectUnexpandedParameterPacks(
- dyn_cast(CapturedPattern)->getInit(), Unexpanded);
- auto NumArgumentsInExpansion =
- getNumArgumentsInExpansionFromUnexpanded(Unexpanded, TemplateArgs);
- if (!NumArgumentsInExpansion)
-continue;
- for (unsigned Arg = 0; Arg < *NumArgumentsInExpansion; ++Arg)
-AddSingleCapture(CapturedPattern, Instantiated++);
-}
- }
- return false;
-}
-
bool Sema::SetupConstraintScope(
FunctionDecl *FD, std::optional> TemplateArgs,
const MultiLevelTemplateArgumentList &MLTAL,
diff --git a/clang/lib/Sema/SemaLambda.cpp b/clang/lib/Sema/SemaLambd
[clang] [clang] Check `std::initializer_list` more strictly (PR #133822)
@@ -0,0 +1,77 @@ +// RUN: %clang_cc1 %s -verify=expected,type-param -std=c++23 -DTYPE_PARAM +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DCONSTANT_PARAM +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DTYPE_TEMPLATE_PARAM +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DDEFAULT_ARG +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DMULTIPLE_PARAMS +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DPARAM_PACK +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DCONSTRAINED_PARAM +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DREQUIRES_CLAUSE +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DNONCLASS_TEMPLATE +// RUN: %clang_cc1 %s -verify=expected,others -std=c++23 -DNONTEMPLATE + zyn0217 wrote: Because otherwise they're redefined https://github.com/llvm/llvm-project/pull/133822 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (PR #133863)
zyn0217 wrote: /cherry-pick dcc2182bc https://github.com/llvm/llvm-project/pull/133863 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Check `std::initializer_list` more strictly (PR #133822)
@@ -12155,16 +12185,16 @@ static ClassTemplateDecl
*LookupStdInitializerList(Sema &S, SourceLocation Loc){
Result.suppressDiagnostics();
// We found something weird. Complain about the first thing we found.
NamedDecl *Found = *Result.begin();
-S.Diag(Found->getLocation(), diag::err_malformed_std_initializer_list);
+S.Diag(Found->getLocation(), diag::err_malformed_std_initializer_list)
+<< diag::MalformedStdInitializerList::BadEntityKind;
+S.Diag(Loc, diag::note_used_here);
return nullptr;
zyn0217 wrote:
Couldn't we use CheckStdInitializerList?
https://github.com/llvm/llvm-project/pull/133822
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Reapply "[cmake] Refactor clang unittest cmake" (PR #134195)
llvmbot wrote: @llvm/pr-subscribers-clang-driver @llvm/pr-subscribers-clang Author: Reid Kleckner (rnk) Changes This reapplies 5ffd9bdb50b57 (#133545) with fixes. The BUILD_SHARED_LIBS=ON build was fixed by adding missing LLVM dependencies to the InterpTests binary in unittests/AST/ByteCode/CMakeLists.txt . --- Patch is 22.08 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/134195.diff 28 Files Affected: - (modified) clang/unittests/AST/ByteCode/CMakeLists.txt (+7-9) - (modified) clang/unittests/AST/CMakeLists.txt (+7-16) - (modified) clang/unittests/ASTMatchers/CMakeLists.txt (+7-15) - (modified) clang/unittests/ASTMatchers/Dynamic/CMakeLists.txt (+5-13) - (modified) clang/unittests/Analysis/CMakeLists.txt (+5-13) - (modified) clang/unittests/Analysis/FlowSensitive/CMakeLists.txt (+5-13) - (modified) clang/unittests/Basic/CMakeLists.txt (+5-13) - (modified) clang/unittests/CMakeLists.txt (+28-4) - (modified) clang/unittests/CodeGen/CMakeLists.txt (+5-10) - (modified) clang/unittests/CrossTU/CMakeLists.txt (+3-9) - (modified) clang/unittests/DirectoryWatcher/CMakeLists.txt (+3-8) - (modified) clang/unittests/Driver/CMakeLists.txt (+7-12) - (modified) clang/unittests/Format/CMakeLists.txt (+3-8) - (modified) clang/unittests/Frontend/CMakeLists.txt (+4-8) - (modified) clang/unittests/Index/CMakeLists.txt (+4-9) - (modified) clang/unittests/InstallAPI/CMakeLists.txt (+3-6) - (modified) clang/unittests/Interpreter/CMakeLists.txt (+12-13) - (modified) clang/unittests/Interpreter/ExceptionTests/CMakeLists.txt (+9-11) - (modified) clang/unittests/Lex/CMakeLists.txt (+4-12) - (modified) clang/unittests/Rewrite/CMakeLists.txt (+3-7) - (modified) clang/unittests/Sema/CMakeLists.txt (+5-13) - (modified) clang/unittests/Serialization/CMakeLists.txt (+6-11) - (modified) clang/unittests/StaticAnalyzer/CMakeLists.txt (+5-13) - (modified) clang/unittests/Support/CMakeLists.txt (+3-8) - (modified) clang/unittests/Tooling/CMakeLists.txt (+11-17) - (modified) clang/unittests/Tooling/Syntax/CMakeLists.txt (+5-10) - (modified) clang/unittests/libclang/CMakeLists.txt (+1-4) - (modified) clang/unittests/libclang/CrashTests/CMakeLists.txt (+1-4) ``diff diff --git a/clang/unittests/AST/ByteCode/CMakeLists.txt b/clang/unittests/AST/ByteCode/CMakeLists.txt index b862fb4834fbd..1469cd6b2a8ea 100644 --- a/clang/unittests/AST/ByteCode/CMakeLists.txt +++ b/clang/unittests/AST/ByteCode/CMakeLists.txt @@ -2,19 +2,17 @@ add_clang_unittest(InterpTests BitcastBuffer.cpp Descriptor.cpp toAPValue.cpp - ) - -clang_target_link_libraries(InterpTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic clangFrontend clangSerialization clangTooling - ) - - target_link_libraries(InterpTests - PRIVATE + LINK_LIBS clangTesting -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) diff --git a/clang/unittests/AST/CMakeLists.txt b/clang/unittests/AST/CMakeLists.txt index bfa6082a6ffa4..f27d34e8a0719 100644 --- a/clang/unittests/AST/CMakeLists.txt +++ b/clang/unittests/AST/CMakeLists.txt @@ -1,10 +1,3 @@ -set(LLVM_LINK_COMPONENTS - FrontendOpenMP - Support - TargetParser - ) - - add_subdirectory(ByteCode) add_clang_unittest(ASTTests @@ -43,10 +36,7 @@ add_clang_unittest(ASTTests TemplateNameTest.cpp TypePrinterTest.cpp UnresolvedSetTest.cpp - ) - -clang_target_link_libraries(ASTTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic @@ -54,11 +44,12 @@ clang_target_link_libraries(ASTTests clangLex clangSerialization clangTooling - ) - -target_link_libraries(ASTTests - PRIVATE + LINK_LIBS clangTesting LLVMTestingAnnotations LLVMTestingSupport -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) diff --git a/clang/unittests/ASTMatchers/CMakeLists.txt b/clang/unittests/ASTMatchers/CMakeLists.txt index 6a1e629d81b65..47bd5c108bb5a 100644 --- a/clang/unittests/ASTMatchers/CMakeLists.txt +++ b/clang/unittests/ASTMatchers/CMakeLists.txt @@ -1,31 +1,23 @@ -set(LLVM_LINK_COMPONENTS - FrontendOpenMP - Support - TargetParser - ) - add_clang_unittest(ASTMatchersTests ASTMatchersInternalTest.cpp ASTMatchersNodeTest.cpp ASTMatchersNarrowingTest.cpp ASTMatchersTraversalTest.cpp GtestMatchersTest.cpp - ) - -clang_target_link_libraries(ASTMatchersTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic clangFrontend clangSerialization clangTooling - ) - -target_link_libraries(ASTMatchersTests - PRIVATE + LINK_LIBS clangTesting LLVMTestingSupport -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) add_subdirectory(Dynamic) diff --git a/clang/unittests/ASTMatchers/Dynamic/CMakeLists.txt b/clang/unittests/ASTMatchers/Dynamic/CMakeLists.txt index 6d0e12bcb0759..b6db7ce62afe7 100644 --- a/clang/unittests/ASTMatchers/Dynamic/CMakeLists.txt +++ b/clang/unittes
[clang] [clang-format] Fix a bug in annotating braces (PR #134039)
llvm-ci wrote: LLVM Buildbot has detected a new failure on builder `sanitizer-x86_64-linux-android` running on `sanitizer-buildbot-android` while building `clang` at step 2 "annotate". Full details are available at: https://lab.llvm.org/buildbot/#/builders/186/builds/7879 Here is the relevant piece of the build log for the reference ``` Step 2 (annotate) failure: 'python ../sanitizer_buildbot/sanitizers/zorg/buildbot/builders/sanitizers/buildbot_selector.py' (failure) ... [ OK ] AddressSanitizer.AtoiAndFriendsOOBTest (2281 ms) [ RUN ] AddressSanitizer.HasFeatureAddressSanitizerTest [ OK ] AddressSanitizer.HasFeatureAddressSanitizerTest (0 ms) [ RUN ] AddressSanitizer.CallocReturnsZeroMem [ OK ] AddressSanitizer.CallocReturnsZeroMem (13 ms) [ DISABLED ] AddressSanitizer.DISABLED_TSDTest [ RUN ] AddressSanitizer.IgnoreTest [ OK ] AddressSanitizer.IgnoreTest (0 ms) [ RUN ] AddressSanitizer.SignalTest [ OK ] AddressSanitizer.SignalTest (189 ms) [ RUN ] AddressSanitizer.ReallocTest [ OK ] AddressSanitizer.ReallocTest (35 ms) [ RUN ] AddressSanitizer.WrongFreeTest [ OK ] AddressSanitizer.WrongFreeTest (129 ms) [ RUN ] AddressSanitizer.LongJmpTest [ OK ] AddressSanitizer.LongJmpTest (0 ms) [ RUN ] AddressSanitizer.ThreadStackReuseTest [ OK ] AddressSanitizer.ThreadStackReuseTest (8 ms) [ DISABLED ] AddressSanitizer.DISABLED_MemIntrinsicUnalignedAccessTest [ DISABLED ] AddressSanitizer.DISABLED_LargeFunctionSymbolizeTest [ DISABLED ] AddressSanitizer.DISABLED_MallocFreeUnwindAndSymbolizeTest [ RUN ] AddressSanitizer.UseThenFreeThenUseTest [ OK ] AddressSanitizer.UseThenFreeThenUseTest (119 ms) [ RUN ] AddressSanitizer.FileNameInGlobalReportTest [ OK ] AddressSanitizer.FileNameInGlobalReportTest (128 ms) [ DISABLED ] AddressSanitizer.DISABLED_StressStackReuseAndExceptionsTest [ RUN ] AddressSanitizer.MlockTest [ OK ] AddressSanitizer.MlockTest (0 ms) [ DISABLED ] AddressSanitizer.DISABLED_DemoThreadedTest [ DISABLED ] AddressSanitizer.DISABLED_DemoStackTest [ DISABLED ] AddressSanitizer.DISABLED_DemoThreadStackTest [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowIn [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowLeft [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowRight [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFHigh [ DISABLED ] AddressSanitizer.DISABLED_DemoOOM [ DISABLED ] AddressSanitizer.DISABLED_DemoDoubleFreeTest [ DISABLED ] AddressSanitizer.DISABLED_DemoNullDerefTest [ DISABLED ] AddressSanitizer.DISABLED_DemoFunctionStaticTest [ DISABLED ] AddressSanitizer.DISABLED_DemoTooMuchMemoryTest [ RUN ] AddressSanitizer.LongDoubleNegativeTest [ OK ] AddressSanitizer.LongDoubleNegativeTest (0 ms) [--] 19 tests from AddressSanitizer (28529 ms total) [--] Global test environment tear-down [==] 22 tests from 2 test suites ran. (28533 ms total) [ PASSED ] 22 tests. YOU HAVE 1 DISABLED TEST Step 24 (run instrumented asan tests [aarch64/aosp_coral-userdebug/AOSP.MASTER]) failure: run instrumented asan tests [aarch64/aosp_coral-userdebug/AOSP.MASTER] (failure) ... [ RUN ] AddressSanitizer.HasFeatureAddressSanitizerTest [ OK ] AddressSanitizer.HasFeatureAddressSanitizerTest (0 ms) [ RUN ] AddressSanitizer.CallocReturnsZeroMem [ OK ] AddressSanitizer.CallocReturnsZeroMem (7 ms) [ DISABLED ] AddressSanitizer.DISABLED_TSDTest [ RUN ] AddressSanitizer.IgnoreTest [ OK ] AddressSanitizer.IgnoreTest (0 ms) [ RUN ] AddressSanitizer.SignalTest [ OK ] AddressSanitizer.SignalTest (327 ms) [ RUN ] AddressSanitizer.ReallocTest [ OK ] AddressSanitizer.ReallocTest (24 ms) [ RUN ] AddressSanitizer.WrongFreeTest [ OK ] AddressSanitizer.WrongFreeTest (257 ms) [ RUN ] AddressSanitizer.LongJmpTest [ OK ] AddressSanitizer.LongJmpTest (0 ms) [ RUN ] AddressSanitizer.ThreadStackReuseTest [ OK ] AddressSanitizer.ThreadStackReuseTest (14 ms) [ DISABLED ] AddressSanitizer.DISABLED_MemIntrinsicUnalignedAccessTest [ DISABLED ] AddressSanitizer.DISABLED_LargeFunctionSymbolizeTest [ DISABLED ] AddressSanitizer.DISABLED_MallocFreeUnwindAndSymbolizeTest [ RUN ] AddressSanitizer.UseThenFreeThenUseTest [ OK ] AddressSanitizer.UseThenFreeThenUseTest (291 ms) [ RUN ] AddressSanitizer.FileNameInGlobalReportTest [ OK ] AddressSanitizer.FileNameInGlobalReportTest (304 ms) [ DISABLED ] AddressSanitizer.DISABLED_StressStackReuseAndExceptionsTest [ RUN ] AddressSanitizer.MlockTest [ OK ] AddressSanitizer.MlockTest (0 ms) [ DISABLED ] AddressSanitizer.DISABLED_DemoThreadedTest [ DISABLED ] AddressSanitizer.DISABLED_DemoStackTest [ DISABLED ] AddressSanitizer.DISABLED_DemoThreadStackTest [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowIn [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowLeft [ DISABLED ] AddressSanitizer.DISABLED_DemoUAFLowRight [ DISABLED ] Addr
[clang] [Clang] Fix a lambda pattern comparison mismatch after ecc7e6ce4 (PR #133863)
llvmbot wrote: /pull-request llvm/llvm-project#134194 https://github.com/llvm/llvm-project/pull/133863 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Reapply "[cmake] Refactor clang unittest cmake" (PR #134195)
https://github.com/rnk created https://github.com/llvm/llvm-project/pull/134195 This reapplies 5ffd9bdb50b57 (#133545) with fixes. The BUILD_SHARED_LIBS=ON build was fixed by adding missing LLVM dependencies to the InterpTests binary in unittests/AST/ByteCode/CMakeLists.txt . >From fe6604dc283a454c212b8e2cd21d94df96fb05d4 Mon Sep 17 00:00:00 2001 From: Reid Kleckner Date: Wed, 2 Apr 2025 18:34:05 + Subject: [PATCH] Reapply "[cmake] Refactor clang unittest cmake" This reapplies 5ffd9bdb50b57 (#133545) with fixes. The BUILD_SHARED_LIBS=ON build was fixed by adding missing LLVM dependencies to the InterpTests binary in unittests/AST/ByteCode/CMakeLists.txt . --- clang/unittests/AST/ByteCode/CMakeLists.txt | 16 -- clang/unittests/AST/CMakeLists.txt| 23 - clang/unittests/ASTMatchers/CMakeLists.txt| 22 - .../ASTMatchers/Dynamic/CMakeLists.txt| 18 +++ clang/unittests/Analysis/CMakeLists.txt | 18 +++ .../Analysis/FlowSensitive/CMakeLists.txt | 18 +++ clang/unittests/Basic/CMakeLists.txt | 18 +++ clang/unittests/CMakeLists.txt| 32 --- clang/unittests/CodeGen/CMakeLists.txt| 15 +++-- clang/unittests/CrossTU/CMakeLists.txt| 12 ++- .../unittests/DirectoryWatcher/CMakeLists.txt | 11 ++- clang/unittests/Driver/CMakeLists.txt | 19 --- clang/unittests/Format/CMakeLists.txt | 11 ++- clang/unittests/Frontend/CMakeLists.txt | 12 +++ clang/unittests/Index/CMakeLists.txt | 13 +++- clang/unittests/InstallAPI/CMakeLists.txt | 9 ++ clang/unittests/Interpreter/CMakeLists.txt| 25 +++ .../Interpreter/ExceptionTests/CMakeLists.txt | 20 ++-- clang/unittests/Lex/CMakeLists.txt| 16 +++--- clang/unittests/Rewrite/CMakeLists.txt| 10 ++ clang/unittests/Sema/CMakeLists.txt | 18 +++ clang/unittests/Serialization/CMakeLists.txt | 17 -- clang/unittests/StaticAnalyzer/CMakeLists.txt | 18 +++ clang/unittests/Support/CMakeLists.txt| 11 ++- clang/unittests/Tooling/CMakeLists.txt| 28 +++- clang/unittests/Tooling/Syntax/CMakeLists.txt | 15 +++-- clang/unittests/libclang/CMakeLists.txt | 5 +-- .../libclang/CrashTests/CMakeLists.txt| 5 +-- 28 files changed, 166 insertions(+), 289 deletions(-) diff --git a/clang/unittests/AST/ByteCode/CMakeLists.txt b/clang/unittests/AST/ByteCode/CMakeLists.txt index b862fb4834fbd..1469cd6b2a8ea 100644 --- a/clang/unittests/AST/ByteCode/CMakeLists.txt +++ b/clang/unittests/AST/ByteCode/CMakeLists.txt @@ -2,19 +2,17 @@ add_clang_unittest(InterpTests BitcastBuffer.cpp Descriptor.cpp toAPValue.cpp - ) - -clang_target_link_libraries(InterpTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic clangFrontend clangSerialization clangTooling - ) - - target_link_libraries(InterpTests - PRIVATE + LINK_LIBS clangTesting -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) diff --git a/clang/unittests/AST/CMakeLists.txt b/clang/unittests/AST/CMakeLists.txt index bfa6082a6ffa4..f27d34e8a0719 100644 --- a/clang/unittests/AST/CMakeLists.txt +++ b/clang/unittests/AST/CMakeLists.txt @@ -1,10 +1,3 @@ -set(LLVM_LINK_COMPONENTS - FrontendOpenMP - Support - TargetParser - ) - - add_subdirectory(ByteCode) add_clang_unittest(ASTTests @@ -43,10 +36,7 @@ add_clang_unittest(ASTTests TemplateNameTest.cpp TypePrinterTest.cpp UnresolvedSetTest.cpp - ) - -clang_target_link_libraries(ASTTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic @@ -54,11 +44,12 @@ clang_target_link_libraries(ASTTests clangLex clangSerialization clangTooling - ) - -target_link_libraries(ASTTests - PRIVATE + LINK_LIBS clangTesting LLVMTestingAnnotations LLVMTestingSupport -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) diff --git a/clang/unittests/ASTMatchers/CMakeLists.txt b/clang/unittests/ASTMatchers/CMakeLists.txt index 6a1e629d81b65..47bd5c108bb5a 100644 --- a/clang/unittests/ASTMatchers/CMakeLists.txt +++ b/clang/unittests/ASTMatchers/CMakeLists.txt @@ -1,31 +1,23 @@ -set(LLVM_LINK_COMPONENTS - FrontendOpenMP - Support - TargetParser - ) - add_clang_unittest(ASTMatchersTests ASTMatchersInternalTest.cpp ASTMatchersNodeTest.cpp ASTMatchersNarrowingTest.cpp ASTMatchersTraversalTest.cpp GtestMatchersTest.cpp - ) - -clang_target_link_libraries(ASTMatchersTests - PRIVATE + CLANG_LIBS clangAST clangASTMatchers clangBasic clangFrontend clangSerialization clangTooling - ) - -target_link_libraries(ASTMatchersTests - PRIVATE + LINK_LIBS clangTesting LLVMTestingSupport -) + LLVM_COMPONENTS + FrontendOpenMP + Support + TargetParser + ) add_subdirectory(Dynamic) diff -
[clang] [llvm] [EquivalenceClasses] Use SmallVector for deterministic iteration order. (PR #134075)
@@ -138,6 +139,9 @@ class EquivalenceClasses {
/// ECValues, it just keeps the key as part of the value.
std::set TheMapping;
+ /// List of all members, used to provide a determinstic iteration order.
+ SmallVector Members;
arsenm wrote:
Can you combine into SetVector?
https://github.com/llvm/llvm-project/pull/134075
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -585,6 +597,23 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned
BuiltinID,
llvm::Value *Env = EmitScalarExpr(E->getArg(0));
return Builder.CreateCall(F, {Env});
}
+ case AMDGPU::BI__builtin_amdgcn_processor_is: {
+assert(CGM.getTriple().isSPIRV() &&
jhuber6 wrote:
Yeah, we 100% should support this for normal IR. We could then replace all of
those `oclc_isa_version` files.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [HIP] Claim `--offload-compress` for `-M` (PR #133456)
https://github.com/yxsamliu updated
https://github.com/llvm/llvm-project/pull/133456
>From be4ea80f2431bff0c017df3aef9f260dddfa9ccc Mon Sep 17 00:00:00 2001
From: "Yaxun (Sam) Liu"
Date: Fri, 28 Mar 2025 11:30:39 -0400
Subject: [PATCH] [HIP] Claim --offloading-compress for -M
Cmake automatically generate dependency files with all compilation
options provided by users. When users use for
HIP compilation, it causes warnings when cmake generates dependency
files. Claim this option to supress warnings.
---
clang/lib/Driver/ToolChains/Clang.cpp | 6 ++
clang/test/Driver/hip-options.hip | 4
2 files changed, 10 insertions(+)
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp
b/clang/lib/Driver/ToolChains/Clang.cpp
index 5f45cf0865b9e..a09b158b75b27 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -1026,6 +1026,12 @@ void Clang::AddPreprocessingOptions(Compilation &C,
const JobAction &JA,
CmdArgs.push_back("-dependency-file");
CmdArgs.push_back(DepFile);
}
+// Cmake generates dependency files using all compilation options specified
+// by users. Claim those not used for dependency files.
+if (JA.isOffloading(Action::OFK_HIP)) {
+ Args.ClaimAllArgs(options::OPT_offload_compress);
+ Args.ClaimAllArgs(options::OPT_no_offload_compress);
+}
bool HasTarget = false;
for (const Arg *A : Args.filtered(options::OPT_MT, options::OPT_MQ)) {
diff --git a/clang/test/Driver/hip-options.hip
b/clang/test/Driver/hip-options.hip
index 0aabc8ad41904..29d23c1b6c8d9 100644
--- a/clang/test/Driver/hip-options.hip
+++ b/clang/test/Driver/hip-options.hip
@@ -242,3 +242,7 @@
// NO-WARN-ATOMIC: clang{{.*}} "-triple" "amdgcn-amd-amdhsa" {{.*}}
"-Werror=atomic-alignment" {{.*}} "-Wno-error=atomic-alignment"
// NO-WARN-ATOMIC-NOT: clang{{.*}} "-triple" "x86_64-unknown-linux-gnu" {{.*}}
"-Werror=atomic-alignment"
// NO-WARN-ATOMIC-NOT: clang{{.*}} "-triple" "x86_64-unknown-linux-gnu" {{.*}}
"-Wno-error=atomic-alignment"
+
+// Check --offload-compress does not cause warning.
+// RUN: %clang -### -Werror --target=x86_64-unknown-linux-gnu -nogpuinc
-nogpulib \
+// RUN: --offload-arch=gfx1100 --offload-compress --offload-host-only -M %s
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
https://github.com/asudarsa updated
https://github.com/llvm/llvm-project/pull/133967
>From 0ab40315cd5ca320d84f87c99b8f23d9b6e8ed36 Mon Sep 17 00:00:00 2001
From: Arvind Sudarsanam
Date: Mon, 31 Mar 2025 14:31:24 -0700
Subject: [PATCH 1/5] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V
backend support inside clang-sycl-linker
This PR does the following:
1. Use SPIR-V backend to do LLVM to SPIR-V translation inside clang-sycl-linker
2. Remove llvm-spirv translator from clang-sycl-linker
Currently, no SPIR-V extensions are enabled for SYCL compilation flow. This
will be updated in subsequent commits.
Thanks
Signed-off-by: Arvind Sudarsanam
---
clang/test/Driver/clang-sycl-linker-test.cpp | 14 +-
clang/test/Driver/sycl-link-spirv-target.cpp | 6 +-
clang/tools/clang-sycl-linker/CMakeLists.txt | 5 +-
.../clang-sycl-linker/ClangSYCLLinker.cpp | 187 +++---
clang/tools/clang-sycl-linker/SYCLLinkOpts.td | 2 +-
5 files changed, 76 insertions(+), 138 deletions(-)
diff --git a/clang/test/Driver/clang-sycl-linker-test.cpp
b/clang/test/Driver/clang-sycl-linker-test.cpp
index 729561bd09cd8..2f860ae74e97d 100644
--- a/clang/test/Driver/clang-sycl-linker-test.cpp
+++ b/clang/test/Driver/clang-sycl-linker-test.cpp
@@ -6,7 +6,7 @@
// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc -o
a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=SIMPLE-FO
// SIMPLE-FO: sycl-device-link: inputs: {{.*}}.bc, {{.*}}.bc libfiles:
output: [[LLVMLINKOUT:.*]].bc
-// SIMPLE-FO-NEXT: "{{.*}}llvm-spirv{{.*}}" {{.*}}-o a.spv [[LLVMLINKOUT]].bc
+// SIMPLE-FO-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv
//
// Test the dry run of a simple case with device library files specified.
// RUN: touch %T/lib1.bc
@@ -14,7 +14,7 @@
// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc
--library-path=%T --device-libs=lib1.bc,lib2.bc -o a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=DEVLIBS
// DEVLIBS: sycl-device-link: inputs: {{.*}}.bc libfiles: {{.*}}lib1.bc,
{{.*}}lib2.bc output: [[LLVMLINKOUT:.*]].bc
-// DEVLIBS-NEXT: "{{.*}}llvm-spirv{{.*}}" {{.*}}-o a.spv [[LLVMLINKOUT]].bc
+// DEVLIBS-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv
//
// Test a simple case with a random file (not bitcode) as input.
// RUN: touch %t.o
@@ -29,13 +29,3 @@
// RUN: not clang-sycl-linker --dry-run -triple=spirv64 %t_1.bc %t_2.bc
--library-path=%T --device-libs=lib1.bc,lib2.bc,lib3.bc -o a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=DEVLIBSERR2
// DEVLIBSERR2: '{{.*}}lib3.bc' SYCL device library file is not found
-//
-// Test if correct set of llvm-spirv options are emitted for windows
environment.
-// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 --is-windows-msvc-env
%t_1.bc %t_2.bc -o a.spv 2>&1 \
-// RUN: | FileCheck %s --check-prefix=LLVMOPTSWIN
-// LLVMOPTSWIN: -spirv-debug-info-version=ocl-100
-spirv-allow-extra-diexpressions -spirv-allow-unknown-intrinsics=llvm.genx.
-spirv-ext=
-//
-// Test if correct set of llvm-spirv options are emitted for linux environment.
-// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc -o
a.spv 2>&1 \
-// RUN: | FileCheck %s --check-prefix=LLVMOPTSLIN
-// LLVMOPTSLIN: -spirv-debug-info-version=nonsemantic-shader-200
-spirv-allow-unknown-intrinsics=llvm.genx. -spirv-ext=
diff --git a/clang/test/Driver/sycl-link-spirv-target.cpp
b/clang/test/Driver/sycl-link-spirv-target.cpp
index 7585ef8b14a59..586adae619165 100644
--- a/clang/test/Driver/sycl-link-spirv-target.cpp
+++ b/clang/test/Driver/sycl-link-spirv-target.cpp
@@ -3,7 +3,7 @@
//
// Test that -Xlinker options are being passed to clang-sycl-linker.
// RUN: touch %t.bc
-// RUN: %clangxx -### --target=spirv64 --sycl-link -Xlinker
--llvm-spirv-path=/tmp \
-// RUN: -Xlinker -triple=spirv64 -Xlinker --library-path=/tmp -Xlinker
--device-libs=lib1.bc,lib2.bc %t.bc 2>&1 \
+// RUN: %clangxx -### --target=spirv64 --sycl-link -Xlinker -triple=spirv64
-Xlinker --library-path=/tmp \
+// RUN: -Xlinker --device-libs=lib1.bc,lib2.bc %t.bc 2>&1 \
// RUN: | FileCheck %s -check-prefix=XLINKEROPTS
-// XLINKEROPTS: "{{.*}}clang-sycl-linker{{.*}}" "--llvm-spirv-path=/tmp"
"-triple=spirv64" "--library-path=/tmp" "--device-libs=lib1.bc,lib2.bc"
"{{.*}}.bc" "-o" "a.out"
+// XLINKEROPTS: "{{.*}}clang-sycl-linker{{.*}}" "-triple=spirv64"
"--library-path=/tmp" "--device-libs=lib1.bc,lib2.bc" "{{.*}}.bc" "-o" "a.out"
diff --git a/clang/tools/clang-sycl-linker/CMakeLists.txt
b/clang/tools/clang-sycl-linker/CMakeLists.txt
index 382c0ca441940..ee89e8b0a5570 100644
--- a/clang/tools/clang-sycl-linker/CMakeLists.txt
+++ b/clang/tools/clang-sycl-linker/CMakeLists.txt
@@ -1,14 +1,17 @@
set(LLVM_LINK_COMPONENTS
${LLVM_TARGETS_TO_BUILD}
+ Analysis
BinaryFormat
BitWriter
Core
IRReader
Linker
+ MC
Option
Object
- TargetParser
Support
+ Target
+ TargetParser
)
set(LLVM_TARGET_DEFIN
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
erichkeane wrote: > This change adds two semi-magical builtins for AMDGPU: > > * `__builtin_amdgcn_processor_is`, which is similar in observable > behaviour with `__builtin_cpu_is`, except that it is never "evaluated" at run > time; > > * `__builtin_amdgcn_is_invocable`, which is behaviourally similar with > `__has_builtin`, except that it is not a macro (i.e. not evaluated at > preprocessing time). > > > Neither of these are `constexpr`, even though when compiling for concrete > (i.e. `gfxXXX` / `gfxXXX-generic`) targets they get evaluated in Clang, so > they shouldn't tear the AST too badly / at all for multi-pass compilation > cases like HIP. They can only be used in specific contexts (as args to > control structures). > > The motivation for adding these is two-fold: > > * as a nice to have, it provides an AST-visible way to incorporate > architecture specific code, rather than having to rely on macros and the > preprocessor, which burn in the choice quite early; > > * as a must have, it allows featureful AMDGCN flavoured SPIR-V to be > produced, where target specific capability is guarded and chosen or discarded > when finalising compilation for a concrete target. > > > I've tried to keep the overall footprint of the change small. The changes to > Sema are a bit unpleasant, but there was a strong desire to have Clang > validate these, and to constrain their uses, and this was the most compact > solution I could come up with (suggestions welcome). > > In the end, I will note there is nothing that is actually AMDGPU specific > here, so it is possible that in the future, assuming interests from other > targets / users, we'd just promote them to generic intrinsics. First read through this, I find myself wondering WHY these aren't constexpr. They seem exactly the sort of thing that folks would like to use `if constexpr` for. https://github.com/llvm/llvm-project/pull/134016 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] 2b7daaf - [sanitizer][CFI] Add support to build CFI with sanitize-coverage (#131296)
Author: Maxim Zhukov
Date: 2025-04-02T16:05:44+03:00
New Revision: 2b7daaf9678181959982b219db0af106f4ef8e3e
URL:
https://github.com/llvm/llvm-project/commit/2b7daaf9678181959982b219db0af106f4ef8e3e
DIFF:
https://github.com/llvm/llvm-project/commit/2b7daaf9678181959982b219db0af106f4ef8e3e.diff
LOG: [sanitizer][CFI] Add support to build CFI with sanitize-coverage (#131296)
Added ability to build together with -fsanitize=cfi and
-fsanitize-coverage=trace-cmp at the same time.
Added:
Modified:
clang/lib/Driver/SanitizerArgs.cpp
clang/test/CodeGen/sanitize-coverage.c
clang/test/Driver/fsanitize-coverage.c
Removed:
diff --git a/clang/lib/Driver/SanitizerArgs.cpp
b/clang/lib/Driver/SanitizerArgs.cpp
index 6e75001585c61..3c7cd562a14e3 100644
--- a/clang/lib/Driver/SanitizerArgs.cpp
+++ b/clang/lib/Driver/SanitizerArgs.cpp
@@ -54,7 +54,8 @@ static const SanitizerMask SupportsCoverage =
SanitizerKind::FuzzerNoLink | SanitizerKind::FloatDivideByZero |
SanitizerKind::SafeStack | SanitizerKind::ShadowCallStack |
SanitizerKind::Thread | SanitizerKind::ObjCCast | SanitizerKind::KCFI |
-SanitizerKind::NumericalStability | SanitizerKind::Vptr;
+SanitizerKind::NumericalStability | SanitizerKind::Vptr |
+SanitizerKind::CFI;
static const SanitizerMask RecoverableByDefault =
SanitizerKind::Undefined | SanitizerKind::Integer |
SanitizerKind::ImplicitConversion | SanitizerKind::Nullability |
diff --git a/clang/test/CodeGen/sanitize-coverage.c
b/clang/test/CodeGen/sanitize-coverage.c
index cb1ed939d4a93..90892fc0dae06 100644
--- a/clang/test/CodeGen/sanitize-coverage.c
+++ b/clang/test/CodeGen/sanitize-coverage.c
@@ -5,6 +5,7 @@
// RUN: %clang %s -target x86_64-unknown-linux-gnu -emit-llvm -S
-fsanitize=thread -fsanitize-coverage=trace-pc,trace-cmp -o - | FileCheck
%s --check-prefixes=CHECK,TSAN
// RUN: %clang %s -target x86_64-unknown-linux-gnu -emit-llvm -S
-fsanitize=undefined -fsanitize-coverage=trace-pc,trace-cmp -o - | FileCheck
%s --check-prefixes=CHECK,UBSAN
// RUN: %clang %s -target x86_64-unknown-linux-gnu -emit-llvm -S
-fsanitize=kcfi -fsanitize-coverage=trace-pc,trace-cmp -o - | FileCheck
%s --check-prefixes=CHECK,KCFI
+// RUN: %clang %s -target x86_64-unknown-linux-gnu -emit-llvm -S
-fsanitize=cfi-fsanitize-coverage=trace-pc,trace-cmp -flto
-fvisibility=default -fno-sanitize-trap=cfi -fno-sanitize-ignorelist
-resource-dir=/dev/null -o - | FileCheck %s --check-prefixes=CHECK,CFI
int x[10];
extern void (*f)(void);
@@ -21,6 +22,7 @@ void foo(int n) {
if (n)
x[n] = 42;
// KCFI-DAG: call void %[[#]]() [ "kcfi"(i32 {{.*}}) ]
+ // CFI-DAG: call void @__ubsan_handle_cfi_check_fail_abort
f();
}
@@ -47,6 +49,7 @@ __attribute__((no_sanitize("coverage"))) void
test_no_sanitize_coverage(int n) {
if (n)
x[n] = 42;
// KCFI-DAG: call void %[[#]]() [ "kcfi"(i32 {{.*}}) ]
+ // CFI-DAG: call void @__ubsan_handle_cfi_check_fail_abort
f();
}
@@ -94,6 +97,14 @@ void test_no_sanitize_kcfi(void) {
f();
}
+// CHECK-LABEL: define dso_local void @test_no_sanitize_cfi(
+__attribute__((no_sanitize("cfi", "coverage")))
+void test_no_sanitize_cfi(void) {
+ // CHECK-NOT: call void @__sanitizer_cov_trace
+ // CFI-NOT: call void @__ubsan_handle_cfi_check_fail_abort
+ f();
+}
+
// CHECK-LABEL: define dso_local void @test_no_sanitize_always_inline(
__attribute__((no_sanitize("coverage")))
void test_no_sanitize_always_inline(int n) {
diff --git a/clang/test/Driver/fsanitize-coverage.c
b/clang/test/Driver/fsanitize-coverage.c
index c2de897f80eeb..dc4c39396d45c 100644
--- a/clang/test/Driver/fsanitize-coverage.c
+++ b/clang/test/Driver/fsanitize-coverage.c
@@ -17,6 +17,7 @@
// RUN: %clang --target=x86_64-linux-gnu -fsanitize=dataflow
-fsanitize-coverage=func,trace-pc %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
// RUN: %clang --target=x86_64-linux-gnu -fsanitize=thread
-fsanitize-coverage=func,trace-pc %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
// RUN: %clang --target=x86_64-linux-gnu -fsanitize=kcfi
-fsanitize-coverage=func,trace-pc %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
+// RUN: %clang --target=x86_64-linux-gnu -fsanitize=cfi
-fsanitize-coverage=func,trace-pc -flto -fvisibility=default
-fno-sanitize-ignorelist -resource-dir=/dev/null %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
// RUN: %clang --target=%itanium_abi_triple -fsanitize=float-divide-by-zero
-fsanitize-coverage=func,trace-pc %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
// RUN: %clang --target=x86_64-linux-gnu
-fsanitize-coverage=func,trace-pc %s -### 2>&1 | FileCheck %s
--check-prefix=CHECK-SANITIZE-COVERAGE-FUNC
// CHECK-SANITIZE-COVERAGE-FUNC: fsanitize-coverage-type=1
[clang] [sanitizer][CFI] Add support to build CFI with sanitize-coverage (PR #131296)
github-actions[bot] wrote: @Mephistophiles Congratulations on having your first Pull Request (PR) merged into the LLVM Project! Your changes will be combined with recent changes from other authors, then tested by our [build bots](https://lab.llvm.org/buildbot/). If there is a problem with a build, you may receive a report in an email or a comment on this PR. Please check whether problems have been caused by your change specifically, as the builds can include changes from many authors. It is not uncommon for your change to be included in a build that fails due to someone else's changes, or infrastructure issues. How to do this, and the rest of the post-merge process, is covered in detail [here](https://llvm.org/docs/MyFirstTypoFix.html#myfirsttypofix-issues-after-landing-your-pr). If your change does cause a problem, it may be reverted, or you can revert it yourself. This is a normal part of [LLVM development](https://llvm.org/docs/DeveloperPolicy.html#patch-reversion-policy). You can fix your changes and open a new PR to merge them again. If you don't get any reports, no action is required from you. Your changes are working as expected, well done! https://github.com/llvm/llvm-project/pull/131296 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -0,0 +1,64 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --check-globals all --version 5 +// RUN: %clang_cc1 -triple amdgcn-amd-amdhsa -target-cpu gfx900 -emit-llvm %s -o - | FileCheck --check-prefix=AMDGCN-GFX900 %s +// RUN: %clang_cc1 -triple amdgcn-amd-amdhsa -target-cpu gfx1010 -emit-llvm %s -o - | FileCheck --check-prefix=AMDGCN-GFX1010 %s +// RUN: %clang_cc1 -triple spirv64-amd-amdhsa -emit-llvm %s -o - | FileCheck --check-prefix=AMDGCNSPIRV %s + yxsamliu wrote: we also need to test the code does not cause error in host compilation, i.e. -triple x86_64 -aux-triple amdgcn or spirv64-amd-amdhsa. we can either do this here or create a sema test. same for below. https://github.com/llvm/llvm-project/pull/134016 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [flang][OpenMP] Extend `do concurrent` mapping to multi-range loops (PR #127634)
https://github.com/ergawy edited https://github.com/llvm/llvm-project/pull/127634 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [ARM][Clang] Make `+nosimd` functional for AArch32 Targets (PR #130623)
https://github.com/Stylie777 updated
https://github.com/llvm/llvm-project/pull/130623
>From 8b2ad7c1475ff4aee065e9feb21469d184320472 Mon Sep 17 00:00:00 2001
From: Jack Styles
Date: Fri, 7 Mar 2025 15:51:34 +
Subject: [PATCH 1/8] [NFC][ARM] Split SIMD identifier away from MVE
Previously, the use of MVE or MVE.FP would be defined
by using the `ARM::AEK_SIMD` identifier. SIMD relates to
the Cortex-A and Cortex-R extension that enables NEON
instructions, which is called MVE for Cortex-M. To enable the
linking of `+simd` and `+nosimd` to `+neon` and `-neon`
when using clang, MVE and MVE.FP can now be defined
using a unique enum identifier, rather than using the existing
identifier for SIMD.
This was originally planned to be merged as part of #130296 but
the changes made `+nosimd` an invalid argument, which, while not
having any functionality, was allowed in previous versions of
LLVM. To avoid regressions being introduced, this has been
combined with the fix for `+nosimd` on AArch32.
---
llvm/include/llvm/TargetParser/ARMTargetParser.def | 10 +-
llvm/include/llvm/TargetParser/ARMTargetParser.h | 1 +
llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp | 2 +-
llvm/lib/Target/ARM/MCTargetDesc/ARMTargetStreamer.cpp | 2 +-
llvm/unittests/TargetParser/TargetParserTest.cpp | 6 +++---
5 files changed, 11 insertions(+), 10 deletions(-)
diff --git a/llvm/include/llvm/TargetParser/ARMTargetParser.def
b/llvm/include/llvm/TargetParser/ARMTargetParser.def
index 6b96c3e83c8c4..e515ab665d117 100644
--- a/llvm/include/llvm/TargetParser/ARMTargetParser.def
+++ b/llvm/include/llvm/TargetParser/ARMTargetParser.def
@@ -224,8 +224,8 @@ ARM_ARCH_EXT_NAME("dotprod", ARM::AEK_DOTPROD, "+dotprod",
"-dotprod")
ARM_ARCH_EXT_NAME("dsp", ARM::AEK_DSP, "+dsp", "-dsp")
ARM_ARCH_EXT_NAME("fp", ARM::AEK_FP, {}, {})
ARM_ARCH_EXT_NAME("fp.dp", ARM::AEK_FP_DP, {}, {})
-ARM_ARCH_EXT_NAME("mve", (ARM::AEK_DSP | ARM::AEK_SIMD), "+mve", "-mve")
-ARM_ARCH_EXT_NAME("mve.fp", (ARM::AEK_DSP | ARM::AEK_SIMD | ARM::AEK_FP),
+ARM_ARCH_EXT_NAME("mve", (ARM::AEK_DSP | ARM::AEK_MVE), "+mve", "-mve")
+ARM_ARCH_EXT_NAME("mve.fp", (ARM::AEK_DSP | ARM::AEK_MVE | ARM::AEK_FP),
"+mve.fp", "-mve.fp")
ARM_ARCH_EXT_NAME("idiv", (ARM::AEK_HWDIVARM | ARM::AEK_HWDIVTHUMB), {}, {})
ARM_ARCH_EXT_NAME("mp", ARM::AEK_MP, {}, {})
@@ -345,12 +345,12 @@ ARM_CPU_NAME("cortex-m33", ARMV8MMainline,
FK_FPV5_SP_D16, false, ARM::AEK_DSP)
ARM_CPU_NAME("star-mc1", ARMV8MMainline, FK_FPV5_SP_D16, false, ARM::AEK_DSP)
ARM_CPU_NAME("cortex-m35p", ARMV8MMainline, FK_FPV5_SP_D16, false,
ARM::AEK_DSP)
ARM_CPU_NAME("cortex-m55", ARMV8_1MMainline, FK_FP_ARMV8_FULLFP16_D16, false,
- (ARM::AEK_DSP | ARM::AEK_SIMD | ARM::AEK_FP | ARM::AEK_FP16))
+ (ARM::AEK_DSP | ARM::AEK_MVE | ARM::AEK_FP | ARM::AEK_FP16))
ARM_CPU_NAME("cortex-m85", ARMV8_1MMainline, FK_FP_ARMV8_FULLFP16_D16, false,
- (ARM::AEK_DSP | ARM::AEK_SIMD | ARM::AEK_FP | ARM::AEK_FP16 |
+ (ARM::AEK_DSP | ARM::AEK_MVE | ARM::AEK_FP | ARM::AEK_FP16 |
ARM::AEK_RAS | ARM::AEK_PACBTI))
ARM_CPU_NAME("cortex-m52", ARMV8_1MMainline, FK_FP_ARMV8_FULLFP16_D16, false,
- (ARM::AEK_DSP | ARM::AEK_SIMD | ARM::AEK_FP | ARM::AEK_FP16 |
+ (ARM::AEK_DSP | ARM::AEK_MVE | ARM::AEK_FP | ARM::AEK_FP16 |
ARM::AEK_RAS | ARM::AEK_PACBTI))
ARM_CPU_NAME("cortex-a32", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
ARM::AEK_CRC)
ARM_CPU_NAME("cortex-a35", ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
ARM::AEK_CRC)
diff --git a/llvm/include/llvm/TargetParser/ARMTargetParser.h
b/llvm/include/llvm/TargetParser/ARMTargetParser.h
index 5dbcfd3d2d693..b2403f42f1b79 100644
--- a/llvm/include/llvm/TargetParser/ARMTargetParser.h
+++ b/llvm/include/llvm/TargetParser/ARMTargetParser.h
@@ -61,6 +61,7 @@ enum ArchExtKind : uint64_t {
AEK_CDECP6 = 1 << 28,
AEK_CDECP7 = 1 << 29,
AEK_PACBTI = 1 << 30,
+ AEK_MVE = 1ULL << 31,
// Unsupported extensions.
AEK_OS = 1ULL << 59,
AEK_IWMMXT = 1ULL << 60,
diff --git a/llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp
b/llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp
index 456c3b5396fc9..956fc9680ee33 100644
--- a/llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp
+++ b/llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp
@@ -12970,7 +12970,7 @@ bool ARMAsmParser::enableArchExtFeature(StringRef Name,
SMLoc &ExtLoc) {
{ARM::AEK_CRYPTO,
{Feature_HasV8Bit},
{ARM::FeatureCrypto, ARM::FeatureNEON, ARM::FeatureFPARMv8}},
- {(ARM::AEK_DSP | ARM::AEK_SIMD | ARM::AEK_FP),
+ {(ARM::AEK_DSP | ARM::AEK_MVE | ARM::AEK_FP),
{Feature_HasV8_1MMainlineBit},
{ARM::HasMVEFloatOps}},
{ARM::AEK_FP,
diff --git a/llvm/lib/Target/ARM/MCTargetDesc/ARMTargetStreamer.cpp
b/llvm/lib/Target/ARM/MCTargetDesc/ARMTargetStreamer.cpp
index b0fa03a35ec04..632dbebf58f04 100644
--- a/llvm/lib/Target/ARM/MCTargetDesc/ARMTargetStreamer.cpp
+++ b/
[clang] [llvm] [EquivalenceClasses] Use SmallVector for deterministic iteration order. (PR #134075)
https://github.com/nikic approved this pull request. This looks reasonable to me. We could keep the existing API with a custom iterator, but given the limited amount of uses this seems ok. https://github.com/llvm/llvm-project/pull/134075 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[libclc] [libclc] Move cbrt to the CLC library; vectorize (PR #133940)
https://github.com/frasercrmck closed https://github.com/llvm/llvm-project/pull/133940 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [EquivalenceClasses] Use SmallVector for deterministic iteration order. (PR #134075)
@@ -138,6 +139,9 @@ class EquivalenceClasses {
/// ECValues, it just keeps the key as part of the value.
std::set TheMapping;
+ /// List of all members, used to provide a determinstic iteration order.
+ SmallVector Members;
fhahn wrote:
This is where things get a little tricky. `ECValue` contain pointers to other
`ECValue`'s (the leader of the class and the next member of the class), so
using something like `SetVector` as the backing storage won't work I think, as
it might reallocate and move the objects.
One thing I tried is using a BumpAllocator for the elements in the class and a
separate DenseMap instead of the std::set:
https://github.com/llvm/llvm-project/commit/5c13864badfd7170c9c7c0a908e3cba8aa0f4fea
I think SetVector (or MapVector) would be slightly worse here, as it stores an
ECValue in both a vector and a denseSet, and ECValue contains 3 pointers, so it
would require quite a bit of extra memory .
The patch could use a `DenseSet`, it just requires a custom DenseMapInfo for
ECValue to just hash/compare the data part. I can do that if desired and
prepare the patch for review
Compile-time wise this looks neutral, but it would simplify things slightly
https://llvm-compile-time-tracker.com/compare.php?from=b0c2ac67a88d3ef86987e2f82115ea0170675a17&to=5c13864badfd7170c9c7c0a908e3cba8aa0f4fea&stat=instructions:u
(the patch just has the unit tests for EQClass removed, because one of the keys
doesn't have DenseMapInfo).
https://github.com/llvm/llvm-project/pull/134075
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
mjklemm wrote: > Should the `nowait(EXPR)` behavior be guarded by the OpenMP version flag? It > seems to me that the tests would currently not respect which OpenMP version > is selected and always generate the error. I agree. This should be under `-fopenmp-version=6.0` https://github.com/llvm/llvm-project/pull/128742 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [EquivalenceClasses] Use SmallVector for deterministic iteration order. (PR #134075)
https://github.com/fhahn updated
https://github.com/llvm/llvm-project/pull/134075
>From 10afe2fb4743c0fccaad2797fffca174736a0997 Mon Sep 17 00:00:00 2001
From: Florian Hahn
Date: Sat, 29 Mar 2025 20:20:39 +
Subject: [PATCH 1/2] [EquivalenceClasses] Use SmallVector for deterministic
iteration order.
Currently iterators over EquivalenceClasses will iterate over std::set,
which guarantees the order specified by the comperator. Unfortunately in
many cases, EquivalenceClasses are used with pointers, so iterating over
std::set of pointers will not be deterministic across runs.
There are multiple places that explicitly try to sort the equivalence
classes before using them to try to get a deterministic order
(LowerTypeTests, SplitModule), but there are others that do not at the
moment and this can result at least in non-determinstic value naming in
Float2Int.
This patch updates EquivalenceClasses to keep track of all members via a
extra SmallVector and removes code from LowerTypeTests and SplitModule
to sort the classes before processing.
Overall it looks like compile-time slightly decreases in most cases, but
close to noise:
https://llvm-compile-time-tracker.com/compare.php?from=7d441d9892295a6eb8aaf481e1715f039f6f224f&to=b0c2ac67a88d3ef86987e2f82115ea0170675a17&stat=instructions
---
.../FlowSensitive/SimplifyConstraints.cpp | 8 ++---
llvm/include/llvm/ADT/EquivalenceClasses.h| 25 +--
llvm/lib/Analysis/VectorUtils.cpp | 8 ++---
.../AArch64/AArch64A57FPLoadBalancing.cpp | 4 +--
llvm/lib/Target/AMDGPU/AMDGPUSplitModule.cpp | 5 +--
llvm/lib/Transforms/IPO/LowerTypeTests.cpp| 29 +++--
llvm/lib/Transforms/Scalar/Float2Int.cpp | 10 +++---
llvm/lib/Transforms/Scalar/LoopDistribute.cpp | 6 ++--
llvm/lib/Transforms/Utils/SplitModule.cpp | 32 +--
.../type-inference.td | 2 +-
.../LowerTypeTests/function-disjoint.ll | 12 +++
.../Transforms/LowerTypeTests/nonstring.ll| 2 +-
llvm/test/tools/llvm-split/preserve-locals.ll | 13
llvm/test/tools/llvm-split/scc-const-alias.ll | 10 +++---
.../test/tools/llvm-split/scc-global-alias.ll | 16 +-
.../TableGen/GlobalISelCombinerEmitter.cpp| 3 +-
16 files changed, 81 insertions(+), 104 deletions(-)
diff --git a/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
b/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
index 69a90334c9df5..3c385ed8ef663 100644
--- a/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
+++ b/clang/lib/Analysis/FlowSensitive/SimplifyConstraints.cpp
@@ -111,8 +111,8 @@ void simplifyConstraints(llvm::SetVector
&Constraints,
FalseAtoms = projectToLeaders(FalseAtoms, EquivalentAtoms);
llvm::DenseMap Substitutions;
-for (auto It = EquivalentAtoms.begin(); It != EquivalentAtoms.end(); ++It)
{
- Atom TheAtom = It->getData();
+for (const auto &E : EquivalentAtoms) {
+ Atom TheAtom = E->getData();
Atom Leader = EquivalentAtoms.getLeaderValue(TheAtom);
if (TrueAtoms.contains(Leader)) {
if (FalseAtoms.contains(Leader)) {
@@ -152,9 +152,9 @@ void simplifyConstraints(llvm::SetVector
&Constraints,
if (Info) {
for (const auto &E : EquivalentAtoms) {
- if (!E.isLeader())
+ if (!E->isLeader())
continue;
- Atom At = *EquivalentAtoms.findLeader(E);
+ Atom At = *EquivalentAtoms.findLeader(*E);
if (TrueAtoms.contains(At) || FalseAtoms.contains(At))
continue;
llvm::SmallVector Atoms =
diff --git a/llvm/include/llvm/ADT/EquivalenceClasses.h
b/llvm/include/llvm/ADT/EquivalenceClasses.h
index 46f186a71f5ce..906971baf74af 100644
--- a/llvm/include/llvm/ADT/EquivalenceClasses.h
+++ b/llvm/include/llvm/ADT/EquivalenceClasses.h
@@ -15,6 +15,7 @@
#ifndef LLVM_ADT_EQUIVALENCECLASSES_H
#define LLVM_ADT_EQUIVALENCECLASSES_H
+#include "llvm/ADT/SmallVector.h"
#include
#include
#include
@@ -138,6 +139,9 @@ class EquivalenceClasses {
/// ECValues, it just keeps the key as part of the value.
std::set TheMapping;
+ /// List of all members, used to provide a determinstic iteration order.
+ SmallVector Members;
+
public:
EquivalenceClasses() = default;
EquivalenceClasses(const EquivalenceClasses &RHS) {
@@ -146,9 +150,10 @@ class EquivalenceClasses {
EquivalenceClasses &operator=(const EquivalenceClasses &RHS) {
TheMapping.clear();
+Members.clear();
for (const auto &E : RHS)
- if (E.isLeader()) {
-member_iterator MI = RHS.member_begin(E);
+ if (E->isLeader()) {
+member_iterator MI = RHS.member_begin(*E);
member_iterator LeaderIt = member_begin(insert(*MI));
for (++MI; MI != member_end(); ++MI)
unionSets(LeaderIt, member_begin(insert(*MI)));
@@ -161,11 +166,10 @@ class EquivalenceClasses {
//
/// iterator* - Provides a way to iterate over all values in the set.
- using iterator =
- type
[libclc] [libclc]: clspv: add a dummy implememtation for mul_hi (PR #134094)
rjodinchr wrote: For the record, clspv implementation is based on OpMulExtended which is SPIR-V specific and not accessible from C. https://github.com/llvm/llvm-project/pull/134094 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Allow parentheses around CTAD declarators (PR #132829)
https://github.com/offsetof updated
https://github.com/llvm/llvm-project/pull/132829
>From 66833f41d26a6c6355ef67b2f9041ba771b690b7 Mon Sep 17 00:00:00 2001
From: offsetof
Date: Mon, 24 Mar 2025 20:51:23 +
Subject: [PATCH 1/3] [clang] Allow parentheses around CTAD declarators
---
clang/docs/ReleaseNotes.rst | 2 ++
.../clang/Basic/DiagnosticSemaKinds.td| 7 +++--
clang/lib/Sema/SemaType.cpp | 9 ++-
...xx1z-class-template-argument-deduction.cpp | 8 +++---
clang/test/SemaCXX/ctad.cpp | 26 +++
5 files changed, 37 insertions(+), 15 deletions(-)
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f919b66dd0e41..24c65dbef5869 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -358,6 +358,8 @@ Bug Fixes to C++ Support
- Fixed a Clang regression in C++20 mode where unresolved dependent call
expressions were created inside non-dependent contexts (#GH122892)
- Clang now emits the ``-Wunused-variable`` warning when some structured
bindings are unused
and the ``[[maybe_unused]]`` attribute is not applied. (#GH125810)
+- Declarations using class template argument deduction with redundant
+ parentheses around the declarator are no longer rejected. (#GH39811)
Bug Fixes to AST Handling
^
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index c77cde297dc32..a19975f6d6bdf 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -2612,10 +2612,9 @@ def err_decltype_auto_initializer_list : Error<
"cannot deduce 'decltype(auto)' from initializer list">;
// C++17 deduced class template specialization types
-def err_deduced_class_template_compound_type : Error<
- "cannot %select{form pointer to|form reference to|form array of|"
- "form function returning|use parentheses when declaring variable with}0 "
- "deduced class template specialization type">;
+def err_deduced_class_template_compound_type
+: Error<"cannot form %select{pointer to|reference to|array of|function "
+"returning}0 deduced class template specialization type">;
def err_deduced_non_class_or_alias_template_specialization_type : Error<
"%select{|function template|variable template|alias template|"
"template template parameter|concept|template}0 %1 requires template "
diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp
index aec33303780a0..16315d93ef8ce 100644
--- a/clang/lib/Sema/SemaType.cpp
+++ b/clang/lib/Sema/SemaType.cpp
@@ -4282,8 +4282,8 @@ static TypeSourceInfo
*GetFullTypeForDeclarator(TypeProcessingState &state,
// If T is 'decltype(auto)', the only declarators we can have are parens
// and at most one function declarator if this is a function declaration.
- // If T is a deduced class template specialization type, we can have no
- // declarator chunks at all.
+ // If T is a deduced class template specialization type, only parentheses
+ // are allowed.
if (auto *DT = T->getAs()) {
const AutoType *AT = T->getAs();
bool IsClassTemplateDeduction = isa(DT);
@@ -4297,11 +4297,6 @@ static TypeSourceInfo
*GetFullTypeForDeclarator(TypeProcessingState &state,
unsigned DiagKind = 0;
switch (DeclChunk.Kind) {
case DeclaratorChunk::Paren:
- // FIXME: Rejecting this is a little silly.
- if (IsClassTemplateDeduction) {
-DiagKind = 4;
-break;
- }
continue;
case DeclaratorChunk::Function: {
if (IsClassTemplateDeduction) {
diff --git a/clang/test/Parser/cxx1z-class-template-argument-deduction.cpp
b/clang/test/Parser/cxx1z-class-template-argument-deduction.cpp
index a1594333abae7..d29eed40b1864 100644
--- a/clang/test/Parser/cxx1z-class-template-argument-deduction.cpp
+++ b/clang/test/Parser/cxx1z-class-template-argument-deduction.cpp
@@ -39,7 +39,7 @@ namespace template_template_arg_pack {
template struct YP {};
struct Z { template struct Q {}; }; // expected-note 2{{here}}
-
+
template using ZId = Z;
template struct A {
@@ -152,7 +152,7 @@ namespace decl {
A a;
A b = 0;
const A c = 0;
- A (parens) = 0; // expected-error {{cannot use parentheses when declaring
variable with deduced class template specialization type}}
+ A (parens) = 0;
A *p = 0; // expected-error {{cannot form pointer to deduced class template
specialization type}}
A &r = *p; // expected-error {{cannot form reference to deduced class
template specialization type}}
A arr[3] = 0; // expected-error {{cannot form array of deduced class
template specialization type}}
@@ -179,7 +179,7 @@ namespace typename_specifier {
}
typename ::A a = 0;
const typename ::A b = 0;
- typename ::A (parens) = 0; // expected-error {{cannot use parentheses when
declaring variable with de
[clang] Hlsl dst function (PR #133828)
@@ -0,0 +1,37 @@
+// RUN: %clang_cc1 -finclude-default-header -triple
dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only
-disable-llvm-passes -verify
+
+float4 test_too_many_arg(float4 p0)
+{
+dst(p0, p0, p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+}
+
+float4 test_no_second_arg(float4 p0)
+{
+return dst(p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+}
+
+float4 test_no_args()
+{
+return dst();
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+}
+
+float4 test_3_components(float3 p0, float3 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector<[...], 3>' to 'vector<[...], 4>' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+}
farzonl wrote:
I think I resolved the last comment on this to quickly thinking the tests were
added. I copied your implementation into godbolt and added a few tests
https://hlsl.godbolt.org/z/exrzrP4j6
All of these should be ` error: call to 'dst' is ambiguous`:
```hlsl
export float4 test(float4 a, double b) {
return dst(a,b);
}
export float4 test(double a, float4 b) {
return dst(a,b);
}
export float4 test(double4 a, float b) {
return dst(a,b);
}
export float4 test(double4 a, float4 b) {
return dst(a,b);
}
```
https://github.com/llvm/llvm-project/pull/133828
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [llvm] [flang-rt] Pass the whole path of libflang_rt.runtime.a to linker on AIX (PR #131041)
DanielCChen wrote: > LGTM. > > It would be safer/clearer to make the `IsFortran` argument change a separate > PR... Good point. I could have made another PR to set `flang-rt` library name only. Will keep that in mind. https://github.com/llvm/llvm-project/pull/131041 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [HLSL] Implement `SpirvType` and `SpirvOpaqueType` (PR #134034)
https://github.com/cassiebeckley updated
https://github.com/llvm/llvm-project/pull/134034
>From 78ac1bc4225b41bc4b9fbd9fd9ab9dc82a2953ca Mon Sep 17 00:00:00 2001
From: Cassandra Beckley
Date: Tue, 1 Apr 2025 23:12:02 -0700
Subject: [PATCH 1/3] [HLSL] Implement `SpirvType` and `SpirvOpaqueType`
This implements the design proposed by [Representing SpirvType in
Clang's Type System](https://github.com/llvm/wg-hlsl/pull/181). It
creates `HLSLInlineSpirvType` as a new `Type` subclass, and
`__hlsl_spirv_type` as a new builtin type template to create such a
type.
This new type is lowered to the `spirv.Type` target extension type, as
described in [Target Extension Types for Inline SPIR-V and Decorated
Types](https://github.com/llvm/wg-hlsl/blob/main/proposals/0017-inline-spirv-and-decorated-types.md).
---
clang/include/clang-c/Index.h | 3 +-
clang/include/clang/AST/ASTContext.h | 5 +
clang/include/clang/AST/ASTNodeTraverser.h| 18 +++
clang/include/clang/AST/PropertiesBase.td | 1 +
clang/include/clang/AST/RecursiveASTVisitor.h | 11 ++
clang/include/clang/AST/Type.h| 142 +-
clang/include/clang/AST/TypeLoc.h | 19 +++
clang/include/clang/AST/TypeProperties.td | 18 +++
clang/include/clang/Basic/BuiltinTemplates.td | 18 ++-
.../clang/Basic/DiagnosticSemaKinds.td| 3 +
clang/include/clang/Basic/TypeNodes.td| 1 +
.../clang/Serialization/ASTRecordReader.h | 2 +
.../clang/Serialization/ASTRecordWriter.h | 14 ++
.../clang/Serialization/TypeBitCodes.def | 1 +
clang/lib/AST/ASTContext.cpp | 59
clang/lib/AST/ASTImporter.cpp | 42 ++
clang/lib/AST/ASTStructuralEquivalence.cpp| 17 +++
clang/lib/AST/ExprConstant.cpp| 1 +
clang/lib/AST/ItaniumMangle.cpp | 40 -
clang/lib/AST/MicrosoftMangle.cpp | 5 +
clang/lib/AST/Type.cpp| 14 ++
clang/lib/AST/TypePrinter.cpp | 48 ++
clang/lib/CodeGen/CGDebugInfo.cpp | 8 +
clang/lib/CodeGen/CGDebugInfo.h | 1 +
clang/lib/CodeGen/CodeGenFunction.cpp | 2 +
clang/lib/CodeGen/CodeGenTypes.cpp| 6 +
clang/lib/CodeGen/ItaniumCXXABI.cpp | 2 +
clang/lib/CodeGen/Targets/SPIR.cpp| 90 ++-
clang/lib/Headers/CMakeLists.txt | 1 +
clang/lib/Headers/hlsl.h | 4 +
clang/lib/Headers/hlsl/hlsl_spirv.h | 30
clang/lib/Sema/SemaExpr.cpp | 1 +
clang/lib/Sema/SemaLookup.cpp | 21 ++-
clang/lib/Sema/SemaTemplate.cpp | 103 -
clang/lib/Sema/SemaTemplateDeduction.cpp | 2 +
clang/lib/Sema/SemaType.cpp | 1 +
clang/lib/Sema/TreeTransform.h| 7 +
clang/lib/Serialization/ASTReader.cpp | 9 ++
clang/lib/Serialization/ASTWriter.cpp | 4 +
.../test/AST/HLSL/Inputs/pch_spirv_type.hlsl | 6 +
clang/test/AST/HLSL/ast-dump-SpirvType.hlsl | 27
clang/test/AST/HLSL/pch_spirv_type.hlsl | 17 +++
clang/test/AST/HLSL/vector-alias.hlsl | 105 +++--
.../inline/SpirvType.alignment.hlsl | 16 ++
.../inline/SpirvType.dx.error.hlsl| 12 ++
clang/test/CodeGenHLSL/inline/SpirvType.hlsl | 68 +
.../inline/SpirvType.incomplete.hlsl | 14 ++
.../inline/SpirvType.literal.error.hlsl | 11 ++
clang/tools/libclang/CIndex.cpp | 5 +
clang/tools/libclang/CXType.cpp | 1 +
.../TableGen/ClangBuiltinTemplatesEmitter.cpp | 72 +++--
51 files changed, 1052 insertions(+), 76 deletions(-)
create mode 100644 clang/lib/Headers/hlsl/hlsl_spirv.h
create mode 100644 clang/test/AST/HLSL/Inputs/pch_spirv_type.hlsl
create mode 100644 clang/test/AST/HLSL/ast-dump-SpirvType.hlsl
create mode 100644 clang/test/AST/HLSL/pch_spirv_type.hlsl
create mode 100644 clang/test/CodeGenHLSL/inline/SpirvType.alignment.hlsl
create mode 100644 clang/test/CodeGenHLSL/inline/SpirvType.dx.error.hlsl
create mode 100644 clang/test/CodeGenHLSL/inline/SpirvType.hlsl
create mode 100644 clang/test/CodeGenHLSL/inline/SpirvType.incomplete.hlsl
create mode 100644 clang/test/CodeGenHLSL/inline/SpirvType.literal.error.hlsl
diff --git a/clang/include/clang-c/Index.h b/clang/include/clang-c/Index.h
index 38e2417dcd181..757f8a3afc758 100644
--- a/clang/include/clang-c/Index.h
+++ b/clang/include/clang-c/Index.h
@@ -3034,7 +3034,8 @@ enum CXTypeKind {
/* HLSL Types */
CXType_HLSLResource = 179,
- CXType_HLSLAttributedResource = 180
+ CXType_HLSLAttributedResource = 180,
+ CXType_HLSLInlineSpirv = 181
};
/**
diff --git a/clang/include/clang/AST/ASTContext.h
b/clang/include/clang/AST/ASTContext.h
index a24f30815e6b9..c62f9f7672010 100644
--- a/clang/include/clang/AST/ASTContext.h
+++ b/clang/i
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
@@ -7,13 +7,14 @@ void foo() {
int main(int argc, char **argv) {
int i;
- #pragma omp target simd nowait( // expected-warning {{extra tokens at the
end of '#pragma omp target simd' are ignored}}
+ #pragma omp target simd nowait( // expected-error {{expected expression}}
// expected-error {{expected ')'}} // expected-note {{to match this '('}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target simd nowait (argc)) // expected-warning {{extra tokens at
the end of '#pragma omp target simd' are ignored}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target simd nowait device (-10u)
for (i = 0; i < argc; ++i) foo();
- #pragma omp target simd nowait (3.14) device (-10u) // expected-warning
{{extra tokens at the end of '#pragma omp target simd' are ignored}}
+ #pragma omp target simd nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating type}}
mjklemm wrote:
```suggestion
#pragma omp target simd nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating-point type}}
```
https://github.com/llvm/llvm-project/pull/128742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
erichkeane wrote:
is the formatting OK on this line? Github is making this look awful weird.
Also-also: This is an infinite loop, right? As you said this is never
evaluated at runtime, the answer would be fixed?
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -13329,6 +13359,9 @@ inline QualType Sema::CheckLogicalOperands(ExprResult &LHS, ExprResult &RHS, // The following is safe because we only use this method for // non-overloadable operands. + if (IsAMDGPUPredicateBI(LHS.get()) && IsAMDGPUPredicateBI(RHS.get())) erichkeane wrote: If we want to diagnose this not in the 'right' place we should, though that is going ot be absolutely fraught with problems. But making it return `void` is not the way to do this. https://github.com/llvm/llvm-project/pull/134016 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
@@ -7,13 +7,13 @@ void foo() {
int main(int argc, char **argv) {
int i;
-#pragma omp target teams distribute parallel for nowait( // expected-warning
{{extra tokens at the end of '#pragma omp target teams distribute parallel for'
are ignored}}
+#pragma omp target teams distribute parallel for nowait( // expected-error
{{expected expression}} // expected-error {{expected ')'}} // expected-note
{{to match this '('}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target teams distribute parallel for nowait (argc)) //
expected-warning {{extra tokens at the end of '#pragma omp target teams
distribute parallel for' are ignored}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target teams distribute parallel for nowait device (-10u)
for (i = 0; i < argc; ++i) foo();
-#pragma omp target teams distribute parallel for nowait (3.14) device (-10u)
// expected-warning {{extra tokens at the end of '#pragma omp target teams
distribute parallel for' are ignored}}
+#pragma omp target teams distribute parallel for nowait (3.14) device (-10u)
// expected-error {{arguments of OpenMP clause 'nowait' with bitwise operators
cannot be of floating type}}
mjklemm wrote:
```suggestion
#pragma omp target teams distribute parallel for nowait (3.14) device (-10u) //
expected-error {{arguments of OpenMP clause 'nowait' with bitwise operators
cannot be of floating-point type}}
```
https://github.com/llvm/llvm-project/pull/128742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
@@ -6,13 +6,11 @@ void foo() {
}
int main(int argc, char **argv) {
-#pragma omp target teams nowait( // expected-warning {{extra tokens at the end
of '#pragma omp target teams' are ignored}}
- foo();
-#pragma omp target teams nowait (argc)) // expected-warning {{extra tokens at
the end of '#pragma omp target teams' are ignored}}
+#pragma omp target teams nowait( // expected-error {{expected expression}} //
expected-error {{expected ')'}} //expected-note {{to match this '('}}
foo();
#pragma omp target teams nowait device (-10u)
foo();
-#pragma omp target teams nowait (3.14) device (-10u) // expected-warning
{{extra tokens at the end of '#pragma omp target teams' are ignored}}
+#pragma omp target teams nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating type}}
mjklemm wrote:
```suggestion
#pragma omp target teams nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating-point type}}
```
https://github.com/llvm/llvm-project/pull/128742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
+
+ for (; __builtin_amdgcn_is_invocable(__builtin_amdgcn_permlane64); ++*p)
break;
+
+**Description**:
+
+When used as the predicate value of the following control structures:
+
+.. code-block:: c++
+
+ if (...)
+ while (...)
+ do { } while (...)
+ for (...)
+
+be it directly, or as arguments to logical operators such as ``!, ||, &&``, the
erichkeane wrote:
How about when used as an initializer in one of those? Consider:
`if (auto b = __builtin_amd_gcn_processor_is("gfx1201"); b &&
another_condition)`
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
@@ -6,13 +6,13 @@ void foo() {
}
int main(int argc, char **argv) {
- #pragma omp target parallel nowait( // expected-warning {{extra tokens at
the end of '#pragma omp target parallel' are ignored}}
- foo();
+ #pragma omp target parallel nowait( // expected-error {{expected
expression}} // expected-error {{expected ')'}} // expected-note {{to match
this '('}}
+foo();
#pragma omp target parallel nowait (argc)) // expected-warning {{extra
tokens at the end of '#pragma omp target parallel' are ignored}}
foo();
#pragma omp target parallel nowait device (-10u)
foo();
- #pragma omp target parallel nowait (3.14) device (-10u) // expected-warning
{{extra tokens at the end of '#pragma omp target parallel' are ignored}}
+ #pragma omp target parallel nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating type}}
mjklemm wrote:
```suggestion
#pragma omp target parallel nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating-point type}}
```
https://github.com/llvm/llvm-project/pull/128742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang] [OpenMP] Support NOWAIT with optional argument (PR #128742)
@@ -7,13 +7,13 @@ void foo() {
int main(int argc, char **argv) {
int i;
- #pragma omp target parallel for nowait( // expected-warning {{extra tokens
at the end of '#pragma omp target parallel for' are ignored}}
+ #pragma omp target parallel for nowait( // expected-error {{expected
expression}} // expected-error {{expected ')'}} // expected-note {{to match
this '('}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target parallel for nowait (argc)) // expected-warning {{extra
tokens at the end of '#pragma omp target parallel for' are ignored}}
for (i = 0; i < argc; ++i) foo();
#pragma omp target parallel for nowait device (-10u)
for (i = 0; i < argc; ++i) foo();
- #pragma omp target parallel for nowait (3.14) device (-10u) //
expected-warning {{extra tokens at the end of '#pragma omp target parallel for'
are ignored}}
+ #pragma omp target parallel for nowait (3.14) device (-10u) //
expected-error {{arguments of OpenMP clause 'nowait' with bitwise operators
cannot be of floating type}}
mjklemm wrote:
```suggestion
#pragma omp target parallel for nowait (3.14) device (-10u) // expected-error
{{arguments of OpenMP clause 'nowait' with bitwise operators cannot be of
floating-point type}}
```
https://github.com/llvm/llvm-project/pull/128742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -6541,6 +6541,22 @@ ExprResult Sema::BuildCallExpr(Scope *Scope, Expr *Fn, SourceLocation LParenLoc, if (Result.isInvalid()) return ExprError(); Fn = Result.get(); + // The __builtin_amdgcn_is_invocable builtin is special, and will be resolved erichkeane wrote: Not a big fan of this AT ALL. This needs to be much less special. https://github.com/llvm/llvm-project/pull/134016 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
https://github.com/asudarsa updated
https://github.com/llvm/llvm-project/pull/133967
>From 0ab40315cd5ca320d84f87c99b8f23d9b6e8ed36 Mon Sep 17 00:00:00 2001
From: Arvind Sudarsanam
Date: Mon, 31 Mar 2025 14:31:24 -0700
Subject: [PATCH 1/6] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V
backend support inside clang-sycl-linker
This PR does the following:
1. Use SPIR-V backend to do LLVM to SPIR-V translation inside clang-sycl-linker
2. Remove llvm-spirv translator from clang-sycl-linker
Currently, no SPIR-V extensions are enabled for SYCL compilation flow. This
will be updated in subsequent commits.
Thanks
Signed-off-by: Arvind Sudarsanam
---
clang/test/Driver/clang-sycl-linker-test.cpp | 14 +-
clang/test/Driver/sycl-link-spirv-target.cpp | 6 +-
clang/tools/clang-sycl-linker/CMakeLists.txt | 5 +-
.../clang-sycl-linker/ClangSYCLLinker.cpp | 187 +++---
clang/tools/clang-sycl-linker/SYCLLinkOpts.td | 2 +-
5 files changed, 76 insertions(+), 138 deletions(-)
diff --git a/clang/test/Driver/clang-sycl-linker-test.cpp
b/clang/test/Driver/clang-sycl-linker-test.cpp
index 729561bd09cd8..2f860ae74e97d 100644
--- a/clang/test/Driver/clang-sycl-linker-test.cpp
+++ b/clang/test/Driver/clang-sycl-linker-test.cpp
@@ -6,7 +6,7 @@
// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc -o
a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=SIMPLE-FO
// SIMPLE-FO: sycl-device-link: inputs: {{.*}}.bc, {{.*}}.bc libfiles:
output: [[LLVMLINKOUT:.*]].bc
-// SIMPLE-FO-NEXT: "{{.*}}llvm-spirv{{.*}}" {{.*}}-o a.spv [[LLVMLINKOUT]].bc
+// SIMPLE-FO-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv
//
// Test the dry run of a simple case with device library files specified.
// RUN: touch %T/lib1.bc
@@ -14,7 +14,7 @@
// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc
--library-path=%T --device-libs=lib1.bc,lib2.bc -o a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=DEVLIBS
// DEVLIBS: sycl-device-link: inputs: {{.*}}.bc libfiles: {{.*}}lib1.bc,
{{.*}}lib2.bc output: [[LLVMLINKOUT:.*]].bc
-// DEVLIBS-NEXT: "{{.*}}llvm-spirv{{.*}}" {{.*}}-o a.spv [[LLVMLINKOUT]].bc
+// DEVLIBS-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv
//
// Test a simple case with a random file (not bitcode) as input.
// RUN: touch %t.o
@@ -29,13 +29,3 @@
// RUN: not clang-sycl-linker --dry-run -triple=spirv64 %t_1.bc %t_2.bc
--library-path=%T --device-libs=lib1.bc,lib2.bc,lib3.bc -o a.spv 2>&1 \
// RUN: | FileCheck %s --check-prefix=DEVLIBSERR2
// DEVLIBSERR2: '{{.*}}lib3.bc' SYCL device library file is not found
-//
-// Test if correct set of llvm-spirv options are emitted for windows
environment.
-// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 --is-windows-msvc-env
%t_1.bc %t_2.bc -o a.spv 2>&1 \
-// RUN: | FileCheck %s --check-prefix=LLVMOPTSWIN
-// LLVMOPTSWIN: -spirv-debug-info-version=ocl-100
-spirv-allow-extra-diexpressions -spirv-allow-unknown-intrinsics=llvm.genx.
-spirv-ext=
-//
-// Test if correct set of llvm-spirv options are emitted for linux environment.
-// RUN: clang-sycl-linker --dry-run -v -triple=spirv64 %t_1.bc %t_2.bc -o
a.spv 2>&1 \
-// RUN: | FileCheck %s --check-prefix=LLVMOPTSLIN
-// LLVMOPTSLIN: -spirv-debug-info-version=nonsemantic-shader-200
-spirv-allow-unknown-intrinsics=llvm.genx. -spirv-ext=
diff --git a/clang/test/Driver/sycl-link-spirv-target.cpp
b/clang/test/Driver/sycl-link-spirv-target.cpp
index 7585ef8b14a59..586adae619165 100644
--- a/clang/test/Driver/sycl-link-spirv-target.cpp
+++ b/clang/test/Driver/sycl-link-spirv-target.cpp
@@ -3,7 +3,7 @@
//
// Test that -Xlinker options are being passed to clang-sycl-linker.
// RUN: touch %t.bc
-// RUN: %clangxx -### --target=spirv64 --sycl-link -Xlinker
--llvm-spirv-path=/tmp \
-// RUN: -Xlinker -triple=spirv64 -Xlinker --library-path=/tmp -Xlinker
--device-libs=lib1.bc,lib2.bc %t.bc 2>&1 \
+// RUN: %clangxx -### --target=spirv64 --sycl-link -Xlinker -triple=spirv64
-Xlinker --library-path=/tmp \
+// RUN: -Xlinker --device-libs=lib1.bc,lib2.bc %t.bc 2>&1 \
// RUN: | FileCheck %s -check-prefix=XLINKEROPTS
-// XLINKEROPTS: "{{.*}}clang-sycl-linker{{.*}}" "--llvm-spirv-path=/tmp"
"-triple=spirv64" "--library-path=/tmp" "--device-libs=lib1.bc,lib2.bc"
"{{.*}}.bc" "-o" "a.out"
+// XLINKEROPTS: "{{.*}}clang-sycl-linker{{.*}}" "-triple=spirv64"
"--library-path=/tmp" "--device-libs=lib1.bc,lib2.bc" "{{.*}}.bc" "-o" "a.out"
diff --git a/clang/tools/clang-sycl-linker/CMakeLists.txt
b/clang/tools/clang-sycl-linker/CMakeLists.txt
index 382c0ca441940..ee89e8b0a5570 100644
--- a/clang/tools/clang-sycl-linker/CMakeLists.txt
+++ b/clang/tools/clang-sycl-linker/CMakeLists.txt
@@ -1,14 +1,17 @@
set(LLVM_LINK_COMPONENTS
${LLVM_TARGETS_TO_BUILD}
+ Analysis
BinaryFormat
BitWriter
Core
IRReader
Linker
+ MC
Option
Object
- TargetParser
Support
+ Target
+ TargetParser
)
set(LLVM_TARGET_DEFIN
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -15576,6 +15609,38 @@ static bool isOverflowingIntegerType(ASTContext &Ctx,
QualType T) {
return Ctx.getIntWidth(T) >= Ctx.getIntWidth(Ctx.IntTy);
}
+static Expr *ExpandAMDGPUPredicateBI(ASTContext &Ctx, CallExpr *CE) {
+ if (!CE->getBuiltinCallee())
+return CXXBoolLiteralExpr::Create(Ctx, false, Ctx.BoolTy,
CE->getExprLoc());
+
+ if (Ctx.getTargetInfo().getTriple().isSPIRV()) {
+CE->setType(Ctx.getLogicalOperationType());
+return CE;
+ }
+
+ bool P = false;
+ auto &TI = Ctx.getTargetInfo();
+
+ if (CE->getDirectCallee()->getName() == "__builtin_amdgcn_processor_is") {
erichkeane wrote:
This isn't sufficient, you actually have to check that it is the builtin. If
you check just the name it could be a different one, as this doesn't give you
the name of namespaces/classes/etc, just the immediate name.
ALSO, `getName` in many cases will assert if this is an operator/other nameless
function.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
erichkeane wrote:
Formatting on this one is weird too. Also, same question again, and again on
4972.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -15576,6 +15609,38 @@ static bool isOverflowingIntegerType(ASTContext &Ctx,
QualType T) {
return Ctx.getIntWidth(T) >= Ctx.getIntWidth(Ctx.IntTy);
}
+static Expr *ExpandAMDGPUPredicateBI(ASTContext &Ctx, CallExpr *CE) {
+ if (!CE->getBuiltinCallee())
+return CXXBoolLiteralExpr::Create(Ctx, false, Ctx.BoolTy,
CE->getExprLoc());
+
+ if (Ctx.getTargetInfo().getTriple().isSPIRV()) {
+CE->setType(Ctx.getLogicalOperationType());
+return CE;
+ }
+
+ bool P = false;
+ auto &TI = Ctx.getTargetInfo();
+
+ if (CE->getDirectCallee()->getName() == "__builtin_amdgcn_processor_is") {
+auto GFX = dyn_cast(CE->getArg(0)->IgnoreParenCasts());
+auto TID = TI.getTargetID();
+if (GFX && TID) {
+ auto N = GFX->getString();
+ P = TI.isValidCPUName(GFX->getString()) && TID->find(N) == 0;
+}
+ } else {
+auto FD = cast(CE->getArg(0)->getReferencedDeclOfCallee());
erichkeane wrote:
so many more missing stars.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
erichkeane wrote:
Re-answering my own question: Later on you seem to be using 'predicate for a
control structure' to mean 'the condition of a if/while/for. However, why is
it problematic to have someone check this and store it in a variable? Why is:
`if (__builtin_amd_gcn_processor_is("gfx1201"))` fine, but:
`bool b = __builtin_amd_gcn_processor_is("gfx1201"); if (b)` a problem?
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
+
+ for (; __builtin_amdgcn_is_invocable(__builtin_amdgcn_permlane64); ++*p)
break;
+
+**Description**:
+
+When used as the predicate value of the following control structures:
+
+.. code-block:: c++
+
+ if (...)
+ while (...)
+ do { } while (...)
+ for (...)
+
+be it directly, or as arguments to logical operators such as ``!, ||, &&``, the
+builtins return a boolean value that:
+
+* indicates whether the current target matches the argument; the argument MUST
+ be a string literal and a valid AMDGPU target
+* indicates whether the builtin function passed as the argument can be invoked
+ by the current target; the argument MUST be either a generic or AMDGPU
+ specific builtin name
+
+Outside of these contexts, the builtins have a ``void`` returning signature
+which prevents their misuse.
erichkeane wrote:
What misuse here? If you want to catch other situations, just diagnose. But
it seems that these values could be useful/possible outside of condition
expressions.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
https://github.com/erichkeane requested changes to this pull request. The user interface design of these builtins is just completely unacceptable, and it has made the sema implementation more difficult than it needs to be. Best I can tell, the answer for even things like `decltype` changes whether it is in a conditional or not, which isn't right. https://github.com/llvm/llvm-project/pull/134016 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -585,6 +597,23 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned
BuiltinID,
llvm::Value *Env = EmitScalarExpr(E->getArg(0));
return Builder.CreateCall(F, {Env});
}
+ case AMDGPU::BI__builtin_amdgcn_processor_is: {
+assert(CGM.getTriple().isSPIRV() &&
+ "__builtin_amdgcn_processor_is should never reach CodeGen for "
+ "concrete targets!");
+StringRef Proc = cast(E->getArg(0))->getString();
+return GetOrInsertAMDGPUPredicate(*this, "llvm.amdgcn.is." + Proc);
+ }
+ case AMDGPU::BI__builtin_amdgcn_is_invocable: {
+assert(CGM.getTriple().isSPIRV() &&
+ "__builtin_amdgcn_is_invocable should never reach CodeGen for "
+ "concrete targets!");
+auto FD = cast(
erichkeane wrote:
```suggestion
auto *FD = cast(
```
With auto, we require a * with all pointer types, looks like you've missed a
few, so please go back and fix the ones you have.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Remove duplicate API (PR #132776)
Jugst3r wrote: Sure, I did what @Endilll suggested. It is kind of unfortunate to have such similar type names :/. I don't think I can use the [[attribute]] C++-style syntax, maybe there is another syntax I am not aware of, but for the mean time I put the deprecated inside the comments. https://github.com/llvm/llvm-project/pull/132776 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang-tools-extra] [clang] support pack expansions for trailing requires clauses (PR #133190)
@@ -78,6 +78,22 @@ class UnresolvedSetImpl;
class VarTemplateDecl;
enum class ImplicitParamKind;
+// Holds a constraint expression along with a pack expansion index, if
+// expanded.
+struct AssociatedConstraint {
+ const Expr *ConstraintExpr = nullptr;
+ int ArgumentPackSubstitutionIndex = -1;
erichkeane wrote:
I was thinking about it... I'd LOVE instead (though this is a 'more work' thing
:D so nothing to do here unless you'd like to ) if we just had a
`std::optional` like type to represent all of these substitution/pack indexes.
The difference here is that the `optional` type would just use a single value
as its `unset` state instead of a separate value, so we don't pay the size cost.
Additionally, we could make all of these unsigned instead of signed ints, so
presumably we get a little bit of (unusable for other reasons!) range back.
Also/also: serialization/etc all ends up being self-solved since they can have
their own serialization/etc.
https://github.com/llvm/llvm-project/pull/133190
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [flang] Complete alignment of -x language modes with gfortran (PR #133775)
https://github.com/DavidTruby updated
https://github.com/llvm/llvm-project/pull/133775
>From aeca7c2c7bf4cc7aef87374f3890f2a42e61d07d Mon Sep 17 00:00:00 2001
From: David Truby
Date: Mon, 31 Mar 2025 19:39:37 +0100
Subject: [PATCH 1/2] [flang] Complete alignment of -x language modes with
gfortran
This fixes an issue where, since the alignment of the -x lanaguage
modes, .f90 files were being preprocessed by default. This patch
completes the alignment of the meaning of the f95-pp-input and f95
language modes to match gfortran, fixing that issue.
---
clang/include/clang/Driver/Types.def | 4 ++--
.../Driver/input-from-stdin/input-from-stdin.f90 | 2 +-
flang/test/Driver/phases.f90 | 12 ++--
flang/test/Driver/pp-fixed-form.f90 | 16
4 files changed, 17 insertions(+), 17 deletions(-)
diff --git a/clang/include/clang/Driver/Types.def
b/clang/include/clang/Driver/Types.def
index 214c5e7a789f9..2b6b3e75c1c97 100644
--- a/clang/include/clang/Driver/Types.def
+++ b/clang/include/clang/Driver/Types.def
@@ -88,8 +88,8 @@ TYPE("assembler-with-cpp", Asm, PP_Asm,
"S", phases
// modules when Flang needs to emit pre-processed files. Therefore, the
// `PP_TYPE` is set to `PP_Fortran` so that the driver is fine with
// "pre-processing a pre-processed file".
-TYPE("f95", PP_Fortran, PP_Fortran, "i",
phases::Preprocess, phases::Compile, phases::Backend, phases::Assemble,
phases::Link)
-TYPE("f95-cpp-input",Fortran, PP_Fortran, nullptr,
phases::Preprocess, phases::Compile, phases::Backend, phases::Assemble,
phases::Link)
+TYPE("f95", Fortran, PP_Fortran, nullptr,
phases::Preprocess, phases::Compile, phases::Backend, phases::Assemble,
phases::Link)
+TYPE("f95-cpp-input",PP_Fortran, PP_Fortran, "i",
phases::Preprocess, phases::Compile, phases::Backend, phases::Assemble,
phases::Link)
TYPE("java", Java, INVALID, nullptr,
phases::Compile, phases::Backend, phases::Assemble, phases::Link)
// LLVM IR/LTO types. We define separate types for IR and LTO because LTO
diff --git a/flang/test/Driver/input-from-stdin/input-from-stdin.f90
b/flang/test/Driver/input-from-stdin/input-from-stdin.f90
index 1fcc0340a64ba..285f0751b35d8 100644
--- a/flang/test/Driver/input-from-stdin/input-from-stdin.f90
+++ b/flang/test/Driver/input-from-stdin/input-from-stdin.f90
@@ -6,7 +6,7 @@
! Input type is implicit
! RUN: cat %s | %flang -E -cpp - | FileCheck %s --check-prefix=PP-NOT-DEFINED
! RUN: cat %s | %flang -DNEW -E -cpp - | FileCheck %s --check-prefix=PP-DEFINED
-! RUN: cat %s | %flang -DNEW -E - | FileCheck %s --check-prefix=PP-DEFINED
+! RUN: cat %s | %flang -DNEW -E - | FileCheck %s --check-prefix=PP-NOT-DEFINED
! RUN: cat %s | %flang -DNEW -E -nocpp - | FileCheck %s
--check-prefix=PP-NOT-DEFINED
! Input type is explicit
diff --git a/flang/test/Driver/phases.f90 b/flang/test/Driver/phases.f90
index b688600dae035..9346773c883ef 100644
--- a/flang/test/Driver/phases.f90
+++ b/flang/test/Driver/phases.f90
@@ -4,15 +4,15 @@
! RUN: %flang -fsyntax-only -ccc-print-phases %s 2>&1 | FileCheck %s
--check-prefix=COMPILE
! RUN: %flang -c -ccc-print-phases %s 2>&1 | FileCheck %s
--check-prefix=EMIT_OBJ
-! PP: +- 0: input, "{{.*}}phases.f90", f95-cpp-input
-! PP-NEXT: 1: preprocessor, {0}, f95
+! PP: +- 0: input, "{{.*}}phases.f90", f95
+! PP-NEXT: 1: preprocessor, {0}, f95-cpp-input
-! COMPILE: +- 0: input, "{{.*}}phases.f90", f95-cpp-input
-! COMPILE-NEXT: 1: preprocessor, {0}, f95
+! COMPILE: +- 0: input, "{{.*}}phases.f90", f95
+! COMPILE-NEXT: 1: preprocessor, {0}, f95-cpp-input
! COMPILE-NEXT: 2: compiler, {1}, none
-! EMIT_OBJ: +- 0: input, "{{.*}}phases.f90", f95-cpp-input
-! EMIT_OBJ-NEXT: 1: preprocessor, {0}, f95
+! EMIT_OBJ: +- 0: input, "{{.*}}phases.f90", f95
+! EMIT_OBJ-NEXT: 1: preprocessor, {0}, f95-cpp-input
! EMIT_OBJ-NEXT: 2: compiler, {1}, ir
! EMIT_OBJ-NEXT: +- 3: backend, {2}, assembler
! EMIT_OBJ-NEXT: 4: assembler, {3}, object
diff --git a/flang/test/Driver/pp-fixed-form.f90
b/flang/test/Driver/pp-fixed-form.f90
index 4695da78763ae..bb869cd3341a7 100644
--- a/flang/test/Driver/pp-fixed-form.f90
+++ b/flang/test/Driver/pp-fixed-form.f90
@@ -1,19 +1,19 @@
!RUN: %flang -save-temps -### %S/Inputs/free-form-test.f90 2>&1 | FileCheck
%s --check-prefix=FREE
-FREE: "-fc1" {{.*}} "-o" "free-form-test.i" {{.*}} "-x" "f95-cpp-input"
"{{.*}}/free-form-test.f90"
-FREE-NEXT: "-fc1" {{.*}} "-ffixed-form" {{.*}} "-x" "f95" "free-form-test.i"
+FREE: "-fc1" {{.*}} "-o" "free-form-test.i" {{.*}} "-x" "f95"
"{{.*}}/free-form-test.f90"
+FREE-NEXT: "-fc1" {{.*}} "-ffixed-form" {{.*}} "-x" "f95-cpp-input"
"free-form-test.i"
!RUN: %flang -save-temps -### %S/Inputs/fixed-form-test.f 2>&1 | FileCheck %s
--check-prefix=FIXED
-FIXED: "-fc1"
[clang] [Clang] Fix dependent local class instantiation bugs (PR #134038)
@@ -4264,7 +4264,8 @@ Sema::InstantiateClassMembers(SourceLocation
PointOfInstantiation,
if (FunctionDecl *Pattern =
Function->getInstantiatedFromMemberFunction()) {
-if (Function->isIneligibleOrNotSelected())
+if (!Instantiation->getDeclContext()->isDependentContext() &&
erichkeane wrote:
Hmm... I don't think we want to go this far, right?We still would like to
aggressively instantiate these so we get diagnostics/etc, even if it is in a
dependent context.
https://github.com/llvm/llvm-project/pull/134038
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [Clang][AMDGPU] Add __builtin_amdgcn_cvt_off_f32_i4 (PR #133741)
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?= ,
Juan Manuel Martinez =?utf-8?q?Caamaño?=
Message-ID:
In-Reply-To:
@@ -743,9 +743,13 @@ GCNTTIImpl::instCombineIntrinsic(InstCombiner &IC,
IntrinsicInst &II) const {
if (!CArg)
break;
-int CArg4BitAsInt = CArg->getValue().trunc(4).getSExtValue();
-float ResVal = 0.0625 * CArg4BitAsInt;
-Constant *Res = ConstantFP::get(II.getType(), ResVal);
+// Tabulated 0.0625 * (sext (CArg & 0xf)).
+constexpr size_t ResValsSize = 16;
+const float ResVals[ResValsSize] = {
shiltian wrote:
or even `constexpr`?
https://github.com/llvm/llvm-project/pull/133741
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -284,6 +284,18 @@ void
CodeGenFunction::AddAMDGPUFenceAddressSpaceMMRA(llvm::Instruction *Inst,
Inst->setMetadata(LLVMContext::MD_mmra, MMRAMetadata::getMD(Ctx, MMRAs));
}
+static Value *GetOrInsertAMDGPUPredicate(CodeGenFunction &CGF, Twine Name) {
+ auto PTy = IntegerType::getInt1Ty(CGF.getLLVMContext());
+
+ auto P = cast(
AlexVlx wrote:
Done.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
@@ -15576,6 +15609,38 @@ static bool isOverflowingIntegerType(ASTContext &Ctx,
QualType T) {
return Ctx.getIntWidth(T) >= Ctx.getIntWidth(Ctx.IntTy);
}
+static Expr *ExpandAMDGPUPredicateBI(ASTContext &Ctx, CallExpr *CE) {
+ if (!CE->getBuiltinCallee())
+return CXXBoolLiteralExpr::Create(Ctx, false, Ctx.BoolTy,
CE->getExprLoc());
+
+ if (Ctx.getTargetInfo().getTriple().isSPIRV()) {
+CE->setType(Ctx.getLogicalOperationType());
+return CE;
+ }
+
+ bool P = false;
+ auto &TI = Ctx.getTargetInfo();
+
+ if (CE->getDirectCallee()->getName() == "__builtin_amdgcn_processor_is") {
+auto GFX = dyn_cast(CE->getArg(0)->IgnoreParenCasts());
+auto TID = TI.getTargetID();
+if (GFX && TID) {
+ auto N = GFX->getString();
+ P = TI.isValidCPUName(GFX->getString()) && TID->find(N) == 0;
+}
+ } else {
+auto FD = cast(CE->getArg(0)->getReferencedDeclOfCallee());
AlexVlx wrote:
Done.
https://github.com/llvm/llvm-project/pull/134016
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm][doc]: Merge the contents of identical entries. (PR #134089)
https://github.com/YLChenZ created
https://github.com/llvm/llvm-project/pull/134089
Closes #133706.
before the patch:
after the patch:
>From 319c55aef5c458ae4ac6c7f3f186d338f6fe2e37 Mon Sep 17 00:00:00 2001
From: YLChenZ
Date: Wed, 2 Apr 2025 22:03:53 +0800
Subject: [PATCH] [llvm][doc]: Merge the contents of identical entries.
---
clang/utils/TableGen/ClangAttrEmitter.cpp | 68 +--
1 file changed, 50 insertions(+), 18 deletions(-)
diff --git a/clang/utils/TableGen/ClangAttrEmitter.cpp
b/clang/utils/TableGen/ClangAttrEmitter.cpp
index 06f6b073240ba..6a2cdcd4ebc8e 100644
--- a/clang/utils/TableGen/ClangAttrEmitter.cpp
+++ b/clang/utils/TableGen/ClangAttrEmitter.cpp
@@ -5144,6 +5144,14 @@ class SpellingList {
Spellings[(size_t)Kind].push_back(Name);
}
+
+ void merge(const SpellingList &Other) {
+for (size_t Kind = 0; Kind < NumSpellingKinds; ++Kind) {
+ Spellings[Kind].insert(Spellings[Kind].end(),
+ Other.Spellings[Kind].begin(),
+ Other.Spellings[Kind].end());
+}
+ }
};
class DocumentationData {
@@ -5301,44 +5309,68 @@ void EmitClangAttrDocs(const RecordKeeper &Records,
raw_ostream &OS) {
return L->getValueAsString("Name") < R->getValueAsString("Name");
}
};
- std::map, CategoryLess>
+
+ std::map,
+ CategoryLess>
SplitDocs;
+
+ // Collect documentation data, grouping by category and heading.
for (const auto *A : Records.getAllDerivedDefinitions("Attr")) {
const Record &Attr = *A;
std::vector Docs =
Attr.getValueAsListOfDefs("Documentation");
+
for (const auto *D : Docs) {
const Record &Doc = *D;
const Record *Category = Doc.getValueAsDef("Category");
// If the category is "InternalOnly", then there cannot be any other
// documentation categories (otherwise, the attribute would be
// emitted into the docs).
- const StringRef Cat = Category->getValueAsString("Name");
- bool InternalOnly = Cat == "InternalOnly";
- if (InternalOnly && Docs.size() > 1)
+ StringRef Cat = Category->getValueAsString("Name");
+ if (Cat == "InternalOnly" && Docs.size() > 1)
PrintFatalError(Doc.getLoc(),
"Attribute is \"InternalOnly\", but has multiple "
"documentation categories");
- if (!InternalOnly)
-SplitDocs[Category].push_back(DocumentationData(
-Doc, Attr, GetAttributeHeadingAndSpellings(Doc, Attr, Cat)));
+ if (Cat == "InternalOnly")
+continue;
+
+ // Generate Heading and Spellings.
+ auto HeadingAndSpellings =
+ GetAttributeHeadingAndSpellings(Doc, Attr, Cat);
+
+ auto &CategoryDocs = SplitDocs[Category];
+
+ // If the heading already exists, merge the documentation.
+ auto It = CategoryDocs.find(HeadingAndSpellings.first);
+ if (It != CategoryDocs.end()) {
+// Merge spellings
+It->second.SupportedSpellings.merge(HeadingAndSpellings.second);
+// Merge content
+It->second.Documentation = &Doc; // Update reference
+ } else {
+// Otherwise, add a new entry.
+CategoryDocs.emplace(HeadingAndSpellings.first,
+ DocumentationData(Doc, Attr,
HeadingAndSpellings));
+ }
}
}
- // Having split the attributes out based on what documentation goes where,
- // we can begin to generate sections of documentation.
- for (auto &I : SplitDocs) {
-WriteCategoryHeader(I.first, OS);
+ // Write out documentation, merging attributes with the same heading.
+ for (auto &CategoryPair : SplitDocs) {
+WriteCategoryHeader(CategoryPair.first, OS);
+
+std::vector MergedDocs;
+for (auto &DocPair : CategoryPair.second)
+ MergedDocs.push_back(std::move(DocPair.second));
-sort(I.second,
- [](const DocumentationData &D1, const DocumentationData &D2) {
- return D1.Heading < D2.Heading;
- });
+std::sort(MergedDocs.begin(), MergedDocs.end(),
+ [](const DocumentationData &D1, const DocumentationData &D2) {
+return D1.Heading < D2.Heading;
+ });
-// Walk over each of the attributes in the category and write out their
-// documentation.
-for (const auto &Doc : I.second)
+// Output each documentation entry.
+for (const auto &Doc : MergedDocs)
WriteDocumentation(Records, Doc, OS);
}
}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
https://github.com/AlexVlx updated
https://github.com/llvm/llvm-project/pull/134016
>From 91eeaf02336e539f14dcb0a79ff15dbe8befe6f1 Mon Sep 17 00:00:00 2001
From: Alex Voicu
Date: Wed, 2 Apr 2025 02:47:42 +0100
Subject: [PATCH 1/8] Add the functional identity and feature queries.
---
clang/docs/LanguageExtensions.rst | 110 ++
clang/include/clang/Basic/BuiltinsAMDGPU.def | 5 +
.../clang/Basic/DiagnosticSemaKinds.td| 10 +
clang/lib/Basic/Targets/SPIR.cpp | 4 +
clang/lib/Basic/Targets/SPIR.h| 4 +
clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp | 29 ++
clang/lib/Sema/SemaExpr.cpp | 157
clang/test/CodeGen/amdgpu-builtin-cpu-is.c| 65
.../CodeGen/amdgpu-builtin-is-invocable.c | 64
.../amdgpu-feature-builtins-invalid-use.cpp | 43 +++
llvm/lib/Target/AMDGPU/AMDGPU.h | 9 +
.../AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp | 207 ++
llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def | 2 +
.../lib/Target/AMDGPU/AMDGPUTargetMachine.cpp | 3 +-
llvm/lib/Target/AMDGPU/CMakeLists.txt | 1 +
...pu-expand-feature-predicates-unfoldable.ll | 28 ++
.../amdgpu-expand-feature-predicates.ll | 359 ++
17 files changed, 1099 insertions(+), 1 deletion(-)
create mode 100644 clang/test/CodeGen/amdgpu-builtin-cpu-is.c
create mode 100644 clang/test/CodeGen/amdgpu-builtin-is-invocable.c
create mode 100644 clang/test/CodeGen/amdgpu-feature-builtins-invalid-use.cpp
create mode 100644 llvm/lib/Target/AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp
create mode 100644
llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates-unfoldable.ll
create mode 100644 llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates.ll
diff --git a/clang/docs/LanguageExtensions.rst
b/clang/docs/LanguageExtensions.rst
index 3b8a9cac6587a..8a7cb75af13e5 100644
--- a/clang/docs/LanguageExtensions.rst
+++ b/clang/docs/LanguageExtensions.rst
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
+
+ for (; __builtin_amdgcn_is_invocable(__builtin_amdgcn_permlane64); ++*p)
break;
+
+**Description**:
+
+When used as the predicate value of the following control structures:
+
+.. code-block:: c++
+
+ if (...)
+ while (...)
+ do { } while (...)
+ for (...)
+
+be it directly, or as arguments to logical operators such as ``!, ||, &&``, the
+builtins return a boolean value that:
+
+* indicates whether the current target matches the argument; the argument MUST
+ be a string literal and a valid AMDGPU target
+* indicates whether the builtin function passed as the argument can be invoked
+ by the current target; the argument MUST be either a generic or AMDGPU
+ specific builtin name
+
+Outside of these contexts, the builtins have a ``void`` returning signature
+which prevents their misuse.
+
+**Example of invalid use**:
+
+.. code-block:: c++
+
+ void kernel(int* p, int x, bool (*pfn)(bool), const char* str) {
+if (__builtin_amdgcn_processor_is("not_an_amdgcn_gfx_id")) return;
+else if (__bui
[libclc] [libclc] Move sinh, cosh & tanh to the CLC library (PR #134063)
https://github.com/frasercrmck closed https://github.com/llvm/llvm-project/pull/134063 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[libclc] d51525b - [libclc] Move lgamma, lgamma_r & tgamma to CLC library (#134053)
Author: Fraser Cormack
Date: 2025-04-02T15:20:32+01:00
New Revision: d51525ba36017507fe95ce48e9670b6f4b267016
URL:
https://github.com/llvm/llvm-project/commit/d51525ba36017507fe95ce48e9670b6f4b267016
DIFF:
https://github.com/llvm/llvm-project/commit/d51525ba36017507fe95ce48e9670b6f4b267016.diff
LOG: [libclc] Move lgamma, lgamma_r & tgamma to CLC library (#134053)
Also enable half-precision variants of tgamma, which were previously
missing.
Note that unlike recent work, these builtins are not vectorized as part
of this commit. Ultimately all three call into lgamma_r, which has heavy
control flow (including switch statements) that would be difficult to
vectorize. Additionally the lgamma_r algorithm is copyrighted to SunPro
so may need a rewrite in the future anyway.
There are no codegen changes (to non-SPIR-V targets) with this commit,
aside from the new half builtins.
Added:
libclc/clc/include/clc/math/clc_lgamma.h
libclc/clc/include/clc/math/clc_lgamma_r.h
libclc/clc/include/clc/math/clc_tgamma.h
libclc/clc/include/clc/math/unary_def_via_fp32.inc
libclc/clc/lib/generic/math/clc_lgamma.cl
libclc/clc/lib/generic/math/clc_lgamma.inc
libclc/clc/lib/generic/math/clc_lgamma_r.cl
libclc/clc/lib/generic/math/clc_lgamma_r.inc
libclc/clc/lib/generic/math/clc_tgamma.cl
Modified:
libclc/clc/lib/generic/SOURCES
libclc/generic/include/clc/math/lgamma_r.h
libclc/generic/lib/math/lgamma.cl
libclc/generic/lib/math/lgamma_r.cl
libclc/generic/lib/math/tgamma.cl
Removed:
libclc/generic/include/clc/math/lgamma_r.inc
libclc/generic/lib/math/lgamma_r.inc
diff --git a/libclc/generic/include/clc/math/lgamma_r.inc
b/libclc/clc/include/clc/math/clc_lgamma.h
similarity index 57%
rename from libclc/generic/include/clc/math/lgamma_r.inc
rename to libclc/clc/include/clc/math/clc_lgamma.h
index 0f78c0e686386..7e516c1f08793 100644
--- a/libclc/generic/include/clc/math/lgamma_r.inc
+++ b/libclc/clc/include/clc/math/clc_lgamma.h
@@ -6,6 +6,15 @@
//
//===--===//
-_CLC_OVERLOAD _CLC_DECL __CLC_GENTYPE lgamma_r(__CLC_GENTYPE x, global
__CLC_INTN *iptr);
-_CLC_OVERLOAD _CLC_DECL __CLC_GENTYPE lgamma_r(__CLC_GENTYPE x, local
__CLC_INTN *iptr);
-_CLC_OVERLOAD _CLC_DECL __CLC_GENTYPE lgamma_r(__CLC_GENTYPE x, private
__CLC_INTN *iptr);
+#ifndef __CLC_MATH_CLC_LGAMMA_H__
+#define __CLC_MATH_CLC_LGAMMA_H__
+
+#define __CLC_BODY
+#define __CLC_FUNCTION __clc_lgamma
+
+#include
+
+#undef __CLC_BODY
+#undef __CLC_FUNCTION
+
+#endif // __CLC_MATH_CLC_LGAMMA_H__
diff --git a/libclc/clc/include/clc/math/clc_lgamma_r.h
b/libclc/clc/include/clc/math/clc_lgamma_r.h
new file mode 100644
index 0..5db55a13b4d82
--- /dev/null
+++ b/libclc/clc/include/clc/math/clc_lgamma_r.h
@@ -0,0 +1,20 @@
+//===--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===--===//
+
+#ifndef __CLC_MATH_CLC_LGAMMA_R_H__
+#define __CLC_MATH_CLC_LGAMMA_R_H__
+
+#define __CLC_FUNCTION __clc_lgamma_r
+#define __CLC_BODY
+
+#include
+
+#undef __CLC_BODY
+#undef __CLC_FUNCTION
+
+#endif // __CLC_MATH_CLC_LGAMMA_R_H__
diff --git a/libclc/clc/include/clc/math/clc_tgamma.h
b/libclc/clc/include/clc/math/clc_tgamma.h
new file mode 100644
index 0..c4e5714616fae
--- /dev/null
+++ b/libclc/clc/include/clc/math/clc_tgamma.h
@@ -0,0 +1,20 @@
+//===--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===--===//
+
+#ifndef __CLC_MATH_CLC_TGAMMA_H__
+#define __CLC_MATH_CLC_TGAMMA_H__
+
+#define __CLC_BODY
+#define __CLC_FUNCTION __clc_tgamma
+
+#include
+
+#undef __CLC_BODY
+#undef __CLC_FUNCTION
+
+#endif // __CLC_MATH_CLC_TGAMMA_H__
diff --git a/libclc/generic/lib/math/lgamma_r.inc
b/libclc/clc/include/clc/math/unary_def_via_fp32.inc
similarity index 62%
rename from libclc/generic/lib/math/lgamma_r.inc
rename to libclc/clc/include/clc/math/unary_def_via_fp32.inc
index 89cb52ea7a78a..aefee6170571d 100644
--- a/libclc/generic/lib/math/lgamma_r.inc
+++ b/libclc/clc/include/clc/math/unary_def_via_fp32.inc
@@ -6,9 +6,6 @@
//
//===--===//
-_CLC_OVERLOAD _CLC_DEF __CLC_GENTYPE lgamma_r(__CLC_GENTYPE x, __CLC_ADDRSPACE
__CLC_INTN *iptr) {
-__CLC_INTN private
[clang] [Clang] Add warning message for C++17 alias template CTAD (PR #133806)
https://github.com/GeorgeKA updated
https://github.com/llvm/llvm-project/pull/133806
>From dd978982b2ab41d3d2a55abb448e653a80158ecd Mon Sep 17 00:00:00 2001
From: George Asante
Date: Mon, 31 Mar 2025 17:41:20 -0400
Subject: [PATCH 1/3] Add warning message for C++17 alias template CTAD
---
clang/include/clang/Basic/DiagnosticSemaKinds.td| 13 ++---
clang/lib/Sema/SemaInit.cpp | 4 +++-
clang/test/SemaCXX/cxx17-compat.cpp | 2 +-
.../cxx1z-class-template-argument-deduction.cpp | 4 ++--
4 files changed, 16 insertions(+), 7 deletions(-)
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index c77cde297dc32..6f71628f19a4a 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -8444,9 +8444,16 @@ let CategoryName = "Lambda Issue" in {
"C++ standards before C++20">, InGroup, DefaultIgnore;
// C++20 class template argument deduction for alias templates.
- def warn_cxx17_compat_ctad_for_alias_templates : Warning<
- "class template argument deduction for alias templates is incompatible with "
- "C++ standards before C++20">, InGroup, DefaultIgnore;
+ def warn_cxx17_compat_ctad_for_alias_templates
+ : Warning<"class template argument deduction for alias templates is "
+"incompatible with "
+"C++ standards before C++20">,
+InGroup,
+DefaultIgnore;
+ def ext_ctad_for_alias_templates_cxx20
+ : ExtWarn<"class template argument deduction for alias templates is a "
+"C++20 extension">,
+InGroup;
}
def err_return_in_captured_stmt : Error<
diff --git a/clang/lib/Sema/SemaInit.cpp b/clang/lib/Sema/SemaInit.cpp
index cea121d576c5c..7229cd42f85d0 100644
--- a/clang/lib/Sema/SemaInit.cpp
+++ b/clang/lib/Sema/SemaInit.cpp
@@ -9896,7 +9896,9 @@ QualType
Sema::DeduceTemplateSpecializationFromInitializer(
if (auto *AliasTemplate = dyn_cast_or_null(
TemplateName.getAsTemplateDecl())) {
Diag(Kind.getLocation(),
- diag::warn_cxx17_compat_ctad_for_alias_templates);
+ getLangOpts().CPlusPlus20
+ ? diag::warn_cxx17_compat_ctad_for_alias_templates
+ : diag::ext_ctad_for_alias_templates_cxx20);
LookupTemplateDecl = AliasTemplate;
auto UnderlyingType = AliasTemplate->getTemplatedDecl()
->getUnderlyingType()
diff --git a/clang/test/SemaCXX/cxx17-compat.cpp
b/clang/test/SemaCXX/cxx17-compat.cpp
index 54ea3384022d4..81b3e1fde5493 100644
--- a/clang/test/SemaCXX/cxx17-compat.cpp
+++ b/clang/test/SemaCXX/cxx17-compat.cpp
@@ -137,7 +137,7 @@ template struct A { A(T); };
template using B = A;
B b = {1};
#if __cplusplus <= 201703L
- // FIXME: diagnose as well
+ // expected-warning@-2 {{class template argument deduction for alias
templates is a C++20 extension}}
#else
// expected-warning@-4 {{class template argument deduction for alias
templates is incompatible with C++ standards before C++20}}
#endif
diff --git a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
index 9aaa13d7ac41a..a7d740e66ba63 100644
--- a/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
+++ b/clang/test/SemaCXX/cxx1z-class-template-argument-deduction.cpp
@@ -113,9 +113,9 @@ namespace dependent {
};
template void f() {
typename T::X tx = 0;
-typename T::Y ty = 0;
+typename T::Y ty = 0; // expected-warning {{class template argument
deduction for alias templates is a C++20 extension}}
}
- template void f();
+ template void f(); // expected-note {{in instantiation of function
template specialization 'dependent::f' requested here}}
template struct C { C(T); };
template C(T) -> C;
>From c48c4a60c1ed8aebff27bf8a7bb108dd3a9ad3ef Mon Sep 17 00:00:00 2001
From: George Asante
Date: Mon, 31 Mar 2025 18:38:06 -0400
Subject: [PATCH 2/3] Switched to using multiclasses for both messages
---
clang/include/clang/Basic/DiagnosticSemaKinds.td | 14 ++
clang/lib/Sema/SemaInit.cpp | 4 ++--
2 files changed, 4 insertions(+), 14 deletions(-)
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index 6f71628f19a4a..79389eebd81f1 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -49,6 +49,8 @@ defm constexpr_ctor_missing_init : CXX20Compat<
defm adl_only_template_id : CXX20Compat<
"use of function template name with no prior declaration in function call "
"with explicit template arguments is">;
+defm ctad_for_alias_templates
+: CXX20Compat<"class template argument deduction for alias templates is">;
// C++23 compatibility with C++20 and earlier.
defm constexpr_static_var : CXX
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
https://github.com/AlexVlx updated
https://github.com/llvm/llvm-project/pull/134016
>From 91eeaf02336e539f14dcb0a79ff15dbe8befe6f1 Mon Sep 17 00:00:00 2001
From: Alex Voicu
Date: Wed, 2 Apr 2025 02:47:42 +0100
Subject: [PATCH 1/9] Add the functional identity and feature queries.
---
clang/docs/LanguageExtensions.rst | 110 ++
clang/include/clang/Basic/BuiltinsAMDGPU.def | 5 +
.../clang/Basic/DiagnosticSemaKinds.td| 10 +
clang/lib/Basic/Targets/SPIR.cpp | 4 +
clang/lib/Basic/Targets/SPIR.h| 4 +
clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp | 29 ++
clang/lib/Sema/SemaExpr.cpp | 157
clang/test/CodeGen/amdgpu-builtin-cpu-is.c| 65
.../CodeGen/amdgpu-builtin-is-invocable.c | 64
.../amdgpu-feature-builtins-invalid-use.cpp | 43 +++
llvm/lib/Target/AMDGPU/AMDGPU.h | 9 +
.../AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp | 207 ++
llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def | 2 +
.../lib/Target/AMDGPU/AMDGPUTargetMachine.cpp | 3 +-
llvm/lib/Target/AMDGPU/CMakeLists.txt | 1 +
...pu-expand-feature-predicates-unfoldable.ll | 28 ++
.../amdgpu-expand-feature-predicates.ll | 359 ++
17 files changed, 1099 insertions(+), 1 deletion(-)
create mode 100644 clang/test/CodeGen/amdgpu-builtin-cpu-is.c
create mode 100644 clang/test/CodeGen/amdgpu-builtin-is-invocable.c
create mode 100644 clang/test/CodeGen/amdgpu-feature-builtins-invalid-use.cpp
create mode 100644 llvm/lib/Target/AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp
create mode 100644
llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates-unfoldable.ll
create mode 100644 llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates.ll
diff --git a/clang/docs/LanguageExtensions.rst
b/clang/docs/LanguageExtensions.rst
index 3b8a9cac6587a..8a7cb75af13e5 100644
--- a/clang/docs/LanguageExtensions.rst
+++ b/clang/docs/LanguageExtensions.rst
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
+
+ for (; __builtin_amdgcn_is_invocable(__builtin_amdgcn_permlane64); ++*p)
break;
+
+**Description**:
+
+When used as the predicate value of the following control structures:
+
+.. code-block:: c++
+
+ if (...)
+ while (...)
+ do { } while (...)
+ for (...)
+
+be it directly, or as arguments to logical operators such as ``!, ||, &&``, the
+builtins return a boolean value that:
+
+* indicates whether the current target matches the argument; the argument MUST
+ be a string literal and a valid AMDGPU target
+* indicates whether the builtin function passed as the argument can be invoked
+ by the current target; the argument MUST be either a generic or AMDGPU
+ specific builtin name
+
+Outside of these contexts, the builtins have a ``void`` returning signature
+which prevents their misuse.
+
+**Example of invalid use**:
+
+.. code-block:: c++
+
+ void kernel(int* p, int x, bool (*pfn)(bool), const char* str) {
+if (__builtin_amdgcn_processor_is("not_an_amdgcn_gfx_id")) return;
+else if (__bui
[clang] [clang] Add SPIR-V to some OpenMP clang tests (PR #133503)
https://github.com/jhuber6 approved this pull request. Alright, LG. Unrelated, but it would be nice if someone who knows SPIR-V better could look at adding similar builtins for the ones we need to get it working through OpenMP. I've landed some merges that should make that work, but it seems that most of the SPIR-V support isn't intended to be used for compute. It's not that hard if we just do the current 'status quo' and emit intrinsics for the target. https://github.com/llvm/llvm-project/pull/133503 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][modules] Determine if the SDK supports builtin modules independent of the target (PR #134005)
@@ -62,6 +63,28 @@ DarwinSDKInfo::RelatedTargetVersionMapping::parseJSON(
Min, Max, MinValue, MaximumDeploymentTarget, std::move(Mapping));
}
+static llvm::Triple::OSType parseOS(const llvm::json::Object &Obj) {
+ // The CanonicalName is the Xcode platform, a version, and an optional
suffix,
+ // e.g. macosx16.0.
vsapsai wrote:
The example doesn't have a suffix, as far as I can tell. Up to you if you want
to make it more representative.
https://github.com/llvm/llvm-project/pull/134005
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][modules] Determine if the SDK supports builtin modules independent of the target (PR #134005)
@@ -2990,26 +2996,18 @@ static bool sdkSupportsBuiltinModules(
return false;
VersionTuple SDKVersion = SDKInfo->getVersion();
- switch (TargetPlatform) {
+ switch (SDKInfo->getOS()) {
// Existing SDKs added support for builtin modules in the fall
// 2024 major releases.
- case Darwin::MacOS:
+ case llvm::Triple::MacOSX:
return SDKVersion >= VersionTuple(15U);
- case Darwin::IPhoneOS:
-switch (TargetEnvironment) {
-case Darwin::MacCatalyst:
vsapsai wrote:
How is it handled in the new version? I don't see a special case for
MacCatalyst in the new code, though I could have missed it.
https://github.com/llvm/llvm-project/pull/134005
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
llvm-ci wrote: LLVM Buildbot has detected a new failure on builder `openmp-offload-amdgpu-runtime-2` running on `rocm-worker-hw-02` while building `clang` at step 7 "Add check check-clang". Full details are available at: https://lab.llvm.org/buildbot/#/builders/10/builds/2698 Here is the relevant piece of the build log for the reference ``` Step 7 (Add check check-clang) failure: test (failure) TEST 'Clang :: Driver/clang-sycl-linker-test.cpp' FAILED Exit Code: 1 Command Output (stderr): -- /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc # RUN: at line 4 + /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc # RUN: at line 5 + /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang-sycl-linker --dry-run -v -triple=spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc -o a.spv 2>&1| /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/FileCheck /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp --check-prefix=SIMPLE-FO # RUN: at line 6 + /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/clang-sycl-linker --dry-run -v -triple=spirv64 /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc -o a.spv + /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/bin/FileCheck /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp --check-prefix=SIMPLE-FO /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp:9:20: error: SIMPLE-FO-NEXT: expected string not found in input // SIMPLE-FO-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv ^ :1:349: note: scanning from here sycl-device-link: inputs: /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc, /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc libfiles: output: /tmp/lit-tmp-faew3ndu/a.spv-b34085.bc ^ :1:349: note: with "LLVMLINKOUT" equal to "/tmp/lit-tmp-faew3ndu/a\\.spv-b34085" sycl-device-link: inputs: /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc, /home/botworker/builds/openmp-offload-amdgpu-runtime-2/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc libfiles: output: /tmp/lit-tmp-faew3ndu/a.spv-b34085.bc
[clang] fixed clang frontend crash with friend class declaration and overload == (PR #133878)
https://github.com/Ankur-0429 updated
https://github.com/llvm/llvm-project/pull/133878
>From 289f8630dccf916b8cf7cc43758bae60df036c5d Mon Sep 17 00:00:00 2001
From: Ankur Ahir
Date: Wed, 2 Apr 2025 00:53:33 -0700
Subject: [PATCH 1/4] clang frontend crash with friend class declaration and
overloaded operator ==
---
clang/docs/ReleaseNotes.rst | 1 +
clang/lib/Sema/SemaDeclCXX.cpp | 3 +--
.../class/class.compare/class.compare.default/p1.cpp | 12
3 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index daad01919ecd4..431ae0863d9e4 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -342,6 +342,7 @@ Bug Fixes to Attribute Support
Bug Fixes to C++ Support
+- Clang Now Supports Implicitly-Defined Comparison Operators for
Friend-Declarations
- Clang now diagnoses copy constructors taking the class by value in template
instantiations. (#GH130866)
- Clang is now better at keeping track of friend function template instance
contexts. (#GH55509)
- Clang now prints the correct instantiation context for diagnostics suppressed
diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp
index 676d53a1f4b45..d1516860dfe71 100644
--- a/clang/lib/Sema/SemaDeclCXX.cpp
+++ b/clang/lib/Sema/SemaDeclCXX.cpp
@@ -9037,8 +9037,7 @@ bool Sema::CheckExplicitlyDefaultedComparison(Scope *S,
FunctionDecl *FD,
return true;
if (llvm::none_of(RD->friends(), [&](const FriendDecl *F) {
- return FD->getCanonicalDecl() ==
- F->getFriendDecl()->getCanonicalDecl();
+ return declaresSameEntity(F->getFriendDecl(), FD);
})) {
Diag(FD->getLocation(), diag::err_defaulted_comparison_not_friend)
<< int(DCK) << int(0) << RD;
diff --git a/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
b/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
index a195e0548152d..5db4f05da552c 100644
--- a/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
+++ b/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
@@ -285,3 +285,15 @@ struct j {
};
bool j::operator==(const j &) const = default;
}
+
+namespace evil2 {
+ struct k {
+ };
+
+ struct l {
+ friend bool operator==(const l& a, const l& b);
+ friend class k;
+ };
+
+ bool operator==(const l& a, const l& b) = default;
+}
\ No newline at end of file
>From 78fecd58986e25147b47220ee9ef0a392c45651e Mon Sep 17 00:00:00 2001
From: Ankur Ahir
Date: Wed, 2 Apr 2025 00:53:33 -0700
Subject: [PATCH 2/4] clang frontend crash with friend class declaration and
overloaded operator ==
---
clang/docs/ReleaseNotes.rst | 1 +
clang/lib/Sema/SemaDeclCXX.cpp | 3 +--
.../class/class.compare/class.compare.default/p1.cpp | 12
3 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index daad01919ecd4..88ed3018c58ea 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -342,6 +342,7 @@ Bug Fixes to Attribute Support
Bug Fixes to C++ Support
+- Clang Now supports Implicitly-Defined Comparison Operators for
Friend-Declarations
- Clang now diagnoses copy constructors taking the class by value in template
instantiations. (#GH130866)
- Clang is now better at keeping track of friend function template instance
contexts. (#GH55509)
- Clang now prints the correct instantiation context for diagnostics suppressed
diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp
index 676d53a1f4b45..d1516860dfe71 100644
--- a/clang/lib/Sema/SemaDeclCXX.cpp
+++ b/clang/lib/Sema/SemaDeclCXX.cpp
@@ -9037,8 +9037,7 @@ bool Sema::CheckExplicitlyDefaultedComparison(Scope *S,
FunctionDecl *FD,
return true;
if (llvm::none_of(RD->friends(), [&](const FriendDecl *F) {
- return FD->getCanonicalDecl() ==
- F->getFriendDecl()->getCanonicalDecl();
+ return declaresSameEntity(F->getFriendDecl(), FD);
})) {
Diag(FD->getLocation(), diag::err_defaulted_comparison_not_friend)
<< int(DCK) << int(0) << RD;
diff --git a/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
b/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
index a195e0548152d..5db4f05da552c 100644
--- a/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
+++ b/clang/test/CXX/class/class.compare/class.compare.default/p1.cpp
@@ -285,3 +285,15 @@ struct j {
};
bool j::operator==(const j &) const = default;
}
+
+namespace evil2 {
+ struct k {
+ };
+
+ struct l {
+ friend bool operator==(const l& a, const l& b);
+ friend class k;
+ };
+
+ bool operator==(const l& a, const l& b) = default;
+}
\ No newline at end of f
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
llvm-ci wrote: LLVM Buildbot has detected a new failure on builder `openmp-offload-sles-build-only` running on `rocm-worker-hw-04-sles` while building `clang` at step 6 "Add check check-clang". Full details are available at: https://lab.llvm.org/buildbot/#/builders/140/builds/20355 Here is the relevant piece of the build log for the reference ``` Step 6 (Add check check-clang) failure: test (failure) TEST 'Clang :: Driver/clang-sycl-linker-test.cpp' FAILED Exit Code: 1 Command Output (stderr): -- /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc # RUN: at line 4 + /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc # RUN: at line 5 + /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang --driver-mode=g++ -emit-llvm -c -target spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp -o /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang-sycl-linker --dry-run -v -triple=spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc -o a.spv 2>&1| /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/FileCheck /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp --check-prefix=SIMPLE-FO # RUN: at line 6 + /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/clang-sycl-linker --dry-run -v -triple=spirv64 /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc -o a.spv + /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/bin/FileCheck /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp --check-prefix=SIMPLE-FO /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.src/clang/test/Driver/clang-sycl-linker-test.cpp:9:20: error: SIMPLE-FO-NEXT: expected string not found in input // SIMPLE-FO-NEXT: SPIR-V Backend: input: [[LLVMLINKOUT]].bc, output: a.spv ^ :1:347: note: scanning from here sycl-device-link: inputs: /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc, /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc libfiles: output: /tmp/lit-tmp-xkwu2lef/a.spv-5691cc.bc ^ :1:347: note: with "LLVMLINKOUT" equal to "/tmp/lit-tmp-xkwu2lef/a\\.spv-5691cc" sycl-device-link: inputs: /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_1.bc, /home/botworker/bbot/builds/openmp-offload-sles-build/llvm.build/tools/clang/test/Driver/Output/clang-sycl-linker-test.cpp.tmp_2.bc libfiles: output: /tmp/lit-tmp-xkwu2lef/a.spv-5691cc.bc
[clang] [SYCL][SPIR-V Backend][clang-sycl-linker] Add SPIR-V backend support inside clang-sycl-linker (PR #133967)
asudarsa wrote: @sarnex, @jhuber6 Thanks much for your kind feedbacks. This is ready to be merged. Can one of you please help? Sincerely https://github.com/llvm/llvm-project/pull/133967 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [AMDGPU][clang][CodeGen][opt] Add late-resolved feature identifying predicates (PR #134016)
https://github.com/AlexVlx updated
https://github.com/llvm/llvm-project/pull/134016
>From 91eeaf02336e539f14dcb0a79ff15dbe8befe6f1 Mon Sep 17 00:00:00 2001
From: Alex Voicu
Date: Wed, 2 Apr 2025 02:47:42 +0100
Subject: [PATCH 01/10] Add the functional identity and feature queries.
---
clang/docs/LanguageExtensions.rst | 110 ++
clang/include/clang/Basic/BuiltinsAMDGPU.def | 5 +
.../clang/Basic/DiagnosticSemaKinds.td| 10 +
clang/lib/Basic/Targets/SPIR.cpp | 4 +
clang/lib/Basic/Targets/SPIR.h| 4 +
clang/lib/CodeGen/TargetBuiltins/AMDGPU.cpp | 29 ++
clang/lib/Sema/SemaExpr.cpp | 157
clang/test/CodeGen/amdgpu-builtin-cpu-is.c| 65
.../CodeGen/amdgpu-builtin-is-invocable.c | 64
.../amdgpu-feature-builtins-invalid-use.cpp | 43 +++
llvm/lib/Target/AMDGPU/AMDGPU.h | 9 +
.../AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp | 207 ++
llvm/lib/Target/AMDGPU/AMDGPUPassRegistry.def | 2 +
.../lib/Target/AMDGPU/AMDGPUTargetMachine.cpp | 3 +-
llvm/lib/Target/AMDGPU/CMakeLists.txt | 1 +
...pu-expand-feature-predicates-unfoldable.ll | 28 ++
.../amdgpu-expand-feature-predicates.ll | 359 ++
17 files changed, 1099 insertions(+), 1 deletion(-)
create mode 100644 clang/test/CodeGen/amdgpu-builtin-cpu-is.c
create mode 100644 clang/test/CodeGen/amdgpu-builtin-is-invocable.c
create mode 100644 clang/test/CodeGen/amdgpu-feature-builtins-invalid-use.cpp
create mode 100644 llvm/lib/Target/AMDGPU/AMDGPUExpandPseudoIntrinsics.cpp
create mode 100644
llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates-unfoldable.ll
create mode 100644 llvm/test/CodeGen/AMDGPU/amdgpu-expand-feature-predicates.ll
diff --git a/clang/docs/LanguageExtensions.rst
b/clang/docs/LanguageExtensions.rst
index 3b8a9cac6587a..8a7cb75af13e5 100644
--- a/clang/docs/LanguageExtensions.rst
+++ b/clang/docs/LanguageExtensions.rst
@@ -4920,6 +4920,116 @@ If no address spaces names are provided, all address
spaces are fenced.
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local")
__builtin_amdgcn_fence(__ATOMIC_SEQ_CST, "workgroup", "local", "global")
+__builtin_amdgcn_processor_is and __builtin_amdgcn_is_invocable
+^^^
+
+``__builtin_amdgcn_processor_is`` and ``__builtin_amdgcn_is_invocable`` provide
+a functional mechanism for programatically querying:
+
+* the identity of the current target processor;
+* the capability of the current target processor to invoke a particular
builtin.
+
+**Syntax**:
+
+.. code-block:: c
+
+ // When used as the predicate for a control structure
+ bool __builtin_amdgcn_processor_is(const char*);
+ bool __builtin_amdgcn_is_invocable(builtin_name);
+ // Otherwise
+ void __builtin_amdgcn_processor_is(const char*);
+ void __builtin_amdgcn_is_invocable(void);
+
+**Example of use**:
+
+.. code-block:: c++
+
+ if (__builtin_amdgcn_processor_is("gfx1201") ||
+ __builtin_amdgcn_is_invocable(__builtin_amdgcn_s_sleep_var))
+__builtin_amdgcn_s_sleep_var(x);
+
+ if (!__builtin_amdgcn_processor_is("gfx906"))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_processor_is("gfx1010") ||
+ __builtin_amdgcn_processor_is("gfx1101"))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ while (__builtin_amdgcn_processor_is("gfx1101")) *p += x;
+
+ do { *p -= x; } while (__builtin_amdgcn_processor_is("gfx1010"));
+
+ for (; __builtin_amdgcn_processor_is("gfx1201"); ++*p) break;
+
+ if
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_wait_event_export_ready))
+__builtin_amdgcn_s_wait_event_export_ready();
+ else if (__builtin_amdgcn_is_invocable(__builtin_amdgcn_s_ttracedata_imm))
+__builtin_amdgcn_s_ttracedata_imm(1);
+
+ do {
+*p -= x;
+ } while
(__builtin_amdgcn_is_invocable(__builtin_amdgcn_global_load_tr_b64_i32));
+
+ for (; __builtin_amdgcn_is_invocable(__builtin_amdgcn_permlane64); ++*p)
break;
+
+**Description**:
+
+When used as the predicate value of the following control structures:
+
+.. code-block:: c++
+
+ if (...)
+ while (...)
+ do { } while (...)
+ for (...)
+
+be it directly, or as arguments to logical operators such as ``!, ||, &&``, the
+builtins return a boolean value that:
+
+* indicates whether the current target matches the argument; the argument MUST
+ be a string literal and a valid AMDGPU target
+* indicates whether the builtin function passed as the argument can be invoked
+ by the current target; the argument MUST be either a generic or AMDGPU
+ specific builtin name
+
+Outside of these contexts, the builtins have a ``void`` returning signature
+which prevents their misuse.
+
+**Example of invalid use**:
+
+.. code-block:: c++
+
+ void kernel(int* p, int x, bool (*pfn)(bool), const char* str) {
+if (__builtin_amdgcn_processor_is("not_an_amdgcn_gfx_id")) return;
+else if (__b
[clang] Hlsl dst function (PR #133828)
@@ -0,0 +1,37 @@
+// RUN: %clang_cc1 -finclude-default-header -triple
dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only
-disable-llvm-passes -verify
+
+float4 test_too_many_arg(float4 p0)
+{
+dst(p0, p0, p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 3 were provided}}
+}
+
+float4 test_no_second_arg(float4 p0)
+{
+return dst(p0);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 1 was provided}}
+}
+
+float4 test_no_args()
+{
+return dst();
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
requires 2 arguments, but 0 were provided}}
+}
+
+float4 test_3_components(float3 p0, float3 p1)
+{
+return dst(p0, p1);
+ // expected-error@-1 {{no matching function for call to 'dst'}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector<[...], 3>' to 'vector<[...], 4>' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+ // expected-note@hlsl/hlsl_intrinsics.h:* {{candidate function not viable:
no known conversion from 'vector' to 'vector' for 1st
argument}}
+}
farzonl wrote:
If this doesn't warn then don't worry we don't need this test case.
https://github.com/llvm/llvm-project/pull/133828
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][CodeGen] Add range metadata for atomic load of boolean type. #131476 (PR #133546)
Jan =?utf-8?q?Górski?= ,
Jan =?utf-8?q?Górski?=
Message-ID:
In-Reply-To:
@@ -590,6 +590,29 @@ static void EmitAtomicOp(CodeGenFunction &CGF, AtomicExpr
*E, Address Dest,
llvm::LoadInst *Load = CGF.Builder.CreateLoad(Ptr);
Load->setAtomic(Order, Scope);
Load->setVolatile(E->isVolatile());
+
+if (CGF.CGM.getCodeGenOpts().OptimizationLevel == 0) {
+ CGF.Builder.CreateStore(Load, Dest);
+ return;
+}
+
+QualType Ty = E->getValueType();
+if (!Ty->isBooleanType()) {
+ CGF.Builder.CreateStore(Load, Dest);
+ return;
efriedma-quic wrote:
Using an early return like this is weird; just stick a normal if statement
around the code that creates the metadata.
https://github.com/llvm/llvm-project/pull/133546
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][CodeGen] Add range metadata for atomic load of boolean type. #131476 (PR #133546)
Jan =?utf-8?q?Górski?= ,
Jan =?utf-8?q?Górski?=
Message-ID:
In-Reply-To:
@@ -590,6 +590,29 @@ static void EmitAtomicOp(CodeGenFunction &CGF, AtomicExpr
*E, Address Dest,
llvm::LoadInst *Load = CGF.Builder.CreateLoad(Ptr);
Load->setAtomic(Order, Scope);
Load->setVolatile(E->isVolatile());
+
+if (CGF.CGM.getCodeGenOpts().OptimizationLevel == 0) {
+ CGF.Builder.CreateStore(Load, Dest);
+ return;
+}
+
+QualType Ty = E->getValueType();
+if (!Ty->isBooleanType()) {
+ CGF.Builder.CreateStore(Load, Dest);
+ return;
+}
+
+llvm::MDBuilder MDHelper(CGF.getLLVMContext());
+llvm::APInt BooleanMin = llvm::APInt(CGF.getContext().getTypeSize(Ty), 0);
+llvm::APInt BooleanEnd = llvm::APInt(CGF.getContext().getTypeSize(Ty), 2);
+
+if (llvm::MDNode *RangeInfo =
+MDHelper.createRange(BooleanMin, BooleanEnd)) {
efriedma-quic wrote:
Please refactor the code from CodeGenFunction::EmitLoadOfScalar, instead of
reinventing the code to compute the range metadata.
https://github.com/llvm/llvm-project/pull/133546
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
