[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
https://github.com/cor3ntin edited https://github.com/llvm/llvm-project/pull/139429 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 %s -verify -Wimplicit-int-conversion
+// RUN: %clang_cc1 %s -verify -Wimplicit-int-conversion
-Wno-implicit-int-conversion-on-negation -DNO_DIAG
+
+#ifdef NO_DIAG
+unsigned char test_no_diag(unsigned char x) {
+return -x; // expected-no-diagnostics
+}
+#else
+unsigned char test_diag(unsigned char x) {
+return -x; // expected-warning {{implicit conversion loses integer
precision: 'int' to 'unsigned char' on negation}}
+}
+#endif
cor3ntin wrote:
```suggestion
// RUN: %clang_cc1 %s -verify=nowarn -Wimplicit-int-conversion
// RUN: %clang_cc1 %s -verify -Wimplicit-int-conversion
-Wno-implicit-int-conversion-on-negation -DNO_DIAG
// nowarn-no-diagnostics
unsigned char test_no_diag(unsigned char x) {
return -x; // expected-warning {{implicit conversion loses integer
precision: 'int' to 'unsigned char' on negation}}
}
```
https://github.com/llvm/llvm-project/pull/139429
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
@@ -12091,6 +12091,11 @@ void Sema::CheckImplicitConversion(Expr *E, QualType
T, SourceLocation CC,
if (SourceMgr.isInSystemMacro(CC))
return;
+if (const auto *UO = dyn_cast(E)) {
+ return DiagnoseImpCast(*this, E, T, CC,
+ diag::warn_impcast_integer_precision_on_negation);
+}
+
cor3ntin wrote:
We should check which operator it is. Did you try to add tests with
```
unsigned char test_diag(unsigned char x) {
return +x;
}```
https://github.com/llvm/llvm-project/pull/139429
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [CLANGD] [NFC] Fix proposed by sanitizer. (PR #140116)
https://github.com/carlosgalvezp approved this pull request. LGTM. In the future, please provide more details about exactly which tool you are using, which you call "sanitizer". For example, what is the name of the tool? "sanitizers" are tools to detect issues at runtime; here you are using a static analyzer. https://github.com/llvm/llvm-project/pull/140116 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
https://github.com/cor3ntin commented: The direction looks good. Can you add more tests, with different integer types? Thanks https://github.com/llvm/llvm-project/pull/139429 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] fix constexpr-unknown handling of self-references. (PR #132990)
https://github.com/cor3ntin approved this pull request. LGTM modulo nit https://github.com/llvm/llvm-project/pull/132990 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] fix constexpr-unknown handling of self-references. (PR #132990)
@@ -3629,17 +3644,17 @@ static bool evaluateVarDeclInit(EvalInfo &Info, const
Expr *E,
Result = VD->getEvaluatedValue();
- // C++23 [expr.const]p8
- // ... For such an object that is not usable in constant expressions, the
- // dynamic type of the object is constexpr-unknown. For such a reference that
- // is not usable in constant expressions, the reference is treated as binding
- // to an unspecified object of the referenced type whose lifetime and that of
- // all subobjects includes the entire constant evaluation and whose dynamic
- // type is constexpr-unknown.
- if (AllowConstexprUnknown) {
-if (!Result)
+ if (!Result) {
+if (AllowConstexprUnknown) {
Result = &Info.CurrentCall->createConstexprUnknownAPValues(VD, Base);
+} else {
+ return false;
+}
cor3ntin wrote:
ping
https://github.com/llvm/llvm-project/pull/132990
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] fix constexpr-unknown handling of self-references. (PR #132990)
https://github.com/cor3ntin edited https://github.com/llvm/llvm-project/pull/132990 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Sycl builtin kernel name (PR #140230)
https://github.com/Fznamznon created
https://github.com/llvm/llvm-project/pull/140230
None
>From eeea84b4989b4497cc804f403258b986c09cf5ab Mon Sep 17 00:00:00 2001
From: Elizabeth Andrews
Date: Thu, 8 Aug 2024 18:05:48 -0700
Subject: [PATCH 1/5] [SYCL] Add support for __builtin_sycl_kernel_name
The builtin takes the kernel name type as it's argument
and returns the mangled name for the kernel caller
function.
---
clang/include/clang/Basic/Builtins.h | 1 +
clang/include/clang/Basic/Builtins.td | 7 +++
.../clang/Basic/DiagnosticSemaKinds.td| 3 +
clang/lib/Basic/Builtins.cpp | 3 +
clang/lib/CodeGen/CGBuiltin.cpp | 20 +++
clang/lib/Sema/SemaChecking.cpp | 33 ++
.../CodeGenSYCL/builtin-sycl-kernel-name.cpp | 60 +++
.../SemaSYCL/builtin-sycl-kernel-name.cpp | 44 ++
8 files changed, 171 insertions(+)
create mode 100644 clang/test/CodeGenSYCL/builtin-sycl-kernel-name.cpp
create mode 100644 clang/test/SemaSYCL/builtin-sycl-kernel-name.cpp
diff --git a/clang/include/clang/Basic/Builtins.h
b/clang/include/clang/Basic/Builtins.h
index 3a5e31de2bc50..0bea100b7e6de 100644
--- a/clang/include/clang/Basic/Builtins.h
+++ b/clang/include/clang/Basic/Builtins.h
@@ -45,6 +45,7 @@ enum LanguageID : uint16_t {
ALL_OCL_LANGUAGES = 0x800, // builtin for OCL languages.
HLSL_LANG = 0x1000,// builtin requires HLSL.
C23_LANG = 0x2000, // builtin requires C23 or later.
+ SYCL_LANG = 0x2000,// builtin requires SYCL.
ALL_LANGUAGES = C_LANG | CXX_LANG | OBJC_LANG, // builtin for all languages.
ALL_GNU_LANGUAGES = ALL_LANGUAGES | GNU_LANG, // builtin requires GNU mode.
ALL_MS_LANGUAGES = ALL_LANGUAGES | MS_LANG // builtin requires MS mode.
diff --git a/clang/include/clang/Basic/Builtins.td
b/clang/include/clang/Basic/Builtins.td
index 11b1e247237a7..13e02898e55aa 100644
--- a/clang/include/clang/Basic/Builtins.td
+++ b/clang/include/clang/Basic/Builtins.td
@@ -4794,6 +4794,13 @@ def GetDeviceSideMangledName : LangBuiltin<"CUDA_LANG"> {
let Prototype = "char const*(...)";
}
+// SYCL
+def SYCLKernelName : LangBuiltin<"SYCL_LANG"> {
+ let Spellings = ["__builtin_sycl_kernel_name"];
+ let Attributes = [NoThrow, CustomTypeChecking];
+ let Prototype = "char const*(...)";
+}
+
// HLSL
def HLSLAddUint64: LangBuiltin<"HLSL_LANG"> {
let Spellings = ["__builtin_hlsl_adduint64"];
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index 6e940a318b61d..d9f7487f2596e 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -12789,6 +12789,9 @@ def err_sycl_entry_point_deduced_return_type : Error<
def warn_cuda_maxclusterrank_sm_90 : Warning<
"maxclusterrank requires sm_90 or higher, CUDA arch provided: %0, ignoring "
"%1 attribute">, InGroup;
+def err_builtin_invalid_argument_count : Error<"builtin requires exactly 1
argument">;
+def err_builtin_invalid_argument : Error<"invalid argument; argument must be a
class or struct type"
+ " with a member type alias named
'type'">;
// VTable pointer authentication errors
def err_non_polymorphic_vtable_pointer_auth : Error<
diff --git a/clang/lib/Basic/Builtins.cpp b/clang/lib/Basic/Builtins.cpp
index 885abdc152e3a..f8275656c5d0c 100644
--- a/clang/lib/Basic/Builtins.cpp
+++ b/clang/lib/Basic/Builtins.cpp
@@ -185,6 +185,9 @@ static bool builtinIsSupported(const llvm::StringTable
&Strings,
/* CUDA Unsupported */
if (!LangOpts.CUDA && BuiltinInfo.Langs == CUDA_LANG)
return false;
+ /* SYCL Unsupported */
+ if (!LangOpts.isSYCL() && BuiltinInfo.Langs == SYCL_LANG)
+return false;
/* CPlusPlus Unsupported */
if (!LangOpts.CPlusPlus && BuiltinInfo.Langs == CXX_LANG)
return false;
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 4fdf2113cb9dc..b7638ed0b27f1 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -29,6 +29,7 @@
#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "clang/Basic/TargetOptions.h"
+#include "clang/Basic/IdentifierTable.h"
#include "clang/Frontend/FrontendDiagnostic.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/Intrinsics.h"
@@ -6219,6 +6220,25 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl
GD, unsigned BuiltinID,
auto Str = CGM.GetAddrOfConstantCString(Name, "");
return RValue::get(Str.getPointer());
}
+ case Builtin::BI__builtin_sycl_kernel_name: {
+ASTContext &Ctx = getContext();
+// Argument to the builtin is a kernel_id_t type trait which is used
+// to retrieve the kernel name type.
+RecordDecl *RD = E->getArg(0)->getType()->castAs()->getDecl();
+IdentifierTable &IdentTable = Ctx.Idents;
+auto Name = DeclarationName(&(IdentTab
[clang] [clang][NFC] Clean up Expr::isTemporaryObject() (PR #140229)
https://github.com/tbaederr updated
https://github.com/llvm/llvm-project/pull/140229
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,sans-serif,Apple Color Emoji,Segoe UI Emoji,Segoe UI Symbol;
font-size: 14px;
line-height: 1.5;
margin: 0;
}
.container { margin: 50px auto; max-width: 600px; text-align: center;
padding: 0 24px; }
a { color: #0366d6; text-decoration: none; }
a:hover { text-decoration: underline; }
h1 { line-height: 60px; font-size: 48px; font-weight: 300; margin: 0px;
text-shadow: 0 1px 0 #fff; }
p { color: rgba(0, 0, 0, 0.5); margin: 20px 0 40px; }
ul { list-style: none; margin: 25px 0; padding: 0; }
li { display: table-cell; font-weight: bold; width: 1%; }
.logo { display: inline-block; margin-top: 35px; }
.logo-img-2x { display: none; }
@media
only screen and (-webkit-min-device-pixel-ratio: 2),
only screen and ( min--moz-device-pixel-ratio: 2),
only screen and ( -o-min-device-pixel-ratio: 2/1),
only screen and (min-device-pixel-ratio: 2),
only screen and (min-resolution: 192dpi),
only screen and (min-resolution: 2dppx) {
.logo-img-1x { display: none; }
.logo-img-2x { display: inline-block; }
}
#suggestions {
margin-top: 35px;
color: #ccc;
}
#suggestions a {
color: #66;
font-weight: 200;
font-size: 14px;
margin: 0 10px;
}
Whoa there!
You have exceeded a secondary rate limit.
Please wait a few minutes before you try again;
in some cases this may take up to an hour.
https://support.github.com/contact";>Contact Support —
https://githubstatus.com";>GitHub Status —
https://twitter.com/githubstatus";>@githubstatus
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Sycl builtin kernel name (PR #140230)
https://github.com/Fznamznon closed https://github.com/llvm/llvm-project/pull/140230 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] Sycl builtin kernel name (PR #140230)
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,h --
clang/test/CodeGenSYCL/builtin-sycl-kernel-name.cpp
clang/test/SemaSYCL/builtin-sycl-kernel-name.cpp
clang/test/SemaSYCL/builtin-sycl-kernel-param-count.cpp
clang/include/clang/Basic/Builtins.h clang/lib/AST/ExprConstant.cpp
clang/lib/Basic/Builtins.cpp clang/lib/CodeGen/CGBuiltin.cpp
clang/lib/CodeGen/CodeGenModule.cpp clang/lib/CodeGen/CodeGenModule.h
clang/lib/CodeGen/CodeGenSYCL.cpp clang/lib/Sema/SemaChecking.cpp
``
View the diff from clang-format here.
``diff
diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp
index 555a488b0..975dee423 100644
--- a/clang/lib/AST/ExprConstant.cpp
+++ b/clang/lib/AST/ExprConstant.cpp
@@ -51,8 +51,8 @@
#include "clang/AST/TypeLoc.h"
#include "clang/Basic/Builtins.h"
#include "clang/Basic/DiagnosticSema.h"
-#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/IdentifierTable.h"
+#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "llvm/ADT/APFixedPoint.h"
#include "llvm/ADT/Sequence.h"
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 71a7b4155..e7836eb08 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -26,10 +26,10 @@
#include "TargetInfo.h"
#include "clang/AST/OSLog.h"
#include "clang/AST/StmtVisitor.h"
+#include "clang/Basic/IdentifierTable.h"
#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "clang/Basic/TargetOptions.h"
-#include "clang/Basic/IdentifierTable.h"
#include "clang/Frontend/FrontendDiagnostic.h"
#include "llvm/IR/InlineAsm.h"
#include "llvm/IR/Intrinsics.h"
@@ -2566,7 +2566,7 @@ EmitKernelNameGlobal(CodeGenModule &CGM, ASTContext &Ctx,
const CallExpr *E) {
SmallString<256> KernelNameSymbol;
llvm::raw_svector_ostream Out(KernelNameSymbol);
- //llvm::raw_string_ostream Out(KernelNameSymbol);
+ // llvm::raw_string_ostream Out(KernelNameSymbol);
MC->mangleCanonicalTypeName(KernelNameType, Out);
llvm::GlobalVariable *GV = new llvm::GlobalVariable(
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index 0a1356a38..4b6b769a2 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -2226,7 +2226,7 @@ static bool CheckBuiltinSyclKernelName(Sema &S, CallExpr
*TheCall) {
QualType ArgTy = TheCall->getArg(0)->getType();
const auto *RT = ArgTy->getAs();
- if(!RT)
+ if (!RT)
return true;
RecordDecl *RD = RT->getDecl();
``
https://github.com/llvm/llvm-project/pull/140230
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [flang] add -floop-interchange and enable it with opt levels (PR #140182)
@@ -421,7 +421,8 @@ static void CheckSubscripts(
static void CheckSubscripts(
semantics::SemanticsContext &context, CoarrayRef &ref) {
- const Symbol &coarraySymbol{ref.GetBase().GetLastSymbol()};
+ const auto &base = ref.GetBase();
+ const Symbol &coarraySymbol{base.GetLastSymbol()};
kiranchandramohan wrote:
Nit: Is this an unrelated change?
https://github.com/llvm/llvm-project/pull/140182
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [flang] add -floop-interchange and enable it with opt levels (PR #140182)
https://github.com/kiranchandramohan commented: Thanks for this PR. Do you have any compilation time and performance data? https://github.com/llvm/llvm-project/pull/140182 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] [Cygwin][MinGW] Internal class in explicitly-instantiation-declarated template should be instantiated (PR #140145)
https://github.com/kikairoya updated
https://github.com/llvm/llvm-project/pull/140145
>From 827e89b8148e7eaa69bbae342cfb2a59d1b09133 Mon Sep 17 00:00:00 2001
From: kikairoya
Date: Mon, 21 Apr 2025 23:30:13 +0900
Subject: [PATCH 1/2] [Cygwin][MinGW] Internal class in
explicitly-instantiation-declarated template should be instantiated
In-code comment says "explicit instantiation decl of the outer class
doesn't affect the inner class" but this behavior seems MSVC
specific, MinGW-GCC and Cygwin-GCC does not.
Clang should honor gcc's behavior.
This change fixes std::string compatibilty and resolves strange link
error (statically linked), strange crash (dynamically linked) using
libstdc++ on Cygwin.
---
clang/lib/Sema/SemaTemplateInstantiate.cpp| 1 +
.../CodeGenCXX/mingw-template-dllexport.cpp | 109 +++---
2 files changed, 93 insertions(+), 17 deletions(-)
diff --git a/clang/lib/Sema/SemaTemplateInstantiate.cpp
b/clang/lib/Sema/SemaTemplateInstantiate.cpp
index d028eea4f8f3e..b7c27b3795f5d 100644
--- a/clang/lib/Sema/SemaTemplateInstantiate.cpp
+++ b/clang/lib/Sema/SemaTemplateInstantiate.cpp
@@ -4247,6 +4247,7 @@ Sema::InstantiateClassMembers(SourceLocation
PointOfInstantiation,
continue;
if (Context.getTargetInfo().getTriple().isOSWindows() &&
+ !Context.getTargetInfo().getTriple().isOSCygMing() &&
TSK == TSK_ExplicitInstantiationDeclaration) {
// On Windows, explicit instantiation decl of the outer class doesn't
// affect the inner class. Typically extern template declarations are
diff --git a/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
b/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
index de112d6da53db..a6047b5955e96 100644
--- a/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
+++ b/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
@@ -6,46 +6,121 @@
#define JOIN2(x, y) x##y
#define JOIN(x, y) JOIN2(x, y)
#define UNIQ(name) JOIN(name, __LINE__)
-#define USEMEMFUNC(class, func) void (class::*UNIQ(use)())() { return
&class::func; }
+#define USEMEMFUNC(class, func) auto UNIQ(use) = &class::func;
template
class c {
+ // MinGW-GCC does not apply 'dllexport' to inline member function in
dll-exported template but clang does from long ago.
void f() {}
+ void g();
+ inline static int u = 0;
+ static int v;
};
+template void c::g() {}
+template int c::v = 0;
+// #1
template class __declspec(dllexport) c;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN1cIiE1fEv
-
+// #2
extern template class __declspec(dllexport) c;
template class c;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN1cIcE1fEv
-
+// #3
extern template class c;
-template class __declspec(dllexport) c;
+template class __declspec(dllexport) c; // expected-warning {{
'dllexport' attribute ignored on explicit instantiation definition }}
-// CHECK-NOT: define {{.*}} dllexport {{.*}} @_ZN1cIdE1fEv
template
struct outer {
- void f();
+ void f() {}
+ void g();
+ inline static int u = 0;
+ static int v;
+ // MinGW-GCC and Clang does not apply 'dllexport' to inner type and its
sub-elements in template class.
struct inner {
-void f();
+void f() {}
+void g();
+inline static int u = 0;
+static int v;
};
};
-template void outer::f() {}
-template void outer::inner::f() {}
+template void outer::g() {}
+template void outer::inner::g() {}
+template int outer::v = 0;
+template int outer::inner::v = 0;
-template class __declspec(dllexport) outer;
+// #4
+template struct __declspec(dllexport) outer;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN5outerIiE1fEv
-// CHECK-NOT: define {{.*}} dllexport {{.*}} @_ZN5outerIiE5inner1fEv
-
-extern template class __declspec(dllimport) outer;
+// #5
+extern template struct __declspec(dllimport) outer;
USEMEMFUNC(outer, f)
+USEMEMFUNC(outer, g)
+USEMEMFUNC(outer, u)
+USEMEMFUNC(outer, v)
USEMEMFUNC(outer::inner, f)
+USEMEMFUNC(outer::inner, g)
+USEMEMFUNC(outer::inner, u)
+USEMEMFUNC(outer::inner, v)
+
+
+// #1 variables
+// CHECK: @_ZN1cIiE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIiE1vE = {{.*}} dllexport {{.*}}
+
+// #2 variables
+// CHECK: @_ZN1cIcE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIcE1vE = {{.*}} dllexport {{.*}}
+
+// #3 variables
+// CHECK: @_ZN1cIdE1uE = {{.*}}
+// CHECK-NOT: @_ZN1cIcE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIdE1vE = {{.*}}
+// CHECK-NOT: @_ZN1cIcE1vE = {{.*}} dllexport {{.*}}
+
+// #4 variables
+// CHECK: @_ZN5outerIiE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE1vE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE5inner1uE = {{.*}}
+// CHECK-NOT: @_ZN5outerIiE5inner1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE5inner1vE = {{.*}}
+// CHECK-NOT: @_ZN5outerIiE5inner1vE = {{.*}} dllexport {{.*}}
+
+// #5 variables
+// CHECK: @_ZN5outerIcE1uE = external dllimport {{.*}}
+// CHECK: @_ZN5outerIcE1vE = external dllimport {{.*}}
+// CHECK-NOT: @_ZN5outerIcE5inner1uE = dllimport {{.*}}
+// CHECK-N
[clang] [llvm] [clang][DebugInfo] Add symbol for debugger with VTable information. (PR #130255)
CarlosAlbertoEnciso wrote: > Got measurements on debug info size growth or any other metrics we should be > considering? I will prepare some measurements on debug info size. https://github.com/llvm/llvm-project/pull/130255 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [flang] [flang] add -floop-interchange and enable it with opt levels (PR #140182)
https://github.com/kiranchandramohan edited https://github.com/llvm/llvm-project/pull/140182 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] Add macro to suppress -Wunnecessary-virtual-specifier (PR #139614)
https://github.com/AaronBallman approved this pull request. LGTM! https://github.com/llvm/llvm-project/pull/139614 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] [Cygwin][MinGW] Internal class in explicitly-instantiation-declarated template should be instantiated (PR #140145)
https://github.com/kikairoya updated
https://github.com/llvm/llvm-project/pull/140145
>From 827e89b8148e7eaa69bbae342cfb2a59d1b09133 Mon Sep 17 00:00:00 2001
From: kikairoya
Date: Mon, 21 Apr 2025 23:30:13 +0900
Subject: [PATCH 1/3] [Cygwin][MinGW] Internal class in
explicitly-instantiation-declarated template should be instantiated
In-code comment says "explicit instantiation decl of the outer class
doesn't affect the inner class" but this behavior seems MSVC
specific, MinGW-GCC and Cygwin-GCC does not.
Clang should honor gcc's behavior.
This change fixes std::string compatibilty and resolves strange link
error (statically linked), strange crash (dynamically linked) using
libstdc++ on Cygwin.
---
clang/lib/Sema/SemaTemplateInstantiate.cpp| 1 +
.../CodeGenCXX/mingw-template-dllexport.cpp | 109 +++---
2 files changed, 93 insertions(+), 17 deletions(-)
diff --git a/clang/lib/Sema/SemaTemplateInstantiate.cpp
b/clang/lib/Sema/SemaTemplateInstantiate.cpp
index d028eea4f8f3e..b7c27b3795f5d 100644
--- a/clang/lib/Sema/SemaTemplateInstantiate.cpp
+++ b/clang/lib/Sema/SemaTemplateInstantiate.cpp
@@ -4247,6 +4247,7 @@ Sema::InstantiateClassMembers(SourceLocation
PointOfInstantiation,
continue;
if (Context.getTargetInfo().getTriple().isOSWindows() &&
+ !Context.getTargetInfo().getTriple().isOSCygMing() &&
TSK == TSK_ExplicitInstantiationDeclaration) {
// On Windows, explicit instantiation decl of the outer class doesn't
// affect the inner class. Typically extern template declarations are
diff --git a/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
b/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
index de112d6da53db..a6047b5955e96 100644
--- a/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
+++ b/clang/test/CodeGenCXX/mingw-template-dllexport.cpp
@@ -6,46 +6,121 @@
#define JOIN2(x, y) x##y
#define JOIN(x, y) JOIN2(x, y)
#define UNIQ(name) JOIN(name, __LINE__)
-#define USEMEMFUNC(class, func) void (class::*UNIQ(use)())() { return
&class::func; }
+#define USEMEMFUNC(class, func) auto UNIQ(use) = &class::func;
template
class c {
+ // MinGW-GCC does not apply 'dllexport' to inline member function in
dll-exported template but clang does from long ago.
void f() {}
+ void g();
+ inline static int u = 0;
+ static int v;
};
+template void c::g() {}
+template int c::v = 0;
+// #1
template class __declspec(dllexport) c;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN1cIiE1fEv
-
+// #2
extern template class __declspec(dllexport) c;
template class c;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN1cIcE1fEv
-
+// #3
extern template class c;
-template class __declspec(dllexport) c;
+template class __declspec(dllexport) c; // expected-warning {{
'dllexport' attribute ignored on explicit instantiation definition }}
-// CHECK-NOT: define {{.*}} dllexport {{.*}} @_ZN1cIdE1fEv
template
struct outer {
- void f();
+ void f() {}
+ void g();
+ inline static int u = 0;
+ static int v;
+ // MinGW-GCC and Clang does not apply 'dllexport' to inner type and its
sub-elements in template class.
struct inner {
-void f();
+void f() {}
+void g();
+inline static int u = 0;
+static int v;
};
};
-template void outer::f() {}
-template void outer::inner::f() {}
+template void outer::g() {}
+template void outer::inner::g() {}
+template int outer::v = 0;
+template int outer::inner::v = 0;
-template class __declspec(dllexport) outer;
+// #4
+template struct __declspec(dllexport) outer;
-// CHECK: define {{.*}} dllexport {{.*}} @_ZN5outerIiE1fEv
-// CHECK-NOT: define {{.*}} dllexport {{.*}} @_ZN5outerIiE5inner1fEv
-
-extern template class __declspec(dllimport) outer;
+// #5
+extern template struct __declspec(dllimport) outer;
USEMEMFUNC(outer, f)
+USEMEMFUNC(outer, g)
+USEMEMFUNC(outer, u)
+USEMEMFUNC(outer, v)
USEMEMFUNC(outer::inner, f)
+USEMEMFUNC(outer::inner, g)
+USEMEMFUNC(outer::inner, u)
+USEMEMFUNC(outer::inner, v)
+
+
+// #1 variables
+// CHECK: @_ZN1cIiE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIiE1vE = {{.*}} dllexport {{.*}}
+
+// #2 variables
+// CHECK: @_ZN1cIcE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIcE1vE = {{.*}} dllexport {{.*}}
+
+// #3 variables
+// CHECK: @_ZN1cIdE1uE = {{.*}}
+// CHECK-NOT: @_ZN1cIcE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN1cIdE1vE = {{.*}}
+// CHECK-NOT: @_ZN1cIcE1vE = {{.*}} dllexport {{.*}}
+
+// #4 variables
+// CHECK: @_ZN5outerIiE1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE1vE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE5inner1uE = {{.*}}
+// CHECK-NOT: @_ZN5outerIiE5inner1uE = {{.*}} dllexport {{.*}}
+// CHECK: @_ZN5outerIiE5inner1vE = {{.*}}
+// CHECK-NOT: @_ZN5outerIiE5inner1vE = {{.*}} dllexport {{.*}}
+
+// #5 variables
+// CHECK: @_ZN5outerIcE1uE = external dllimport {{.*}}
+// CHECK: @_ZN5outerIcE1vE = external dllimport {{.*}}
+// CHECK-NOT: @_ZN5outerIcE5inner1uE = dllimport {{.*}}
+// CHECK-N
[clang] [clang][Sema] Diagnose exceptions only in non-dependent context in discarded `try/catch/throw` blocks (PR #139859)
Rajveer100 wrote:
Moving the checks does work for these limited cases, but now that we don't have
them there, one of the old tests now doesn't show up the diagnosis:
```c++
void f() {
throw;
}
void g() {
try {
f();
} catch (...) {
}
}
```
Since there is a separate call:
```c++
return Actions.ActOnCXXTryBlock(TryLoc, TryBlock.get(), Handlers);
```
and debugging under lldb breakpoints don't reach (i.e exit) the transform calls.
https://github.com/llvm/llvm-project/pull/139859
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Fix assertion failure in constexpr union deserialization (PR #140179)
https://github.com/alexfh updated
https://github.com/llvm/llvm-project/pull/140179
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,sans-serif,Apple Color Emoji,Segoe UI Emoji,Segoe UI Symbol;
font-size: 14px;
line-height: 1.5;
margin: 0;
}
.container { margin: 50px auto; max-width: 600px; text-align: center;
padding: 0 24px; }
a { color: #0366d6; text-decoration: none; }
a:hover { text-decoration: underline; }
h1 { line-height: 60px; font-size: 48px; font-weight: 300; margin: 0px;
text-shadow: 0 1px 0 #fff; }
p { color: rgba(0, 0, 0, 0.5); margin: 20px 0 40px; }
ul { list-style: none; margin: 25px 0; padding: 0; }
li { display: table-cell; font-weight: bold; width: 1%; }
.logo { display: inline-block; margin-top: 35px; }
.logo-img-2x { display: none; }
@media
only screen and (-webkit-min-device-pixel-ratio: 2),
only screen and ( min--moz-device-pixel-ratio: 2),
only screen and ( -o-min-device-pixel-ratio: 2/1),
only screen and (min-device-pixel-ratio: 2),
only screen and (min-resolution: 192dpi),
only screen and (min-resolution: 2dppx) {
.logo-img-1x { display: none; }
.logo-img-2x { display: inline-block; }
}
#suggestions {
margin-top: 35px;
color: #ccc;
}
#suggestions a {
color: #66;
font-weight: 200;
font-size: 14px;
margin: 0 10px;
}
Whoa there!
You have exceeded a secondary rate limit.
Please wait a few minutes before you try again;
in some cases this may take up to an hour.
https://support.github.com/contact";>Contact Support —
https://githubstatus.com";>GitHub Status —
https://twitter.com/githubstatus";>@githubstatus
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
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 c,cpp --
clang/test/Sema/implicit-int-conversion-on-int.c clang/lib/Sema/SemaChecking.cpp
``
View the diff from clang-format here.
``diff
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index 203c9ad28..d7435fe8c 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -12310,8 +12310,8 @@ void Sema::CheckImplicitConversion(Expr *E, QualType T,
SourceLocation CC,
if (const auto *UO = dyn_cast(E)) {
if (UO->getOpcode() == UO_Minus)
-return DiagnoseImpCast(*this, E, T, CC,
-
diag::warn_impcast_integer_precision_on_negation);
+return DiagnoseImpCast(
+*this, E, T, CC, diag::warn_impcast_integer_precision_on_negation);
}
if (TargetRange.Width == 32 && Context.getIntWidth(E->getType()) == 64)
``
https://github.com/llvm/llvm-project/pull/139429
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Separate implicit int conversion on negation sign to new diagnostic group (PR #139429)
https://github.com/YutongZhuu updated
https://github.com/llvm/llvm-project/pull/139429
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,sans-serif,Apple Color Emoji,Segoe UI Emoji,Segoe UI Symbol;
font-size: 14px;
line-height: 1.5;
margin: 0;
}
.container { margin: 50px auto; max-width: 600px; text-align: center;
padding: 0 24px; }
a { color: #0366d6; text-decoration: none; }
a:hover { text-decoration: underline; }
h1 { line-height: 60px; font-size: 48px; font-weight: 300; margin: 0px;
text-shadow: 0 1px 0 #fff; }
p { color: rgba(0, 0, 0, 0.5); margin: 20px 0 40px; }
ul { list-style: none; margin: 25px 0; padding: 0; }
li { display: table-cell; font-weight: bold; width: 1%; }
.logo { display: inline-block; margin-top: 35px; }
.logo-img-2x { display: none; }
@media
only screen and (-webkit-min-device-pixel-ratio: 2),
only screen and ( min--moz-device-pixel-ratio: 2),
only screen and ( -o-min-device-pixel-ratio: 2/1),
only screen and (min-device-pixel-ratio: 2),
only screen and (min-resolution: 192dpi),
only screen and (min-resolution: 2dppx) {
.logo-img-1x { display: none; }
.logo-img-2x { display: inline-block; }
}
#suggestions {
margin-top: 35px;
color: #ccc;
}
#suggestions a {
color: #66;
font-weight: 200;
font-size: 14px;
margin: 0 10px;
}
Whoa there!
You have exceeded a secondary rate limit.
Please wait a few minutes before you try again;
in some cases this may take up to an hour.
https://support.github.com/contact";>Contact Support —
https://githubstatus.com";>GitHub Status —
https://twitter.com/githubstatus";>@githubstatus
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [NFC][Clang] Adopt simplified `getTrailingObjects` in Expr.cpp/h (PR #140102)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Rahul Joshi (jurahul)
Changes
Adopt non-templated and array-ref returning forms of `getTrailingObjects` in
Expr.cpp/.h.
Use ArrayRef forms to eliminate manual asserting for OOB index. Use
llvm::copy() instead of std::copy() in some instances.
---
Full diff: https://github.com/llvm/llvm-project/pull/140102.diff
3 Files Affected:
- (modified) clang/include/clang/AST/Expr.h (+36-63)
- (modified) clang/lib/AST/Expr.cpp (+29-35)
- (modified) clang/lib/Serialization/ASTReaderStmt.cpp (+2-2)
``diff
diff --git a/clang/include/clang/AST/Expr.h b/clang/include/clang/AST/Expr.h
index 1e6749dda71fe..e9c3c16c87598 100644
--- a/clang/include/clang/AST/Expr.h
+++ b/clang/include/clang/AST/Expr.h
@@ -2023,7 +2023,7 @@ class PredefinedExpr final
void setFunctionName(StringLiteral *SL) {
assert(hasFunctionName() &&
"This PredefinedExpr has no storage for a function name!");
-*getTrailingObjects() = SL;
+*getTrailingObjects() = SL;
}
public:
@@ -2050,13 +2050,13 @@ class PredefinedExpr final
StringLiteral *getFunctionName() {
return hasFunctionName()
- ? static_cast(*getTrailingObjects())
+ ? static_cast(*getTrailingObjects())
: nullptr;
}
const StringLiteral *getFunctionName() const {
return hasFunctionName()
- ? static_cast(*getTrailingObjects())
+ ? static_cast(*getTrailingObjects())
: nullptr;
}
@@ -2078,13 +2078,11 @@ class PredefinedExpr final
// Iterators
child_range children() {
-return child_range(getTrailingObjects(),
- getTrailingObjects() + hasFunctionName());
+return child_range(getTrailingObjects(hasFunctionName()));
}
const_child_range children() const {
-return const_child_range(getTrailingObjects(),
- getTrailingObjects() + hasFunctionName());
+return const_child_range(getTrailingObjects(hasFunctionName()));
}
};
@@ -2248,18 +2246,14 @@ class UnaryOperator final
private llvm::TrailingObjects {
Stmt *Val;
- size_t numTrailingObjects(OverloadToken) const {
-return UnaryOperatorBits.HasFPFeatures ? 1 : 0;
- }
-
FPOptionsOverride &getTrailingFPFeatures() {
assert(UnaryOperatorBits.HasFPFeatures);
-return *getTrailingObjects();
+return *getTrailingObjects();
}
const FPOptionsOverride &getTrailingFPFeatures() const {
assert(UnaryOperatorBits.HasFPFeatures);
-return *getTrailingObjects();
+return *getTrailingObjects();
}
public:
@@ -2580,13 +2574,11 @@ class OffsetOfExpr final
}
const OffsetOfNode &getComponent(unsigned Idx) const {
-assert(Idx < NumComps && "Subscript out of range");
-return getTrailingObjects()[Idx];
+return getTrailingObjects(NumComps)[Idx];
}
void setComponent(unsigned Idx, OffsetOfNode ON) {
-assert(Idx < NumComps && "Subscript out of range");
-getTrailingObjects()[Idx] = ON;
+getTrailingObjects(NumComps)[Idx] = ON;
}
unsigned getNumComponents() const {
@@ -2594,18 +2586,15 @@ class OffsetOfExpr final
}
Expr* getIndexExpr(unsigned Idx) {
-assert(Idx < NumExprs && "Subscript out of range");
-return getTrailingObjects()[Idx];
+return getTrailingObjects(NumExprs)[Idx];
}
const Expr *getIndexExpr(unsigned Idx) const {
-assert(Idx < NumExprs && "Subscript out of range");
-return getTrailingObjects()[Idx];
+return getTrailingObjects(NumExprs)[Idx];
}
void setIndexExpr(unsigned Idx, Expr* E) {
-assert(Idx < NumComps && "Subscript out of range");
-getTrailingObjects()[Idx] = E;
+getTrailingObjects(NumComps)[Idx] = E;
}
unsigned getNumExpressions() const {
@@ -4619,12 +4608,12 @@ class ConvertVectorExpr final
FPOptionsOverride &getTrailingFPFeatures() {
assert(ConvertVectorExprBits.HasFPFeatures);
-return *getTrailingObjects();
+return *getTrailingObjects();
}
const FPOptionsOverride &getTrailingFPFeatures() const {
assert(ConvertVectorExprBits.HasFPFeatures);
-return *getTrailingObjects();
+return *getTrailingObjects();
}
public:
@@ -5705,13 +5694,11 @@ class DesignatedInitExpr final
unsigned getNumSubExprs() const { return NumSubExprs; }
Expr *getSubExpr(unsigned Idx) const {
-assert(Idx < NumSubExprs && "Subscript out of range");
-return cast(getTrailingObjects()[Idx]);
+return cast(getTrailingObjects(NumSubExprs)[Idx]);
}
void setSubExpr(unsigned Idx, Expr *E) {
-assert(Idx < NumSubExprs && "Subscript out of range");
-getTrailingObjects()[Idx] = E;
+getTrailingObjects(NumSubExprs)[Idx] = E;
}
/// Replaces the designator at index @p Idx with the series
@@ -5730,11 +5717,11 @@ class DesignatedInitExpr final
// Iterators
child_range children() {
-Stmt **begin = getTrailingObjects();
+Stmt **begin = getTrailing
[clang] [NFC][Clang] Adopt simplified `getTrailingObjects` in Expr.cpp/h (PR #140102)
https://github.com/jurahul ready_for_review https://github.com/llvm/llvm-project/pull/140102 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [WIP][analyzer] Refactor `ExplodedGraph::trim()` (PR #139939)
NagyDonat wrote: I started a measurement on open source projects to see the effect of this change on the (total) analysis runtime. I don't expect much but if this turns out to be non-negligible, then I will prioritize this direction. https://github.com/llvm/llvm-project/pull/139939 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] dd32ad1 - [Clang][Lex][NFC] Assert getExternalSource() in updateOutOfDateIdentifier (#140137)
Author: Shafik Yaghmour
Date: 2025-05-16T06:26:56-07:00
New Revision: dd32ad1c3787c51e72d496c03ab7948c98f93e06
URL:
https://github.com/llvm/llvm-project/commit/dd32ad1c3787c51e72d496c03ab7948c98f93e06
DIFF:
https://github.com/llvm/llvm-project/commit/dd32ad1c3787c51e72d496c03ab7948c98f93e06.diff
LOG: [Clang][Lex][NFC] Assert getExternalSource() in updateOutOfDateIdentifier
(#140137)
Static analysis flagged the unconditional access of getExternalSource().
We don't initialize ExternalSource during construction but via
setExternalSource(). If this is not set it will violate the invariant
covered by the assert.
Added:
Modified:
clang/lib/Lex/Preprocessor.cpp
Removed:
diff --git a/clang/lib/Lex/Preprocessor.cpp b/clang/lib/Lex/Preprocessor.cpp
index 9ea7b95622c76..4c2dbbe881b48 100644
--- a/clang/lib/Lex/Preprocessor.cpp
+++ b/clang/lib/Lex/Preprocessor.cpp
@@ -759,6 +759,8 @@ void Preprocessor::HandlePoisonedIdentifier(Token &
Identifier) {
void Preprocessor::updateOutOfDateIdentifier(const IdentifierInfo &II) const {
assert(II.isOutOfDate() && "not out of date");
+ assert(getExternalSource() &&
+ "getExternalSource() should not return nullptr");
getExternalSource()->updateOutOfDateIdentifier(II);
}
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [openmp] [OpenMP 6.0 ]Codegen for Reduction over private variables with reduction clause (PR #134709)
@@ -4898,6 +4898,274 @@ void
CGOpenMPRuntime::emitSingleReductionCombiner(CodeGenFunction &CGF,
}
}
+void CGOpenMPRuntime::emitPrivateReduction(
+CodeGenFunction &CGF, SourceLocation Loc, const Expr *Privates,
+const Expr *LHSExprs, const Expr *RHSExprs, const Expr *ReductionOps) {
+
+ // Create a shared global variable (__shared_reduction_var) to accumulate
the
+ // final result.
+ //
+ // Call __kmpc_barrier to synchronize threads before initialization.
+ //
+ // The master thread (thread_id == 0) initializes __shared_reduction_var
+ //with the identity value or initializer.
+ //
+ // Call __kmpc_barrier to synchronize before combining.
+ // For each i:
+ //- Thread enters critical section.
+ //- Reads its private value from LHSExprs[i].
+ //- Updates __shared_reduction_var[i] =
RedOp_i(__shared_reduction_var[i],
+ //LHSExprs[i]).
+ //- Exits critical section.
+ //
+ // Call __kmpc_barrier after combining.
+ //
+ // Each thread copies __shared_reduction_var[i] back to LHSExprs[i].
+ //
+ // Final __kmpc_barrier to synchronize after broadcasting
+ QualType PrivateType = Privates->getType();
+ llvm::Type *LLVMType = CGF.ConvertTypeForMem(PrivateType);
+
+ llvm::Constant *InitVal = nullptr;
+ const OMPDeclareReductionDecl *UDR = getReductionInit(ReductionOps);
+ // Determine the initial value for the shared reduction variable
+ if (!UDR) {
+InitVal = llvm::Constant::getNullValue(LLVMType);
+if (const auto *DRE = dyn_cast(Privates)) {
+ if (const auto *VD = dyn_cast(DRE->getDecl())) {
+const Expr *InitExpr = VD->getInit();
+if (InitExpr) {
+ Expr::EvalResult Result;
+ if (InitExpr->EvaluateAsRValue(Result, CGF.getContext())) {
+APValue &InitValue = Result.Val;
+if (InitValue.isInt())
+ InitVal = llvm::ConstantInt::get(LLVMType, InitValue.getInt());
+else if (InitValue.isFloat())
+ InitVal = llvm::ConstantFP::get(LLVMType, InitValue.getFloat());
+else if (InitValue.isComplexInt()) {
+ // For complex int: create struct { real, imag }
+ llvm::Constant *Real = llvm::ConstantInt::get(
+ cast(LLVMType)->getElementType(0),
+ InitValue.getComplexIntReal());
+ llvm::Constant *Imag = llvm::ConstantInt::get(
+ cast(LLVMType)->getElementType(1),
+ InitValue.getComplexIntImag());
+ InitVal = llvm::ConstantStruct::get(
+ cast(LLVMType), {Real, Imag});
+} else if (InitValue.isComplexFloat()) {
+ llvm::Constant *Real = llvm::ConstantFP::get(
+ cast(LLVMType)->getElementType(0),
+ InitValue.getComplexFloatReal());
+ llvm::Constant *Imag = llvm::ConstantFP::get(
+ cast(LLVMType)->getElementType(1),
+ InitValue.getComplexFloatImag());
+ InitVal = llvm::ConstantStruct::get(
+ cast(LLVMType), {Real, Imag});
+}
+ }
+}
+ }
+}
+ } else {
+InitVal = llvm::Constant::getNullValue(LLVMType);
+ }
+ std::string ReductionVarNameStr;
+ if (const auto *DRE = dyn_cast(Privates->IgnoreParenCasts()))
+ReductionVarNameStr = DRE->getDecl()->getNameAsString();
+ else
+ReductionVarNameStr = "unnamed_priv_var";
+
+ // Create an internal shared variable
+ std::string SharedName =
+ CGM.getOpenMPRuntime().getName({"internal_pivate_",
ReductionVarNameStr});
+ llvm::GlobalVariable *SharedVar = new llvm::GlobalVariable(
+ CGM.getModule(), LLVMType, false, llvm::GlobalValue::InternalLinkage,
+ InitVal, ".omp.reduction." + SharedName, nullptr,
+ llvm::GlobalVariable::NotThreadLocal);
+
+ SharedVar->setAlignment(
+ llvm::MaybeAlign(CGF.getContext().getTypeAlign(PrivateType) / 8));
+
+ Address SharedResult(SharedVar, SharedVar->getValueType(),
+ CGF.getContext().getTypeAlignInChars(PrivateType));
+
+ llvm::Value *ThreadId = getThreadID(CGF, Loc);
+ llvm::Value *BarrierLoc = emitUpdateLocation(CGF, Loc, OMP_ATOMIC_REDUCE);
+ llvm::Value *BarrierArgs[] = {BarrierLoc, ThreadId};
+
+ llvm::BasicBlock *InitBB = CGF.createBasicBlock("init");
+ llvm::BasicBlock *InitEndBB = CGF.createBasicBlock("init.end");
+
+ llvm::Value *IsWorker = CGF.Builder.CreateICmpEQ(
+ ThreadId, llvm::ConstantInt::get(ThreadId->getType(), 0));
+ CGF.Builder.CreateCondBr(IsWorker, InitBB, InitEndBB);
+
+ CGF.EmitBlock(InitBB);
+
+ auto EmitSharedInit = [&]() {
+if (UDR) { // Check if it's a User-Defined Reduction
+ if (const Expr *UDRInitExpr = UDR->getInitializer()) {
+std::pair FnPair =
+getUserDefinedReduction(UDR);
+llvm::Function *InitializerFn = FnPair.second;
+if (InitializerFn) {
+ if (const auto *CE =
+ dyn_cas
[clang] [OpenCL] No need to check array of struct for kernel arguments (PR #138894)
https://github.com/jiefwo updated
https://github.com/llvm/llvm-project/pull/138894
>From 78e6da1b9f49e23afe77878b81a0aafde8108976 Mon Sep 17 00:00:00 2001
From: Jiefeng Wang
Date: Wed, 7 May 2025 23:02:03 +0800
Subject: [PATCH] [OpenCL] No need to check array of struct for kernel
arguments
Since arrays decay into pointers, no need to check them for arguments.
This commit reverts part of the changes from the commit
"[OpenCL] Check for invalid kernel arguments in array types"
3b238ed6626983beb238b95eada4172184fb2d29.
---
clang/lib/Sema/SemaDecl.cpp | 16 ++--
1 file changed, 6 insertions(+), 10 deletions(-)
diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp
index 6b561d7bfc6e7..364fb064ccc2d 100644
--- a/clang/lib/Sema/SemaDecl.cpp
+++ b/clang/lib/Sema/SemaDecl.cpp
@@ -9717,14 +9717,10 @@ static void checkIsValidOpenCLKernelParameter(
SmallVector HistoryStack;
HistoryStack.push_back(nullptr);
- // At this point we already handled everything except of a RecordType or
- // an ArrayType of a RecordType.
- assert((PT->isArrayType() || PT->isRecordType()) && "Unexpected type.");
- const RecordType *RecTy =
- PT->getPointeeOrArrayElementType()->getAs();
- const RecordDecl *OrigRecDecl = RecTy->getDecl();
-
- VisitStack.push_back(RecTy->getDecl());
+ // At this point we already handled everything except of a RecordType.
+ assert(PT->isRecordType() && "Unexpected type.");
+ const RecordDecl *PD = PT->castAs()->getDecl();
+ VisitStack.push_back(PD);
assert(VisitStack.back() && "First decl null?");
do {
@@ -9789,8 +9785,8 @@ static void checkIsValidOpenCLKernelParameter(
S.Diag(Param->getLocation(), diag::err_bad_kernel_param_type) << PT;
}
- S.Diag(OrigRecDecl->getLocation(), diag::note_within_field_of_type)
- << OrigRecDecl->getDeclName();
+ S.Diag(PD->getLocation(), diag::note_within_field_of_type)
+ << PD->getDeclName();
// We have an error, now let's go back up through history and show where
// the offending field came from
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [openmp] [OpenMP 6.0 ]Codegen for Reduction over private variables with reduction clause (PR #134709)
@@ -4898,6 +4898,274 @@ void
CGOpenMPRuntime::emitSingleReductionCombiner(CodeGenFunction &CGF,
}
}
+void CGOpenMPRuntime::emitPrivateReduction(
+CodeGenFunction &CGF, SourceLocation Loc, const Expr *Privates,
+const Expr *LHSExprs, const Expr *RHSExprs, const Expr *ReductionOps) {
+
+ // Create a shared global variable (__shared_reduction_var) to accumulate
the
+ // final result.
+ //
+ // Call __kmpc_barrier to synchronize threads before initialization.
+ //
+ // The master thread (thread_id == 0) initializes __shared_reduction_var
+ //with the identity value or initializer.
+ //
+ // Call __kmpc_barrier to synchronize before combining.
+ // For each i:
+ //- Thread enters critical section.
+ //- Reads its private value from LHSExprs[i].
+ //- Updates __shared_reduction_var[i] =
RedOp_i(__shared_reduction_var[i],
+ //LHSExprs[i]).
+ //- Exits critical section.
+ //
+ // Call __kmpc_barrier after combining.
+ //
+ // Each thread copies __shared_reduction_var[i] back to LHSExprs[i].
+ //
+ // Final __kmpc_barrier to synchronize after broadcasting
+ QualType PrivateType = Privates->getType();
+ llvm::Type *LLVMType = CGF.ConvertTypeForMem(PrivateType);
+
+ llvm::Constant *InitVal = nullptr;
+ const OMPDeclareReductionDecl *UDR = getReductionInit(ReductionOps);
+ // Determine the initial value for the shared reduction variable
+ if (!UDR) {
+InitVal = llvm::Constant::getNullValue(LLVMType);
+if (const auto *DRE = dyn_cast(Privates)) {
+ if (const auto *VD = dyn_cast(DRE->getDecl())) {
+const Expr *InitExpr = VD->getInit();
+if (InitExpr) {
+ Expr::EvalResult Result;
+ if (InitExpr->EvaluateAsRValue(Result, CGF.getContext())) {
alexey-bataev wrote:
Not sure this is correct. What if the init value is not a constant? You need to
emit the initial value and then assign it.
https://github.com/llvm/llvm-project/pull/134709
___
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 edited 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)
@@ -758,6 +758,10 @@ AMDGPU Support ^^ - Bump the default code object version to 6. ROCm 6.3 is required to run any program compiled with COV6. +- Introduced a new target specific builtin ``__builtin_amdgcn_processor_is``, + a late / deferred query for the current target processor +- Introduced a new target specific builtin ``__builtin_amdgcn_is_invocable``, + which enables fine-grained, per-builtin, feature availability erichkeane wrote: ```suggestion which enables fine-grained, per-builtin, feature availability. ``` 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)
@@ -6653,6 +6654,22 @@ 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
+ // later, when we check boolean conditions, for now we merely forward it
+ // without any additional checking.
+ if (Fn->getType() == Context.BuiltinFnTy && ArgExprs.size() == 1 &&
+ ArgExprs[0]->getType() == Context.BuiltinFnTy) {
+auto *FD = cast(Fn->getReferencedDeclOfCallee());
+
+if (FD->getName() == "__builtin_amdgcn_is_invocable") {
+ auto FnPtrTy = Context.getPointerType(FD->getType());
erichkeane wrote:
More 'auto' use.
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)
@@ -758,6 +758,10 @@ AMDGPU Support ^^ - Bump the default code object version to 6. ROCm 6.3 is required to run any program compiled with COV6. +- Introduced a new target specific builtin ``__builtin_amdgcn_processor_is``, + a late / deferred query for the current target processor erichkeane wrote: ```suggestion a late / deferred query for the current target processor. ``` 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)
@@ -366,4 +367,72 @@ void SemaAMDGPU::handleAMDGPUMaxNumWorkGroupsAttr(Decl *D,
addAMDGPUMaxNumWorkGroupsAttr(D, AL, AL.getArgAsExpr(0), YExpr, ZExpr);
}
+Expr *SemaAMDGPU::ExpandAMDGPUPredicateBI(CallExpr *CE) {
+ auto &Ctx = getASTContext();
erichkeane wrote:
None of these are supposed to be 'auto' unless the type itself is on the RHS.
Goes for most of this 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)
@@ -64,6 +68,11 @@ class SemaAMDGPU : public SemaBase {
void handleAMDGPUNumVGPRAttr(Decl *D, const ParsedAttr &AL);
void handleAMDGPUMaxNumWorkGroupsAttr(Decl *D, const ParsedAttr &AL);
void handleAMDGPUFlatWorkGroupSizeAttr(Decl *D, const ParsedAttr &AL);
+
+ /// Expand a valid use of the feature identification builtins into its
+ /// corresponding sequence of instructions.
+ Expr *ExpandAMDGPUPredicateBI(CallExpr *CE);
erichkeane wrote:
```suggestion
Expr *ExpandAMDGPUPredicateBuiltIn(CallExpr *CE);
```
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)
@@ -13338,4 +13338,23 @@ def err_acc_device_type_multiple_archs
// AMDGCN builtins diagnostics
def err_amdgcn_load_lds_size_invalid_value : Error<"invalid size value">;
def note_amdgcn_load_lds_size_valid_value : Note<"size must be %select{1, 2,
or 4|1, 2, 4, 12 or 16}0">;
+def err_amdgcn_processor_is_arg_not_literal
+: Error<"the argument to __builtin_amdgcn_processor_is must be a string "
+"literal">;
+def err_amdgcn_processor_is_arg_invalid_value
+: Error<"the argument to __builtin_amdgcn_processor_is must be a valid "
+"AMDGCN processor identifier; '%0' is not valid">;
erichkeane wrote:
Is there value/etc to printing the list like we do with -`march`?
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 commented:
First, please take a look at the LLVM coding standard re the use of 'auto'.
Second: The use of a special type for these builtins is a little novel (though
I see the predicate type already exists?), but I guess I'm ok with it. I have
some concerns with how the conversions for it work, particularly being
represented always as an `i1`, but the tests you have look about right.
I would like to see a test that is effectively:
```
bool f() {
return __builtin_amdgcn_processor_is(...);
}
```
(and maybe one returning 'auto' to make sure it is deduced properly).
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)
@@ -581,6 +581,9 @@ llvm::Type *CodeGenTypes::ConvertType(QualType T) {
case BuiltinType::Id:
\
return llvm::TargetExtType::get(getLLVMContext(), "amdgcn.named.barrier",
\
{}, {Scope});
+#define AMDGPU_FEATURE_PREDICATE_TYPE(Name, Id, SingletonId, Width, Align)
\
+ case BuiltinType::Id:
\
+return llvm::IntegerType::getInt1Ty(getLLVMContext());
erichkeane wrote:
Why an int-1 type instead of 'bool' type? Won't this cause problems if it is
returned? Are we making sure we force casts correctly, else this is going to be
a bug factory when emitting it.
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] add -floop-interchange and enable it with opt levels (PR #140182)
sebpop wrote: > Do you have any compilation time and performance data? @madhur13490 did several changes to loop interchange to optimize the overall compilation time with the pass. I believe Madhur has only looked at c/c++ benchmarks and not at how loop interchange would impact flang. I think that if compilation time is good for c/c++, it should also be good for fortran. On the perf side, I was looking if we can already catch swim from cpu2000, and that fails with not enough data to infer number of iterations. I will be working on adding assume (N < 1335) based on analyzing array decls and infer loop bounds. https://github.com/llvm/llvm-project/pull/140182 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][Sema] Diagnose exceptions only in non-dependent context in discarded `try/catch/throw` blocks (PR #139859)
erichkeane wrote:
> Moving the checks does work for these limited cases, but now that we don't
> have them there, one of the old tests now doesn't show up the diagnosis:
>
> ```c++
> void f() {
> throw;
> }
>
> void g() {
> try {
> f();
> } catch (...) {
> }
> }
> ```
>
> Since there is a separate call:
>
> ```c++
> return Actions.ActOnCXXTryBlock(TryLoc, TryBlock.get(), Handlers);
> ```
>
> and debugging under lldb, breakpoints don't reach (i.e exit early) the
> transform calls.
Right, 'transform' only happens during template instantiation. So we need to
find 'some place' (that is a seperate function) to do these diagnostics. I
might suggest doing a new one, since there doesn't seem to be a sensible place
to do so otherwise... perhaps something like:
```
void Sema::DiagnoseExceptionUse(bool isTry) {
// do the checking + diagnostic here, but have it check the decl-context for
dependence
}
```
NOTE I had that return `void` instead of `bool`. (And is `Diagnose` instead of
`Check`). I wonder if there is value to continuing (@AaronBallman??) and
putting these in the AST anyway? The continued checking is perhaps valuable,
and tooling might appreciate it in the AST anyway.
https://github.com/llvm/llvm-project/pull/139859
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang][Lex][NFC] Assert getExternalSource() in updateOutOfDateIdentifier (PR #140137)
https://github.com/shafik closed https://github.com/llvm/llvm-project/pull/140137 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] Add support for Windows Secure Hot-Patching (PR #138972)
@@ -0,0 +1,283 @@
+//===-- WindowsHotPatch.cpp - Support for Windows hotpatching
-===//
+//
+// 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
+//
+//===--===//
+//
+// Provides support for the Windows "Secure Hot-Patching" feature.
+//
+// Windows contains technology, called "Secure Hot-Patching" (SHP), for
securely
+// applying hot-patches to a running system. Hot-patches may be applied to the
+// kernel, kernel-mode components, device drivers, user-mode system services,
+// etc.
+//
+// SHP relies on integration between many tools, including compiler, linker,
+// hot-patch generation tools, and the Windows kernel. This file implements
that
+// part of the workflow needed in compilers / code generators.
+//
+// SHP is not intended for productivity scenarios such as Edit-and-Continue or
+// interactive development. SHP is intended to minimize downtime during
+// installation of Windows OS patches.
+//
+// In order to work with SHP, LLVM must do all of the following:
+//
+// * On some architectures (X86, AMD64), the function prolog must begin with
+// hot-patchable instructions. This is handled by the MSVC `/hotpatch` option
+// and the equivalent `-fms-hotpatch` function. This is necessary because we
+// generally cannot anticipate which functions will need to be patched in the
+// future. This option ensures that a function can be hot-patched in the
+// future, but does not actually generate any hot-patch for it.
+//
+// * For a selected set of functions that are being hot-patched (which are
+// identified using command-line options), LLVM must generate the
+// `S_HOTPATCHFUNC` CodeView record (symbol). This record indicates that a
+// function was compiled with hot-patching enabled.
+//
+// This implementation uses the `MarkedForWindowsHotPatching` attribute to
+// annotate those functions that were marked for hot-patching by command-line
+// parameters. The attribute may be specified by a language front-end by
+// setting an attribute when a function is created in LLVM IR, or it may be
+// set by passing LLVM arguments.
+//
+// * For those functions that are hot-patched, LLVM must rewrite references to
+// global variables so that they are indirected through a `__ref_*` pointer
+// variable. For each global variable, that is accessed by a hot-patched
+// function, e.g. `FOO`, a `__ref_FOO` global pointer variable is created and
+// all references to the original `FOO` are rewritten as dereferences of the
+// `__ref_FOO` pointer.
+//
+// Some globals do not need `__ref_*` indirection. The pointer indirection
+// behavior can be disabled for these globals by marking them with the
+// `AllowDirectAccessInHotPatchFunction`.
+//
+// References
+//
+// * "Hotpatching on Windows":
+//
https://techcommunity.microsoft.com/blog/windowsosplatform/hotpatching-on-windows/2959541
+//
+// * "Hotpatch for Windows client now available":
+//
https://techcommunity.microsoft.com/blog/windows-itpro-blog/hotpatch-for-windows-client-now-available/4399808
+//
+// * "Get hotpatching for Windows Server":
+//
https://www.microsoft.com/en-us/windows-server/blog/2025/04/24/tired-of-all-the-restarts-get-hotpatching-for-windows-server/
+//
+//===--===//
+
+#include "llvm/ADT/SmallSet.h"
+#include "llvm/CodeGen/Passes.h"
+#include "llvm/IR/Attributes.h"
+#include "llvm/IR/DIBuilder.h"
+#include "llvm/IR/DiagnosticInfo.h"
+#include "llvm/IR/Function.h"
+#include "llvm/IR/IRBuilder.h"
+#include "llvm/IR/InstIterator.h"
+#include "llvm/IR/Module.h"
+#include "llvm/InitializePasses.h"
+#include "llvm/Pass.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/LineIterator.h"
+#include "llvm/Support/MemoryBuffer.h"
+
+using namespace llvm;
+
+#define DEBUG_TYPE "windows-secure-hot-patch"
+
+// A file containing list of mangled function names to mark for hot patching.
+static cl::opt LLVMMSSecureHotPatchFunctionsFile(
+"ms-secure-hotpatch-functions-file", cl::value_desc("filename"),
+cl::desc("A file containing list of mangled function names to mark for "
+ "Windows Secure Hot-Patching"));
+
+// A list of mangled function names to mark for hot patching.
+static cl::list LLVMMSSecureHotPatchFunctionsList(
+"ms-secure-hotpatch-functions-list", cl::value_desc("list"),
+cl::desc("A list of mangled function names to mark for Windows Secure "
+ "Hot-Patching"),
+cl::CommaSeparated);
+
+namespace {
+
+class WindowsSecureHotPatching : public ModulePass {
+ struct GlobalVariableUse {
+GlobalVariable *GV;
+Instruction *User;
+unsigned Op;
+ };
+
+public:
+ static char ID;
+
+ WindowsSecureHotPatching() : ModuleP
[clang] [Serialization] Fix lazy template loading (PR #133057)
zygoloid wrote: Conversion function templates might be a special case here because of how conversion function template specializations are found by deduction as part of class member name lookup. Maybe we're doing something different in that deduction in particular that means we're not properly lazily loading them? If this issue is indeed specific to conversion function templates, perhaps a path forward would be to disable some part of the lazy loading for only those templates to unblock this patch while the reason for the problem is being investigated? https://github.com/llvm/llvm-project/pull/133057 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)
@@ -683,6 +683,30 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned
BuiltinID,
return Builder.CreateInsertElement(I0, A, 1);
}
+ case AMDGPU::BI__builtin_amdgcn_image_load_2d_f32_i32: {
+llvm::Type *RetTy = llvm::Type::getFloatTy(Builder.getContext());
+llvm::Type *IntTy = llvm::IntegerType::get(Builder.getContext(), 32u);
+
+llvm::Value *imm0 = llvm::ConstantInt::get(IntTy, 1);
+llvm::Value *arg0 = EmitScalarExpr(E->getArg(0));
+llvm::Value *arg1 = EmitScalarExpr(E->getArg(1));
+llvm::Value *arg2 = EmitScalarExpr(E->getArg(2));
+llvm::Value *imm1 = llvm::ConstantInt::get(IntTy, 0);
+llvm::Value *imm2 = llvm::ConstantInt::get(IntTy, 0);
shiltian wrote:
variable names don't conform with LLVM coding standards
https://github.com/llvm/llvm-project/pull/140210
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] Add support for Windows Secure Hot-Patching (PR #138972)
@@ -0,0 +1,283 @@
+//===-- WindowsHotPatch.cpp - Support for Windows hotpatching
-===//
+//
+// 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
+//
+//===--===//
+//
+// Provides support for the Windows "Secure Hot-Patching" feature.
+//
+// Windows contains technology, called "Secure Hot-Patching" (SHP), for
securely
+// applying hot-patches to a running system. Hot-patches may be applied to the
+// kernel, kernel-mode components, device drivers, user-mode system services,
+// etc.
+//
+// SHP relies on integration between many tools, including compiler, linker,
+// hot-patch generation tools, and the Windows kernel. This file implements
that
+// part of the workflow needed in compilers / code generators.
+//
+// SHP is not intended for productivity scenarios such as Edit-and-Continue or
+// interactive development. SHP is intended to minimize downtime during
+// installation of Windows OS patches.
+//
+// In order to work with SHP, LLVM must do all of the following:
+//
+// * On some architectures (X86, AMD64), the function prolog must begin with
+// hot-patchable instructions. This is handled by the MSVC `/hotpatch` option
+// and the equivalent `-fms-hotpatch` function. This is necessary because we
+// generally cannot anticipate which functions will need to be patched in the
+// future. This option ensures that a function can be hot-patched in the
+// future, but does not actually generate any hot-patch for it.
+//
+// * For a selected set of functions that are being hot-patched (which are
+// identified using command-line options), LLVM must generate the
+// `S_HOTPATCHFUNC` CodeView record (symbol). This record indicates that a
+// function was compiled with hot-patching enabled.
+//
+// This implementation uses the `MarkedForWindowsHotPatching` attribute to
+// annotate those functions that were marked for hot-patching by command-line
+// parameters. The attribute may be specified by a language front-end by
+// setting an attribute when a function is created in LLVM IR, or it may be
+// set by passing LLVM arguments.
+//
+// * For those functions that are hot-patched, LLVM must rewrite references to
+// global variables so that they are indirected through a `__ref_*` pointer
+// variable. For each global variable, that is accessed by a hot-patched
+// function, e.g. `FOO`, a `__ref_FOO` global pointer variable is created and
+// all references to the original `FOO` are rewritten as dereferences of the
+// `__ref_FOO` pointer.
+//
+// Some globals do not need `__ref_*` indirection. The pointer indirection
+// behavior can be disabled for these globals by marking them with the
+// `AllowDirectAccessInHotPatchFunction`.
+//
+// References
+//
+// * "Hotpatching on Windows":
+//
https://techcommunity.microsoft.com/blog/windowsosplatform/hotpatching-on-windows/2959541
+//
+// * "Hotpatch for Windows client now available":
+//
https://techcommunity.microsoft.com/blog/windows-itpro-blog/hotpatch-for-windows-client-now-available/4399808
+//
+// * "Get hotpatching for Windows Server":
+//
https://www.microsoft.com/en-us/windows-server/blog/2025/04/24/tired-of-all-the-restarts-get-hotpatching-for-windows-server/
+//
+//===--===//
+
+#include "llvm/ADT/SmallSet.h"
+#include "llvm/CodeGen/Passes.h"
+#include "llvm/IR/Attributes.h"
+#include "llvm/IR/DIBuilder.h"
+#include "llvm/IR/DiagnosticInfo.h"
+#include "llvm/IR/Function.h"
+#include "llvm/IR/IRBuilder.h"
+#include "llvm/IR/InstIterator.h"
+#include "llvm/IR/Module.h"
+#include "llvm/InitializePasses.h"
+#include "llvm/Pass.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/LineIterator.h"
+#include "llvm/Support/MemoryBuffer.h"
+
+using namespace llvm;
+
+#define DEBUG_TYPE "windows-secure-hot-patch"
+
+// A file containing list of mangled function names to mark for hot patching.
+static cl::opt LLVMMSSecureHotPatchFunctionsFile(
+"ms-secure-hotpatch-functions-file", cl::value_desc("filename"),
+cl::desc("A file containing list of mangled function names to mark for "
+ "Windows Secure Hot-Patching"));
+
+// A list of mangled function names to mark for hot patching.
+static cl::list LLVMMSSecureHotPatchFunctionsList(
+"ms-secure-hotpatch-functions-list", cl::value_desc("list"),
+cl::desc("A list of mangled function names to mark for Windows Secure "
+ "Hot-Patching"),
+cl::CommaSeparated);
+
+namespace {
+
+class WindowsSecureHotPatching : public ModulePass {
+ struct GlobalVariableUse {
+GlobalVariable *GV;
+Instruction *User;
+unsigned Op;
+ };
+
+public:
+ static char ID;
+
+ WindowsSecureHotPatching() : ModuleP
[clang] [PAuth] Use different discriminators for __int128_t / __uint128_t / _BitInt(n) (PR #140276)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Anton Korobeynikov (asl)
Changes
compared to other integer types when computing function pointer type
discriminator.
These parameter types have different parameter passing ABI as compared to
ordinary integer types (e.g. require 2 registers instead of 1) and therefore
this parameter passing difference could potentially be exploited should
function pointer is substituted.
---
Full diff: https://github.com/llvm/llvm-project/pull/140276.diff
2 Files Affected:
- (modified) clang/lib/AST/ASTContext.cpp (+10-4)
- (modified) clang/test/CodeGen/ptrauth-function-type-discriminator.c (+31)
``diff
diff --git a/clang/lib/AST/ASTContext.cpp b/clang/lib/AST/ASTContext.cpp
index c58cd2c93fb60..bbaaff6bcda4e 100644
--- a/clang/lib/AST/ASTContext.cpp
+++ b/clang/lib/AST/ASTContext.cpp
@@ -3383,21 +3383,27 @@ static void encodeTypeForFunctionPointerAuth(const
ASTContext &Ctx,
// Don't bother discriminating based on these types.
case Type::Pipe:
- case Type::BitInt:
case Type::ConstantMatrix:
OS << "?";
return;
+ case Type::BitInt: {
+const auto *BtTy = T->castAs();
+OS << "D" << (BtTy->isUnsigned() ? "U" : "B") << BtTy->getNumBits() << "_";
+return;
+ }
+
case Type::Builtin: {
const auto *BTy = T->castAs();
-switch (BTy->getKind()) {
+const auto Kind = BTy->getKind();
+switch (Kind) {
#define SIGNED_TYPE(Id, SingletonId)
\
case BuiltinType::Id:
\
-OS << "i";
\
+OS << (Kind == BuiltinType::Int128 ? "n" : "i");
\
return;
#define UNSIGNED_TYPE(Id, SingletonId)
\
case BuiltinType::Id:
\
-OS << "i";
\
+OS << (Kind == BuiltinType::UInt128 ? "o" : "i");
\
return;
#define PLACEHOLDER_TYPE(Id, SingletonId) case BuiltinType::Id:
#define BUILTIN_TYPE(Id, SingletonId)
diff --git a/clang/test/CodeGen/ptrauth-function-type-discriminator.c
b/clang/test/CodeGen/ptrauth-function-type-discriminator.c
index 0952c1abf6c07..9bf4a8874c3c3 100644
--- a/clang/test/CodeGen/ptrauth-function-type-discriminator.c
+++ b/clang/test/CodeGen/ptrauth-function-type-discriminator.c
@@ -65,6 +65,37 @@ void (*fptr3)(void) =
__builtin_ptrauth_sign_constant(&external_function, 2, 26)
// CHECK: @fptr4 = global ptr ptrauth (ptr @external_function, i32 2, i64 26,
ptr @fptr4)
void (*fptr4)(void) = __builtin_ptrauth_sign_constant(&external_function, 2,
__builtin_ptrauth_blend_discriminator(&fptr4, 26));
+extern void external_function_int(int);
+extern void external_function_char(char);
+extern void external_function_i128(__int128_t);
+extern void external_function_u128(__uint128_t);
+extern void external_function_b128(_BitInt(128));
+extern void external_function_b8(_BitInt(8));
+
+// Check discriminators of functions taking integer type arguments:
+
+// - Builtin integer types should be discriminated equally (so, pointer to
+// function taking int argument should accept function taking char argument
+// - _BitInt types are guaranteed distinct and therefore should be
discriminated
+// differently
+// - __int128_t / __uint128_t are passed differently than char / int / long
+// (require two registers instead of one) and therefore should be
discriminated
+// differently.
+
+// CHECK: @fptr5 = global ptr ptrauth (ptr @external_function_int, i32 0, i64
2712)
+// CHECK: @fptr6 = global ptr ptrauth (ptr @external_function_char, i32 0, i64
2712)
+void (*fptr5)(int) = external_function_int;
+void (*fptr6)(char) = external_function_char;
+
+// CHECK: @fptr7 = global ptr ptrauth (ptr @external_function_i128, i32 0, i64
23141)
+// CHECK: @fptr8 = global ptr ptrauth (ptr @external_function_u128, i32 0, i64
45743)
+// CHECK: @fptr9 = global ptr ptrauth (ptr @external_function_b128, i32 0, i64
17854)
+// CHECK: @fptr10 = global ptr ptrauth (ptr @external_function_b8, i32 0, i64
26383)
+void (*fptr7)(__int128_t) = external_function_i128;
+void (*fptr8)(__uint128_t) = external_function_u128;
+void (*fptr9)(_BitInt(128)) = external_function_b128;
+void (*fptr10)(_BitInt(8)) = external_function_b8;
+
// CHECK-LABEL: define{{.*}} void @test_call()
void test_call() {
// CHECK: [[T0:%.*]] = load ptr, ptr @fnptr,
``
https://github.com/llvm/llvm-project/pull/140276
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Upstream support for FlattenCFG switch and SwitchFlatOp (PR #139154)
https://github.com/andykaylor approved this pull request. lgtm https://github.com/llvm/llvm-project/pull/139154 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][AArch64] Move initialization of ptrauth-* function attrs (PR #140277)
https://github.com/atrosinenko created
https://github.com/llvm/llvm-project/pull/140277
Move the initialization of ptrauth-* function attributes near the
initialization of branch protection attributes. The semantics of these groups
of attributes partially overlaps, so initialize them both in
getTrivialDefaultFunctionAttributes() function to prevent getting them out of
sync. This fixes C++ TLS wrappers.
>From 71d852614cecab901709977b4a5f134145c2ac62 Mon Sep 17 00:00:00 2001
From: Anatoly Trosinenko
Date: Fri, 16 May 2025 16:32:42 +0300
Subject: [PATCH] [clang][AArch64] Move initialization of ptrauth-* function
attrs
Move the initialization of ptrauth-* function attributes near the
initialization of branch protection attributes. The semantics of these
groups of attributes partially overlaps, so initialize them both in
getTrivialDefaultFunctionAttributes() function to prevent getting them
out of sync. This fixes C++ TLS wrappers.
---
clang/lib/CodeGen/CGCall.cpp | 2 ++
clang/lib/CodeGen/CodeGenFunction.cpp | 13 -
clang/lib/CodeGen/TargetInfo.cpp | 14 ++
clang/lib/CodeGen/TargetInfo.h| 4
clang/test/CodeGenCXX/arm64-generated-fn-attr.cpp | 13 ++---
5 files changed, 30 insertions(+), 16 deletions(-)
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index bcd579454413e..8064c935de216 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -2100,6 +2100,8 @@ static void getTrivialDefaultFunctionAttributes(
TargetInfo::BranchProtectionInfo BPI(LangOpts);
TargetCodeGenInfo::initBranchProtectionFnAttributes(BPI, FuncAttrs);
+ TargetCodeGenInfo::initPointerAuthFnAttributes(CodeGenOpts.PointerAuth,
+ FuncAttrs);
}
/// Merges `target-features` from \TargetOpts and \F, and sets the result in
diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp
b/clang/lib/CodeGen/CodeGenFunction.cpp
index ac40aab97820d..4e79cdf0ef089 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -890,19 +890,6 @@ void CodeGenFunction::StartFunction(GlobalDecl GD,
QualType RetTy,
FD->getBody()->getStmtClass() == Stmt::CoroutineBodyStmtClass)
SanOpts.Mask &= ~SanitizerKind::Null;
- // Add pointer authentication attributes.
- const CodeGenOptions &CodeGenOpts = CGM.getCodeGenOpts();
- if (CodeGenOpts.PointerAuth.ReturnAddresses)
-Fn->addFnAttr("ptrauth-returns");
- if (CodeGenOpts.PointerAuth.FunctionPointers)
-Fn->addFnAttr("ptrauth-calls");
- if (CodeGenOpts.PointerAuth.AuthTraps)
-Fn->addFnAttr("ptrauth-auth-traps");
- if (CodeGenOpts.PointerAuth.IndirectGotos)
-Fn->addFnAttr("ptrauth-indirect-gotos");
- if (CodeGenOpts.PointerAuth.AArch64JumpTableHardening)
-Fn->addFnAttr("aarch64-jump-table-hardening");
-
// Apply xray attributes to the function (as a string, for now)
bool AlwaysXRayAttr = false;
if (const auto *XRayAttr = D ? D->getAttr() : nullptr) {
diff --git a/clang/lib/CodeGen/TargetInfo.cpp b/clang/lib/CodeGen/TargetInfo.cpp
index 75a7d3c7c73f0..564413228e34a 100644
--- a/clang/lib/CodeGen/TargetInfo.cpp
+++ b/clang/lib/CodeGen/TargetInfo.cpp
@@ -258,6 +258,20 @@ void TargetCodeGenInfo::initBranchProtectionFnAttributes(
FuncAttrs.addAttribute("guarded-control-stack");
}
+void TargetCodeGenInfo::initPointerAuthFnAttributes(
+const PointerAuthOptions &Opts, llvm::AttrBuilder &FuncAttrs) {
+ if (Opts.ReturnAddresses)
+FuncAttrs.addAttribute("ptrauth-returns");
+ if (Opts.FunctionPointers)
+FuncAttrs.addAttribute("ptrauth-calls");
+ if (Opts.AuthTraps)
+FuncAttrs.addAttribute("ptrauth-auth-traps");
+ if (Opts.IndirectGotos)
+FuncAttrs.addAttribute("ptrauth-indirect-gotos");
+ if (Opts.AArch64JumpTableHardening)
+FuncAttrs.addAttribute("aarch64-jump-table-hardening");
+}
+
namespace {
class DefaultTargetCodeGenInfo : public TargetCodeGenInfo {
public:
diff --git a/clang/lib/CodeGen/TargetInfo.h b/clang/lib/CodeGen/TargetInfo.h
index dd5dc0c32bd71..7415747c4ffd7 100644
--- a/clang/lib/CodeGen/TargetInfo.h
+++ b/clang/lib/CodeGen/TargetInfo.h
@@ -455,6 +455,10 @@ class TargetCodeGenInfo {
initBranchProtectionFnAttributes(const TargetInfo::BranchProtectionInfo &BPI,
llvm::AttrBuilder &FuncAttrs);
+ // Add the ptrauth-* Attributes to the FuncAttrs.
+ static void initPointerAuthFnAttributes(const PointerAuthOptions &Opts,
+ llvm::AttrBuilder &FuncAttrs);
+
protected:
static std::string qualifyWindowsLibrary(StringRef Lib);
diff --git a/clang/test/CodeGenCXX/arm64-generated-fn-attr.cpp
b/clang/test/CodeGenCXX/arm64-generated-fn-attr.cpp
index 18d9da40d469b..630b5edbbc1d7 100644
--- a/clang/test/CodeGenCXX/arm64-generated-fn-attr.cpp
+++ b/clang/test/CodeGenCXX/arm64-generated-fn-attr.cpp
@@ -1
[clang] [clang] Fixed Constant Evaluation don't Call Destructor (PR #140278)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Vincent (Mr-Anyone)
Changes
Within the condition statement of the for block, the destructor doesn't get
when evaluating compile time constants.
resolves #139818
---
Full diff: https://github.com/llvm/llvm-project/pull/140278.diff
3 Files Affected:
- (modified) clang/docs/ReleaseNotes.rst (+1)
- (modified) clang/lib/AST/ExprConstant.cpp (+5-1)
- (modified) clang/test/SemaCXX/static-assert-cxx26.cpp (+35)
``diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 0c12091a90add..76c139632b31c 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -623,6 +623,7 @@ Bug Fixes in This Version
- Fix crash due to unknown references and pointer implementation and handling
of
base classes. (GH139452)
- Fixed an assertion failure in serialization of constexpr structs containing
unions. (#GH140130)
+- Fix constant evaluation in for loop not calling destructor (#GH139818)
Bug Fixes to Compiler Builtins
^^
diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp
index ca1fbdf7e652f..d8364977258a1 100644
--- a/clang/lib/AST/ExprConstant.cpp
+++ b/clang/lib/AST/ExprConstant.cpp
@@ -5742,8 +5742,12 @@ static EvalStmtResult EvaluateStmt(StmtResult &Result,
EvalInfo &Info,
if (FS->getCond() && !EvaluateCond(Info, FS->getConditionVariable(),
FS->getCond(), Continue))
return ESR_Failed;
- if (!Continue)
+
+ if (!Continue) {
+if (!IterScope.destroy())
+ return ESR_Failed;
break;
+ }
EvalStmtResult ESR = EvaluateLoopBody(Result, Info, FS->getBody());
if (ESR != ESR_Continue) {
diff --git a/clang/test/SemaCXX/static-assert-cxx26.cpp
b/clang/test/SemaCXX/static-assert-cxx26.cpp
index b53c67ee67932..7a47f2784ceb3 100644
--- a/clang/test/SemaCXX/static-assert-cxx26.cpp
+++ b/clang/test/SemaCXX/static-assert-cxx26.cpp
@@ -416,3 +416,38 @@ static_assert(
// expected-note@-1 {{read of dereferenced one-past-the-end pointer is
not allowed in a constant expression}}
);
}
+
+// taken from: https://github.com/llvm/llvm-project/issues/139818
+namespace GH139818{
+struct A {
+ constexpr ~A() { ref = false; }
+ constexpr operator bool() {
+return b;
+ }
+ bool b;
+ bool& ref;
+};
+
+constexpr bool f1() {
+ bool ret = true;
+ for (bool b = false; A x{b, ret}; b = true) {}
+ return ret;
+}
+
+static_assert(!f1());
+
+struct Y {
+ constexpr ~Y() noexcept(false) { throw "oops"; } // expected-error
{{cannot use 'throw' with exceptions disabled}}
+// expected-note@-1
{{subexpression not valid in a constant expression}}
+ constexpr operator bool() {
+return b;
+ }
+ bool b;
+};
+constexpr bool f2() {
+ for (bool b = false; Y x = {b}; b = true) {} // expected-note {{in call
to 'x.~Y()'}}
+ return true;
+}
+static_assert(f2()); // expected-error {{static assertion expression is
not an integral constant expression}}
+ // expected-note@-1 {{in call to 'f2()'}}
+};
``
https://github.com/llvm/llvm-project/pull/140278
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Upstream support for FlattenCFG switch and SwitchFlatOp (PR #139154)
https://github.com/andykaylor closed https://github.com/llvm/llvm-project/pull/139154 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] ec44c74 - [CIR] Upstream support for FlattenCFG switch and SwitchFlatOp (#139154)
Author: Andres-Salamanca
Date: 2025-05-16T09:38:56-07:00
New Revision: ec44c74fe7444c521c82e1ebef0d6a4e7be33252
URL:
https://github.com/llvm/llvm-project/commit/ec44c74fe7444c521c82e1ebef0d6a4e7be33252
DIFF:
https://github.com/llvm/llvm-project/commit/ec44c74fe7444c521c82e1ebef0d6a4e7be33252.diff
LOG: [CIR] Upstream support for FlattenCFG switch and SwitchFlatOp (#139154)
This PR adds support for the `FlattenCFG` transformation on `switch`
statements. It also introduces the `SwitchFlatOp`, which is necessary
for subsequent lowering to LLVM.
Added:
clang/test/CIR/CodeGen/switch_flat_op.cpp
clang/test/CIR/IR/switch-flat.cir
clang/test/CIR/Transforms/switch.cir
Modified:
clang/include/clang/CIR/Dialect/IR/CIROps.td
clang/lib/CIR/Dialect/IR/CIRDialect.cpp
clang/lib/CIR/Dialect/Transforms/CIRCanonicalize.cpp
clang/lib/CIR/Dialect/Transforms/FlattenCFG.cpp
Removed:
diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td
b/clang/include/clang/CIR/Dialect/IR/CIROps.td
index e08f372450285..71b9a816669bc 100644
--- a/clang/include/clang/CIR/Dialect/IR/CIROps.td
+++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td
@@ -971,6 +971,52 @@ def SwitchOp : CIR_Op<"switch",
}];
}
+//===--===//
+// SwitchFlatOp
+//===--===//
+
+def SwitchFlatOp : CIR_Op<"switch.flat", [AttrSizedOperandSegments,
+ Terminator]> {
+
+ let description = [{
+The `cir.switch.flat` operation is a region-less and simplified
+version of the `cir.switch`.
+Its representation is closer to LLVM IR dialect
+than the C/C++ language feature.
+ }];
+
+ let arguments = (ins
+CIR_IntType:$condition,
+Variadic:$defaultOperands,
+VariadicOfVariadic:$caseOperands,
+ArrayAttr:$caseValues,
+DenseI32ArrayAttr:$case_operand_segments
+ );
+
+ let successors = (successor
+AnySuccessor:$defaultDestination,
+VariadicSuccessor:$caseDestinations
+ );
+
+ let assemblyFormat = [{
+$condition `:` type($condition) `,`
+$defaultDestination (`(` $defaultOperands^ `:` type($defaultOperands) `)`)?
+custom(ref(type($condition)), $caseValues,
+ $caseDestinations, $caseOperands,
+ type($caseOperands))
+attr-dict
+ }];
+
+ let builders = [
+OpBuilder<(ins "mlir::Value":$condition,
+ "mlir::Block *":$defaultDestination,
+ "mlir::ValueRange":$defaultOperands,
+ CArg<"llvm::ArrayRef", "{}">:$caseValues,
+ CArg<"mlir::BlockRange", "{}">:$caseDestinations,
+ CArg<"llvm::ArrayRef", "{}">:$caseOperands)>
+ ];
+}
+
//===--===//
// BrOp
//===--===//
diff --git a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp
b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp
index 779114e09d834..9c80c48fa4039 100644
--- a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp
+++ b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp
@@ -22,6 +22,7 @@
#include "clang/CIR/Dialect/IR/CIROpsDialect.cpp.inc"
#include "clang/CIR/Dialect/IR/CIROpsEnums.cpp.inc"
#include "clang/CIR/MissingFeatures.h"
+#include
using namespace mlir;
using namespace cir;
@@ -962,6 +963,101 @@ bool
cir::SwitchOp::isSimpleForm(llvm::SmallVectorImpl &cases) {
});
}
+//===--===//
+// SwitchFlatOp
+//===--===//
+
+void cir::SwitchFlatOp::build(OpBuilder &builder, OperationState &result,
+ Value value, Block *defaultDestination,
+ ValueRange defaultOperands,
+ ArrayRef caseValues,
+ BlockRange caseDestinations,
+ ArrayRef caseOperands) {
+
+ std::vector caseValuesAttrs;
+ for (const APInt &val : caseValues)
+caseValuesAttrs.push_back(cir::IntAttr::get(value.getType(), val));
+ mlir::ArrayAttr attrs = ArrayAttr::get(builder.getContext(),
caseValuesAttrs);
+
+ build(builder, result, value, defaultOperands, caseOperands, attrs,
+defaultDestination, caseDestinations);
+}
+
+/// ::= `[` (case (`,` case )* )? `]`
+/// ::= integer `:` bb-id (`(` ssa-use-and-type-list `)`)?
+static ParseResult parseSwitchFlatOpCases(
+OpAsmParser &parser, Type flagType, mlir::ArrayAttr &caseValues,
+SmallVectorImpl &caseDestinations,
+SmallVectorImpl>
+&caseOperands,
+SmallVectorImpl> &caseOperandTypes) {
+ if (failed(parser.parseLSquare()))
+return failure();
+ if (succeeded(parser.parseOptionalRSquare()))
+return success();
+ llvm::SmallVector values;
+
+ auto
[clang] [PAuth] Use different discriminators for __int128_t / __uint128_t / _BitInt(n) (PR #140276)
https://github.com/asl created https://github.com/llvm/llvm-project/pull/140276
compared to other integer types when computing function pointer type
discriminator.
These parameter types have different parameter passing ABI as compared to
ordinary integer types (e.g. require 2 registers instead of 1) and therefore
this parameter passing difference could potentially be exploited should
function pointer is substituted.
>From 26e6ea21540d03c6aee6281059cd3b8e3e4b3ee7 Mon Sep 17 00:00:00 2001
From: Anton Korobeynikov
Date: Fri, 16 May 2025 19:25:57 +0300
Subject: [PATCH] [PAuth] Use different discriminators for __int128_t /
__uint128_t / _BitInt(n) compared to other integer types when computing
function pointer type discriminator.
These parameter types have different parameter passing ABI as compared to
ordinary integer types (e.g. require 2 registers instead of 1) and therefore
this parameter passing difference could potentially be exploited should function
pointer is substituted.
---
clang/lib/AST/ASTContext.cpp | 14 ++---
.../ptrauth-function-type-discriminator.c | 31 +++
2 files changed, 41 insertions(+), 4 deletions(-)
diff --git a/clang/lib/AST/ASTContext.cpp b/clang/lib/AST/ASTContext.cpp
index c58cd2c93fb60..bbaaff6bcda4e 100644
--- a/clang/lib/AST/ASTContext.cpp
+++ b/clang/lib/AST/ASTContext.cpp
@@ -3383,21 +3383,27 @@ static void encodeTypeForFunctionPointerAuth(const
ASTContext &Ctx,
// Don't bother discriminating based on these types.
case Type::Pipe:
- case Type::BitInt:
case Type::ConstantMatrix:
OS << "?";
return;
+ case Type::BitInt: {
+const auto *BtTy = T->castAs();
+OS << "D" << (BtTy->isUnsigned() ? "U" : "B") << BtTy->getNumBits() << "_";
+return;
+ }
+
case Type::Builtin: {
const auto *BTy = T->castAs();
-switch (BTy->getKind()) {
+const auto Kind = BTy->getKind();
+switch (Kind) {
#define SIGNED_TYPE(Id, SingletonId)
\
case BuiltinType::Id:
\
-OS << "i";
\
+OS << (Kind == BuiltinType::Int128 ? "n" : "i");
\
return;
#define UNSIGNED_TYPE(Id, SingletonId)
\
case BuiltinType::Id:
\
-OS << "i";
\
+OS << (Kind == BuiltinType::UInt128 ? "o" : "i");
\
return;
#define PLACEHOLDER_TYPE(Id, SingletonId) case BuiltinType::Id:
#define BUILTIN_TYPE(Id, SingletonId)
diff --git a/clang/test/CodeGen/ptrauth-function-type-discriminator.c
b/clang/test/CodeGen/ptrauth-function-type-discriminator.c
index 0952c1abf6c07..9bf4a8874c3c3 100644
--- a/clang/test/CodeGen/ptrauth-function-type-discriminator.c
+++ b/clang/test/CodeGen/ptrauth-function-type-discriminator.c
@@ -65,6 +65,37 @@ void (*fptr3)(void) =
__builtin_ptrauth_sign_constant(&external_function, 2, 26)
// CHECK: @fptr4 = global ptr ptrauth (ptr @external_function, i32 2, i64 26,
ptr @fptr4)
void (*fptr4)(void) = __builtin_ptrauth_sign_constant(&external_function, 2,
__builtin_ptrauth_blend_discriminator(&fptr4, 26));
+extern void external_function_int(int);
+extern void external_function_char(char);
+extern void external_function_i128(__int128_t);
+extern void external_function_u128(__uint128_t);
+extern void external_function_b128(_BitInt(128));
+extern void external_function_b8(_BitInt(8));
+
+// Check discriminators of functions taking integer type arguments:
+
+// - Builtin integer types should be discriminated equally (so, pointer to
+// function taking int argument should accept function taking char argument
+// - _BitInt types are guaranteed distinct and therefore should be
discriminated
+// differently
+// - __int128_t / __uint128_t are passed differently than char / int / long
+// (require two registers instead of one) and therefore should be
discriminated
+// differently.
+
+// CHECK: @fptr5 = global ptr ptrauth (ptr @external_function_int, i32 0, i64
2712)
+// CHECK: @fptr6 = global ptr ptrauth (ptr @external_function_char, i32 0, i64
2712)
+void (*fptr5)(int) = external_function_int;
+void (*fptr6)(char) = external_function_char;
+
+// CHECK: @fptr7 = global ptr ptrauth (ptr @external_function_i128, i32 0, i64
23141)
+// CHECK: @fptr8 = global ptr ptrauth (ptr @external_function_u128, i32 0, i64
45743)
+// CHECK: @fptr9 = global ptr ptrauth (ptr @external_function_b128, i32 0, i64
17854)
+// CHECK: @fptr10 = global ptr ptrauth (ptr @external_function_b8, i32 0, i64
26383)
+void (*fptr7)(__int128_t) = external_function_i128;
+void (*fptr8)(__uint128_t) = external_function_u128;
+void (*fptr9)(_BitInt(128)) = external_function_b128;
+void (*fptr10)(_BitInt(8)) = external_function_b8;
+
// CHECK-
[clang] [llvm] [NFCI][Sanitizer] Convert SpecialCaseList::Sections from StringMap to vector. (PR #140127)
https://github.com/qinkunbao updated
https://github.com/llvm/llvm-project/pull/140127
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,sans-serif,Apple Color Emoji,Segoe UI Emoji,Segoe UI Symbol;
font-size: 14px;
line-height: 1.5;
margin: 0;
}
.container { margin: 50px auto; max-width: 600px; text-align: center;
padding: 0 24px; }
a { color: #0366d6; text-decoration: none; }
a:hover { text-decoration: underline; }
h1 { line-height: 60px; font-size: 48px; font-weight: 300; margin: 0px;
text-shadow: 0 1px 0 #fff; }
p { color: rgba(0, 0, 0, 0.5); margin: 20px 0 40px; }
ul { list-style: none; margin: 25px 0; padding: 0; }
li { display: table-cell; font-weight: bold; width: 1%; }
.logo { display: inline-block; margin-top: 35px; }
.logo-img-2x { display: none; }
@media
only screen and (-webkit-min-device-pixel-ratio: 2),
only screen and ( min--moz-device-pixel-ratio: 2),
only screen and ( -o-min-device-pixel-ratio: 2/1),
only screen and (min-device-pixel-ratio: 2),
only screen and (min-resolution: 192dpi),
only screen and (min-resolution: 2dppx) {
.logo-img-1x { display: none; }
.logo-img-2x { display: inline-block; }
}
#suggestions {
margin-top: 35px;
color: #ccc;
}
#suggestions a {
color: #66;
font-weight: 200;
font-size: 14px;
margin: 0 10px;
}
Whoa there!
You have exceeded a secondary rate limit.
Please wait a few minutes before you try again;
in some cases this may take up to an hour.
https://support.github.com/contact";>Contact Support —
https://githubstatus.com";>GitHub Status —
https://twitter.com/githubstatus";>@githubstatus
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Fixed Constant Evaluation don't Call Destructor (PR #140278)
https://github.com/Mr-Anyone created
https://github.com/llvm/llvm-project/pull/140278
Within the condition statement of the for block, the destructor doesn't get
when evaluating compile time constants.
resolves #139818
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,sans-serif,Apple Color Emoji,Segoe UI Emoji,Segoe UI Symbol;
font-size: 14px;
line-height: 1.5;
margin: 0;
}
.container { margin: 50px auto; max-width: 600px; text-align: center;
padding: 0 24px; }
a { color: #0366d6; text-decoration: none; }
a:hover { text-decoration: underline; }
h1 { line-height: 60px; font-size: 48px; font-weight: 300; margin: 0px;
text-shadow: 0 1px 0 #fff; }
p { color: rgba(0, 0, 0, 0.5); margin: 20px 0 40px; }
ul { list-style: none; margin: 25px 0; padding: 0; }
li { display: table-cell; font-weight: bold; width: 1%; }
.logo { display: inline-block; margin-top: 35px; }
.logo-img-2x { display: none; }
@media
only screen and (-webkit-min-device-pixel-ratio: 2),
only screen and ( min--moz-device-pixel-ratio: 2),
only screen and ( -o-min-device-pixel-ratio: 2/1),
only screen and (min-device-pixel-ratio: 2),
only screen and (min-resolution: 192dpi),
only screen and (min-resolution: 2dppx) {
.logo-img-1x { display: none; }
.logo-img-2x { display: inline-block; }
}
#suggestions {
margin-top: 35px;
color: #ccc;
}
#suggestions a {
color: #66;
font-weight: 200;
font-size: 14px;
margin: 0 10px;
}
Whoa there!
You have exceeded a secondary rate limit.
Please wait a few minutes before you try again;
in some cases this may take up to an hour.
https://support.github.com/contact";>Contact Support —
https://githubstatus.com";>GitHub Status —
https://twitter.com/githubstatus";>@githubstatus
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang] Add new warning: not eliding copy on return (missed NRVO) (PR #139973)
grigorypas wrote: @mizvekov Could you please merge if it is good to go? I don't have write access. https://github.com/llvm/llvm-project/pull/139973 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [NFCI][cfi] Refactor into 'SanitizerInfoFromCFICheckKind' (PR #140117)
vitalybuka wrote: LGTM https://github.com/llvm/llvm-project/pull/140117 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [cfi] Enable -fsanitize-annotate-debug-info functionality for CFI checks (PR #139809)
https://github.com/vitalybuka approved this pull request. https://github.com/llvm/llvm-project/pull/139809 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][CGRecordLayout] Remove dependency on isZeroSize (PR #96422)
rnk wrote: Sorry to hear about the issue. I think the way this is supposed to work is that the clang frontend target ABIInfo is supposed to carefully construct struct types that cause the backend to pass the C struct type correctly in registers and memory. See for example clang/lib/CodeGen/Targets/X86.cpp [classifyReturnType](https://github.com/llvm/llvm-project/blob/7674d6fa9e45c1748d0dd49430aa472028d44a2d/clang/lib/CodeGen/Targets/X86.cpp#L470), which has pretty extensive logic. The AMDGPU one ultimately passes the LLVM struct type directly: https://github.com/llvm/llvm-project/blob/main/clang/lib/CodeGen/Targets/AMDGPU.cpp#L175 The AArch64.cpp backend insulates itself from the LLVM IR struct types that the frontend generates by creating arrays for HVAs: https://github.com/llvm/llvm-project/blob/main/clang/lib/CodeGen/Targets/AArch64.cpp#L453 It seems like AMDGPU should be doing something similar. https://github.com/llvm/llvm-project/pull/96422 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [SPIRV] Add PreLegalizer pattern matching for `faceforward` (PR #139959)
@@ -98,21 +110,98 @@ void applySPIRVDistance(MachineInstr &MI,
MachineRegisterInfo &MRI,
SPIRVGlobalRegistry *GR =
MI.getMF()->getSubtarget().getSPIRVGlobalRegistry();
- auto RemoveAllUses = [&](Register Reg) {
-SmallVector UsesToErase(
-llvm::make_pointer_range(MRI.use_instructions(Reg)));
-
-// calling eraseFromParent to early invalidates the iterator.
-for (auto *MIToErase : UsesToErase) {
- GR->invalidateMachineInstr(MIToErase);
- MIToErase->eraseFromParent();
-}
- };
- RemoveAllUses(SubDestReg); // remove all uses of FSUB Result
+ removeAllUses(SubDestReg, MRI, GR); // remove all uses of FSUB Result
GR->invalidateMachineInstr(SubInstr);
SubInstr->eraseFromParent(); // remove FSUB instruction
}
+/// This match is part of a combine that
+/// rewrites select(fcmp(dot(I, Ng), 0), N, 0 - N) to faceforward(N, I, Ng)
+/// (vXf32 (g_select
+/// (g_fcmp
+///(g_intrinsic dot(vXf32 I) (vXf32 Ng)
+/// 0)
+/// (vXf32 N)
+/// (vXf32 g_fsub (0) (vXf32 N
+/// ->
+/// (vXf32 (g_intrinsic faceforward
+/// (vXf32 N) (vXf32 I) (vXf32 Ng)))
+///
+bool matchSelectToFaceForward(MachineInstr &MI, MachineRegisterInfo &MRI) {
+ if (MI.getOpcode() != TargetOpcode::G_SELECT)
+return false;
+
+ // Check if select's condition is a comparison between a dot product and 0.
+ Register CondReg = MI.getOperand(1).getReg();
+ MachineInstr *CondInstr = MRI.getVRegDef(CondReg);
+ if (!CondInstr || CondInstr->getOpcode() != TargetOpcode::G_FCMP)
+return false;
+
+ Register DotReg = CondInstr->getOperand(2).getReg();
+ MachineInstr *DotInstr = MRI.getVRegDef(DotReg);
+ if (DotInstr->getOpcode() != TargetOpcode::G_FMUL &&
+ (DotInstr->getOpcode() != TargetOpcode::G_INTRINSIC ||
+ cast(DotInstr)->getIntrinsicID() != Intrinsic::spv_fdot))
+return false;
+
+ Register CondZeroReg = CondInstr->getOperand(3).getReg();
+ MachineInstr *CondZeroInstr = MRI.getVRegDef(CondZeroReg);
+ if (CondZeroInstr->getOpcode() != TargetOpcode::G_FCONSTANT ||
+ !CondZeroInstr->getOperand(1).getFPImm()->isZero())
+return false;
+
+ // Check if select's false operand is the negation of the true operand.
+ Register TrueReg = MI.getOperand(2).getReg();
+ Register FalseReg = MI.getOperand(3).getReg();
+ MachineInstr *FalseInstr = MRI.getVRegDef(FalseReg);
+ if (FalseInstr->getOpcode() != TargetOpcode::G_FNEG)
+return false;
+ if (TrueReg != FalseInstr->getOperand(1).getReg())
+return false;
kmpeng wrote:
Yeah, I was also wondering if this way of checking would be a problem since
there's multiple ways to represent 2 operands being negations (another example
being `G_FSUB`). I'm also still learning and have not been able to come up with
a generic way to check this. @farzonl Do you have any thoughts?
https://github.com/llvm/llvm-project/pull/139959
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [DirectX][SPIRV] Fix the lowering of dot4add (PR #140315)
https://github.com/inbelic approved this pull request. https://github.com/llvm/llvm-project/pull/140315 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
@@ -70,18 +70,116 @@ class MustacheTemplateFile : public Template {
MustacheTemplateFile(StringRef TemplateStr) : Template(TemplateStr) {}
};
+static std::unique_ptr NamespaceTemplate = nullptr;
+
+static std::unique_ptr RecordTemplate = nullptr;
+
+static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ return Error::success();
+}
+
Error MustacheHTMLGenerator::generateDocs(
StringRef RootDir, StringMap> Infos,
const clang::doc::ClangDocContext &CDCtx) {
+ if (auto Err = setupTemplateFiles(CDCtx))
+return Err;
+ // Track which directories we already tried to create.
+ StringSet<> CreatedDirs;
+ // Collect all output by file name and create the necessary directories.
+ StringMap> FileToInfos;
+ for (const auto &Group : Infos) {
+doc::Info *Info = Group.getValue().get();
+
+SmallString<128> Path;
+sys::path::native(RootDir, Path);
+sys::path::append(Path, Info->getRelativeFilePath(""));
+if (!CreatedDirs.contains(Path)) {
+ if (std::error_code EC = sys::fs::create_directories(Path))
+return createStringError(EC, "failed to create directory '%s'.",
+ Path.c_str());
ilovepi wrote:
Ah, good catch. I had missed that.
https://github.com/llvm/llvm-project/pull/138061
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [OpenACC][CIR] Implement beginning of 'copy' lowering for compute con… (PR #140304)
https://github.com/bcardosolopes commented: A question in the comments but otherwise PR looks good: adds boilerplate and initial cases, deferring more work for incremental PRs, way to go. https://github.com/llvm/llvm-project/pull/140304 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] [Cygwin][MinGW] Internal class in explicitly-instantiation-declarated template should be instantiated (PR #140145)
jeremyd2019 wrote: They've failed linking things due to missing `std::istream::sentry` - I've started a new run https://github.com/jeremyd2019/llvm-mingw/actions/runs/15078849158 with ``` #if defined(__MINGW32__) || defined(__CYGWIN__) extern template class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS basic_istream::sentry; #endif ``` rather than explicitly listing the member(s). https://github.com/llvm/llvm-project/pull/140145 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [DirectX][SPIRV] Fix the lowering of dot4add (PR #140315)
@@ -1,17 +1,19 @@ -// RUN: %clang_cc1 -finclude-default-header -triple \ -// RUN: dxil-pc-shadermodel6.3-compute %s -emit-llvm -disable-llvm-passes -o - | \ -// RUN: FileCheck %s -DTARGET=dx -// RUN: %clang_cc1 -finclude-default-header -triple \ -// RUN: spirv-pc-vulkan-compute %s -emit-llvm -disable-llvm-passes -o - | \ -// RUN: FileCheck %s -DTARGET=spv +// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.3-compute %s -emit-llvm -o - | FileCheck %s -DTARGET=dx inbelic wrote: ```suggestion // RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.4-compute %s -emit-llvm -o - | FileCheck %s -DTARGET=dx ``` Should this error on the lower version at this point as well, or, is it fine to rely on the error generated when lowering further? https://github.com/llvm/llvm-project/pull/140315 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
@@ -70,18 +70,116 @@ class MustacheTemplateFile : public Template {
MustacheTemplateFile(StringRef TemplateStr) : Template(TemplateStr) {}
};
+static std::unique_ptr NamespaceTemplate = nullptr;
+
+static std::unique_ptr RecordTemplate = nullptr;
+
+static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ return Error::success();
+}
+
Error MustacheHTMLGenerator::generateDocs(
StringRef RootDir, StringMap> Infos,
const clang::doc::ClangDocContext &CDCtx) {
+ if (auto Err = setupTemplateFiles(CDCtx))
+return Err;
+ // Track which directories we already tried to create.
+ StringSet<> CreatedDirs;
+ // Collect all output by file name and create the necessary directories.
+ StringMap> FileToInfos;
+ for (const auto &Group : Infos) {
+doc::Info *Info = Group.getValue().get();
+
+SmallString<128> Path;
+sys::path::native(RootDir, Path);
+sys::path::append(Path, Info->getRelativeFilePath(""));
+if (!CreatedDirs.contains(Path)) {
+ if (std::error_code EC = sys::fs::create_directories(Path))
+return createStringError(EC, "failed to create directory '%s'.",
+ Path.c_str());
ilovepi wrote:
ah, no dice. We hit a static assert in Format.h
```
error: static assertion failed due to requirement
'std::is_scalar_v': format can't be used with non fundamental
/ non pointer type
85 | static_assert(std::is_scalar_v
```
https://github.com/llvm/llvm-project/pull/138061
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clangd] Add tweak to override pure virtuals (PR #139348)
https://github.com/marcogmaia edited https://github.com/llvm/llvm-project/pull/139348 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] suggest headers on undeclared errors (PR #140247)
https://github.com/jongmyeong-choi edited https://github.com/llvm/llvm-project/pull/140247 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [clang][ExprConst] allow single element access of vector object to be constant expression (PR #101126)
ziqingluo-90 wrote:
```
typedef __attribute__((__ext_vector_type__(2))) float float2;
struct vec2 {
float2 _v;
public:
constexpr vec2(float x, float y) {
_v.x = x;
_v.y = y;
}
};
constexpr struct vec2 f() {
return vec2(1.0, 1.0);
}
int main() {
vec2 S = f();
}
```
Clang crashes: https://godbolt.org/z/sx74s1bqd
https://github.com/llvm/llvm-project/pull/101126
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [SPIRV] Add PreLegalizer pattern matching for `faceforward` (PR #139959)
https://github.com/kmpeng updated
https://github.com/llvm/llvm-project/pull/139959
>From be7df595c1df0f310a1941da3d3ac3529270241e Mon Sep 17 00:00:00 2001
From: kmpeng
Date: Tue, 29 Apr 2025 16:46:08 -0700
Subject: [PATCH 1/5] WIP - debug `matchSelectToFaceForward`
---
.../lib/Headers/hlsl/hlsl_intrinsic_helpers.h | 9 +-
llvm/lib/Target/SPIRV/SPIRVCombine.td | 9 +-
.../SPIRV/SPIRVPreLegalizerCombiner.cpp | 100 ++
...egalizercombiner-select-to-faceforward.mir | 33 ++
.../SPIRV/hlsl-intrinsics/faceforward.ll | 44 ++--
5 files changed, 153 insertions(+), 42 deletions(-)
create mode 100644
llvm/test/CodeGen/SPIRV/GlobalISel/InstCombine/prelegalizercombiner-select-to-faceforward.mir
diff --git a/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
b/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
index 4eb7b8f45c85a..68dc6d96f89c1 100644
--- a/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
+++ b/clang/lib/Headers/hlsl/hlsl_intrinsic_helpers.h
@@ -127,11 +127,12 @@ template constexpr vector lit_impl(T
NDotL, T NDotH, T M) {
}
template constexpr T faceforward_impl(T N, T I, T Ng) {
-#if (__has_builtin(__builtin_spirv_faceforward))
- return __builtin_spirv_faceforward(N, I, Ng);
-#else
+// #if (__has_builtin(__builtin_spirv_faceforward))
+// return __builtin_spirv_faceforward(N, I, Ng);
+// #else
+// return select(dot(I, Ng) < 0, N, -N);
+// #endif
return select(dot(I, Ng) < 0, N, -N);
-#endif
}
template constexpr T ldexp_impl(T X, T Exp) {
diff --git a/llvm/lib/Target/SPIRV/SPIRVCombine.td
b/llvm/lib/Target/SPIRV/SPIRVCombine.td
index 6f726e024de52..78a57146e61d2 100644
--- a/llvm/lib/Target/SPIRV/SPIRVCombine.td
+++ b/llvm/lib/Target/SPIRV/SPIRVCombine.td
@@ -15,8 +15,15 @@ def vector_length_sub_to_distance_lowering : GICombineRule <
(apply [{ applySPIRVDistance(*${root}, MRI, B); }])
>;
+def vector_select_to_faceforward_lowering : GICombineRule <
+ (defs root:$root),
+ (match (wip_match_opcode G_SELECT):$root,
+ [{ return matchSelectToFaceForward(*${root}, MRI); }]),
+ (apply [{ applySPIRVFaceForward(*${root}, MRI, B); }])
+>;
+
def SPIRVPreLegalizerCombiner
: GICombiner<"SPIRVPreLegalizerCombinerImpl",
- [vector_length_sub_to_distance_lowering]> {
+ [vector_length_sub_to_distance_lowering,
vector_select_to_faceforward_lowering]> {
let CombineAllMethodName = "tryCombineAllImpl";
}
diff --git a/llvm/lib/Target/SPIRV/SPIRVPreLegalizerCombiner.cpp
b/llvm/lib/Target/SPIRV/SPIRVPreLegalizerCombiner.cpp
index c96ee6b02491a..d323453e68828 100644
--- a/llvm/lib/Target/SPIRV/SPIRVPreLegalizerCombiner.cpp
+++ b/llvm/lib/Target/SPIRV/SPIRVPreLegalizerCombiner.cpp
@@ -113,6 +113,106 @@ void applySPIRVDistance(MachineInstr &MI,
MachineRegisterInfo &MRI,
SubInstr->eraseFromParent(); // remove FSUB instruction
}
+/// This match is part of a combine that
+/// rewrites select(fcmp(dot(I, Ng), 0), N, 0 - N) to faceforward(N, I, Ng)
+/// (vXf32 (g_select
+/// (g_fcmp
+///(g_intrinsic dot(vXf32 I) (vXf32 Ng)
+/// 0)
+/// (vXf32 N)
+/// (vXf32 g_fsub (0) (vXf32 N
+/// ->
+/// (vXf32 (g_intrinsic faceforward
+/// (vXf32 N) (vXf32 I) (vXf32 Ng)))
+///
+bool matchSelectToFaceForward(MachineInstr &MI, MachineRegisterInfo &MRI) {
+ if (MI.getOpcode() != TargetOpcode::G_SELECT)
+return false;
+
+ // Check if the condition is a comparison between a dot product and zero
+ Register CondReg = MI.getOperand(1).getReg();
+ MachineInstr *CondInstr = MRI.getVRegDef(CondReg);
+ if (!CondInstr || CondInstr->getOpcode() != TargetOpcode::G_FCMP)
+return false;
+
+ if (!CondInstr->getOperand(2).getReg())
+return false;
+ Register DotReg = CondInstr->getOperand(2).getReg();
+ MachineInstr *DotInstr = MRI.getVRegDef(DotReg);
+ // fmul?
+ if (DotInstr->getOpcode() != TargetOpcode::G_INTRINSIC ||
cast(DotInstr)->getIntrinsicID() != Intrinsic::spv_fdot)
+return false;
+ // // is using getImm() correct?
+ // if (!CondInstr->getOperand(3).isImm() ||
+ // CondInstr->getOperand(3).getImm() != 0)
+ // return false;
+
+ // // Check if the false operand is the negation of the true operand
+ // Register TrueReg = MI.getOperand(2).getReg();
+ // Register FalseReg = MI.getOperand(3).getReg();
+ // MachineInstr *FalseInstr = MRI.getVRegDef(FalseReg);
+ // // getImm() correct?
+ // if (FalseInstr->getOpcode() != TargetOpcode::G_FSUB ||
+ // FalseInstr->getOperand(1).getImm() != 0 ||
+ // FalseInstr->getOperand(2).getReg() != TrueReg)
+ // return false;
+
+ return true;
+}
+void applySPIRVFaceForward(MachineInstr &MI, MachineRegisterInfo &MRI,
+ MachineIRBuilder &B) {
+
+ // Extract the operands for N, I, and Ng from the match criteria.
+ Register CondReg = MI.getOperand(1).getReg();
+ MachineInstr *CondInstr = MRI.getVRegDef(CondReg);
+ Register DotReg = Co
[clang] [llvm] [SPIRV] Add PreLegalizer pattern matching for `faceforward` (PR #139959)
@@ -98,21 +110,98 @@ void applySPIRVDistance(MachineInstr &MI,
MachineRegisterInfo &MRI,
SPIRVGlobalRegistry *GR =
MI.getMF()->getSubtarget().getSPIRVGlobalRegistry();
- auto RemoveAllUses = [&](Register Reg) {
-SmallVector UsesToErase(
-llvm::make_pointer_range(MRI.use_instructions(Reg)));
-
-// calling eraseFromParent to early invalidates the iterator.
-for (auto *MIToErase : UsesToErase) {
- GR->invalidateMachineInstr(MIToErase);
- MIToErase->eraseFromParent();
-}
- };
- RemoveAllUses(SubDestReg); // remove all uses of FSUB Result
+ removeAllUses(SubDestReg, MRI, GR); // remove all uses of FSUB Result
GR->invalidateMachineInstr(SubInstr);
SubInstr->eraseFromParent(); // remove FSUB instruction
}
+/// This match is part of a combine that
+/// rewrites select(fcmp(dot(I, Ng), 0), N, 0 - N) to faceforward(N, I, Ng)
+/// (vXf32 (g_select
+/// (g_fcmp
+///(g_intrinsic dot(vXf32 I) (vXf32 Ng)
+/// 0)
+/// (vXf32 N)
+/// (vXf32 g_fsub (0) (vXf32 N
+/// ->
+/// (vXf32 (g_intrinsic faceforward
+/// (vXf32 N) (vXf32 I) (vXf32 Ng)))
+///
+bool matchSelectToFaceForward(MachineInstr &MI, MachineRegisterInfo &MRI) {
+ if (MI.getOpcode() != TargetOpcode::G_SELECT)
+return false;
+
+ // Check if select's condition is a comparison between a dot product and 0.
+ Register CondReg = MI.getOperand(1).getReg();
+ MachineInstr *CondInstr = MRI.getVRegDef(CondReg);
+ if (!CondInstr || CondInstr->getOpcode() != TargetOpcode::G_FCMP)
+return false;
+
+ Register DotReg = CondInstr->getOperand(2).getReg();
+ MachineInstr *DotInstr = MRI.getVRegDef(DotReg);
+ if (DotInstr->getOpcode() != TargetOpcode::G_FMUL &&
kmpeng wrote:
You're right, this would pose a problem. Made an update to check that the dot
operands are not vectors. Let me know if that looks good.
https://github.com/llvm/llvm-project/pull/139959
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] libcxx: std::ostream::sentry should be exported (PR #140169)
https://github.com/jeremyd2019 closed https://github.com/llvm/llvm-project/pull/140169 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] libcxx: std::ostream::sentry should be exported (PR #140169)
https://github.com/jeremyd2019 reopened https://github.com/llvm/llvm-project/pull/140169 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [DirectX][SPIRV] Fix the lowering of dot4add (PR #140315)
@@ -1,17 +1,19 @@ -// RUN: %clang_cc1 -finclude-default-header -triple \ -// RUN: dxil-pc-shadermodel6.3-compute %s -emit-llvm -disable-llvm-passes -o - | \ -// RUN: FileCheck %s -DTARGET=dx -// RUN: %clang_cc1 -finclude-default-header -triple \ -// RUN: spirv-pc-vulkan-compute %s -emit-llvm -disable-llvm-passes -o - | \ -// RUN: FileCheck %s -DTARGET=spv +// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.3-compute %s -emit-llvm -o - | FileCheck %s -DTARGET=dx bogner wrote: Yes, this should error. However, the definitions are already marked with `_HLSL_AVAILABILITY(shadermodel, 6.4)` so I'm not sure why this *isn't* erroring. I'll look into this. https://github.com/llvm/llvm-project/pull/140315 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
https://github.com/ilovepi updated
https://github.com/llvm/llvm-project/pull/138061
>From de0222f0459bb6c5832ccdbe70b19f5ea8b0cd98 Mon Sep 17 00:00:00 2001
From: Paul Kirth
Date: Wed, 30 Apr 2025 08:09:41 -0700
Subject: [PATCH] [clang-doc] Add HTMLMustacheGenerator methods
Split from #133161. This patch fills in the implementation for a number
of the MustacheHTMLGenerator methods. Many of these APIs are just
stubbed out, and will have their implementation filled in by later
patches.
Co-authored-by: Peter Chou
---
.../clang-doc/HTMLMustacheGenerator.cpp | 95 +++
.../unittests/clang-doc/CMakeLists.txt| 1 +
.../clang-doc/HTMLMustacheGeneratorTest.cpp | 37 +++-
3 files changed, 130 insertions(+), 3 deletions(-)
diff --git a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
index 6ba0325685599..366deb55b77b9 100644
--- a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
+++ b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
@@ -70,18 +70,113 @@ class MustacheTemplateFile : public Template {
MustacheTemplateFile(StringRef TemplateStr) : Template(TemplateStr) {}
};
+static std::unique_ptr NamespaceTemplate = nullptr;
+
+static std::unique_ptr RecordTemplate = nullptr;
+
+static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ return Error::success();
+}
+
Error MustacheHTMLGenerator::generateDocs(
StringRef RootDir, StringMap> Infos,
const clang::doc::ClangDocContext &CDCtx) {
+ if (auto Err = setupTemplateFiles(CDCtx))
+return Err;
+ // Track which directories we already tried to create.
+ StringSet<> CreatedDirs;
+ // Collect all output by file name and create the necessary directories.
+ StringMap> FileToInfos;
+ for (const auto &Group : Infos) {
+doc::Info *Info = Group.getValue().get();
+
+SmallString<128> Path;
+sys::path::native(RootDir, Path);
+sys::path::append(Path, Info->getRelativeFilePath(""));
+if (!CreatedDirs.contains(Path)) {
+ if (std::error_code EC = sys::fs::create_directories(Path))
+return createStringError(EC, "failed to create directory '%s'.",
+ Path.c_str());
+ CreatedDirs.insert(Path);
+}
+
+sys::path::append(Path, Info->getFileBaseName() + ".html");
+FileToInfos[Path].push_back(Info);
+ }
+
+ for (const auto &Group : FileToInfos) {
+std::error_code FileErr;
+raw_fd_ostream InfoOS(Group.getKey(), FileErr, sys::fs::OF_None);
+if (FileErr)
+ return createFileOpenError(Group.getKey(), FileErr);
+
+for (const auto &Info : Group.getValue())
+ if (Error Err = generateDocForInfo(Info, InfoOS, CDCtx))
+return Err;
+ }
return Error::success();
}
+static json::Value extractValue(const NamespaceInfo &I,
+const ClangDocContext &CDCtx) {
+ Object NamespaceValue = Object();
+ return NamespaceValue;
+}
+
+static json::Value extractValue(const RecordInfo &I,
+const ClangDocContext &CDCtx) {
+ Object RecordValue = Object();
+ return RecordValue;
+}
+
+static Error setupTemplateValue(const ClangDocContext &CDCtx, json::Value &V,
+Info *I) {
+ return createStringError(inconvertibleErrorCode(),
+ "setupTemplateValue is unimplemented");
+}
+
Error MustacheHTMLGenerator::generateDocForInfo(Info *I, raw_ostream &OS,
const ClangDocContext &CDCtx) {
+ switch (I->IT) {
+ case InfoType::IT_namespace: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+NamespaceTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_record: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+// Serialize the JSON value to the output stream in a readable format.
+RecordTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_enum:
+OS << "IT_enum\n";
+break;
+ case InfoType::IT_function:
+OS << "IT_Function\n";
+break;
+ case InfoType::IT_typedef:
+OS << "IT_typedef\n";
+break;
+ case InfoType::IT_default:
+return createStringError(inconvertibleErrorCode(), "unexpected InfoType");
+ }
return Error::success();
}
Error MustacheHTMLGenerator::createResources(ClangDocContext &CDCtx) {
+ for (const auto &FilePath : CDCtx.UserStylesheets)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
+ for (const auto &FilePath : CDCtx.JsScripts)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
return Error::success();
}
diff --git a/clang-tools-extra/unittests/clang-doc/CMakeLists.txt
b/clang-tools-extra/unittests/clang-doc/CMakeLists.txt
index 81c18e6014072..fd14d85c63485 10
[clang] [llvm] [SPIRV] Add PreLegalizer pattern matching for `faceforward` (PR #139959)
https://github.com/kmpeng edited https://github.com/llvm/llvm-project/pull/139959 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [SPIRV] Add PreLegalizer pattern matching for `faceforward` (PR #139959)
https://github.com/kmpeng edited https://github.com/llvm/llvm-project/pull/139959 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
https://github.com/andykaylor created
https://github.com/llvm/llvm-project/pull/140322
There was a problem introduced today where sometimes the CIR for functions with
no arguments would be generated with phantom arguments. This was causing
intermittent test failures.
The problem appears to have been that we were using two different Profile
implementations to generate the FoldingSetNodeID for CIRGenFunctionInfo so
occaissionally when we tried to look for a pre-existing entry for a function
with no arguments it would incorrectly match a CIRGenFunctionInfo entry that
had arguments.
To prevent this from happening again, I rewrote one of the two Profile
functions to call the other.
>From 303469f78c951eb25d57b1309ad00e9670a7d475 Mon Sep 17 00:00:00 2001
From: Andy Kaylor
Date: Fri, 16 May 2025 17:07:16 -0700
Subject: [PATCH] [CIR] Fix problem with phantom function arguments
There was a problem introduced today where sometimes the CIR for functions
with no arguments would be generated with phantom arguments. This was
causing intermittent test failures.
The problem appears to have been that we were using two different Profile
implementations to generate the FoldingSetNodeID for CIRGenFunctionInfo
so occaissionally when we tried to look for a pre-existing entry for
a function with no arguments it would incorrectly match a
CIRGenFunctionInfo entry that had arguments.
To prevent this from happening again, I rewrote one of the two Profile
functions to call the other.
---
clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h | 12 ++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
index 1e06599575fbd..b74460b09a44e 100644
--- a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
+++ b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
+// static Profile function requires.
+llvm::SmallVector argTypes;
+for (const ArgInfo &argInfo : arguments())
+ argTypes.push_back(argInfo.type);
+
+Profile(id, required, getReturnType(), argTypes);
}
llvm::ArrayRef arguments() const {
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
llvmbot wrote:
@llvm/pr-subscribers-clangir
Author: Andy Kaylor (andykaylor)
Changes
There was a problem introduced today where sometimes the CIR for functions with
no arguments would be generated with phantom arguments. This was causing
intermittent test failures.
The problem appears to have been that we were using two different Profile
implementations to generate the FoldingSetNodeID for CIRGenFunctionInfo so
occaissionally when we tried to look for a pre-existing entry for a function
with no arguments it would incorrectly match a CIRGenFunctionInfo entry that
had arguments.
To prevent this from happening again, I rewrote one of the two Profile
functions to call the other.
---
Full diff: https://github.com/llvm/llvm-project/pull/140322.diff
1 Files Affected:
- (modified) clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h (+10-2)
``diff
diff --git a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
index 1e06599575fbd..b74460b09a44e 100644
--- a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
+++ b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
+// static Profile function requires.
+llvm::SmallVector argTypes;
+for (const ArgInfo &argInfo : arguments())
+ argTypes.push_back(argInfo.type);
+
+Profile(id, required, getReturnType(), argTypes);
}
llvm::ArrayRef arguments() const {
``
https://github.com/llvm/llvm-project/pull/140322
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Andy Kaylor (andykaylor)
Changes
There was a problem introduced today where sometimes the CIR for functions with
no arguments would be generated with phantom arguments. This was causing
intermittent test failures.
The problem appears to have been that we were using two different Profile
implementations to generate the FoldingSetNodeID for CIRGenFunctionInfo so
occaissionally when we tried to look for a pre-existing entry for a function
with no arguments it would incorrectly match a CIRGenFunctionInfo entry that
had arguments.
To prevent this from happening again, I rewrote one of the two Profile
functions to call the other.
---
Full diff: https://github.com/llvm/llvm-project/pull/140322.diff
1 Files Affected:
- (modified) clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h (+10-2)
``diff
diff --git a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
index 1e06599575fbd..b74460b09a44e 100644
--- a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
+++ b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
+// static Profile function requires.
+llvm::SmallVector argTypes;
+for (const ArgInfo &argInfo : arguments())
+ argTypes.push_back(argInfo.type);
+
+Profile(id, required, getReturnType(), argTypes);
}
llvm::ArrayRef arguments() const {
``
https://github.com/llvm/llvm-project/pull/140322
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
andykaylor wrote:
This problem will be fixed when I refactor the code to eliminate the ArgInfo
class, which now contains only the type.
https://github.com/llvm/llvm-project/pull/140322
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
ilovepi wrote: ### Merge activity * **May 16, 8:28 PM EDT**: A user started a stack merge that includes this pull request via [Graphite](https://app.graphite.dev/github/pr/llvm/llvm-project/138061). https://github.com/llvm/llvm-project/pull/138061 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
https://github.com/bcardosolopes approved this pull request. LGTM. Is there an easy to add a test for this? I understand sometimes it's hard, but just double checking. https://github.com/llvm/llvm-project/pull/140322 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
https://github.com/ilovepi updated
https://github.com/llvm/llvm-project/pull/138061
>From 187b130d43c4470768b3885cb9c333b9edfb3897 Mon Sep 17 00:00:00 2001
From: Paul Kirth
Date: Wed, 30 Apr 2025 08:09:41 -0700
Subject: [PATCH] [clang-doc] Add HTMLMustacheGenerator methods
Split from #133161. This patch fills in the implementation for a number
of the MustacheHTMLGenerator methods. Many of these APIs are just
stubbed out, and will have their implementation filled in by later
patches.
Co-authored-by: Peter Chou
---
.../clang-doc/HTMLMustacheGenerator.cpp | 95 +++
.../unittests/clang-doc/CMakeLists.txt| 1 +
.../clang-doc/HTMLMustacheGeneratorTest.cpp | 37 +++-
3 files changed, 130 insertions(+), 3 deletions(-)
diff --git a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
index 6ba0325685599..366deb55b77b9 100644
--- a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
+++ b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
@@ -70,18 +70,113 @@ class MustacheTemplateFile : public Template {
MustacheTemplateFile(StringRef TemplateStr) : Template(TemplateStr) {}
};
+static std::unique_ptr NamespaceTemplate = nullptr;
+
+static std::unique_ptr RecordTemplate = nullptr;
+
+static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ return Error::success();
+}
+
Error MustacheHTMLGenerator::generateDocs(
StringRef RootDir, StringMap> Infos,
const clang::doc::ClangDocContext &CDCtx) {
+ if (auto Err = setupTemplateFiles(CDCtx))
+return Err;
+ // Track which directories we already tried to create.
+ StringSet<> CreatedDirs;
+ // Collect all output by file name and create the necessary directories.
+ StringMap> FileToInfos;
+ for (const auto &Group : Infos) {
+doc::Info *Info = Group.getValue().get();
+
+SmallString<128> Path;
+sys::path::native(RootDir, Path);
+sys::path::append(Path, Info->getRelativeFilePath(""));
+if (!CreatedDirs.contains(Path)) {
+ if (std::error_code EC = sys::fs::create_directories(Path))
+return createStringError(EC, "failed to create directory '%s'.",
+ Path.c_str());
+ CreatedDirs.insert(Path);
+}
+
+sys::path::append(Path, Info->getFileBaseName() + ".html");
+FileToInfos[Path].push_back(Info);
+ }
+
+ for (const auto &Group : FileToInfos) {
+std::error_code FileErr;
+raw_fd_ostream InfoOS(Group.getKey(), FileErr, sys::fs::OF_None);
+if (FileErr)
+ return createFileOpenError(Group.getKey(), FileErr);
+
+for (const auto &Info : Group.getValue())
+ if (Error Err = generateDocForInfo(Info, InfoOS, CDCtx))
+return Err;
+ }
return Error::success();
}
+static json::Value extractValue(const NamespaceInfo &I,
+const ClangDocContext &CDCtx) {
+ Object NamespaceValue = Object();
+ return NamespaceValue;
+}
+
+static json::Value extractValue(const RecordInfo &I,
+const ClangDocContext &CDCtx) {
+ Object RecordValue = Object();
+ return RecordValue;
+}
+
+static Error setupTemplateValue(const ClangDocContext &CDCtx, json::Value &V,
+Info *I) {
+ return createStringError(inconvertibleErrorCode(),
+ "setupTemplateValue is unimplemented");
+}
+
Error MustacheHTMLGenerator::generateDocForInfo(Info *I, raw_ostream &OS,
const ClangDocContext &CDCtx) {
+ switch (I->IT) {
+ case InfoType::IT_namespace: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+NamespaceTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_record: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+// Serialize the JSON value to the output stream in a readable format.
+RecordTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_enum:
+OS << "IT_enum\n";
+break;
+ case InfoType::IT_function:
+OS << "IT_Function\n";
+break;
+ case InfoType::IT_typedef:
+OS << "IT_typedef\n";
+break;
+ case InfoType::IT_default:
+return createStringError(inconvertibleErrorCode(), "unexpected InfoType");
+ }
return Error::success();
}
Error MustacheHTMLGenerator::createResources(ClangDocContext &CDCtx) {
+ for (const auto &FilePath : CDCtx.UserStylesheets)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
+ for (const auto &FilePath : CDCtx.JsScripts)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
return Error::success();
}
diff --git a/clang-tools-extra/unittests/clang-doc/CMakeLists.txt
b/clang-tools-extra/unittests/clang-doc/CMakeLists.txt
index 81c18e6014072..fd14d85c63485 10
[clang-tools-extra] 6a0e626 - [clang-doc] Add HTMLMustacheGenerator methods (#138061)
Author: Paul Kirth
Date: 2025-05-16T17:30:45-07:00
New Revision: 6a0e626af58bdf8ae8c07cfe8aec94f4565734a3
URL:
https://github.com/llvm/llvm-project/commit/6a0e626af58bdf8ae8c07cfe8aec94f4565734a3
DIFF:
https://github.com/llvm/llvm-project/commit/6a0e626af58bdf8ae8c07cfe8aec94f4565734a3.diff
LOG: [clang-doc] Add HTMLMustacheGenerator methods (#138061)
Split from #133161. This patch fills in the implementation for a number
of the MustacheHTMLGenerator methods. Many of these APIs are just
stubbed out, and will have their implementation filled in by later
patches.
Co-authored-by: Peter Chou
Added:
Modified:
clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
clang-tools-extra/unittests/clang-doc/CMakeLists.txt
clang-tools-extra/unittests/clang-doc/HTMLMustacheGeneratorTest.cpp
Removed:
diff --git a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
index 6ba0325685599..366deb55b77b9 100644
--- a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
+++ b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
@@ -70,18 +70,113 @@ class MustacheTemplateFile : public Template {
MustacheTemplateFile(StringRef TemplateStr) : Template(TemplateStr) {}
};
+static std::unique_ptr NamespaceTemplate = nullptr;
+
+static std::unique_ptr RecordTemplate = nullptr;
+
+static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ return Error::success();
+}
+
Error MustacheHTMLGenerator::generateDocs(
StringRef RootDir, StringMap> Infos,
const clang::doc::ClangDocContext &CDCtx) {
+ if (auto Err = setupTemplateFiles(CDCtx))
+return Err;
+ // Track which directories we already tried to create.
+ StringSet<> CreatedDirs;
+ // Collect all output by file name and create the necessary directories.
+ StringMap> FileToInfos;
+ for (const auto &Group : Infos) {
+doc::Info *Info = Group.getValue().get();
+
+SmallString<128> Path;
+sys::path::native(RootDir, Path);
+sys::path::append(Path, Info->getRelativeFilePath(""));
+if (!CreatedDirs.contains(Path)) {
+ if (std::error_code EC = sys::fs::create_directories(Path))
+return createStringError(EC, "failed to create directory '%s'.",
+ Path.c_str());
+ CreatedDirs.insert(Path);
+}
+
+sys::path::append(Path, Info->getFileBaseName() + ".html");
+FileToInfos[Path].push_back(Info);
+ }
+
+ for (const auto &Group : FileToInfos) {
+std::error_code FileErr;
+raw_fd_ostream InfoOS(Group.getKey(), FileErr, sys::fs::OF_None);
+if (FileErr)
+ return createFileOpenError(Group.getKey(), FileErr);
+
+for (const auto &Info : Group.getValue())
+ if (Error Err = generateDocForInfo(Info, InfoOS, CDCtx))
+return Err;
+ }
return Error::success();
}
+static json::Value extractValue(const NamespaceInfo &I,
+const ClangDocContext &CDCtx) {
+ Object NamespaceValue = Object();
+ return NamespaceValue;
+}
+
+static json::Value extractValue(const RecordInfo &I,
+const ClangDocContext &CDCtx) {
+ Object RecordValue = Object();
+ return RecordValue;
+}
+
+static Error setupTemplateValue(const ClangDocContext &CDCtx, json::Value &V,
+Info *I) {
+ return createStringError(inconvertibleErrorCode(),
+ "setupTemplateValue is unimplemented");
+}
+
Error MustacheHTMLGenerator::generateDocForInfo(Info *I, raw_ostream &OS,
const ClangDocContext &CDCtx) {
+ switch (I->IT) {
+ case InfoType::IT_namespace: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+NamespaceTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_record: {
+json::Value V =
+extractValue(*static_cast(I), CDCtx);
+if (auto Err = setupTemplateValue(CDCtx, V, I))
+ return Err;
+// Serialize the JSON value to the output stream in a readable format.
+RecordTemplate->render(V, OS);
+break;
+ }
+ case InfoType::IT_enum:
+OS << "IT_enum\n";
+break;
+ case InfoType::IT_function:
+OS << "IT_Function\n";
+break;
+ case InfoType::IT_typedef:
+OS << "IT_typedef\n";
+break;
+ case InfoType::IT_default:
+return createStringError(inconvertibleErrorCode(), "unexpected InfoType");
+ }
return Error::success();
}
Error MustacheHTMLGenerator::createResources(ClangDocContext &CDCtx) {
+ for (const auto &FilePath : CDCtx.UserStylesheets)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
+ for (const auto &FilePath : CDCtx.JsScripts)
+if (Error Err = copyFile(FilePath, CDCtx.OutDirectory))
+ return Err;
return Error::success();
}
diff --g
[clang-tools-extra] [clang-doc] Add HTMLMustacheGenerator methods (PR #138061)
https://github.com/ilovepi closed https://github.com/llvm/llvm-project/pull/138061 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] [clang-doc] Add helpers for Template config (PR #138062)
https://github.com/ilovepi updated
https://github.com/llvm/llvm-project/pull/138062
>From 442fe2fd1353c540bc0ec39ec913e7e414def6bc Mon Sep 17 00:00:00 2001
From: Paul Kirth
Date: Wed, 30 Apr 2025 08:10:20 -0700
Subject: [PATCH] [clang-doc] Add helpers for Template config
This patch adds or fills in some helper functions related to template
setup when initializing the mustache backend. It was split from #133161.
Co-authored-by: Peter Chou
---
.../clang-doc/HTMLMustacheGenerator.cpp | 45 ++
.../clang-doc/support/CMakeLists.txt | 4 +-
clang-tools-extra/clang-doc/support/Utils.cpp | 61 +++
clang-tools-extra/clang-doc/support/Utils.h | 26
.../unittests/clang-doc/CMakeLists.txt| 12
.../clang-doc/HTMLMustacheGeneratorTest.cpp | 11 +++-
.../unittests/clang-doc/config.h.cmake| 6 ++
7 files changed, 162 insertions(+), 3 deletions(-)
create mode 100644 clang-tools-extra/clang-doc/support/Utils.cpp
create mode 100644 clang-tools-extra/clang-doc/support/Utils.h
create mode 100644 clang-tools-extra/unittests/clang-doc/config.h.cmake
diff --git a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
index 366deb55b77b9..f7e53fc64196a 100644
--- a/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
+++ b/clang-tools-extra/clang-doc/HTMLMustacheGenerator.cpp
@@ -18,6 +18,7 @@
#include "llvm/Support/Error.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/Mustache.h"
+#include "llvm/Support/Path.h"
using namespace llvm;
using namespace llvm::json;
@@ -74,7 +75,51 @@ static std::unique_ptr
NamespaceTemplate = nullptr;
static std::unique_ptr RecordTemplate = nullptr;
+static Error
+setupTemplate(std::unique_ptr &Template,
+ StringRef TemplatePath,
+ std::vector> Partials) {
+ auto T = MustacheTemplateFile::createMustacheFile(TemplatePath);
+ if (Error Err = T.takeError())
+return Err;
+ Template = std::move(T.get());
+ for (const auto [Name, FileName] : Partials) {
+if (auto Err = Template->registerPartialFile(Name, FileName))
+ return Err;
+ }
+ return Error::success();
+}
+
static Error setupTemplateFiles(const clang::doc::ClangDocContext &CDCtx) {
+ // Template files need to use the native path when they're opened,
+ // but have to be used in Posix style when used in HTML.
+ auto ConvertToNative = [](std::string &&Path) -> std::string {
+SmallString<128> PathBuf(Path);
+llvm::sys::path::native(PathBuf);
+return PathBuf.str().str();
+ };
+
+ std::string NamespaceFilePath =
+ ConvertToNative(CDCtx.MustacheTemplates.lookup("namespace-template"));
+ std::string ClassFilePath =
+ ConvertToNative(CDCtx.MustacheTemplates.lookup("class-template"));
+ std::string CommentFilePath =
+ ConvertToNative(CDCtx.MustacheTemplates.lookup("comments-template"));
+ std::string FunctionFilePath =
+ ConvertToNative(CDCtx.MustacheTemplates.lookup("function-template"));
+ std::string EnumFilePath =
+ ConvertToNative(CDCtx.MustacheTemplates.lookup("enum-template"));
+ std::vector> Partials = {
+ {"Comments", CommentFilePath},
+ {"FunctionPartial", FunctionFilePath},
+ {"EnumPartial", EnumFilePath}};
+
+ if (Error Err = setupTemplate(NamespaceTemplate, NamespaceFilePath,
Partials))
+return Err;
+
+ if (Error Err = setupTemplate(RecordTemplate, ClassFilePath, Partials))
+return Err;
+
return Error::success();
}
diff --git a/clang-tools-extra/clang-doc/support/CMakeLists.txt
b/clang-tools-extra/clang-doc/support/CMakeLists.txt
index a4f7993d5c9d8..f470a613b95d9 100644
--- a/clang-tools-extra/clang-doc/support/CMakeLists.txt
+++ b/clang-tools-extra/clang-doc/support/CMakeLists.txt
@@ -6,4 +6,6 @@ set(LLVM_LINK_COMPONENTS
add_clang_library(clangDocSupport STATIC
File.cpp
- )
\ No newline at end of file
+ Utils.cpp
+ )
+
diff --git a/clang-tools-extra/clang-doc/support/Utils.cpp
b/clang-tools-extra/clang-doc/support/Utils.cpp
new file mode 100644
index 0..f1d193379afa6
--- /dev/null
+++ b/clang-tools-extra/clang-doc/support/Utils.cpp
@@ -0,0 +1,61 @@
+//===--===//
+//
+// 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
+//
+//===--===//
+
+#include "Utils.h"
+#include "llvm/ADT/SmallString.h"
+#include "llvm/ADT/StringRef.h"
+#include "llvm/Support/FileSystem.h"
+#include "llvm/Support/Path.h"
+
+using namespace llvm;
+
+SmallString<128> appendPathNative(StringRef Base, StringRef Path) {
+ SmallString<128> Default;
+ sys::path::native(Base, Default);
+ sys::path::append(Default, Path);
+ return Default;
+}
+
+SmallString<128> appendPathPosix
[clang-tools-extra] [clang-doc] Add helpers for Template config (PR #138062)
https://github.com/ilovepi edited https://github.com/llvm/llvm-project/pull/138062 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
https://github.com/el-ev approved this pull request. LGTM https://github.com/llvm/llvm-project/pull/140322 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
bcardosolopes wrote:
Oh, I see, thanks for the extra context!
https://github.com/llvm/llvm-project/pull/140322
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
andykaylor wrote: > LGTM. Is there an easy to add a test for this? I understand sometimes it's > hard, but just double checking. The only thing I can think of would be a unit test that calculates folding set IDs for a bunch of similar, but slightly different, function signatures to verify that we don't incorrectly match them. That's probably worth doing, but I wouldn't want to hold this fix for it. https://github.com/llvm/llvm-project/pull/140322 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libcxx] [Cygwin][MinGW] Internal class in explicitly-instantiation-declarated template should be instantiated (PR #140145)
jeremyd2019 wrote: trying to use this libc++ ci is less than helpful: the jobs keep failing for no apparent reason (cancelled? docker not running?), and whatever seems to be supposed to rerun the failed jobs is not all that reliable about doing so. Hopefully my llvm-mingw fork will show this technique works, and we can be reassured that being in an ifdef that isn't true for any other platform shouldn't break any other platform either. https://github.com/llvm/llvm-project/pull/140145 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] f7ef8dc - [CIR] Fix problem with phantom function arguments (#140322)
Author: Andy Kaylor
Date: 2025-05-16T18:01:12-07:00
New Revision: f7ef8dcbad9d5ac6cf2a8917d5ce03564fdfd0d8
URL:
https://github.com/llvm/llvm-project/commit/f7ef8dcbad9d5ac6cf2a8917d5ce03564fdfd0d8
DIFF:
https://github.com/llvm/llvm-project/commit/f7ef8dcbad9d5ac6cf2a8917d5ce03564fdfd0d8.diff
LOG: [CIR] Fix problem with phantom function arguments (#140322)
There was a problem introduced today where sometimes the CIR for
functions with no arguments would be generated with phantom arguments.
This was causing intermittent test failures.
The problem appears to have been that we were using two different
Profile implementations to generate the FoldingSetNodeID for
CIRGenFunctionInfo so occaissionally when we tried to look for a
pre-existing entry for a function with no arguments it would incorrectly
match a CIRGenFunctionInfo entry that had arguments.
To prevent this from happening again, I rewrote one of the two Profile
functions to call the other.
Added:
Modified:
clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
Removed:
diff --git a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
index 1e06599575fbd..b74460b09a44e 100644
--- a/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
+++ b/clang/lib/CIR/CodeGen/CIRGenFunctionInfo.h
@@ -112,8 +112,16 @@ class CIRGenFunctionInfo final
// NOLINTNEXTLINE(readability-identifier-naming)
void Profile(llvm::FoldingSetNodeID &id) {
-id.AddBoolean(required.getOpaqueData());
-getReturnType().Profile(id);
+// It's unfortunate that we are looping over the arguments twice (here and
+// in the static Profile function we call from here), but if the Profile
+// functions get out of sync, we can end up with incorrect function
+// signatures, and we don't have the argument types in the format that the
+// static Profile function requires.
+llvm::SmallVector argTypes;
+for (const ArgInfo &argInfo : arguments())
+ argTypes.push_back(argInfo.type);
+
+Profile(id, required, getReturnType(), argTypes);
}
llvm::ArrayRef arguments() const {
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [CIR] Fix problem with phantom function arguments (PR #140322)
https://github.com/andykaylor closed https://github.com/llvm/llvm-project/pull/140322 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Driver] Use llvm::is_contained (NFC) (PR #140310)
https://github.com/tgymnich approved this pull request. https://github.com/llvm/llvm-project/pull/140310 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Frontend] Avoid creating a temporary instance of std::string (NFC) (PR #140326)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Kazu Hirata (kazutakahirata)
Changes
Since getLastArgValue returns StringRef, and the constructor of
SmallString accepts StringRef, we do not need to go through a
temporary instance of std::string.
---
Full diff: https://github.com/llvm/llvm-project/pull/140326.diff
1 Files Affected:
- (modified) clang/lib/Frontend/CompilerInvocation.cpp (+1-2)
``diff
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index fd48e425a5c21..3c23073fc6a8c 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2055,8 +2055,7 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
// The memory profile runtime appends the pid to make this name more unique.
const char *MemProfileBasename = "memprof.profraw";
if (Args.hasArg(OPT_fmemory_profile_EQ)) {
-SmallString<128> Path(
-std::string(Args.getLastArgValue(OPT_fmemory_profile_EQ)));
+SmallString<128> Path(Args.getLastArgValue(OPT_fmemory_profile_EQ));
llvm::sys::path::append(Path, MemProfileBasename);
Opts.MemoryProfileOutput = std::string(Path);
} else if (Args.hasArg(OPT_fmemory_profile))
``
https://github.com/llvm/llvm-project/pull/140326
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Frontend] Avoid creating a temporary instance of std::string (NFC) (PR #140326)
https://github.com/kazutakahirata created
https://github.com/llvm/llvm-project/pull/140326
Since getLastArgValue returns StringRef, and the constructor of
SmallString accepts StringRef, we do not need to go through a
temporary instance of std::string.
>From 0f87807ef3159de0350dd85b4f5f8dd289c797f4 Mon Sep 17 00:00:00 2001
From: Kazu Hirata
Date: Fri, 16 May 2025 17:58:06 -0700
Subject: [PATCH] [Frontend] Avoid creating a temporary instance of std::string
(NFC)
Since getLastArgValue returns StringRef, and the constructor of
SmallString accepts StringRef, we do not need to go through a
temporary instance of std::string.
---
clang/lib/Frontend/CompilerInvocation.cpp | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp
b/clang/lib/Frontend/CompilerInvocation.cpp
index fd48e425a5c21..3c23073fc6a8c 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -2055,8 +2055,7 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
// The memory profile runtime appends the pid to make this name more unique.
const char *MemProfileBasename = "memprof.profraw";
if (Args.hasArg(OPT_fmemory_profile_EQ)) {
-SmallString<128> Path(
-std::string(Args.getLastArgValue(OPT_fmemory_profile_EQ)));
+SmallString<128> Path(Args.getLastArgValue(OPT_fmemory_profile_EQ));
llvm::sys::path::append(Path, MemProfileBasename);
Opts.MemoryProfileOutput = std::string(Path);
} else if (Args.hasArg(OPT_fmemory_profile))
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] suggest headers on undeclared errors (#120388) (PR #140247)
@@ -26,14 +26,18 @@ void no_get_1() {
auto [a0, a1] = A(); // expected-error {{decomposes into 3 elements}}
auto [b0, b1] = B(); // expected-error {{decomposes into 3 elements}}
}
- auto [a0, a1, a2] = A(); // expected-error {{undeclared identifier 'get'}}
expected-note {{in implicit initialization of binding declaration 'a0'}}
+ auto [a0, a1, a2] = A(); // expected-error {{undeclared identifier 'get'}} \
+ // expected-note {{perhaps `#include ` is
needed?}} \
shafik wrote:
Yeah, this does not seem helpful, it feels actively harmful actually.
https://github.com/llvm/llvm-project/pull/140247
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [-Wunsafe-buffer-usage] Fix false warnings when const sized array is safely accessed (array [idx %const]) (PR #140113)
malavikasamak wrote: CC @dtarditi https://github.com/llvm/llvm-project/pull/140113 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [Clang] Fix an assertion in the resolution of perfect matches (PR #140073)
cor3ntin wrote: @alexfh Thanks a lot. I have a fix, a PR will be made once I run the tests https://github.com/llvm/llvm-project/pull/140073 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
