https://github.com/davemgreen commented:
This should be left on. People can use -mcpu=neoverse-v2+norng if they don't
have it. Lots of features can be disabled for many different reasons. If Grace
doesn't have it then maybe that should apply just to Grace, and we split the
configs into two?
T
https://github.com/TianYe717 updated
https://github.com/llvm/llvm-project/pull/162005
>From 28819c6782e4d6b351b617460b79e4436cc2644a Mon Sep 17 00:00:00 2001
From: Ye Tian <[email protected]>
Date: Sun, 2 Nov 2025 21:45:55 +0800
Subject: [PATCH 1/4] [Headers][X86] Improve PALIGNR helper: unify
a
https://github.com/ranapratap55 created
https://github.com/llvm/llvm-project/pull/166720
None
>From 039276238f48949ab7254a05dbe851507ebf107c Mon Sep 17 00:00:00 2001
From: N Rana Pratap Reddy
Date: Thu, 6 Nov 2025 12:10:50 +0530
Subject: [PATCH] [AMDGPU] Modifies builtin def to take _Float16 f
https://github.com/zeyi2 updated
https://github.com/llvm/llvm-project/pull/166571
>From 7b24432b9ec88c5a5dd66fcc2179356ffb0a8db7 Mon Sep 17 00:00:00 2001
From: mtx
Date: Wed, 5 Nov 2025 22:28:54 +0800
Subject: [PATCH 1/4] [clang-tidy] Rename `cert-flp30-c` to
`bugprone-avoid-float-loop-counter
@@ -470,10 +490,164 @@ TEST_F(DefineOutlineTest, InCppFile) {
}
}
)cpp"},
+
+ // Criterion 1: Distance
+ {
+ R"cpp(
+struct Foo {
+void ignored1(); // Too far away
+void ignored2(
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
@@ -309,6 +309,9 @@ EligibleRegion getEligiblePoints(llvm::StringRef Code,
llvm::StringRef FullyQualifiedName,
const LangOptions &LangOpts);
+std::string getNamespaceAtPosition(llvm::StringRef Code, const Positi
@@ -548,59 +601,193 @@ class DefineOutline : public Tweak {
return std::move(*Effect);
}
- // Returns the most natural insertion point for \p QualifiedName in \p
- // Contents. This currently cares about only the namespace proximity, but in
- // feature it should also
https://github.com/HighCommander4 requested changes to this pull request.
Thanks for working on this nice improvement, and once again apologies for the
long turnaround time on this review.
https://github.com/llvm/llvm-project/pull/128164
___
cfe-commi
https://github.com/HighCommander4 edited
https://github.com/llvm/llvm-project/pull/128164
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/localspook approved this pull request.
https://github.com/llvm/llvm-project/pull/166571
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -3,8 +3,9 @@
cert-flp30-c
-This check flags ``for`` loops where the induction expression has a
-floating-point type.
+The `cert-flp30-c` check is an aliaes, please see
localspook wrote:
```suggestion
The `cert-flp30-c` check is an alias, pleas
ericxu233 wrote:
> I just meant the AVX512F/VL variants vcvtps2ph_mask / vcvtps2ph256_mask /
> __builtin_ia32_vcvtps2ph512_mask - those should just require additional
> handling for the getNumArgs() == 4 masked variant.
>
> But I'm happy for this just to focus on the F16C cases if you prefer t
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Kazu Hirata (kazutakahirata)
Changes
In C++17, static constexpr members are implicitly inline, so they no
longer require an out-of-line definition.
Identified with readability-redundant-declaration.
---
Full diff: https://github.com/llvm
https://github.com/kazutakahirata created
https://github.com/llvm/llvm-project/pull/166711
In C++17, static constexpr members are implicitly inline, so they no
longer require an out-of-line definition.
Identified with readability-redundant-declaration.
>From 59c55943831637e2da8fc6aafa9b2d5689
https://github.com/Susikrishna updated
https://github.com/llvm/llvm-project/pull/166332
>From 16414b96e0ca5ee44b0332bc387ecbd0046eacd4 Mon Sep 17 00:00:00 2001
From: Susikrishna
Date: Tue, 4 Nov 2025 12:49:38 +0530
Subject: [PATCH 1/6] [Clang] Use ExtWarn for static local variable in extern
in
ChuanqiXu9 wrote:
> Also, can you please compile real-world modules (e.g. libc++) and show how
> this affects the PCM size?
In my simple test, it didn't affect the PCM size. It makes sense in the theory
too. Since on the one hand, the number of MacroID and PreprocessedEntityID is
much less th
@@ -151,14 +151,14 @@ struct UnsafeQualTypeDenseMapInfo {
};
/// An ID number that refers to a macro in an AST file.
-using MacroID = uint32_t;
+using MacroID = uint64_t;
ChuanqiXu9 wrote:
Previously I did this for DeclID and TypeIdx since they are used most.
https://github.com/TianYe717 updated
https://github.com/llvm/llvm-project/pull/162005
>From 28819c6782e4d6b351b617460b79e4436cc2644a Mon Sep 17 00:00:00 2001
From: Ye Tian <[email protected]>
Date: Sun, 2 Nov 2025 21:45:55 +0800
Subject: [PATCH 1/3] [Headers][X86] Improve PALIGNR helper: unify
a
@@ -8956,7 +8948,7 @@ LLVM_DUMP_METHOD void ASTReader::dump() {
llvm::errs() << "*** PCH/ModuleFile Remappings:\n";
dumpModuleIDMap("Global bit offset map", GlobalBitOffsetsMap);
dumpModuleIDMap("Global source location entry map", GlobalSLocEntryMap);
- dumpModuleIDMap("
@@ -9748,36 +9755,40 @@ MacroInfo *ASTReader::getMacro(MacroID ID) {
return nullptr;
}
- ID -= NUM_PREDEF_MACRO_IDS;
- if (!MacrosLoaded[ID]) {
-GlobalMacroMapType::iterator I
- = GlobalMacroMap.find(ID + NUM_PREDEF_MACRO_IDS);
-assert(I != GlobalMacroMap.
@@ -6718,11 +6698,23 @@ SourceRange ASTReader::ReadSkippedRange(unsigned
GlobalIndex) {
return Range;
}
+unsigned
+ASTReader::translatePreprocessedEntityIDToIndex(PreprocessedEntityID ID) const
{
+ unsigned ModuleFileIndex = ID >> 32;
+ assert(ModuleFileIndex && "not tra
@@ -2261,16 +2262,22 @@ MacroInfo *ASTReader::ReadMacroRecord(ModuleFile &F,
uint64_t Offset) {
PreprocessedEntityID
ASTReader::getGlobalPreprocessedEntityID(ModuleFile &M,
- unsigned LocalID) const {
+
https://github.com/ChuanqiXu9 updated
https://github.com/llvm/llvm-project/pull/166346
>From 09f3de8c2f1b3cef99a6834d1df0674b48b95511 Mon Sep 17 00:00:00 2001
From: Chuanqi Xu
Date: Tue, 4 Nov 2025 17:41:04 +0800
Subject: [PATCH] [clang] [Serialization] No transitive change for MacroID and
Pre
rez5427 wrote:
I think register coalescer is pretty much the same thing as gcc's early_remat.
The llvm's register coalescer decide to remat this, because the return register
is been used. And gcc's early_remat decide not to remat this. I put part of the
gcc's log in here:
cast_and_load_1.c.31
https://github.com/zeyi2 updated
https://github.com/llvm/llvm-project/pull/166571
>From 7b24432b9ec88c5a5dd66fcc2179356ffb0a8db7 Mon Sep 17 00:00:00 2001
From: mtx
Date: Wed, 5 Nov 2025 22:28:54 +0800
Subject: [PATCH 1/3] [clang-tidy] Rename `cert-flp30-c` to
`bugprone-avoid-float-loop-counter
https://github.com/cyndyishida closed
https://github.com/llvm/llvm-project/pull/166640
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: Cyndy Ishida
Date: 2025-11-05T20:38:09-08:00
New Revision: e700f157026bf8b4d58f936c5db8f152e269d77f
URL:
https://github.com/llvm/llvm-project/commit/e700f157026bf8b4d58f936c5db8f152e269d77f
DIFF:
https://github.com/llvm/llvm-project/commit/e700f157026bf8b4d58f936c5db8f152e269d77f.diff
https://github.com/darkbuck updated
https://github.com/llvm/llvm-project/pull/165519
>From d1ca8b3a8868e3197685966f7b110332366e3d27 Mon Sep 17 00:00:00 2001
From: Michael Liao
Date: Sat, 18 Oct 2025 19:46:39 -0400
Subject: [PATCH] [CUDA] Add device-side kernel launch support
- CUDA's dynamic p
topperc wrote:
> 2 only happens when users use __attribute__((target(""))), but user
> cannot use a if it doesn't exist at all. As long as we define a new
> fastcc implementation together when defining the new feature, there won't be
> a problem.
It can also happen in LTO if differen
@@ -396,6 +396,118 @@ void loan_from_previous_iteration(MyObj safe, bool
condition) {
} // expected-note {{destroyed here}}
}
+//===--===//
+// Basic Definite Use-After-Return (Return-Stack-Addr
@@ -1202,5 +1202,242 @@ TEST_F(LifetimeAnalysisTest, LivenessOutsideLoop) {
EXPECT_THAT(Origins({"p"}), MaybeLiveAt("p1"));
}
+TEST_F(LifetimeAnalysisTest, SimpleReturnStackAddress) {
+ SetupTest(R"(
+MyObj* target() {
+ MyObj s;
+ MyObj* p = &s;
+ POINT(
@@ -396,6 +396,118 @@ void loan_from_previous_iteration(MyObj safe, bool
condition) {
} // expected-note {{destroyed here}}
}
+//===--===//
+// Basic Definite Use-After-Return (Return-Stack-Addr
https://github.com/usx95 commented:
Took an initial cursory look. Will look again soon.
https://github.com/llvm/llvm-project/pull/165370
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/usx95 edited https://github.com/llvm/llvm-project/pull/165370
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
wangpc-pp wrote:
> I was under the impression glibc doesn't even support this yet. Patches were
> posted but they didn't land as of when I last checked a few months ago.
Oops, I just checked that, you are right. We need to wait for that.
https://github.com/llvm/llvm-project/pull/166540
___
https://github.com/wangpc-pp converted_to_draft
https://github.com/llvm/llvm-project/pull/166540
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
wangpc-pp wrote:
> Reverse ping - I'd like to see this move forward.
I don't know if we have a consensus about the value here, is it OK to set
values to 64 for both RV32 and RV64?
https://github.com/llvm/llvm-project/pull/162986
___
cfe-commits maili
https://github.com/ColinKinloch updated
https://github.com/llvm/llvm-project/pull/161737
>From 98912f07d2b6f3a1d8c79d2e38407dc1d9ba2905 Mon Sep 17 00:00:00 2001
From: Colin Kinloch
Date: Thu, 2 Oct 2025 22:01:40 +0100
Subject: [PATCH 1/7] [clang][Sema] Add fortify warnings for `unistd.h`
Defin
zsrkmyn wrote:
@ellishg @evodius96 All check has passed now. Could you help merge it? many
thanks!
https://github.com/llvm/llvm-project/pull/165358
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Roland McGrath (frobtech)
Changes
No other runtimes can yet be built for the arm-fuchsia target,
but this one can be. There is no OS-specific code in the arm
builtins needed for Fuchsia.
---
Full diff: https://github.com/llvm/llvm-projec
https://github.com/frobtech ready_for_review
https://github.com/llvm/llvm-project/pull/166686
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frobtech created
https://github.com/llvm/llvm-project/pull/166686
No other runtimes can yet be built for the arm-fuchsia target,
but this one can be. There is no OS-specific code in the arm
builtins needed for Fuchsia.
>From b06c985167de498a18f1363c0c49dd55c180d022 Mon Sep
evelez7 wrote:
> do you have an example where its wrong in that case?
It's fixed there. See this link where you can see the indentation for all the
classes in the global namespace.
https://erickvelez.com/clang-doc-mustache-output/pr164314/
https://github.com/llvm/llvm-project/pull/166672
ilovepi wrote:
> See https://erickvelez.com/clang-doc-mustache-output/pr166672/ for display
> diff. The doc entities in the namespace page are no longer indented.
What am I supposed to be looking at that is wrong? Are you saying there is
extra indentation? The page looks OK to me. Or is that a
https://github.com/rapidsna edited
https://github.com/llvm/llvm-project/pull/164737
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/t-rasmud approved this pull request.
LGTM!
https://github.com/llvm/llvm-project/pull/165650
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/a-tarasyuk updated
https://github.com/llvm/llvm-project/pull/165919
>From 2b9ad0f1609fa7926448451750eccfcfa080b520 Mon Sep 17 00:00:00 2001
From: Oleksandr Tarasiuk
Date: Fri, 31 Oct 2025 22:54:53 +0200
Subject: [PATCH 1/3] [Clang] fix false-positive lambda shadow diagnostics
https://github.com/a-tarasyuk updated
https://github.com/llvm/llvm-project/pull/166180
>From b97e99f73cddc4ca5c525dbd9e781679902adaec Mon Sep 17 00:00:00 2001
From: Oleksandr Tarasiuk
Date: Mon, 3 Nov 2025 17:21:38 +0200
Subject: [PATCH 1/2] [Clang] fix confusing diagnostics for lambdas with
i
https://github.com/ilovepi approved this pull request.
Surprised this is dead. I thought the HTML still used it, but if tests are
passing, then its probably OK.
https://github.com/llvm/llvm-project/pull/166595
___
cfe-commits mailing list
cfe-commits@
@@ -272,6 +286,11 @@ GetCountedByAttrOnIncompletePointee(QualType Ty, NamedDecl
**ND) {
if (!PointeeTy->isIncompleteType(ND))
return {};
+ // If counted_by is on void*, it was already validated at declaration time
+ // as a GNU extension. No need to re-check GNU mode
Radovan =?utf-8?q?Božić?= ,
Radovan =?utf-8?q?Božić?= ,
Radovan =?utf-8?q?Božić?=
Message-ID:
In-Reply-To:
philnik777 wrote:
> I believe it is intentionally not defined. It's relying on `CHECK_FILE` which
> is defined as:
>
> ```
> #ifdef IO_DEBUG
> # define CHECK_FILE(FILE, RET) do {
@@ -58,6 +58,52 @@ static RValue emitBuiltinBitOp(CIRGenFunction &cgf, const
CallExpr *e,
return RValue::get(result);
}
+namespace {
+struct WidthAndSignedness {
+ unsigned Width;
+ bool Signed;
+};
+} // namespace
+
+static WidthAndSignedness
+getIntegerWidthAndSignednes
@@ -2503,6 +2503,118 @@ mlir::LogicalResult
CIRToLLVMCmpOpLowering::matchAndRewrite(
return cmpOp.emitError() << "unsupported type for CmpOp: " << type;
}
+mlir::LogicalResult CIRToLLVMBinOpOverflowOpLowering::matchAndRewrite(
+cir::BinOpOverflowOp op, OpAdaptor adaptor
xlauko wrote:
Update to follow test structure of other CIR tests:
- add LLVM check of expected LLVM
- add OGCG checks of expected llvm output of original codegen that should match
through CIR produced LLVM
https://github.com/llvm/llvm-project/pull/166643
_
@@ -58,6 +58,52 @@ static RValue emitBuiltinBitOp(CIRGenFunction &cgf, const
CallExpr *e,
return RValue::get(result);
}
+namespace {
+struct WidthAndSignedness {
+ unsigned Width;
+ bool Signed;
+};
+} // namespace
+
+static WidthAndSignedness
+getIntegerWidthAndSignednes
@@ -2503,6 +2503,118 @@ mlir::LogicalResult
CIRToLLVMCmpOpLowering::matchAndRewrite(
return cmpOp.emitError() << "unsupported type for CmpOp: " << type;
}
+mlir::LogicalResult CIRToLLVMBinOpOverflowOpLowering::matchAndRewrite(
+cir::BinOpOverflowOp op, OpAdaptor adaptor
Zararest wrote:
@lhames could you, please, review these changes or give a hint who I should tag
(My tag is based on a mention in the llvm discussion from the first comment of
this PR)
https://github.com/llvm/llvm-project/pull/166645
___
cfe-commits ma
@@ -382,5 +410,279 @@ void CIRGenFunction::exitCXXTryStmt(const CXXTryStmt &s,
bool isFnTryBlock) {
return;
}
- cgm.errorNYI("exitCXXTryStmt: Required catch");
+ // Emit the structure of the EH dispatch for this catch.
+ emitCatchDispatchBlock(*this, catchScope, tryO
https://github.com/adams381 edited
https://github.com/llvm/llvm-project/pull/166643
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/llvm-beanz approved this pull request.
https://github.com/llvm/llvm-project/pull/166617
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -8281,3 +8281,53 @@ void CodeGenModule::moveLazyEmissionStates(CodeGenModule
*NewBuilder) {
NewBuilder->ABI->MangleCtx = std::move(ABI->MangleCtx);
}
+
+bool CodeGenModule::classNeedsVectorDestructor(const CXXRecordDecl *RD) {
+ if (!Context.getTargetInfo().emitVectorDe
https://github.com/bogner approved this pull request.
https://github.com/llvm/llvm-project/pull/166617
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -8281,3 +8281,53 @@ void CodeGenModule::moveLazyEmissionStates(CodeGenModule
*NewBuilder) {
NewBuilder->ABI->MangleCtx = std::move(ABI->MangleCtx);
}
+
+bool CodeGenModule::classNeedsVectorDestructor(const CXXRecordDecl *RD) {
+ if (!Context.getTargetInfo().emitVectorDe
@@ -150,7 +150,7 @@ class VTableComponent {
bool isRTTIKind() const { return isRTTIKind(getKind()); }
- GlobalDecl getGlobalDecl() const {
+ GlobalDecl getGlobalDecl(bool HasVectorDeletingDtors) const {
efriedma-quic wrote:
Oh, I see. I guess this is ok
@@ -3146,7 +3198,8 @@ bool CXXDestructorDecl::isCalledByDelete(const
FunctionDecl *OpDel) const {
// delete operator, as that destructor is never called, unless the
// destructor is virtual (see [expr.delete]p8.1) because then the
// selected operator depends on the dyna
@@ -1206,6 +1206,9 @@ void CodeGenFunction::EmitNewArrayInitializer(
EmitCXXAggrConstructorCall(Ctor, NumElements, CurPtr, CCE,
/*NewPointerIsChecked*/true,
CCE->requiresZeroInitialization());
+if (CGM.getCon
@@ -370,6 +370,18 @@ class ASTContext : public RefCountedBase {
mutable llvm::DenseSet DestroyingOperatorDeletes;
mutable llvm::DenseSet TypeAwareOperatorNewAndDeletes;
+ /// Global and array operators delete are only required for MSVC deleting
+ /// destructors support.
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Ivan Shumakov (Zararest)
Changes
There is an issue with using PBQP register allocator in clang that doesn't have
AArch64 target:
https://groups.google.com/g/llvm-dev/c/ZIvatINBENo
This change links PBQP regalloc to all LLVM targets.
---
github-actions[bot] wrote:
Thank you for submitting a Pull Request (PR) to the LLVM Project!
This PR will be automatically labeled and the relevant teams will be notified.
If you wish to, you can add reviewers by using the "Reviewers" section on this
page.
If this is not working for you, it
https://github.com/Zararest created
https://github.com/llvm/llvm-project/pull/166645
There is an issue with using PBQP register allocator in clang that doesn't have
AArch64 target:
https://groups.google.com/g/llvm-dev/c/ZIvatINBENo
This change links PBQP regalloc to all LLVM targets.
>From ae
https://github.com/jansvoboda11 approved this pull request.
https://github.com/llvm/llvm-project/pull/166640
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
llvmbot wrote:
@llvm/pr-subscribers-clang
@llvm/pr-subscribers-clangir
Author: None (adams381)
Changes
This implements the builtins that handle overflow.
---
Patch is 37.73 KiB, truncated to 20.00 KiB below, full version:
https://github.com/llvm/llvm-project/pull/166643.diff
5 Files A
github-actions[bot] wrote:
Thank you for submitting a Pull Request (PR) to the LLVM Project!
This PR will be automatically labeled and the relevant teams will be notified.
If you wish to, you can add reviewers by using the "Reviewers" section on this
page.
If this is not working for you, it
https://github.com/adams381 created
https://github.com/llvm/llvm-project/pull/166643
This implements the builtins that handle overflow.
>From 7abb2167def4dfceeb96c4b231a94fd8308c6e08 Mon Sep 17 00:00:00 2001
From: Adam Smith
Date: Wed, 5 Nov 2025 13:04:40 -0800
Subject: [PATCH] [CIR] Upstream
@@ -382,5 +410,279 @@ void CIRGenFunction::exitCXXTryStmt(const CXXTryStmt &s,
bool isFnTryBlock) {
return;
}
- cgm.errorNYI("exitCXXTryStmt: Required catch");
+ // Emit the structure of the EH dispatch for this catch.
+ emitCatchDispatchBlock(*this, catchScope, tryO
@@ -382,5 +410,279 @@ void CIRGenFunction::exitCXXTryStmt(const CXXTryStmt &s,
bool isFnTryBlock) {
return;
}
- cgm.errorNYI("exitCXXTryStmt: Required catch");
+ // Emit the structure of the EH dispatch for this catch.
+ emitCatchDispatchBlock(*this, catchScope, tryO
https://github.com/benlangmuir approved this pull request.
https://github.com/llvm/llvm-project/pull/164345
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,110 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
arsenm wrote:
This is failing on the broken v-to-s copy for the asm input. I think
@@ -0,0 +1,85 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
+
+define i32 @bcnt032_not_for_vregs(i32 %val0) {
+; CHECK-LABEL: bcnt032_not_for_vregs:
+; CHECK:
@@ -264,7 +264,9 @@ def S_BREV_B64 : SOP1_64 <"s_brev_b64",
} // End isReMaterializable = 1, isAsCheapAsAMove = 1
let Defs = [SCC] in {
-def S_BCNT0_I32_B32 : SOP1_32 <"s_bcnt0_i32_b32">;
+def S_BCNT0_I32_B32 : SOP1_32 <"s_bcnt0_i32_b32",
+ [(set i32:$sdst, (UniformBinFrag 32
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Cyndy Ishida (cyndyishida)
Changes
Before this patch, compilations using modules &
`-mspeculative-load-hardening` failed because it was not consistent whether
this option impacts module compatibility.
This repairs it by always treating it
https://github.com/cyndyishida created
https://github.com/llvm/llvm-project/pull/166640
Before this patch, compilations using modules &
`-mspeculative-load-hardening` failed because it was not consistent whether
this option impacts module compatibility.
This repairs it by always treating it as
https://github.com/SergejSalnikov edited
https://github.com/llvm/llvm-project/pull/166631
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SergejSalnikov edited
https://github.com/llvm/llvm-project/pull/166631
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
llvm-ci wrote:
LLVM Buildbot has detected a new failure on builder
`clang-ppc64le-linux-test-suite` running on `ppc64le-clang-test-suite` while
building `clang,lldb` at step 6 "test-build-unified-tree-check-all".
Full details are available at:
https://lab.llvm.org/buildbot/#/builders/95/build
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Sergej Salnikov (SergejSalnikov)
Changes
---
Full diff: https://github.com/llvm/llvm-project/pull/166631.diff
1 Files Affected:
- (modified) clang/lib/Frontend/PrintPreprocessedOutput.cpp (+1-1)
``diff
diff --git a/clang/lib
https://github.com/SergejSalnikov created
https://github.com/llvm/llvm-project/pull/166631
None
>From a53cdabdd7fcf83c2377cf78f2848c1f795c7b2c Mon Sep 17 00:00:00 2001
From: Sergej Salnikov
Date: Wed, 5 Nov 2025 21:25:04 +0100
Subject: [PATCH] Correctly handle negative line changes
---
clang
linuxrocks123 wrote:
@arsenm okay, comments dealt with. If you'd like to preserve the global isel
error, I should revert or partially revert the last commit. Otherwise, LGTM?
https://github.com/llvm/llvm-project/pull/164847
___
cfe-commits mailing l
https://github.com/linuxrocks123 updated
https://github.com/llvm/llvm-project/pull/164847
>From 96825875e1c8bad9a9b3b28fe943a5469f94c575 Mon Sep 17 00:00:00 2001
From: Patrick Simmons
Date: Thu, 23 Oct 2025 11:50:32 -0500
Subject: [PATCH 01/13] Initial work
---
clang/include/clang/Basic/Built
@@ -0,0 +1,96 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
+
+define amdgpu_ps i32 @bcnt032_not_for_vregs(i64 %val) {
+; CHECK-LABEL: bcnt032_not_for_vregs:
+;
@@ -0,0 +1,96 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
+
+define amdgpu_ps i32 @bcnt032_not_for_vregs(i64 %val) {
linuxrocks123 wrote:
Don
@@ -0,0 +1,110 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
linuxrocks123 wrote:
@arsenm
```LLVM ERROR: unable to map instruction: %8:sreg_32
@@ -0,0 +1,96 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn -mcpu=gfx900 < %s | FileCheck %s
+
+define amdgpu_ps i32 @bcnt032_not_for_vregs(i64 %val) {
linuxrocks123 wrote:
Don
https://github.com/ahmednoursphinx updated
https://github.com/llvm/llvm-project/pull/166178
>From 3cdf74e3eee0e206a3588acb1b4e53fd66eb9517 Mon Sep 17 00:00:00 2001
From: ahmed
Date: Mon, 3 Nov 2025 17:16:29 +0200
Subject: [PATCH 01/10] fix: rematerialize smaller predicate masks
---
llvm/lib/T
https://github.com/ahmednoursphinx updated
https://github.com/llvm/llvm-project/pull/166178
>From 3cdf74e3eee0e206a3588acb1b4e53fd66eb9517 Mon Sep 17 00:00:00 2001
From: ahmed
Date: Mon, 3 Nov 2025 17:16:29 +0200
Subject: [PATCH 1/9] fix: rematerialize smaller predicate masks
---
llvm/lib/Tar
1 - 100 of 396 matches
Mail list logo
