@@ -5354,6 +5368,22 @@ AMDGPURegisterBankInfo::getInstrMapping(const
MachineInstr &MI) const {
}
case Intrinsic::amdgcn_pops_exiting_wave_id:
return getDefaultMappingSOP(MI);
+case Intrinsic::amdgcn_tensor_load_to_lds_d2:
+case Intrinsic::amdgcn_tensor_st
seanm wrote:
Another interesting case:
```c++
static int testToWindowsExtendedPath()
{
#ifdef _WIN32
// lots of real, non constant work
return ret;
#else
return 0;
#endif
}
```
For some platforms this can be constexpr; for other platforms (Windows), it
cannot. It was transformed to `cons
rjmccall wrote:
Right, but the code still contains some kind of check to decide which version
of the library function to call. Maybe it's implicit by multiversioned
functions or something, but there's *something* in the source code, because if
the user just writes a kernel that does nothing bu
https://github.com/xgupta edited
https://github.com/llvm/llvm-project/pull/146772
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/efriedma-quic edited
https://github.com/llvm/llvm-project/pull/144970
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/efriedma-quic edited
https://github.com/llvm/llvm-project/pull/144970
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/yronglin updated
https://github.com/llvm/llvm-project/pull/146224
>From 43e2dc670d7c9ed5e23b5d26dff1e273c84b5a53 Mon Sep 17 00:00:00 2001
From: yronglin
Date: Thu, 3 Jul 2025 02:15:22 +0800
Subject: [PATCH] [C23][Parser] Diagnostic for attribute declaration where
statement i
Lukas =?utf-8?q?Döllerer?= ,
Lukas =?utf-8?q?Döllerer?= ,
Lukas =?utf-8?q?Döllerer?=
Message-ID:
In-Reply-To:
kripken wrote:
@Lukasdoe You may be interested in this Binaryen PR which adds passes for
branch hint debugging:
https://github.com/WebAssembly/binaryen/pull/7695
For example you cou
https://github.com/jkorous-apple edited
https://github.com/llvm/llvm-project/pull/145784
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/Sirraide commented:
One more thing I thought of: What about C++ edge cases, e.g. both Clang and GCC
accept this... which doesn’t quite seem right to me
(https://godbolt.org/z/K58eEvG9P):
```c++
typeof(int){} x; // Probably parsed as typeof(int{})
```
https://github.com/llvm/
@@ -1000,8 +1001,10 @@ ExprResult Parser::ParseCastExpression(CastParseKind
ParseKind,
Token Replacement;
CastExpressionIdValidator Validator(
/*Next=*/Tok,
-/*AllowTypes=*/isTypeCast != TypeCastState::NotTypeCast,
-/*AllowNonTypes=*/isTypeCast
@@ -3580,6 +3580,37 @@ def int_amdgcn_fdiv_fast : DefaultAttrsIntrinsic<
[IntrNoMem, IntrSpeculatable]
>;
+class AMDGPUTensorLoadStore:
+ Intrinsic<
+[],
+[llvm_v4i32_ty, // D# group 0
+ llvm_v8i32_ty, // D# group 1
+ llvm_v4i32_ty, // D# group 2
+ llvm_
https://github.com/AmrDeveloper updated
https://github.com/llvm/llvm-project/pull/146755
>From ba1109f6c42bc902d9dc1f0613ce0b4959e8233e Mon Sep 17 00:00:00 2001
From: AmrDeveloper
Date: Wed, 2 Jul 2025 20:26:39 +0200
Subject: [PATCH] [CIR] Upstream SubstNonTypeTemplateParmExpr support for
Comp
@@ -621,6 +621,32 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned
BuiltinID,
llvm::Function *F = CGM.getIntrinsic(IID, {LoadTy});
return Builder.CreateCall(F, {Addr});
}
+ case AMDGPU::BI__builtin_amdgcn_tensor_load_to_lds:
+ case AMDGPU::BI__builtin_amdg
https://github.com/AaronBallman updated
https://github.com/llvm/llvm-project/pull/146394
>From 94cd71d65fe27cdde0c39416a0e2e709af98ed0c Mon Sep 17 00:00:00 2001
From: Aaron Ballman
Date: Mon, 30 Jun 2025 13:26:57 -0400
Subject: [PATCH 1/3] [C23] Fix typeof handling in enum declarations
We have
sivadeilra wrote:
@arsenm, @efriedma-quic, @namazso , do you have more feedback on this PR? I'd
like to get this merged and move on to the next pieces for supporting Windows
hot-patching.
https://github.com/llvm/llvm-project/pull/144745
___
cfe-comm
@@ -4210,6 +4215,7 @@ class Parser : public CodeCompletionHandler {
/// \endverbatim
ExprResult ParseParenExpression(ParenParseOption &ExprType,
bool stopIfCastExpr, bool isTypeCast,
+ bool ParenKnownToBeNon
@@ -2850,7 +2870,8 @@ Parser::ParseParenExpression(ParenParseOption &ExprType,
bool stopIfCastExpr,
isFoldOperator(NextToken().getKind())) {
ExprType = ParenParseOption::FoldExpr;
return ParseFoldExpression(ExprResult(), T);
- } else if (isTypeCast) {
+
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 h,cpp,c --
clang/include/clang/Parse/Parser.h clang/lib/Pars
seanm wrote:
> > This actually compiles, though with a `-Wduplicate-decl-specifier` warning.
>
> I failed to reproduce this...
I don't have a minimal repro case, I was building
[VTK](https://gitlab.kitware.com/vtk/vtk). It's pretty simple to build it
yourself if you're familiar with CMake. I
@@ -292,6 +292,21 @@ class ClangTidyDiagnosticConsumer : public
DiagnosticConsumer {
void HandleDiagnostic(DiagnosticsEngine::Level DiagLevel,
const Diagnostic &Info) override;
+ void BeginSourceFile(const LangOptions &LangOpts,
vbv
@@ -5354,6 +5368,22 @@ AMDGPURegisterBankInfo::getInstrMapping(const
MachineInstr &MI) const {
}
case Intrinsic::amdgcn_pops_exiting_wave_id:
return getDefaultMappingSOP(MI);
+case Intrinsic::amdgcn_tensor_load_to_lds_d2:
+case Intrinsic::amdgcn_tensor_st
https://github.com/tgymnich edited
https://github.com/llvm/llvm-project/pull/146636
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
HazardyKnusperkeks wrote:
> > While it is less code, I find a bit harder to understand and the code gen
> > is far worse: https://gcc.godbolt.org/z/KzG4YnTh3
>
> `IsBlank` is misleading because of `std::isblank` whereas `Blanks.contains`
> is not, so the latter has better readability for me. A
@@ -2332,6 +2332,12 @@ void CodeGenFunction::EmitOMPPrivateLoopCounters(
for (const Expr *E : S.counters()) {
const auto *VD = cast(cast(E)->getDecl());
const auto *PrivateVD = cast(cast(*I)->getDecl());
+// Privatize original counter variable (e.g., __beginN, __e
@@ -0,0 +1,419 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
UTC_ARGS: --clang-args ['-fopenmp', '-std=c++20'] --function-signature
--include-generated-funcs --replace-value-regex
"__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]
lenary wrote:
What is your intention here around the LLVM 21 release cycle, knowing this is a
big feature and we are 1 week from the branch date
https://github.com/llvm/llvm-project/pull/146534
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
h
AlexVlx wrote:
> So your users today are building for generic AMDGPU but using builtins that
> are only available on a specific processor release? Presumably that code is
> protected _somehow_ and their programs are not simply crashing at runtime. Is
> that something you'd be able to leverage
https://github.com/efriedma-quic updated
https://github.com/llvm/llvm-project/pull/144970
>From b1f7402423af22b0ea8cdf0b3d3734bcce68a636 Mon Sep 17 00:00:00 2001
From: Eli Friedman
Date: Thu, 19 Jun 2025 18:29:49 -0700
Subject: [PATCH 1/2] [clang] Consistently handle consteval constructors for
https://github.com/jkorous-apple approved this pull request.
I want to check that my mental model of InSourceFile is correct but otherwise
LGTM.
https://github.com/llvm/llvm-project/pull/145784
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
h
@@ -292,6 +292,21 @@ class ClangTidyDiagnosticConsumer : public
DiagnosticConsumer {
void HandleDiagnostic(DiagnosticsEngine::Level DiagLevel,
const Diagnostic &Info) override;
+ void BeginSourceFile(const LangOptions &LangOpts,
+
https://github.com/tJener created
https://github.com/llvm/llvm-project/pull/146761
Before this commit, when `LineJoiner` joins a line with affected leading
whitespace, it would drop the knowledge of this entirely. However, when the
`AffectedRangeManager` is computing the affected lines, the le
https://github.com/Sirraide edited
https://github.com/llvm/llvm-project/pull/146394
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Sirraide wrote:
> > ```c++
> > typeof(int){} x; // Probably parsed as typeof(int{})
> > ```
>
> Actually, I’m not sure what we think this is supposed to be; I haven’t
> checked.
Actually, that’s just a compound literal; I forgot we supported those in C++.
https://github.com/llvm/llvm-project/
https://github.com/dbartol updated
https://github.com/llvm/llvm-project/pull/145784
>From 0be65986e1e2adf973a032936afa9cf48841055b Mon Sep 17 00:00:00 2001
From: Dave Bartolomeo
Date: Wed, 25 Jun 2025 17:45:50 -0400
Subject: [PATCH 1/3] EndSourceFile() for preprocessor before diagnostic client
https://github.com/Bigcheese approved this pull request.
lgtm
https://github.com/llvm/llvm-project/pull/146766
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: Martin Storsjö
Date: 2025-07-02T23:58:22+03:00
New Revision: 551d6ddaa3810749ecae33f65759870b78b9a86a
URL:
https://github.com/llvm/llvm-project/commit/551d6ddaa3810749ecae33f65759870b78b9a86a
DIFF:
https://github.com/llvm/llvm-project/commit/551d6ddaa3810749ecae33f65759870b78b9a86a.diff
mstorsjo wrote:
I pushed a fix to add `REQUIRES: x86-registered-target` in
551d6ddaa3810749ecae33f65759870b78b9a86a.
https://github.com/llvm/llvm-project/pull/146643
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bi
efriedma-quic wrote:
> we only know the concrete target when we are finalising, which happens at a
> completely different time-point, on possibly a different machine;
This is precisely why we want the frontend diagnostic: if we diagnose the bad
cases in the frontend, later stages are guarantee
@@ -9125,9 +9126,25 @@ bool
LValueExprEvaluator::VisitCompoundLiteralExpr(const CompoundLiteralExpr *E) {
assert((!Info.getLangOpts().CPlusPlus || E->isFileScope()) &&
"lvalue compound literal in c++?");
- // Defer visiting the literal until the lvalue-to-rvalue con
https://github.com/efriedma-quic edited
https://github.com/llvm/llvm-project/pull/137163
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/efriedma-quic approved this pull request.
LGTM
https://github.com/llvm/llvm-project/pull/137163
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/cristianassaiante updated
https://github.com/llvm/llvm-project/pull/145059
>From ab6063493744ef5a1ee92fd249bf8d86b7299fad Mon Sep 17 00:00:00 2001
From: Cristian Assaiante
Date: Fri, 20 Jun 2025 16:56:23 +0200
Subject: [PATCH 1/7] Adding -opt-disable and a test for it
---
c
401 - 443 of 443 matches
Mail list logo
