jansvoboda11 wrote:
The `unique_ptr` in `ASTUnit` always stores a **copy** of the `shared_ptr` in
`CompilerInvocation`.
https://github.com/llvm/llvm-project/pull/137675
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi
@@ -59,19 +60,46 @@ class InProcessModuleCache : public ModuleCache {
InMemoryModuleCache InMemory;
public:
- InProcessModuleCache(ModuleCacheMutexes &Mutexes) : Mutexes(Mutexes) {}
+ InProcessModuleCache(ModuleCacheEntries &Entries) : Entries(Entries) {}
void prepare
https://github.com/jansvoboda11 commented:
The diagnostic is confusing - it tells the user they used an invalid
combination of flags, but shows flags they did not use at all. I think we can
improve on this and report when a flag is deduced/assumed/defaulted.
https://github.com/llvm/llvm-projec
@@ -861,9 +864,13 @@ class CompilerInstance : public ModuleLoader {
///
/// Explicitly-specified \c VFS takes precedence over the VFS of this
instance
/// when creating the clone and also prevents \c FileManager sharing.
- std::unique_ptr cloneForModuleCompile(
- S
jansvoboda11 wrote:
I think so. Even if this gets diagnosed more clearly by the driver, users can
bypass that using `-Xclang`. Even for me as a Clang developer this message
would be confusing, especially in more complex scenarios (e.g. with response
files involved).
https://github.com/llvm/ll
jansvoboda11 wrote:
I'm going to land this later this week unless tagged people have objections.
https://github.com/llvm/llvm-project/pull/132984
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cf
jansvoboda11 wrote:
Ping.
https://github.com/llvm/llvm-project/pull/135737
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/135737
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/136601
The `DiagnosticConsumer` interface is not thread-safe. To enable thread-safety
of `CompilerInstance` objects cloned from the same parent, this PR allows
passing an explicit `DiagnosticConsumer` to `cloneFo
@@ -3349,9 +3351,14 @@ ASTReader::ReadControlBlock(ModuleFile &F,
.getModuleCache()
.getInMemoryModuleCache()
.isPCMFinal(F.FileName);
- if (isDiagnosedResult(
@@ -3349,9 +3351,14 @@ ASTReader::ReadControlBlock(ModuleFile &F,
.getModuleCache()
.getInMemoryModuleCache()
.isPCMFinal(F.FileName);
- if (isDiagnosedResult(
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/128959
>From cb389c006e5ca28d5baef73625776a3e0e58b8af Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Wed, 26 Feb 2025 15:12:43 -0800
Subject: [PATCH 1/2] [clang][deps] Propagate the whole service
Adding any kin
https://github.com/jansvoboda11 edited
https://github.com/llvm/llvm-project/pull/128959
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/130395
Currently, `LockFileManager` assumes the owner of the lock file creates an
output file. This is problematic for at least three reasons:
1. It is orthogonal to the main purpose of this class - mutual exclus
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/129751
>From 5f119a3a4ae8642d6ab0498c1cf6b39d5099c835 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Thu, 27 Feb 2025 15:23:18 -0800
Subject: [PATCH 1/5] [clang][modules][deps] Add mutex as an alternative to
fi
@@ -0,0 +1,66 @@
+//===--===//
+//
+// 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: Apac
@@ -95,6 +96,10 @@ class DependencyScanningService {
return SharedCache;
}
+ ModuleCacheMutexes &getSharedModuleCacheMutexes() {
jansvoboda11 wrote:
We don't _need_ it, but we currently call
`std::make_shared` for each TU, which is unnecessary. We
co
@@ -3394,6 +3394,10 @@ def fno_inline_functions : Flag<["-"],
"fno-inline-functions">, Group;
def fno_inline : Flag<["-"], "fno-inline">, Group,
Visibility<[ClangOption, CC1Option]>;
+def fpartial_inlining : Flag<["-"], "fpartial-inlining">, Group,
jansvoboda
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130627
>From 88b5bbd1303480dc40db3ea76d0f831a69ff1957 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Sat, 8 Mar 2025 21:51:51 -0800
Subject: [PATCH 1/5] [Support] Return `LockFileManager` errors right away
---
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/130395
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
jansvoboda11 wrote:
I don't know what the desired properties of this feature are exactly, but I'm
concerned that this does not guarantee the resulting command line only refers
to inputs in the sysroot. Imagine that we run this without the header search
and VFS pruning optimizations - the comma
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130627
>From 88b5bbd1303480dc40db3ea76d0f831a69ff1957 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Sat, 8 Mar 2025 21:51:51 -0800
Subject: [PATCH 1/2] [Support] Return `LockFileManager` errors right away
---
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/130627
This patch removes some internal state out of `LockFileManager` by moving the
locking code from the constructor into new member function `tryLock()` which
returns the errors right away. This simplifies and
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/106274
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130395
>From a9969c43be752ccaa70959fd4527a57dc2981881 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Thu, 6 Mar 2025 17:29:34 -0800
Subject: [PATCH 1/3] [Support] Remove output file checks from
`LockFileManager
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/128959
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/130627
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/130823
I would like to ensure that `CompilerInvocation` and its constituents are not
mutated late during the compilation. This PR moves adjustment of the
`HeaderSearchOptions::ModulesSkip{DiagnosticOptions,Header
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/130825
This PR makes the `HeaderSearchOptions` object referenced by `HeaderSearch`
constant. Depends on #130823.
>From a38a64dd092ba4d476d91f91178af8fc6d96606b Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Th
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130627
>From 88b5bbd1303480dc40db3ea76d0f831a69ff1957 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Sat, 8 Mar 2025 21:51:51 -0800
Subject: [PATCH 1/6] [Support] Return `LockFileManager` errors right away
---
@@ -158,41 +160,40 @@ class RemoveUniqueLockFileOnSignal {
} // end anonymous namespace
LockFileManager::LockFileManager(StringRef FileName)
-{
- this->FileName = FileName;
- if (std::error_code EC = sys::fs::make_absolute(this->FileName)) {
-std::string S("failed to obt
Author: Jan Svoboda
Date: 2025-03-11T13:06:10-07:00
New Revision: 2f403ee4871465aa7fbb5e540bdc791466704f24
URL:
https://github.com/llvm/llvm-project/commit/2f403ee4871465aa7fbb5e540bdc791466704f24
DIFF:
https://github.com/llvm/llvm-project/commit/2f403ee4871465aa7fbb5e540bdc791466704f24.diff
L
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/130834
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/131193
This PR adds new `ModuleCache` interface to Clang's implicitly-built modules
machinery. The main motivation for this change is to create a second
implementation that uses a more efficient kind of `llvm::Ad
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/130989
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130989
>From cd0b9a8b878c4ecd929552c836f3d4665e926ef4 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Wed, 12 Mar 2025 09:37:27 -0700
Subject: [PATCH 1/2] [Support] Introduce new `AdvisoryLock` interface
This PR
https://github.com/jansvoboda11 edited
https://github.com/llvm/llvm-project/pull/129751
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/129751
>From 5f119a3a4ae8642d6ab0498c1cf6b39d5099c835 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Thu, 27 Feb 2025 15:23:18 -0800
Subject: [PATCH 1/3] [clang][modules][deps] Add mutex as an alternative to
fi
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/130834
>From 834eac8d3ce286172456e2a177d0e5284ade5b46 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Tue, 11 Mar 2025 09:17:50 -0700
Subject: [PATCH 1/2] [Support] Do not remove lock file on failure
---
clang/
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/106271
>From e4c1f9254257c92789a924c70b06399c0ea794bd Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Tue, 27 Aug 2024 09:34:01 -0700
Subject: [PATCH 1/6] [clang] `TargetInfo` does not own `TargetOptions`
---
c
jansvoboda11 wrote:
This makes sense in principle, but I'm wondering whether
https://github.com/llvm/llvm-project/pull/100681 is the better approach. Seems
like having the client call `finish()` is simpler than doing that
"automatically" in `CompilerInstance::ExecuteAction()` and then trying t
https://github.com/jansvoboda11 approved this pull request.
https://github.com/llvm/llvm-project/pull/127110
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,164 @@
+//===- ModuleMapFile.h - Parsing and representation -*- C++
-*-===//
+//
+// 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: Ap
@@ -0,0 +1,164 @@
+//===- ModuleMapFile.h - Parsing and representation -*- C++
-*-===//
+//
+// 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: Ap
@@ -0,0 +1,164 @@
+//===- ModuleMapFile.h - Parsing and representation -*- C++
-*-===//
+//
+// 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: Ap
@@ -3157,25 +2140,18 @@ bool ModuleMap::parseModuleMapFile(FileEntryRef File,
bool IsSystem,
assert((!Offset || *Offset <= Buffer->getBufferSize()) &&
"invalid buffer offset");
- // Parse this module map file.
- Lexer L(SourceMgr.getLocForStartOfFile(ID), MMapLan
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/128959
Shared state between dependency scanning workers is managed by the dependency
scanning service.
Right now, the members are individually threaded through the worker, action,
and collector. This makes any c
https://github.com/jansvoboda11 approved this pull request.
LGTM, but let's wait for the other reviewers.
https://github.com/llvm/llvm-project/pull/119740
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/li
https://github.com/jansvoboda11 approved this pull request.
LGTM, thanks!
https://github.com/llvm/llvm-project/pull/128446
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -129,6 +129,26 @@ static void optimizeDiagnosticOpts(DiagnosticOptions &Opts,
Opts.Remarks.clear();
}
+static void optimizeCWD(FileSystemOptions &FSOpts, CodeGenOptions &CGOpts,
+const std::string &CWD) {
jansvoboda11 wrote:
In LL
@@ -492,8 +492,23 @@ static std::string getModuleContextHash(const ModuleDeps
&MD,
auto &FSOpts = const_cast(CI.getFileSystemOpts());
if (CWD && !IgnoreCWD)
HashBuilder.add(*CWD);
- else
+ else {
FSOpts.WorkingDir.clear();
+auto &CGOpts = const_cast(CI.getCo
https://github.com/jansvoboda11 requested changes to this pull request.
Leaving the correctness of the change up for other reviewers. My only gripe is
the `const_cast`.
https://github.com/llvm/llvm-project/pull/128446
___
cfe-commits mailing list
cfe-
@@ -492,8 +492,23 @@ static std::string getModuleContextHash(const ModuleDeps
&MD,
auto &FSOpts = const_cast(CI.getFileSystemOpts());
jansvoboda11 wrote:
Please create a follow-up patch to #124786 to remove the `const_cast` of
`FileSystemOptions` here.
http
@@ -129,6 +129,26 @@ static void optimizeDiagnosticOpts(DiagnosticOptions &Opts,
Opts.Remarks.clear();
}
+static void optimizeCWD(FileSystemOptions &FSOpts, CodeGenOptions &CGOpts,
+const std::string &CWD) {
+ FSOpts.WorkingDir.clear();
+ if (CGOpts
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/134404
This is crucial when recovering from fatal loader errors. Without it, the
`Lexer` keeps yielding more tokens and the compiler may access invalid
`ASTReader` state.
rdar://133388373
>From 62a5616b139c5bfc
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/133467
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -569,7 +569,7 @@
CrossTranslationUnitContext::ASTLoader::loadFromDump(StringRef ASTDumpPath) {
return ASTUnit::LoadFromASTFile(
ASTDumpPath, CI.getPCHContainerOperations()->getRawReader(),
ASTUnit::LoadEverything, Diags, CI.getFileSystemOpts(),
- CI.getHe
@@ -48,7 +48,7 @@ void ASTMergeAction::ExecuteAction() {
/*ShouldOwnClient=*/true));
std::unique_ptr Unit = ASTUnit::LoadFromASTFile(
ASTFiles[I], CI.getPCHContainerReader(), ASTUnit::LoadEverything,
Diags,
-CI.getFileSys
Author: Jan Svoboda
Date: 2025-04-04T10:20:44-07:00
New Revision: ea0869ccb9a80e235f6f47eaa47fb1d5888d660a
URL:
https://github.com/llvm/llvm-project/commit/ea0869ccb9a80e235f6f47eaa47fb1d5888d660a
DIFF:
https://github.com/llvm/llvm-project/commit/ea0869ccb9a80e235f6f47eaa47fb1d5888d660a.diff
L
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/134404
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/132984
>From 46de2dd2ceb82510f1e6719a34ad67164943983d Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Fri, 28 Mar 2025 09:02:12 -0700
Subject: [PATCH] [clang] Do not share ownership of `HeaderSearchOptions`
---
Author: Jan Svoboda
Date: 2025-03-20T09:37:10-07:00
New Revision: 15ad84fffd76e845ce48f7cb8a917eb6532b93c9
URL:
https://github.com/llvm/llvm-project/commit/15ad84fffd76e845ce48f7cb8a917eb6532b93c9
DIFF:
https://github.com/llvm/llvm-project/commit/15ad84fffd76e845ce48f7cb8a917eb6532b93c9.diff
L
Author: Jan Svoboda
Date: 2025-03-28T09:34:19-07:00
New Revision: 277ab85d1ccf80750f5193495c0665808c2863de
URL:
https://github.com/llvm/llvm-project/commit/277ab85d1ccf80750f5193495c0665808c2863de
DIFF:
https://github.com/llvm/llvm-project/commit/277ab85d1ccf80750f5193495c0665808c2863de.diff
L
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/134284
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/134284
None
>From c333f2bd8dc4ae9497f7def29d3f3989f309a991 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Thu, 3 Apr 2025 10:37:34 -0700
Subject: [PATCH] [clang][deps] Avoid unchecked error assertion
---
clan
jansvoboda11 wrote:
> > multi-project workspace builds cannot avoid hitting this kind of issue
> > (demonstrated by the test) and they are left with a build error they cannot
> > fix.
>
> Is this because a multi-project workspace results in separate build-products
> directories but a single m
https://github.com/jansvoboda11 edited
https://github.com/llvm/llvm-project/pull/138920
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/138920
>From 306a9ceab9aa77ccb2752e847339c5a67e672549 Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Wed, 7 May 2025 10:10:47 -0700
Subject: [PATCH 1/2] [clang][modules][deps] Rebuild implicit modules when
thei
@@ -0,0 +1,77 @@
+// Test that modifications to a common header (imported from both a PCH and a
TU)
+// cause rebuilds of dependent modules imported from the TU on incremental
build.
+
+// RUN: rm -rf %t
+// RUN: split-file %s %t
+
+//--- module.modulemap
+module mod_common { he
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/137363
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
jansvoboda11 wrote:
(Just adding a note here for my records that this speeds up from-scratch scans
by 1.4% and incremental _null_ scans by 67% on my benchmark.)
https://github.com/llvm/llvm-project/pull/137363
___
cfe-commits mailing list
cfe-commits@
https://github.com/jansvoboda11 approved this pull request.
https://github.com/llvm/llvm-project/pull/138955
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/138983
Timestamps are an implementation detail of the cross-process module cache
implementation. This PR hide it from the `ModuleCache` API, which simplifies
the in-process implementation.
>From 8d594dd6ffdf2601
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/139091
Force-validation of user headers was implemented in acb803e8 to deal with files
changing during build. The dependency scanner guarantees an immutable file
system during single build session, so the validat
@@ -0,0 +1,178 @@
+// Test that modifications to a common header (imported from both a PCH and a
TU)
+// cause rebuilds of dependent modules imported from the TU on incremental
build.
+
+// RUN: rm -rf %t
+// RUN: split-file %s %t
+
+//--- module.modulemap
+module mod_common { h
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/139091
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/133827
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/138920
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 approved this pull request.
https://github.com/llvm/llvm-project/pull/139410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/139584
The `DiagnosticOptions` class is currently intrusively reference-counted, which
makes reasoning about its lifetime very difficult in some cases. For example,
`CompilerInvocation` owns the `DiagnosticOption
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/138983
>From 5ae3871b981d22583ce1b168cbbe3429317d89eb Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Thu, 24 Apr 2025 15:10:23 -0700
Subject: [PATCH] [clang][modules] Timestamp-less validation API
---
clang-to
https://github.com/jansvoboda11 edited
https://github.com/llvm/llvm-project/pull/138983
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
jansvoboda11 wrote:
Test change LGTM, I was probably trying to get to a minimal test-case and ended
up with something that was relying on implementation details.
https://github.com/llvm/llvm-project/pull/132853
___
cfe-commits mailing list
cfe-commits
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/139751
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/jansvoboda11 created
https://github.com/llvm/llvm-project/pull/139751
This PR adds the `%sdk/SDKSettings.json` file to the PCM input file table, so
that the PCM gets invalidated when the file changes. This is necessary for
availability checks to work correctly.
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/139751
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,s
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/139584
Rate limit · GitHub
body {
background-color: #f6f8fa;
color: #24292e;
font-family: -apple-system,BlinkMacSystemFont,Segoe
UI,Helvetica,Arial,s
https://github.com/jansvoboda11 closed
https://github.com/llvm/llvm-project/pull/138983
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -2032,6 +2032,7 @@ class SourceManagerForFile {
// as they are created in `createSourceManagerForFile` so that they can be
// deleted in the reverse order as they are created.
std::unique_ptr FileMgr;
+ std::unique_ptr DiagOpts;
jansvoboda11 wrote:
I
@@ -107,6 +107,7 @@ class ASTUnit {
private:
std::unique_ptr LangOpts;
+ std::shared_ptr DiagOpts;
jansvoboda11 wrote:
I was hoping it could be, but the situation is a bit weird. The documentation
for `ASTUnit::LoadFromCommandLine()` says the `Diagnostics
@@ -837,6 +838,7 @@ class ASTUnit {
static std::unique_ptr LoadFromCommandLine(
const char **ArgBegin, const char **ArgEnd,
std::shared_ptr PCHContainerOps,
+ std::shared_ptr DiagOpts,
jansvoboda11 wrote:
Explained above.
https://github.com
https://github.com/jansvoboda11 updated
https://github.com/llvm/llvm-project/pull/135813
>From 5cab1328fd689783ac4117083e231e7f0a49c3db Mon Sep 17 00:00:00 2001
From: Jan Svoboda
Date: Wed, 9 Apr 2025 13:11:38 -0700
Subject: [PATCH] [clang][lex] Introduce new single-module-parse mode
---
clan
Author: Jan Svoboda
Date: 2025-05-19T10:40:55-07:00
New Revision: 7a242387c950c7060143da6da0e6fb91f36bb458
URL:
https://github.com/llvm/llvm-project/commit/7a242387c950c7060143da6da0e6fb91f36bb458
DIFF:
https://github.com/llvm/llvm-project/commit/7a242387c950c7060143da6da0e6fb91f36bb458.diff
L
Author: Jan Svoboda
Date: 2025-05-19T12:11:11-07:00
New Revision: 175f8a444b296ba956505a5760805f05a175111b
URL:
https://github.com/llvm/llvm-project/commit/175f8a444b296ba956505a5760805f05a175111b
DIFF:
https://github.com/llvm/llvm-project/commit/175f8a444b296ba956505a5760805f05a175111b.diff
L
Author: Jan Svoboda
Date: 2025-05-19T12:57:54-07:00
New Revision: 72b2219b3e2319e29831e4e9b07c440444f3add6
URL:
https://github.com/llvm/llvm-project/commit/72b2219b3e2319e29831e4e9b07c440444f3add6
DIFF:
https://github.com/llvm/llvm-project/commit/72b2219b3e2319e29831e4e9b07c440444f3add6.diff
L
jansvoboda11 wrote:
This change makes sense to me. I have two questions:
1. Can we add a test that checks that the block is omitted? (maybe via
`llvm-bcanalyzer -dump`)
2. Can we make the `explicit-build.cpp` test more robust? (possibly by writing
out import signatures)
https://github.com/llvm
jansvoboda11 wrote:
FWIW I tried running this patch through my `clang-scan-deps` benchmark on an
internal Apple project and saw 16% increase in instruction count.
https://github.com/llvm/llvm-project/pull/140867
___
cfe-commits mailing list
cfe-commit
jansvoboda11 wrote:
Are those in the upstream repo? If so, where are they coming from? I have a
sparse checkout, so it's possible I missed some non-Clang and non-LLDB usages
of `DiagnosticOptions`.
https://github.com/llvm/llvm-project/pull/139584
___
jansvoboda11 wrote:
> @jansvoboda11 I've revered your PR due buildbot failures above (and my local
> build failures with the same error messages). I'm happy to try your revised
> patch to see if it build cleanly. Thanks!
I already forward-fixed both failures: d25f95fdbc5314f30618912e18f00ad4dd
jansvoboda11 wrote:
Reverted the revert, since the build now fails due to the forward-fixes not
being reverted...
https://github.com/llvm/llvm-project/pull/139584
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/m
1401 - 1500 of 1565 matches
Mail list logo
