https://github.com/NagyDonat commented:
This is a very promising change! I can recall some analyzer reports that just
dumped lots of macro code without any formatting.
Could you add a few testcases that would demonstrate the effects of your
change? (The tests of the analyzer are in `clang/test
@@ -68,6 +68,11 @@ struct PathDiagnosticConsumerOptions {
/// without re-compiling the program under analysis.
bool ShouldDisplayMacroExpansions = false;
+ /// Whether to include clang-formatted macros during macro expansions
+ /// or to keep it as it was stored, the def
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/156046
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -143,68 +143,49 @@ def BitwiseShiftChecker : Checker<"BitwiseShift">,
]>,
Documentation;
-def CallAndMessageModeling : Checker<"CallAndMessageModeling">,
- HelpText<"Responsible for essential modeling and assumptions after a "
- "function/method call. For ins
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/156073
In CallAndMessageChecker the initialization of bug types was highly obfuscated
(even compared to other `mutable std::unique_ptr` hacks). This commit cleans up
this situation and removes a totally superfluous
NagyDonat wrote:
> Based on your review of the code do you have any suggestions on how I should
> go about tightening the heuristic?
Your code puts pointers (that were allocated by `new` or `new[]`) into a
relinquished state when they are passed to a constructor call. Currently this
is overly
NagyDonat wrote:
> > > I don't think I have the time to do a proper focused review, so I'd at
> > > this point forward you to @NagyDonat to finish the reviews.
> >
> >
> > @steakhal The "Requested change" review status is still active from your
> > old review. Is it just a github glitch?
>
>
@@ -3068,12 +3124,225 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart owning pointer name
+static bool isSmartOwningPtrName(StringRef Name) {
+ return Name =
NagyDonat wrote:
This change itself seems to be correct, but I have doubts about the overall
viability of the implementation of this `alpha` checker.
In particular I don't like that it maintains its independent model about things
that are modeled more accurately by other checkers. For example
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/152751
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat commented:
Thanks for updating the commit!
I reviewed the tests and added some minor suggestions in the implementation.
Among my earlier suggestions [the visibility of
VisitSymbol](https://github.com/llvm/llvm-project/pull/152751/files#r2301055454)
and the [comple
@@ -1096,6 +1104,54 @@ class StopTrackingCallback final : public SymbolVisitor {
return true;
}
};
+
+/// EscapeTrackedCallback - A SymbolVisitor that marks allocated symbols as
+/// escaped.
+///
+/// This visitor is used to suppress false positive leak reports when smar
@@ -3068,12 +3124,225 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart owning pointer name
+static bool isSmartOwningPtrName(StringRef Name) {
+ return Name =
@@ -0,0 +1,232 @@
+// RUN: %clang_analyze_cc1 -verify -analyzer-output=text %s \
+// RUN: -analyzer-checker=core \
+// RUN: -analyzer-checker=cplusplus \
+// RUN: -analyzer-checker=unix \
+// RUN: -analyzer-checker=unix.Malloc
NagyDonat wrote:
```suggesti
@@ -0,0 +1,232 @@
+// RUN: %clang_analyze_cc1 -verify -analyzer-output=text %s \
+// RUN: -analyzer-checker=core \
+// RUN: -analyzer-checker=cplusplus \
+// RUN: -analyzer-checker=unix \
+// RUN: -analyzer-checker=unix.Malloc
+
+#include "Inputs/system-header-simulator-fo
@@ -3068,12 +3124,225 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart owning pointer name
+static bool isSmartOwningPtrName(StringRef Name) {
+ return Name =
@@ -3068,12 +3124,225 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart owning pointer name
+static bool isSmartOwningPtrName(StringRef Name) {
+ return Name =
@@ -3074,6 +3077,43 @@ void MallocChecker::checkPostCall(const CallEvent &Call,
(*PostFN)(this, C.getState(), Call, C);
return;
}
+
+ ProgramStateRef State = C.getState();
+
+ if (const auto *Ctor = dyn_cast(&Call)) {
+// Ensure we are constructing a concrete ob
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/155131
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -218,3 +239,138 @@ void caller() {
(void)n;
} // no-warning: No potential memory leak here, because that's been already
reported.
} // namespace symbol_reaper_lifetime
+
+
+// Minimal RAII class that properly deletes its pointer.
+class Bar {
+public:
+ explicit Bar(int
@@ -3074,6 +3077,43 @@ void MallocChecker::checkPostCall(const CallEvent &Call,
(*PostFN)(this, C.getState(), Call, C);
return;
}
+
+ ProgramStateRef State = C.getState();
+
+ if (const auto *Ctor = dyn_cast(&Call)) {
+// Ensure we are constructing a concrete ob
https://github.com/NagyDonat commented:
> Im sorry about messing up the naming convention, I havent contributed too
> much to open source stuff.
No harm done, this is part of the plan :slightly_smiling_face: Every
contributor must learn these sometime and this time it was your turn to do so.
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/155237
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
NagyDonat wrote:
> AFAIK, the check now will not crash if `enable_if` with incorrect signature
> is encountered.
>
> * If first arg is not `bool` - bail out
>
> * If parameter count is not 1 or to - bail out
>
> * If 2nd arg is incorrect - do not warn (As I understand from the tes
https://github.com/NagyDonat commented:
I reviewed the code changes within this commit, and overall I'm very grateful
for this contribution, because it overcomes lots of technical challenges to
introduce a promising heuristic.
As this code is unavoidable complex, I think it's especially import
@@ -3068,12 +3124,242 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart pointer name
+static bool isSmartPtrName(StringRef Name) {
+ return Name == "unique_ptr
@@ -3068,12 +3124,242 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart pointer name
+static bool isSmartPtrName(StringRef Name) {
+ return Name == "unique_ptr
@@ -3068,12 +3124,242 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart pointer name
+static bool isSmartPtrName(StringRef Name) {
+ return Name == "unique_ptr
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/152751
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -3068,12 +3124,242 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart pointer name
+static bool isSmartPtrName(StringRef Name) {
+ return Name == "unique_ptr
@@ -3194,7 +3480,6 @@ void MallocChecker::checkEscapeOnReturn(const ReturnStmt
*S,
if (!Sym)
// If we are returning a field of the allocated struct or an array element,
// the callee could still free the memory.
-// TODO: This logic should be a part of generic sy
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/152751
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -3068,12 +3124,242 @@ void MallocChecker::checkDeadSymbols(SymbolReaper
&SymReaper,
C.addTransition(state->set(RS), N);
}
+// Helper function to check if a name is a recognized smart pointer name
+static bool isSmartPtrName(StringRef Name) {
+ return Name == "unique_ptr
@@ -1096,6 +1104,54 @@ class StopTrackingCallback final : public SymbolVisitor {
return true;
}
};
+
+/// EscapeTrackedCallback - A SymbolVisitor that marks allocated symbols as
+/// escaped.
+///
+/// This visitor is used to suppress false positive leak reports when smar
NagyDonat wrote:
Oh, I see, that commit was merged on Saturday, while I started to develop this
commit (IIRC) on Thursday.
Compared to that change, the only difference provided by my trivial patch is
that it silences fixits for `enable_if` implementations that are not standard
but take a bool
NagyDonat wrote:
Thanks for working on this :slightly_smiling_face: !
Is your code ready for a first round of review, or are you still planning
significant changes?
By the way, please change the title of this PR to e.g. "Fix unique_ptr
aggregate initialization false positives" or something si
@@ -756,3 +756,33 @@ abs(const number &v) {
}
}
+
+// NOLINTBEGIN
+namespace custom {
+template struct enable_if { };
+
+template struct enable_if { typedef T type; };
+
+template
+using enable_if_t = typename enable_if::type;
+
+} // namespace custom
+// NOLINTEND
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/155237
This commit ensures that the modernize-use-constraints check ignores templates
that happen to be named `enable_if` or `enable_if_t` if they are not declared
in the namespace `std`.
This patch motivated by a
NagyDonat wrote:
> Probably for this checker another approach can work: Store only if a string
> is known to be null-terminated, or known to contain no zero at all (probably
> other state like "likely does not contain terminating zero" or "likely
> contains zero only at the end" can be used).
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/146664
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,295 @@
+//=== MissingTerminatingZeroChecker.cpp -*- 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
@@ -1,11 +1,27 @@
// RUN: %clang_analyze_cc1 -triple amdgcn-unknown-unknown \
-// RUN: -analyzer-checker=core -verify %s
+// RUN: -Wno-incompatible-library-redeclaration \
+// RUN: -analyzer-checker=core,unix.cstring.BadSizeArg -verify %s
NagyDonat wrote:
NagyDonat wrote:
> Would it make sense to extend the warning to any pointer cast, not only
> `struct` types?
I feel that it is reasonable to limit this to `struct`s, especially right now
as we introduce this as a replacement for the checker in the Static Analyzer.
Let's publish this check thi
https://github.com/NagyDonat approved this pull request.
This commit still LGTM and it would be nice to get it merged.
@PiotrZSL @EugeneZelenko @vbvictor Are you satisfied with the current state of
the commit? Do you have any concerns that are not answered yet?
https://github.com/llvm/llvm-pro
@@ -3156,8 +3156,14 @@ void MallocChecker::checkPreCall(const CallEvent &Call,
for (unsigned I = 0, E = Call.getNumArgs(); I != E; ++I) {
SVal ArgSVal = Call.getArgSVal(I);
if (isa(ArgSVal)) {
- SymbolRef Sym = ArgSVal.getAsSymbol();
- if (!Sym)
+ const
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/152138
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/152163
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
NagyDonat wrote:
> Interpret this in the good way. I trust you, I skimmed through the PR and
> looks good. If you deem to get a second opinion because it needs some in
> depth review go for it. Otherwise I'm fine merging.
Thanks for the confidence! (I already asked Endre to review, so I'll pr
@@ -0,0 +1,295 @@
+//=== MissingTerminatingZeroChecker.cpp -*- 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
NagyDonat wrote:
> I'm not too interested reviewing this.
Completely understandable, I'm uploading quite many boring cleanup commits
nowdays :) I'm adding you as a reviewer on them to give you an opportunity to
object if you see any glaring issues, but I don't expect you to dig into the
detai
NagyDonat wrote:
Yet another suggestion for the name: `misc-override-with-different-visibility`.
(The word `override` is shorter than `virtual-function` and still clearly
identifies the situation where this check is relevant.) However I don't want to
make this bikeshedding even longer, so I ca
NagyDonat wrote:
Rerunning the CI checks didn't change the fact that the Github automation runs
the tests on a merge between my commit (which would pass the CI) with
74af2cec7bbd307d1dcb7f9f3cdf339d551a1f9f (where the CI fails) instead of a more
recent main revision (where the CI would pass).
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/152163
From 6b9489e246333eda92b3aa67f87096ad0a17bfd6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Tue, 5 Aug 2025 16:46:49 +0200
Subject: [PATCH] [analyzer] Remove impossible BugType from CStrin
@@ -25,25 +25,39 @@ namespace ento {
namespace retaincountchecker {
class RefCountBug : public BugType {
+ StringRef ReportMessage;
+
public:
- enum RefCountBugKind {
-UseAfterRelease,
-ReleaseNotOwned,
-DeallocNotOwned,
-FreeNotOwned,
-OverAutorelease,
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/152138
From bb4a8fb1a587050ee5b463304bbc3a47b01a526b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Mon, 4 Aug 2025 19:38:05 +0200
Subject: [PATCH 1/3] [NFC][analyzer] Conversion to CheckerFamily:
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/152138
From bb4a8fb1a587050ee5b463304bbc3a47b01a526b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Mon, 4 Aug 2025 19:38:05 +0200
Subject: [PATCH 1/2] [NFC][analyzer] Conversion to CheckerFamily:
NagyDonat wrote:
The failure of the CI jobs is unrelated to this commit; it seems that the tests
were broken on _main_ at the moment when the CI jobs tested this commit. As
main was stabilized since that time, re-running the CI jobs should be
successful.
https://github.com/llvm/llvm-project/p
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/152163
CStringChecker had an AdditionOverflow bug type which was intended for a
situation where the analyzer concludes that the addition of two size/length
values overflows `size_t`.
I strongly suspect that the ana
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/152138
This commit converts RetainCountChecker to the new checker family framework
that was introduced in the commit
6833076a5d9f5719539a24e900037da5a3979289
This commit also performs some minor cleanup around the p
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/151370
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -1219,6 +1219,16 @@ MemRegionManager::getElementRegion(QualType elementType,
NonLoc Idx,
const ASTContext &Ctx) {
QualType T = Ctx.getCanonicalType(elementType).getUnqualifiedType();
+ // The address space must be preserved because som
https://github.com/NagyDonat approved this pull request.
LGTM, with one very minor nitpick in a comment.
https://github.com/llvm/llvm-project/pull/150417
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/lis
@@ -0,0 +1,21 @@
+const int global_const = 42;
+
+struct TestStruct {
+ const int x;
+ int y;
+};
+
+void immutable_violation_examples() {
+ *(int *)&global_const = 100; // warn: Trying to write to immutable memory
+
+ const int local_const = 42;
+ *(int *)&local_const = 43;
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/150417
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/151370
From 690e311be23ef506c1316d076991d3b557718474 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Wed, 30 Jul 2025 19:40:10 +0200
Subject: [PATCH 1/3] [analyzer] Retain address space information
@@ -1219,6 +1219,15 @@ MemRegionManager::getElementRegion(QualType elementType,
NonLoc Idx,
const ASTContext &Ctx) {
QualType T = Ctx.getCanonicalType(elementType).getUnqualifiedType();
+ // The address space must be preserved because som
@@ -1219,6 +1219,16 @@ MemRegionManager::getElementRegion(QualType elementType,
NonLoc Idx,
const ASTContext &Ctx) {
QualType T = Ctx.getCanonicalType(elementType).getUnqualifiedType();
+ // The address space must be preserved because som
@@ -1219,6 +1219,16 @@ MemRegionManager::getElementRegion(QualType elementType,
NonLoc Idx,
const ASTContext &Ctx) {
QualType T = Ctx.getCanonicalType(elementType).getUnqualifiedType();
+ // The address space must be preserved because som
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/151370
From 690e311be23ef506c1316d076991d3b557718474 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Wed, 30 Jul 2025 19:40:10 +0200
Subject: [PATCH 1/2] [analyzer] Retain address space information
@@ -0,0 +1,168 @@
+//=== StoreToImmutableChecker.cpp - Store to immutable memory ---*- 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 +1,23 @@
+// Global const variable
NagyDonat wrote:
I would prefer removing this comment and "// Const struct member" because they
just state the obvious.
https://github.com/llvm/llvm-project/pull/150417
___
c
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/144901
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/151370
The factory method `MemRegionManager::getElementRegion()` is the main way of
constructing `ElementRegion` objects, which are widespread in the analysis and
may represent array elements (as lvalues), pointer a
@@ -0,0 +1,18 @@
+// RUN: %clang_analyze_cc1 -triple amdgcn-unknown-unknown \
+// RUN: -analyzer-checker=core,unix.Malloc,debug.ExprInspection -verify \
+// RUN: -Wno-incompatible-pointer-types -Wno-unused-comparison %s
+
+// expected-no-diagnostics
+//
+// By default, pointers ar
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/150971
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -124,6 +124,12 @@ New checks
pointer and store it as class members without handle the copy and move
constructors and the assignments.
+- New :doc:`bugprone-invalid-enum-default-initialization
NagyDonat wrote:
> LLVM documentation (https://llvm.org/docs
https://github.com/NagyDonat edited
https://github.com/llvm/llvm-project/pull/150417
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,32 @@
+// RUN: %clang_analyze_cc1 -analyzer-checker=alpha.core.StoreToImmutable %s
+
+// Global const variable
+const int global_const = 42;
NagyDonat wrote:
@steakhal You are right that globals and locals are handled differently _by the
checker imple
@@ -29,11 +30,59 @@ class StoreToImmutableChecker : public Checker
{
void checkBind(SVal Loc, SVal Val, const Stmt *S, CheckerContext &C) const;
private:
+ bool isInitializationContext(const Stmt *S, CheckerContext &C) const;
bool isEffectivelyConstRegion(const MemRegio
NagyDonat wrote:
These buildbots demonstrate the size and diversity of this project by bringing
up random errors in subsystems that are completely unrelated to the actual PR
:roll_eyes:
Perhaps it would be nice to reduce the scope of testing and avoid noise from
subprojects that are complete
@@ -8274,6 +8275,15 @@ inline void QualType::removeLocalVolatile() {
removeLocalFastQualifiers(Qualifiers::Volatile);
}
+inline QualType QualType::removeNonAddressSpaceQualifiers() {
+ if (getQualifiers().hasTargetSpecificAddressSpace()) {
+removeLocalFastQualifiers();
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/151136
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/151171
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -28,23 +28,20 @@ using namespace ento;
namespace {
class StackAddrEscapeChecker
-: public Checker,
- check::EndFunction> {
+: public CheckerFamily,
+ check::EndFunction> {
mutable IdentifierInfo *dispatch_semaphore_tI
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/151136
From 2ca9d38d8aa41108cc2ab26edf1a12e4dd7177bb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Tue, 29 Jul 2025 14:14:33 +0200
Subject: [PATCH 1/2] [analyzer] Conversion to CheckerFamily:
St
@@ -124,6 +124,12 @@ New checks
pointer and store it as class members without handle the copy and move
constructors and the assignments.
+- New :doc:`bugprone-invalid-enum-default-initialization
NagyDonat wrote:
Unfortunately the github pull request UI ge
@@ -142,34 +142,18 @@ void CFErrorFunctionChecker::checkASTDecl(const
FunctionDecl *D,
//===--===//
namespace {
+class NSOrCFErrorDerefChecker
+: public CheckerFamily> {
+ mutable IdentifierInfo *NSErro
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/151171
From 036f1e61900c19b0142c20580a0bc5894f509921 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Tue, 29 Jul 2025 17:18:34 +0200
Subject: [PATCH 1/2] [analyzer] Conversion to CheckerFamily:
NS
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/151141
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/151171
This commit converts the class `NSOrCFErrorDerefChecker` to the checker family
framework and simplifies some parts of the implementation (e.g. removes two
very trivial subclasses of `BugType`).
This commit i
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/151141
This commit eliminates some corrupted variants of the once-widespread `mutable
std::unique_ptr` antipattern from the checker file
`BasicObjCFoundationChecks.cpp`.
Previous purges probably missed these becaus
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/151136
This commit converts the class StackAddrEscapeChecker to the checker family
framework and slightly simplifies the implementation.
This commit is almost NFC, the only technically "functional" change is that it
@@ -584,7 +579,7 @@ ProgramStateRef
CStringChecker::CheckLocation(CheckerContext &C,
// These checks are either enabled by the CString out-of-bounds checker
// explicitly or implicitly by the Malloc checker.
// In the latter case we only do modeling but do not emit
https://github.com/NagyDonat updated
https://github.com/llvm/llvm-project/pull/150971
From 3422f5acd8fe090a542f63b2e380fe32e9fcf45b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Don=C3=A1t=20Nagy?=
Date: Mon, 28 Jul 2025 16:22:07 +0200
Subject: [PATCH 1/2] [NFC][analyzer] Conversion to CheckerFamily
@@ -584,7 +579,7 @@ ProgramStateRef
CStringChecker::CheckLocation(CheckerContext &C,
// These checks are either enabled by the CString out-of-bounds checker
// explicitly or implicitly by the Malloc checker.
// In the latter case we only do modeling but do not emit
@@ -845,17 +825,11 @@ void CStringChecker::emitOutOfBoundsBug(CheckerContext &C,
ProgramStateRef State, const Stmt *S,
StringRef WarningMsg) const {
if (ExplodedNode *N = C.generateErrorNode(State
https://github.com/NagyDonat closed
https://github.com/llvm/llvm-project/pull/150935
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/NagyDonat created
https://github.com/llvm/llvm-project/pull/150971
This commit converts the class CStringChecker to the checker family framework
and slightly simplifies the implementation.
This commit is NFC and preserves the confused garbage descriptions and
categories of
NagyDonat wrote:
The CI fails because my commit https://github.com/llvm/llvm-project/pull/150240
was merged after the creation of this commit and introduced one additional
occurrence of "Attempt to free released memory". It will be fixed soon.
https://github.com/llvm/llvm-project/pull/150935
_
https://github.com/NagyDonat approved this pull request.
The commit looks good to me; thanks for cleaning this up!
https://github.com/llvm/llvm-project/pull/150935
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/m
1 - 100 of 1359 matches
Mail list logo
