[valgrind] [Bug 452758] Valgrind does not read properly DWARF5 as generated by Clang14

2022-07-14 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=452758

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 CC||dushis...@mail.ru

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 444399] disInstr(arm64): unhandled instruction 0xC87F2D89 (LD{,A}XP and ST{,L}XP)

2022-07-05 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=444399

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 CC||dushis...@mail.ru

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 436606] -std=c++17 mask connect-3arg-lambda check

2021-05-05 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=436606

--- Comment #1 from Evgeniy A. Dushistov  ---
Looks like this is expected behavior of clang/libclang:

for code like this: 

#define signals public  
class QObject { 
signals:
void destroyed();   
public: 
template
static void connect(QObject *, void (QObject::*mf)(void), F f) {}   
void deleteLater(); 
};  

void test(QObject *o1, QObject *o2) 
{   
QObject::connect(o1, &QObject::destroyed,   
[=] { o2->deleteLater(); });
}   

clang dumps different AST.  

In case of -std=c++14 (clang++ -std=c++14 -Xclang -ast-dump -c test.cpp)
3rd argument (lambda) passed to QObject::connect looks like this:   
CXXConstructExpr 0x5589e3278970  '(lambda at
test.cpp:14:21)':'(lambda at test.cpp:14:21)' 'void ((lambda at test.cpp:14:21)
&&) noexcept' elidable 
  `-MaterializeTemporaryExpr 0x5589e3275cd0  '(lambda
at test.cpp:14:21)' xvalue  
`-LambdaExpr

in case of -std=c++17 lambda looks like this:   
   `-LambdaExpr 0x5641f55db880  '(lambda at
test.cpp:14:21)'
|-CXXRecordDecl 0x5641f55db420  col:21 implicit class
definition

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 436606] New: -std=c++17 mask connect-3arg-lambda check

2021-05-04 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=436606

Bug ID: 436606
   Summary: -std=c++17  mask connect-3arg-lambda check
   Product: clazy
   Version: unspecified
  Platform: Other
OS: Linux
Status: REPORTED
  Severity: normal
  Priority: NOR
 Component: general
  Assignee: unassigned-b...@kde.org
  Reporter: dushis...@mail.ru
CC: smart...@kde.org
  Target Milestone: ---

SUMMARY
If compile_commands.json contains '-std=c++17' or if I use
-extra-arg='-std=c++17' then warning about usage of 3 argument connect
disappear

STEPS TO REPRODUCE
Build clazy from source code (90fc09d3047792e585d07a78a79f9648f7d5fb1c) and run
./build/bin/clazy-standalone -extra-arg='-std=c++17'
-extra-arg='-I/usr/include/qt' -extra-arg='-I/usr/lib/clang/11.1.0/include'
tests/connect-3arg-lambda/main.cpp

OBSERVED RESULT
No warnings about 'Pass a context object as 3rd connect parameter'

EXPECTED RESULT
Warnings with text 'Pass a context object as 3rd connect parameter'

SOFTWARE/OS VERSIONS
Arch Linux, Qt Version: 5.15.2+kde+r192

ADDITIONAL INFORMATION

I tried clazy/tests/run_tests.py .
By default all tests passed,
but if I replace -std=c++14 with -std=c++17,
then such tests failed:
connect-3arg-lambda
reserve-candidates
qt6-qlatin1stringchar-to-u
range-loop-detach
qstring-allocations
container-anti-pattern
detaching-member
qstring-allocations

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 423780] clazy-fully-qualified-moc-types not detecting missing namespace

2021-04-15 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=423780

--- Comment #2 from Evgeniy A. Dushistov  ---
I made some experiments, and looks like clazy should be run against output of
moc, not against header by it self:

```
// boo.h
#pragma once

#include 

namespace MyNameSpace {

struct MyGadget {
Q_GADGET
};

class MyObject : public QObject
{   
Q_OBJECT
Q_PROPERTY(MyGadget myprop READ myprop CONSTANT); // Wrong, needs
namespace
public:
MyGadget myprop() const;
};
}
```

clazy-standalone --extra-arg="-I/usr/lib/clang/11.1.0/include"
--checks=level0,level1,level2   -p ./build/compile_commands.json boo.h

reports nothing,
but

```
$ clazy-standalone --extra-arg="-I/usr/lib/clang/11.1.0/include"
--checks=level0,level1,level2   -p ./build/compile_commands.json
build/moc_boo.cpp
In file included from moc_boo.cpp:10:
build/../boo.h:11:5: warning: Q_PROPERTY of type MyGadget should use full
qualification (MyNameSpace::MyGadget) [-Wclazy-fully-qualified-moc-types]
class MyObject : public QObject
^
1 warning generated.
```

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 423780] clazy-fully-qualified-moc-types not detecting missing namespace

2021-04-15 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=423780

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 CC||dushis...@mail.ru

--- Comment #1 from Evgeniy A. Dushistov  ---
I get the same problem.

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 370713] new check: Q_INVOKABLEs returning QObject

2020-08-25 Thread Evgeniy A. Dushistov
https://bugs.kde.org/show_bug.cgi?id=370713

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 CC||dushis...@mail.ru

--- Comment #1 from Evgeniy A. Dushistov  ---
For me it was unexpected that class marked as Q_GADGET can not be returned via
Q_INVOKABLE

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] clazy-empty-qstringliteral for uic generated files

2018-12-13 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #11 from Evgeniy A. Dushistov  ---
Sorry for noise, it was cache issue,
after clean rebuild ui_ files were generated without empty QStringLiteral

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] clazy-empty-qstringliteral for uic generated files

2018-12-13 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 Resolution|FIXED   |---
 Status|RESOLVED|REOPENED

--- Comment #10 from Evgeniy A. Dushistov  ---
>  if (preProcessorVisitor && preProcessorVisitor->qtVersion() >= 51200)

At now I updated my system Qt to 5.12 and:

>ui_ValueCancelDialog.h:42:42: warning: Use QString instead of an empty 
>QStringLiteral [-Wclazy-empty-qstringliteral]

So may be change 51200 to something like 5 ?

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] clazy-empty-qstringliteral for uic generated files

2018-10-09 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #8 from Evgeniy A. Dushistov  ---
Thank you, now after long period of time I have zero warnings in my projects.

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] clazy-empty-qstringliteral for uic generated files

2018-10-09 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

Evgeniy A. Dushistov  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---

--- Comment #5 from Evgeniy A. Dushistov  ---
The problem is still here,

I suppose this is because of:

+// Since 5.12 uic no longer uses QStringLiteral("")
+if (preProcessorVisitor && preProcessorVisitor->qtVersion() >= 051200)
+return false;

In my case preProcessorVisitor->qtVersion() return 51102,
but this is in decimal representation, in octet representation it is 0143636,
and os it it bigger then 051200.

I suppose the right fix will be 51200, without '0' prefix.

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] clazy-empty-qstringliteral for uic generated files

2018-10-09 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

--- Comment #4 from Evgeniy A. Dushistov  ---
Thank you.

-- 
You are receiving this mail because:
You are watching all bug changes.

[clazy] [Bug 399544] New: clazy-empty-qstringliteral for uic generated files

2018-10-09 Thread Evgeniy A . Dushistov
https://bugs.kde.org/show_bug.cgi?id=399544

Bug ID: 399544
   Summary: clazy-empty-qstringliteral for uic generated files
   Product: clazy
   Version: unspecified
  Platform: Other
OS: Linux
Status: REPORTED
  Severity: normal
  Priority: NOR
 Component: general
  Assignee: unassigned-b...@kde.org
  Reporter: dushis...@mail.ru
CC: smart...@kde.org
  Target Milestone: ---

SUMMARY

clazy-empty-qstringliteral is great, but it produces too many warnings for uic
generated files.

STEPS TO REPRODUCE
1. Create ".ui" file
2. Change stylesheet to something, and then clear it.
3. You will get in generated by uic c++ code, something like:

ActionConfirmContextMenu->setStyleSheet(QStringLiteral(""));

OBSERVED RESULT


EXPECTED RESULT

It would be great if clazy automatically disable `clazy-empty-qstringliteral` 
for headers file that starts with "ui_" and ends with ".h"

SOFTWARE VERSIONS
(available in About System)
Qt Version: 5.11.2

ADDITIONAL INFORMATION

-- 
You are receiving this mail because:
You are watching all bug changes.