bf4ac5521d
astyle fix
2009-08-28 23:29:08 +03:00
3b2744e04a
borland c++: fixed compiler errors
2009-08-28 10:22:40 +02:00
8dcb846627
Fixed #614 (If you see this, there is a bug ###### Token::Match() - varid was 0)
...
http://sourceforge.net/apps/trac/cppcheck/ticket/614
2009-08-26 02:44:19 +07:00
311fb1a1f8
src/checkmemoryleak.{h,cpp}: minor improvements.
...
Corrections for commit 0d6b6e840d
2009-08-26 02:32:10 +07:00
ff3d1db480
Replace "const unsigned int" to "unsigned int" in methods parameters.
...
No functional change.
2009-08-26 02:26:12 +07:00
522f4c18c8
astyle formatting
2009-08-23 17:17:57 +02:00
9da2ecf4f4
Refactoring the unit testing. Wrote special tests for CheckMemoryLeakInFunction::findleak
2009-08-23 15:48:25 +02:00
0d6b6e840d
Improve performance by using bsearch(3) instead of a std::set
...
The function name list is already in memory so we can avoid copying it into a std::set
if we sort the list alphabetically. This way no binary search tree has to be allocated.
Before (best of 5 runs):
./testrunner > /dev/null 0,21s user 0,00s system 95% cpu 0,221 total
After (best of 5 runs):
./testrunner > /dev/null 0,20s user 0,00s system 95% cpu 0,210 total
2009-08-22 20:24:26 +02:00
8b46172bcf
Memory leaks: Use varid instead of varname
2009-08-19 19:42:07 +02:00
85682ed429
Fixed #591 (False positive: Deallocating a deallocated pointer)
2009-08-18 20:49:08 +02:00
2cdb0abb82
Refactoring: Use the Token::link instead of loop
2009-08-17 22:23:37 +02:00
52856fcbeb
ettlmartin: improved the MathLib::IsInt
2009-08-16 18:45:26 +02:00
522f783ada
CheckMemoryLeak: try to speed up call_func().
...
Using std::set::find() instead of sequence of Match() calls.
2009-08-16 22:02:11 +07:00
ecdbcbce3d
Fixed #566 (False positive when assigning the return value of realloc(NULL,..) to a pointer variable holding a freed memory address)
2009-08-10 22:04:28 +02:00
ce8c5b0236
Fixed #555 (False positive: [CuTest.c:25]: (error) Memory leak: len)
2009-08-05 21:18:16 +02:00
cb57bd6678
Listing error messages in XML format
2009-08-04 21:36:55 +02:00
9e348ca739
Fixed #554 (resource leak false positive)
2009-08-04 21:34:14 +02:00
769119c725
Fix ticket #552 (False positive with --all when realloc is used with two variables.)
...
http://sourceforge.net/apps/trac/cppcheck/ticket/552
2009-08-02 23:30:43 +03:00
10f6678cf9
TODO test case realloc5 added and "Checkmemoryleak: simplifycode result for" debug info added.
2009-08-02 22:23:47 +03:00
b7ffcf53ba
Fixed #511 (false positive, memory leak when using 'var = ({});')
2009-07-31 23:42:21 +02:00
050b85c5eb
memory leak for struct member: fixed false positive when the struct pointer is stored away
2009-07-29 11:38:20 +02:00
65d6bce165
Replace Match() to operator==() from std::string.
...
No functional change.
2009-07-28 00:13:44 +07:00
5ce151983e
Replace simpleMatch() to operator==() from std::string.
...
No functional change.
2009-07-28 00:13:11 +07:00
8643936e0c
Fixed ticket #500 (Tokenizer: simplify the "(p != NULL)" conditions)
...
Also teach simplifyIfNot() to handle variables like Foo::var.
https://sourceforge.net/apps/trac/cppcheck/ticket/500
2009-07-25 18:23:03 +07:00
9136d8cf80
Tokenizer: simplify return statements - remove redundant parantheses
2009-07-24 21:55:35 +02:00
6bb85703f6
memory leaks: fixed false positive for struct members - better handling of if/else
2009-07-24 09:05:40 +02:00
565ac2fca6
Fixed #493 (Memory leak: False positive when using the exit(0))
2009-07-23 16:30:30 +02:00
3243933c90
Fixed #494 (False positives of checking leaks struct members)
...
Better handling of function calls
2009-07-22 08:30:51 +02:00
6f646246ee
dereferencing deallocated memory: it is ok to take the address to deallocated memory but it is not ok to dereference the address
2009-07-21 13:04:13 +02:00
d2278b5ce0
Fix some of the issues from #478 (Warnings in Visual Studio build with -W4)
...
http://sourceforge.net/apps/trac/cppcheck/ticket/478
2009-07-20 23:24:23 +03:00
74699cab66
memory leak for struct member: made the checking more sensitive. Skip bail-out execution paths were the member is properly deallocated
2009-07-20 14:39:24 +02:00
ddaea3244d
memleak: implemented simple checking for leaking struct members
2009-07-19 16:51:31 +02:00
2ba0897ecb
src/checkmemoryleak.cpp: added fchmod() function to white list.
2009-07-18 23:48:48 +07:00
a0a526a205
Fixed ticket #464 (resource leak not detected when getc is used on the file handle)
...
http://sourceforge.net/apps/trac/cppcheck/ticket/464
2009-07-18 18:32:55 +07:00
4f2520bef3
doxygen: made the todos visible in the doxygen output
2009-07-14 08:17:12 +02:00
913789605f
refactoring: Refactoring of the leak-checking classes
2009-07-13 15:07:26 +02:00
7d35447d80
refactoring: renamed the severity "all" to "possible error"
2009-07-13 10:16:31 +02:00
2a02041fd8
memleak: fixed todo test case
2009-07-07 15:22:37 +02:00
9beb73824b
Fixed #446 (memory leak false positive when variable is static)
2009-07-06 12:20:13 +02:00
8b0e481d46
Refactoring: Rename member functions to follow naming guidelines.
2009-07-05 23:16:43 +03:00
61f4a29758
minor fix of test case
2009-06-23 20:40:27 +02:00
7f8b40d135
calling unknown function -> use
2009-06-23 20:37:34 +02:00
5c36d33943
astyle formatting
2009-06-21 21:03:58 +02:00
13e805f332
Fixed ticket #399 (Add detection for resource leaks after open() usage)
...
http://sourceforge.net/apps/trac/cppcheck/ticket/399
2009-06-21 22:01:43 +07:00
8715ba1458
CheckMemoryLeakInFunction: More sensitive checking when the code calls an unknown function
2009-06-21 14:12:59 +02:00
f28dec1f5a
detect memory leak when all is given when calling an unknown function
2009-06-21 13:48:39 +02:00
d1d9c9a9c9
CheckMemoryLeakInClass: Only show the memory leak message when --all is used
2009-06-20 14:17:56 +02:00
453a5cdd47
Fixed ticket #411 (false positive: resource leak in while loop)
2009-06-20 13:58:30 +02:00
5d0d21d668
Fixed a false positive (memory leak)
...
This false positive was first discovered in the VLC sourcecode. The TestMemoryLeakInFunction::simple11 test case is a simplified test code of that code.
2009-06-17 21:42:49 +02:00
ed98f1b367
Fixed #131 (Regression: leak reported by 1.27 is not reported by 1.29)
...
The leak will be reported as a definite bug again. Not just a possible bug.
cppcheck don't know if the execution path is possible, but it knows that if the execution path is taken the memory leak will be certain.
2009-06-16 19:05:05 +02:00
Reijo Tomperi
danmar
Slava Semushin
Daniel Marjamäki
Benjamin Kramer