cppcheck / News: Recent posts


Cppcheck-1.58 has been released.

Commandline/Settings changes:
- Added --include to the cppcheck command line client. This forces inclusion of the given file. This can for instance be used instead of --append and will then allow you to use #define etc also.
- The threads handling has been improved. Using -jN now works in windows also.

- NULL pointers: Improved checking of default function argument values.... read more

Posted by Daniel Marjamäki 2013-01-12


Cppcheck-1.57 has been released

Commandline/Settings changes:

* Support for Java and C# code has been removed.
* New option --language= (alias: -x=) to enforce treating code as a specific language


* Support GCC binary number syntax (0b...)
* Fixed search order for include paths
* Several error messages have been improved to be more understandable and consistent
* Rewrote enum parsing
* Simplification of well-known math.h functions
* Check for variable being assigned a value never used handles now also reassignments
* Fixed several false negatives in buffer overrun check
* Detect more inadequate comparisons of boolean variables
* Better detection of mismatching and assignments and comparisons with bitwise operations ... read more

Posted by Daniel Marjamäki 2012-11-04


Cppcheck-1.56 has been released.

Commandline/Settings changes:
- Code is considered to be C++11 or C11 compatible by default, added new standards to --std= option: c89, c11, c++03
- '--inconclusive' has been made "official" and shown in help dialog

- Uninstanciated templates are no longer removed from token list
- Support 'using namespace std;' - 'std ::' added to C++ standard library types when 'using namespace std;' is found
- Several error messages have been improved to be more understandable and consistent
- Checking of preprocessor configurations that are considered invalid, because the macro used in #if is also used in the code, is now skipped.... read more

Posted by Daniel Marjamäki 2012-09-01

cppcheck 1.56 pre-release

We have released a 1.56 pre - release.

There are still known bugs that we intend to fix.

But if you still want to download and test this then feel free to do it. If you see some problem then please report it so we can try to fix it before we release the real 1.56.

The pre-release is available here:

Posted by Daniel Marjamäki 2012-08-26


Cppcheck can now analyse source files that are UTF-16 encoded

Faster checking. Some projects, for instance sqlite, is much quicker now. The difference when checking other projects is not big.

Removed deprecated command line options -s, --style, -a, --all, --auto-dealloc, --test-2-pass.

Improved C code analysis. Better analysis of unknown types etc.

New check: comparison of modulo results that are always true/false. For instance: if (x % 5 == 5)... read more

Posted by Daniel Marjamäki 2012-06-30


Command line:
* Added --relative-path

Improved checks to detect more bugs:
* improved checking for buffer overruns on dynamically allocated memory (#3569)
* when variable is compared against a value in condition, the variable value is known in the conditional block.
* null pointer checking - handle more standard functions and improved checking of loops
- more standard functions are checked (#1171, #410)
- improved checking of loops (#1927)
- dangerous conditions (#3518)
* detect more bugs related to std::find and string::find usage (#3162)
* detect more sign extension problems (#3637)
* detect redundant .c_str() for strings (#1079)
* invalid pointer cast from float* to double* (#1255)
* more strict about int vs. bool (#1877)
* Preprocessor: extract more configurations => more code is checked
* Unused functions - fixed false negatives for functions that returns reference
* Better checking of sizeof
* Improved "unused private function" checking... read more

Posted by Daniel Marjamäki 2012-04-25


No major new features. No major changes to the command line flags.

We have fixed many problems such as false positives, segmentation faults, etc.

We improved existing checks to detect more errors:
* type checking for format string parameters
* uninitialized variables
* bad std::string usage
* buffer overflow checking
* iterator used on different containers
* doublefree of pointer
* null pointers
* bad usage of isdigit/isgraph/isalpha/etc
* wrong usage of std::find
* more deprecated functions... read more

Posted by Daniel Marjamäki 2012-02-11


New checks

* wrong value passed to isgraph
* missuse of std::cout
* some warnings when sizeof is used on pointer variable
* endless loop because of unsigned comparison
* better check for missing break in switch (redundant strcpy)
* using && instead of & in expression
* wrong usage of printf/scanf/etc (format string and parameters mismatch)
* useless calls of STL functions (wrong usage)
* detect suspicicous use of semicolon after if/for/while (--inconclusive)
* added checking of readlink() usage
* boolean comparison with string literals... read more

Posted by Daniel Marjamäki 2011-12-10


New checks:
* actual initialization order of member variables is not the same as the written order. Use --inconclusive and --enable=style to enable this check.
* when first comparison is true, the 2nd comparison is always true. Example: '(x>5 && x!=1)'. Use --enable=style to enable this check.

we fixed many false positives and false negatives in existing checks.

For details, see this list of fixed tickets:

Posted by Anonymous 2011-10-08


There was many fixes of false positives and serious problems like crashes, hangs, etc.

New checks:

* Check for std::auto_ptr misuse (related to strict ownership)
* Read array and then immediately check if the array index is within limits
* assign pointer to int/long
* assign bool to pointer
* duplicate "break" statements in switch
* matching "if" and "else if" conditions when using bitwise and
* matching assigment and condition when using bitwise and
* test if unsigned value is less than zero

Posted by Anonymous 2011-08-14


Many bug fixes and improvements.

2 more checks were added:
* strcmp condition is always true
* using sizeof with constant

Posted by Anonymous 2011-06-12


Release notes for 1.48

There are no major new features in 1.48. There are more and better checks.

New check: Wrong usage of ! operator in conditions.
Example: if (!x == 0) {

New check: Use "throw" without arguments to rethrow exceptions.

New check: Comparison of substring with string literal will always/never match because size doesn't match.

New check: Postfix increment of boolean

New check: Clarify condition with parantheses (when there are assignment + comparison)
Example: if (a = b > 0) {... read more

Posted by Anonymous 2011-04-09


Release notes for 1.47

It is now possible to exclude files and folders from the analysis. Use -i on the command line (i=ignore).

Custom rules can now be created using regular expressions. To read more about creating custom rules, see http://sourceforge.net/projects/cppcheck/files/Articles/

A new XML format is launched. To use this format, the --xml-version=2 is used. The new xml format is incompatible with the xml format used in previous versions. The old XML format will still be used unless --xml-version=2 is given, so your existing scripts, tools and plugins should still work.... read more

Posted by Anonymous 2011-02-06


Release notes for 1.46

This release has improvements and bug fixes.

We fixed 153 tickets, and that is a somewhat "usual" number for a Cppcheck release.

The report has been improved. New severities were added to make the messages more informational. The possible severities are now:
* error
* warning
* style
* performance

This has no effect on the command line flags nor the xml report. The command line flags and the xml report is fully compatible with previous versions.... read more

Posted by Anonymous 2010-12-12



* project file

Command line:

* report: use system dependent path separator
* --enable: the except* is not available anymore
* --enable: the "unusedFunctions" was renamed to "unusedFunction"


* New check: Initialization of a variable by itself
* New check: dangerous usage of scanf
* More checking for obsolete and dangerous functions
* Optimisations
* C++0x updates: enums, strings... read more

Posted by Anonymous 2010-10-03


I broke out the headers checking into a separate project. Its primary purpose is to detect unnecessary includes.

The headers checking was among the first Cppcheck checks. It has been disabled for years because it doesn't work with the Cppcheck data.
Adapting the Cppcheck data for the headers checking would be complicated and it would involve disabling the Preprocessor - and that is not acceptable. Therefore it was broken out into a separate project.... read more

Posted by Anonymous 2010-07-26


* command line: a -D command was added
* GUI: minor updates

New checks:
* (style) Redundant assignment (missing break in a switch block)
* (style) Empty catch blocks
* (error) Usage of iterator pointing to invalid memory after reserve
* (error) Detect common realloc-mistake that can cause memory leaks.

Improved checks:
* Memory leaks
* Buffer overruns
* Mismatching allocation / deallocation
* Bad iterators
* Uninitialized variables
* Null pointer dereference
* Unused variable

Posted by Anonymous 2010-07-10


This release has many fixes and enhancements.

We deprecated the inconclusive checks. The "possible error" and "possible style" are not available anymore. From now on we will only work on fully conclusive checks.

Command line:
* Since we removed all inconclusive results, the "--all", "-a" and "--auto-dealloc" are deprecated. These flags can still be given but they don't have any effect.... read more

Posted by Anonymous 2010-05-08


Another hotfix release. This is more properly made than 1.41-2. With proper tagging and updating of all version numbers. But it has the same hotfix as 1.41-2

Posted by Anonymous 2010-03-09


A hotfix release

Posted by Anonymous 2010-03-09


This release has many bug fixes and minor improvements.

Command line:
* You can now specify files to check through a file instead of through command line

New checks:
* simple checking of function constness
* readded check for unused variable
* duplicate typedefs
* STL: suspicious condition when checking result of find

Core improvements:
* rewrote a few error messages, to make them easier to understand
* better handling of pointer aliasing
* improved handling of templates
* various improvements in the leaks checking
* many improvements in the enum handling
* improved handling of typedefs
* more checking for suspicious pointer arithmetic... read more

Posted by Anonymous 2010-03-07


Cppcheck-1.40 was released today

New checks:
* new style check: assignment operator does not return *this
* new style check: passing negative size to memset, etc
* new performance check: use empty() instead of checking size() against zero

Rewritten checks:
* rewrote the checking for null pointers
* rewrote the checking for uninitialized variables

Improved checks:
* improved the checking for invalid usage of auto variables
* improved the checking for leaks
* fixed a few checks so they don't bail out when finding an error - the checking continues and more errors may be detected.
* rewrote various error messages to make them easier to understand... read more

Posted by Anonymous 2010-01-17

cppcheck-1.40 delayed

Cppcheck-1.40 will not be released tomorrow as planned.

It will probably be released next weekend instead, on January 17th.

The reason for the delay is to allow us to make more improvements.

Posted by Anonymous 2010-01-09


Cppcheck 1.39 was released today.

Many fixes, refactorings and improvements.

New features:
* new flag --enable that you'll use to enable various checks. The --all and --style are deprecated
* check for uninitialized variable
* check for exception safety problems
* more checking of STL iterators

Posted by Anonymous 2009-12-06

cppcheck-1.39 delay

cppcheck-1.39 was planned to be released today.

But it is delayed. It will instead be released on december 6th.

Posted by Anonymous 2009-11-29

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.

No, thanks