date | project | content | link |
|---|
| 00:09 today | KDE | Commit by arichardson :: r 1296508 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/ (9 files in 4 dirs): ( link) Fix potential crash due to bad pointer | # |
| 00:09 today | KDE | Commit by arichardson :: r 1296507 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/ (3 files in 2 dirs): ( link) correctly increment address when reading array
Also improved debug messages | # |
| 00:09 today | KDE | Commit by arichardson :: r 1296506 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/array/ (4 files): ( link) ArrayData::setLength() now takes an unsigned int | # |
| 00:09 today | KDE | Commit by arichardson :: r 1296505 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/ (4 files in 3 dirs): ( link) make DataInformation::name() non-virtual and add setter | # |
| 19:07 yesterday | KDE | Commit by arichardson :: r 1296485 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/datainformation.cpp: ( link) Fix AdditionalData being deleted by mistake
Only one of those conditions has to be true, not all three | # |
| 17:23 yesterday | KDE | Commit by arichardson :: r 1296482 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/ (3 files): ( link) Fix crash when setting additionalData in DataInformation
The crash happened due to bad if scoping
Also add a new method unsetValidationError(), which is slightly more efficient that setValidationError(QString()) | # |
| 17:23 yesterday | KDE | Commit by arichardson :: r 1296480 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/ (4 files in 2 dirs): ( link) Fix memory leaks in tests | # |
| 17:23 yesterday | KDE | Commit by arichardson :: r 1296481 okteta/trunk/KDE/kdesdk/okteta/core/ (3 files): ( link) ByteArrayModel: use delete[] not delete
The buffer passed will most likely be allocated with new Byte[] therefore we must use delete[], Buffers allocated with malloc() already caused undefined behaviour before and using new Byte; makes no sense Therefore this change should be safe (found with valgrind) | # |
| 17:23 yesterday | KDE | Commit by arichardson :: r 1296479 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/script/scriptlogger.cpp: ( link) fix memory leak in ScriptLogger::clear() | # |
| 17:23 yesterday | KDE | Commit by arichardson :: r 1296478 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/ (27 files in 9 dirs): ( link) Move PrimitiveDataType enum + toString() functions to own file
Also create a wrapper struct to prevent implicit conversions This wrapper will be removed once okteta requires C++11 | # |
| 17:22 yesterday | KDE | Commit by arichardson :: r 1296477 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/script/ (scriptlogger.cpp scriptlogger.h): ( link) Fix crashes due to dangling pointer in ScriptLogger | # |
| 17:22 yesterday | KDE | Commit by arichardson :: r 1296476 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/ (3 files in 2 dirs): ( link) print error messages from parseStructureNames to stderr
If this was not done they would be lost. | # |
| 17:22 yesterday | KDE | Commit by arichardson :: r 1296475 okteta/trunk/KDE/kdesdk/okteta/ (12 files in 5 dirs): ( link) make Scriptlogger return a QDebug for easier message formatting | # |
| 03:01 yesterday | KDE | Commit by arichardson :: r 1296348 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/ (3 files in 2 dirs): ( link) return null when cast fails also in release mode
This fixes basicdatainformationtest when compiled in realease mode. Typical case of premature optimization. One virtual function call which just returns true/false + one boolean check will definitely not be a bottleneck. And since it catches potential programming errors better leave it like that.
BUG: 300527 | # |
| 03:01 yesterday | KDE | Commit by arichardson :: r 1296346 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/ (20 files in 6 dirs): ( link) remove AdditionalData from DataInformation public API
This simplifies all the code using it | # |
| 03:01 yesterday | KDE | Commit by arichardson :: r 1296347 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/ (4 files in 3 dirs): ( link) fix crash when no QScriptEngine is set
This only happens in structures without any dynamic elements | # |
| 03:00 yesterday | KDE | Commit by arichardson :: r 1296345 okteta/trunk/KDE/kdesdk/okteta/kasten/ (20 files in 9 dirs): ( link) Refactor Parsers to not modify any member variables and improve error messages. | # |
| 03:00 yesterday | KDE | Commit by arichardson :: r 1296344 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/ (16 files in 3 dirs): ( link) Use default typeName() implementation where applicable | # |
| 03:22 Tuesday | KDE | Commit by arichardson :: r 1296005 okteta/trunk/KDE/kdesdk/okteta/kasten/controllers/view/structures/datatypes/array/arraydatainformation.cpp: ( link) Fix assert failure due to wrong calculation
It is expected that primitive arrays always start at a full byte Here we mask the value with ~7 which should be slightly faster than subtracting (8 - bitOffset)
BUG: 300408 | # |
| 03:07 Tuesday | KDE | Commit by arichardson :: r 1296003 okteta/branches/KDE/4.8/kdesdk/okteta/kasten/controllers/view/structures/datatypes/array/arraydatainformation.cpp: ( link) Fix assert failure due to wrong calculation
It is expected that primitive arrays always start at a full byte Here we mask the value with ~7 which should be slightly faster than subtracting (8 - bitOffset)
BUG: 300408 | # |