| Fixed spurious exception throw when a QSSLDevice was opened and shut without transferring any data. Also fixed an earlier botched replacement of the atomic integer implementation on POSIX - this fixes all but one test in the test suite on POSIX. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Added QThread::isBeingCancelled(). |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Misc changes to the build config |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Added a shell script to install dev libraries for you on Ubuntu/Debian. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Updated test results. Now fully working on VS2010 on Windows x64 (at long last!). Will start work on Linux tomorrow. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Fixed race condition in QChildProcess where both pipes weren't being fully drained on QIODeviceS::waitForData() - this could cause one pipe to stall. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Stunningly enough, QPipe - the oldest code in TnFOX not touched since 2002 - somehow STILL had a bug in it. Due to Windows pipes being badly implemented, you can't wait on them to signal like everything else in Windows or POSIX so you have to run them via async i/o. This means leaving a null byte read running most of time so the event object is correctly anchored to there being data available. Somehow or other I managed to delete the crucial line which fires this null byte async read during my last surgery on QPipe, so occasionally QIODeviceS::waitForData() was returning when there was no data, thus causing a hang during QPipe::readBlock(). |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Little misc fixes |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| The TestDLL test had stopped building itself correctly ages ago. Fixed now. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Fixed a long standing bug in AllTests, the automated test runner, whereby it choked if the test apps outputted more data than fitted into pipe's buffer. A handy side effect is that running the automated test suite is now VERY considerably faster! |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Restored the non-SSE based version of QHostAddress as the SSE optimised version suffers from alignment issues. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Fixed Visual Studio 2010 support (it was very broken) as well as fixing compilation errors on POSIX for GCC v4.4.5. For some odd reason there is still a segfault on process init on Linux whereas on Windows there are checked iterator exceptions tripping due to new checking in VS2010. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Fixed bad tabbing in last commit |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Added note about GIT clone being weird and not fetching submodules automatically without extra command line options. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Generic fxbitscanr() just didn't work :(. Fixed, plus added 64 bit version. |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
over 2 years ago
|
| Fixed up CRLF |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
almost 3 years ago
|
| Fixed nedmalloc not being included from github |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
almost 3 years ago
|
| Added CLOCK_MONOTONIC based POSIX nanosecond timer support to FXProcess::getNsCount(). |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
almost 3 years ago
|
| Hatched nedmalloc out into its own GIT repo (at long last! And thanks EVER so much to the git subtree split tool at http://github.com/apenwarr/git-subtree/) |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
almost 3 years ago
|
| Deleted nedmalloc from TnFOX |
Niall Douglasas Niall Douglas (s [underscore] sourceforge {at} nedprod [dot] com)
|
More...
|
almost 3 years ago
|