Just attempted to build from master at github. A problem introduced in one of the July 26 commits?
[120/984] Run test ‘/Users/louis/bui...te/Frameworks/buffer/test_buffer’… FAILED: /Users/louis/build/TextMate/Frameworks/buffer/test_buffer && touch /Users/louis/build/TextMate/Frameworks/buffer/test_buffer.run test_buffer(10037,0x109f85000) malloc: *** error for object 0x7feb10d13ed0: pointer being freed was not allocated *** set a breakpoint in malloc_error_break to debug /bin/sh: line 1: 10037 Abort trap: 6 /Users/louis/build/TextMate/Frameworks/buffer/test_buffer ninja: build stopped: subcommand failed.
On 27 Jul 2013, at 13:42, Louis Zulli wrote:
Just attempted to build from master at github. A problem introduced in one of the July 26 commits?
[120/984] Run test ‘/Users/louis/bui...te/Frameworks/buffer/test_buffer’… FAILED: /Users/louis/build/TextMate/Frameworks/buffer/test_buffer […]
Is this reproducible? If so, please run test_buffer with gdb and get a stack dump.
Basically most tests are now running concurrently (and not on main thread). There might be some code somewhere that isn’t capable of doing that, but I haven’t found it yet — that is, I suspect it’s the use of NSSpellChecker (which the buffer tests do indirectly use) and I plan to move that to main thread, but currently there are things depending on the framework that have tests that can’t yet handle GCD code…
Anyway, long story short; try re-run the build, and most of the time it should work fine. If you can provide more details about what’s causing the issue, that’d be great.
-
Allan Odgaard -
Louis Zulli