#4606 closed Bugs (fixed)
[flyweight] test_intermod_holder regression test failure on Windows
| Reported by: | Owned by: | Ion Gaztañaga | |
|---|---|---|---|
| Milestone: | To Be Determined | Component: | flyweight |
| Version: | Boost Development Trunk | Severity: | Problem |
| Keywords: | Cc: |
Description
When running the regression tests on Windows XP i see the test_intermod_holder.exe regression test process hanging around (it doesn't complete, and the test system doesn't terminate it).
If i run the test in the VC9 debugger, it goes into the call to pIWbemLocator->ConnectServer in get_wmi_class_attribute (win32_api.hpp) and never returns (at least not in a reasonable timeframe). All of the Windows based test runners seem to have the same issue.
I don't know why this happens though.
Change History (3)
comment:1 by , 12 years ago
| Component: | flyweight → interprocess |
|---|---|
| Owner: | changed from to |
comment:2 by , 12 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Updated Interprocess so that intermodule_singleton uses windows native shared memory.
comment:3 by , 12 years ago
| Component: | interprocess → flyweight |
|---|
Changed the component back to flyweight, since this most accurately matches with the ticket's original description and further development. As of changeset 65927 intermodule_holder is not based on boost::interprocess::named_mutex and relies instead on Boost.Interprocess internal intermodule_holder, which is not affected by the hanging problem on DLL loading time. Thanks to Ion for his work on solving this out.
Changed component to interprocess, as per discussed in http://thread.gmane.org/gmane.comp.lib.boost.devel/208308