On Thu, 15 Nov 2018, Gabriel Ivăncescu wrote:
I don't know where these failures are from, my patches only update listbox.c, they're not even from the same file nor have anything to do with synchronization or w/e.
The patch series modifies dlls/user32/listbox.c, that is user32.dll, which, when running in Wine, could impact any of the user32 tests (*). So to make sure we don't miss new failures, when a patch modifies a dll the TestBot now reruns all of that dll's tests.
Of course it ignores existing errors but this is defined as errors that happened in the last WineTest run. But in this case the "WaitForSingleObject" and "destroy child on thread exit" errors are somewhat random and did not happen in the last WineTest run.
For the standard 32 bit tests, I find them on the 12th but not on the 13th or 14th:
12: https://test.winehq.org/data/ebae298aa4d2711fef35d4ac60c6012438f36d61/linux_... 13: https://test.winehq.org/data/ead7e637c0d18760acd446d686ad18526e76e0f0/linux_... 14: https://test.winehq.org/data/d9c7d4147b569553bc97ef57c6200002fe81565e/linux_...
Random errors are the bane of WineTest :-(
(*) It could even break things in other dlls in fact.