shell32/brsfolder.c:225: Test failed: The new folder did not get the name foo

[Michael Mc Donnell]

On Sun, Oct 16, 2011 at 8:40 PM, Dan Kegel <dank at kegel.com> wrote:
> On Tue, Oct 11, 2011 at 2:08 PM, Michael Mc Donnell
> <michael at mcdonnell.dk> wrote:
>> On Sun, Oct 9, 2011 at 11:02 AM, Dan Kegel <dank at kegel.com> wrote:
>>> The buildbot noticed this failure on 64 bits with git today:
>>>
>>> brsfolder.c:225: Test failed: The new folder did not get the name foo
>>> brsfolder.c:232: Test failed: SHBrowseForFolder did not return the
>>> pidl for the new folder. Expected
>>> 'Z:\home\bob\wineslave.dir\sandbox\slave\runtests-default-x86_64\build\dlls\shell32\tests\test_click_make_new_folder_button\foo'
>>> got 'Z:\home\bob\wineslave.dir\sandbox\slave\runtests-default-x86_64\build\dlls\shell32\tests\test_click_make_new_folder_button\New
>>> Folder'
>>
>> The test_click_make_new_folder_button test might be timing sensitive
>> because it simulates user input to a dialog via a timer. It presses a
>> button, selects text, inputs new folder name, and finally closes
>> dialog. I suspect that it might fail if the machine is heavily loaded
>> or the focus is lost to another program. That could result in the the
>> new folder name not being inputted, which would explain why the folder
>> does not get the new name.
>>
>> I think one possible remedy would be to somehow pump the message queue
>> instead of using a timer. I'm not exactly sure how to do this, and I'm
>> unfortunately too busy with my thesis to look into this.
>>
>>> This is the first time the buildbot has ever seen this failure, I think.
>>> http://test.winehq.org/data/tests/shell32:brsfolder.html
>>> shows crashes for this test on NT, but no other failures.
>>
>> I think this might be a separate issue because it seems to happen in
>> the test_selection. I didn't write this test, so I don't have a good
>> idea what might be the problem. That part was written by André
>> Hentschel so he might have a good idea (I've CC'ed him).
>
> It might be worse than that.  There are valgrind errors
> ( see http://bugs.winehq.org/show_bug.cgi?id=26074
> and http://bugs.winehq.org/show_bug.cgi?id=28737 ),
> and just now I saw a 64 bit crash in brsfolder,
> http://buildbot.kegel.com:8010/builders/runtests-default-x86_64/builds/521/steps/shell_3/logs/stdio
>
> I'm blacklisting shell32/brsfolder as flaky for now.
> - Dan

That's a good idea. Thanks for opening those bug reports and running
valgrind. Also, it's nice that the buildbot is catching these subtle
bugs.