Unknown issue with test.winehq.org

[Paul Vriens]

Alexandre Julliard wrote:
> Paul Vriens <paul.vriens.wine at gmail.com> writes:
> 
>> exception: 42 tests executed (0 marked as todo, 0 failures), 0 skipped.
>> exception: 279 tests executed (0 marked as todo, 5 failures), 0 skipped.
>>
>> The first one triggers the 'end of this test' in dissect. The second
>> one is thus not accepted as dissect expects a 'start' line (and a line
>> in between actually).
>>
>> The question now is: should dissect handle these cases or is this
>> particular test-case wrong?
> 
> It's because some of the tests are run in a child process. dissect
> should probably handle it, ideally by reporting the sum of the test
> counts as the results for the whole test.
> 
So what's the difference between this one and for example the following ones 
running in child processes:

kernel32:process start dlls/kernel32/tests/process.c 1.10
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/process.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
process: 396 tests executed (0 marked as todo, 6 failures), 0 skipped.

or:

msvcrt:file start dlls/msvcrt/tests/file.c 1.45
tests/file.c: 3 tests executed (0 marked as todo, 0 failures), 0 skipped.
tests/file.c: 1 tests executed (0 marked as todo, 0 failures), 0 skipped.
file: 361 tests executed (0 marked as todo, 0 failures), 0 skipped.

or better yet:

shell32:shellpath start dlls/shell32/tests/shellpath.c 1.16
tests/shellpath.c: 10 tests executed (0 marked as todo, 4 failures), 0 skipped.
tests/shellpath.c: 4 tests executed (0 marked as todo, 0 failures), 0 skipped.
shellpath: 452 tests executed (0 marked as todo, 0 failures), 0 skipped.

Obviously the start of line is a filename for these ones, whereas the start of 
the line for the exception one is the test itself.

Should we consolidate the above ones as well? I think yes, cause the shell32 
example is reflected as 0 errors on test.winehq.org.

Cheers,

Paul.