Francois Gouget fgouget@codeweavers.com writes:
This way we exercise almost as much of the API as in interactive mode without disturbing users, and the tests still run fast. The slowest one is the winmm test which takes 16 seconds on my machine which supports all the tested sound formats (but has only one sound card). Given that winetest's timeout is two minutes this gives a wide margin.
The capture test takes about 80 seconds here with the ALSA driver, which is a bit long. It also fails with the OSS driver:
capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=512, should be 800 capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=512, should be 800 capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=512, should be 800 capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=1024, should be 1600 capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags etc.
Also the wave test makes a lot of noise here, both with ALSA and OSS drivers.
On Fri, 6 Aug 2004, Alexandre Julliard wrote: [...]
The capture test takes about 80 seconds here with the ALSA driver, which is a bit long.
That's strange. Could you run it (with or without the patch but without the -q flag) and send me the output? I guess it's either that Alsa has multiple capture devices or that there's a bug somewhere.
If needed we can still speed up the test by reducing the capture time, e.g. from 0.1 second to 0.05 seconds.
It also fails with the OSS driver:
capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=512, should be 800
[...]
This looks like a bug that needs to be fixed. To confirm, could you run the test in interactive mode without the patch? If you get the same errors then it means there's a bug in wineoss.
[...]
Also the wave test makes a lot of noise here, both with ALSA and OSS drivers.
Yes it makes some sort of clicking noises here too, but iirc only in Alsa or OSS, not both (I don't remember which and my tree is too hacked to test right now). I'll recheck on Windows, and then in Wine. Currently I suspect it's a bug in one of: wineoss, the OSS driver, winealsa or the Alsa driver.
Francois Gouget fgouget@codeweavers.com writes:
That's strange. Could you run it (with or without the patch but without the -q flag) and send me the output? I guess it's either that Alsa has multiple capture devices or that there's a bug somewhere.
If needed we can still speed up the test by reducing the capture time, e.g. from 0.1 second to 0.05 seconds.
This is what I get without the patch:
capture.c:341:found 1 WaveIn devices capture.c:265: 0: "SB16 Wave In" (winealsa: default) 1.0 (2:260): channels=2 formats=fffff capture.c:310:Testing invalid format: 2 MHz sample rate capture.c:265: WAVE_MAPPER: "Wine wave in mapper" (not supported) 0.1 (255:1): channels=2 formats=fffff capture.c:310:Testing invalid format: 2 MHz sample rate capture.c: 1375 tests executed, 0 marked as todo, 0 failures.
With the patch I get a bunch of "Recording for 0.1 second" traces that seem to appear at about 1 second intervals, so I guess the problem is that the specified duration is not handled correctly.
It also fails with the OSS driver: capture.c:128: Test failed: WaitForSingleObject failed for header capture.c:129: Test failed: WHDR_DONE not set in frag.dwFlags capture.c:130: Test failed: frag.dwBytesRecorded=512, should be 800
[...]
This looks like a bug that needs to be fixed. To confirm, could you run the test in interactive mode without the patch? If you get the same errors then it means there's a bug in wineoss.
Yes, I get the same errors without the patch.
-
Alexandre Julliard -
Francois Gouget