Re: Wine users with a CD-ROM drive, please test
From: <Joerg-Cyril.Hoehle(a)t-systems.com>
this is my second call for participation, as only Jeff Zaroyko responded to the first one.
I wrote a testsuite for mcicda (play music from CD-ROM using the MCI). Wine's testbot job #4330 at https://testbot.winehq.org/JobDetails.pl?Key=4330 contains both a patch and binary of what I'd initially want to test about mcida.dll, solely lacking compatibility fixes that testing on YOUR machine will show to be necessary.
I can attach a .iso to each VM and run your tests. Would that help? Keep in mind that almost none of the VMs have sound support. Ge.
Hi, David c/o paulo lesgaz wrote:
mcicda.c:425: Test failed: SEEK to 00203A43 position last + length: MCIERR_HARDWARE mcicda.c:436: Test failed: PLAY from 1 notify: MCIERR_HARDWARE Same as Jeff, please see if one of you can identify which prior MCI command causes this.
mcicda.c:275: Test failed: status position initially 00:02:33 mcicda.c:280: Test failed: status position start 00:02:33 Interesting, thank you. Will change the tests. I once read there must be a 2 second gap but it seems it need not be exactly 2 seconds.
Greg Geldorp wrote:
I can attach a .iso to each VM and run your tests. Would that help? AFAIK, a .iso contains a data CD only, not music tracks. If that's right, then it would not be very interesting, as data-only CDs are already skipped.
Keep in mind that almost none of the VMs have sound support. Thanks for reminding me to temporarily disable sound on a native system and see how that changes test results. From what I remember, disabling sound in the preferences (XP then wants to reboot) achieved the same results as sound-less machines in the MCI tests.
Regards, Jörg Höhle
Test results from Windows 7 x64: I tried with original audio CD, and a burned copy, mcicda_test.exe and mcicda_test64.exe, and they all give: mcicda.c:270: CD length 46:20:53 mcicda.c:325: track #1 length 118107ms mcicda.c:392: Tests skipped: Got no mixed data+audio CD. mcicda.c:410: last track length: 04m:15s:37frames mcicda.c:418: last track position: 42m:07s:16frames mcicda.c:425: Test failed: SEEK to 0035162E position last + length: MCIERR_HARDWARE mcicda.c:436: Test failed: PLAY from 1 notify: MCIERR_HARDWARE mcicda.c:439: Tests skipped: Cannot manage to play track 1. mcicda: 64 tests executed (0 marked as todo, 2 failures), 2 skipped.
David c/o paulo lesgaz wrote:
mcicda.c:425: Test failed: SEEK to 00203A43 position last + length: MCIERR_HARDWARE mcicda.c:436: Test failed: PLAY from 1 notify: MCIERR_HARDWARE Same as Jeff, please see if one of you can identify which prior MCI command causes this.
Do you have any instructions how I can identify this?
Hi, Thank you for your test results so far.
mcicda.c:425: Test failed: SEEK to 00203A43 position last + length: MCIERR_HARDWARE mcicda.c:436: Test failed: PLAY from 1 notify: MCIERR_HARDWARE Same as Jeff, please see if one of you can identify which prior MCI command causes this. Do you have any instructions how I can identify this?
This sequence of commands worked for Jeff: open cdaudio alias c status c type track 1 set c time format tmsf seek c to start play c from 1 to 2 wait The sequence of commands that the tests use is below. I don't see anything there that explains MCIERR_HARDWARE. Perhaps it's a timing issue and the suite would work after adding a couple of Sleep()? Copy&Paste of this sequence into the mcishell (see bug 20232, comment #10) works for me. Does it generate MCIERR_HARDWARE on your machine? Can you try and eliminate commands until you reach a minimal set that generates the error (start with removing capability, info and notify)? open cdaudio alias c notify shareable capability c has video notify capability c can play capability c capability c 16385 capability c device type status c time format info c product wait notify status c media present status c mode status c mode status c ready info c product identity info c identity info c upc status c number of tracks status c length status c position status c position start notify status c position start track 1 notify play c from 00:02:00 to 00:01:00 notify resume c seek c wait seek c to start notify # # Perhaps use seek c to start notify wait set c video audio all off set c time format ms status c position start status c position start track 1 status c notify wait status c length track 1 set c door closed notify status c type track 1 # # skipping data track tests set c time format msf status c length track 20 # # Replace 20 with number of tracks here status c position track 20 # # Perhaps add status c mode (still seeking?) seek c to 24:55:74 # # Replace with position last track start + length seek c to 24:56:00 set c time format tmsf play c from 1 Thank you for your help, Jörg Höhle
Jörg-Cyril : may be the driver doesn't support TMSF, and from/to fields must be provided in MSF would you mind trying that ? A+ -- -- Eric Pouech
Hi, Eric Pouech wrote:
may be the driver doesn't support TMSF, and from/to fields must be provided in MSF would you mind trying that ? Me? The others ;) The 5 machines I've had access to all pass the tests.
I somehow doubt it'll help because Jeff Zaroyko wrote:
set c time format tmsf seek c to start play c from 1 to 2 wait This works - I hear music. So TMSF works for Jeff, even though the testsuite yields MCIERR_HARDWARE there. Furthermore the failing Seek before Play uses MSF in the proposed testsuite.
Nevertheless worth a try. Any hint may work. Perhaps also add the "wait" flag to the tests: seek c to start notify WAIT Perhaps the cause is one of the unusual commands: set c video audio all off set c door closed notify # # Perhaps add status c mode after that? On some machines, inserting an audio CD starts the media player. Perhaps this interacts with the tests? Thank you for your help, Jörg Höhle
Hi, http://testbot.winehq.org/JobDetails.pl?Key=4395 contains a new patch and binary where I eliminated a few commands that might cause the MCIERR_HARDWARE observed with the second Seek and first Play, as well as a few more calls to "status c mode" to verify the current state. Thank you for testing, Jörg Höhle
You will attached the results of the tests on my real vista box As the last time, I used the original CD of Police: Greatest hits David ________________________________ De : "Joerg-Cyril.Hoehle(a)t-systems.com" <Joerg-Cyril.Hoehle(a)t-systems.com> À : wine-devel(a)winehq.org Cc : tijnema(a)gmail.com; jeffzaroyko(a)gmail.com; jeremielapuree(a)yahoo.fr; knan-wine(a)anduin.net Envoyé le : Mar 10 août 2010, 17h 50min 39s Objet : Wine users with a CD-ROM drive, please test Hi, http://testbot.winehq.org/JobDetails.pl?Key=4395 contains a new patch and binary where I eliminated a few commands that might cause the MCIERR_HARDWARE observed with the second Seek and first Play, as well as a few more calls to "status c mode" to verify the current state. Thank you for testing, Jörg Höhle
Hi, testbot.winehq.org job #4404 now contains another patch and binary. I removed the call to set c door closed estimating that this was the most likely among the three suspects to cause the error (I also reduced the Sleep() time). Please test yet once again. I've not changed anything about the 00:02:33 start position which I'll leave in until testbot gathers more data from users. Thank you very much, Jörg Höhle
On Wed, 11 Aug 2010, Joerg-Cyril.Hoehle(a)t-systems.com wrote:
Hi,
testbot.winehq.org job #4404 now contains another patch and binary. I removed the call to set c door closed estimating that this was the most likely among the three suspects to cause the error (I also reduced the Sleep() time).
Yay, no more MCIERR_HARDWARE! Some others, though. Win7 64bit. Z:\test>mcicda_test_4404.exe mcicda.c:287: CD length 60:42:06 mcicda.c:351: track #1 length 245334ms mcicda.c:423: Tests skipped: Got no mixed data+audio CD. mcicda.c:441: last track length: 06m:46s:25frames mcicda.c:449: last track position: 53m:57s:56frames mcicda.c:541: Test failed: status mode after play is playing mcicda.c:548: Test failed: status mode after play is playing mcicda: 105 tests executed (0 marked as todo, 2 failures), 1 skipped.
Please test yet once again. I've not changed anything about the 00:02:33 start position which I'll leave in until testbot gathers more data from users.
I had that one too, for one out of five test cds or so. -- -erik http://useofwords.blogspot.com/
Hi,
mcicda.c:541: Test failed: status mode after play is playing mcicda.c:548: Test failed: status mode after play is playing mcicda: 105 tests executed (0 marked as todo, 2 failures), 1 skipped.
Here I don't know what to do. Even after the former Sleep(3900ms), status mode reported playing. Perhaps some drives just enjoy spinning? If you look at the source, you'll see it does set c time format tmsf play c from 1 to 1:00:00:21 notify Sleep status c mode Within almost 4 seconds time, I'd have expect the drive to spin up, play for 280ms and stop again. I'm tempted to a) leave the test as is, so that test.winehq.org will slowly show how often this happens among users. Or do people think I should rather b) silence the error and don't care. After all, it tells me nothing about how to implement. BTW, in the next version, the 00:02:33 start position will not be flagged as error any more. Regards, Jörg Höhle
Hi, I found out that some machines ignore set c audio all off which was in the testsuite, while others react to it and even update the system preferences volume mixer control if it happens to be open. I don't know whether that difference depends on the HW (CD-ROM driver) or the OS. You can find it out by posting your results here. I'm now using set c audio all on (which should be the default) such that all users running the testsuite get to hear something. https://testbot.winehq.org/JobDetails.pl?Key=4434 mcicda.c:612: Test failed: STOP all returned MCIERR_HARDWARE (I fixed that, but not yet on testbot). Regards, Jörg Höhle
On Fri, Aug 13, 2010 at 11:03 PM, <Joerg-Cyril.Hoehle(a)t-systems.com> wrote:
I'm now using set c audio all on (which should be the default) such that all users running the testsuite get to hear something.
https://testbot.winehq.org/JobDetails.pl?Key=4434 mcicda.c:612: Test failed: STOP all returned MCIERR_HARDWARE (I fixed that, but not yet on testbot).
I hear the playback, no more failures here with the same media as tested before, except for the one you mention 612 which only occurred with a blank cd. Nice work!
On Fri, Aug 13, 2010 at 3:03 PM, <Joerg-Cyril.Hoehle(a)t-systems.com> wrote:
Hi,
I found out that some machines ignore set c audio all off which was in the testsuite, while others react to it and even update the system preferences volume mixer control if it happens to be open.
I don't know whether that difference depends on the HW (CD-ROM driver) or the OS. You can find it out by posting your results here.
I'm now using set c audio all on (which should be the default) such that all users running the testsuite get to hear something.
https://testbot.winehq.org/JobDetails.pl?Key=4434 mcicda.c:612: Test failed: STOP all returned MCIERR_HARDWARE (I fixed that, but not yet on testbot).
Regards, Jörg Höhle
That fixed playback. One thing, with my Beethoven CD: mcicda.c:300: Test failed: status position initially 00:03:00 mcicda.c:306: Test failed: status position start 00:03:00 mcicda.c:350: Test failed: status position initially 3001ms IIRC, it can be anything, but should atleast be 2 seconds. - Matijn
Matijn Woudt wrote:
On Fri, Aug 13, 2010 at 3:03 PM, <Joerg-Cyril.Hoehle(a)t-systems.com> wrote:
Hi,
I found out that some machines ignore set c audio all off which was in the testsuite, while others react to it and even update the system preferences volume mixer control if it happens to be open.
I don't know whether that difference depends on the HW (CD-ROM driver) or the OS. You can find it out by posting your results here.
I'm now using set c audio all on (which should be the default) such that all users running the testsuite get to hear something.
https://testbot.winehq.org/JobDetails.pl?Key=4434 mcicda.c:612: Test failed: STOP all returned MCIERR_HARDWARE (I fixed that, but not yet on testbot).
Regards, Jörg Höhle
That fixed playback. One thing, with my Beethoven CD: mcicda.c:300: Test failed: status position initially 00:03:00 mcicda.c:306: Test failed: status position start 00:03:00 mcicda.c:350: Test failed: status position initially 3001ms
IIRC, it can be anything, but should atleast be 2 seconds.
It is. 3001 ms is one ms more than three seconds.... James McKenzie
participants (8)
-
Eric Pouech -
Erik Inge Bolsø -
Greg Geldorp -
James McKenzie -
Jeff Zaroyko -
Joerg-Cyril.Hoehle＠t-systems.com -
Matijn Woudt -
paulo lesgaz