Ben Saylor wrote:
> On Thursday 17 July 2003 1:47 am, Ben Saylor wrote:
> > I know you developers are overloaded, but here are a few things I've
> > discovered as I attempt to diagnose the problem..
> >
> > After installing the Audiophile once on a Windows machine, MIDI works
> > with 0.9.5, even if I power off and unplug the unit from USB and AC.
> > It seems that the firmware only has to be loaded once.
> >
> > I get sound (but always at 48khz) with 0.9.5 iff I do the following:
> > 1. make sure ALSA modules are unloaded
> > 2. unplug the audiophile from the computer
> > 3. plug the audiophile into the computer
> > snd-usb-audio and other modules are loaded automatically.
> > --However, MIDI doesn't work if I do this.
> >
> > If I load the ALSA modules any other way than the above, MIDI works,
> > but audio is either static or distorted. I also get a bunch of these
> > in /var/log/debug when I play a sound file (44100 Hz, 16 bit, stereo)
> > with aplay:
> >
> > Jul 17 01:14:32 purebox kernel: usb.c: usb-check-bandwidth would have
> > FAILED: was 938, would be 940, bustime = 2 us
> > ...
> > etc. running tail -f /var/log/debug shows that 10 of these lines are
> > spit out per second during playback. They don't appear if I let the
> > modules load automatically.
> >
> > Ben
>
> Another discovery and a temporary solution that works for me, at least..
>
> snd_usb_audio_probe() gets called twice, I'm guessing first for the
> audio interface and second for the midi interface.
yes
> I found that if snd-usb-audio is not loaded and I plug in the
> Audiophile and let hotplug take care of loading the modules (the
> only way audio will work), then usb_set_configuration() fails
> during the second call to snd_usb_audio_probe(), causing the probe
> to abort. So no midi.
usb_set_configuration() should never be called twice for the same
device.
> This patch causes usb_set_configuration() to be called only during the
> first probe. I have no idea what's really causing the problem, but for
> now this gives me audio and midi at the same time. :)
Your patch tries to do the right thing, but I don't think a static
variable named "frogbreath" would be suitable for inclusion in the
kernel (without a funny comment explaing its name. ;-)
Takashi, could you please move the call to usb_set_configuration()
(l.2726) to a place where it's called only for the first _probe()
call, e.g. into the "if (! chip) {" block at line 2756?
Regards,
Clemens
-------------------------------------------------------
This SF.net email is sponsored by: VM Ware
With VMware you can run multiple operating systems on a single machine.
WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the
same time. Free trial click here: http://www.vmware.com/wl/offer/345/0
_______________________________________________
Alsa-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/alsa-user
