Subject: Re: [linux-audio-dev] RME HDSP output
From: Paul Davis (pbd_AT_op.net)
Date: Wed Jun 26 2002 - 17:45:22 EEST
>Ok here is some really weird behavior. While playing with some sessions
>last night, I ended up loading the HDSP module first, then my ice1712
>module. I needed the ice1712 module to set the clock of the Delta1010,
>to sync to wordclock, so the 1010ai box would sync to the adat in. And
>darned if the output from the HDSP started working. Not really sure why.
>It happened at the end of the night, so I didn't get much time to play
>with it, so there maybe some other issues. Ill be doing some tracking
>tonight, so I will see what I can find. Any suggestions on what to look
>at, to discover why this happens?
i don't have any clues, except that i am now remembering that this
happened to me as well the "first" time i got things to work. it came
out of the blue, and then it kept working until last week. its really
puzzling.
i would try to recreate what you did, after a cold boot and
disconnecting the i/o box (then reconnecting. test it each step of the
way. the usual, in other words. maybe disable the sound startup and do
the modules 1 by 1 by hand (the hammerfall-mem module can still go in
early - it doesn't matter, i think).
>I tried the "insmod snd-hdsp.o snd_enable=yes" trick, but it didn't
>work. In fact the snd_enable option only took 0 or 1 as values. Also
>when loaded, as stated previously, aplay and jack worked fine, no reload
>of the module was necessary.
if the i/o box has not been disconnected from the host, the firmware
stays loaded, so the second time the module is loaded (even across
reboots of the computer), things work correctly. as i said, i've fixed
this bug - it was a clear error in the code.
--p
This archive was generated by hypermail 2b28 : Wed Jun 26 2002 - 17:58:00 EEST