On Thu, 23 May 2002, Takashi Iwai wrote: > > - this suggests autodetection should be possible. > > yes. but i think we can keep it as a module option, because, as > Thomas mentioned, there is almost no cards with dual codecs.
Sure. Agree. > oh, you hit a bug. > the second codec is not shown, because the primary and secondary codec > try to create the same file. > the attached patch will fix this problem. > the second codec will be shown in the file > /proc/asound/card0/ac97#0-1. Funny, what I can see now is pc145:/proc/asound/card0# ls -l total 69 -rw-r--r-- 1 root root 0 May 23 11:27 ac97#0 -rw-r--r-- 1 root root 0 May 23 11:27 ac97#0 -rw-r--r-- 1 root root 0 May 23 11:27 ac97#0regs -rw-r--r-- 1 root root 0 May 23 11:27 ac97#0regs But the source code looks good. > does alsamixer show entries for the second codec? > they must be there even without the patch below. Well, I have first to compile all libraries and utils and put it on my embedeed system. My choice is the moment is either: 1. backport your changes to ALSA5. 2. Port my application to ALSA9. This is VoIP application running in memory mapped mode, very close to the hardware. I am using things like mmap_control->status.frag_io on the play side (fragments are 80 samples long) to determine where to write the next soound data (1-2 fragments before the currently processed one). Also access to mmap_control->status.status is important. On the capture side I first prepare the buffer overwriting it with a "magick" values, and then read new samples as they are transferred by DMA to the buffer. In this way I can achieve 1-2 ms latency. I would prefer to port all to ALSA9, especially if the support for the driver will continue to be so good. I would prefer to use alsa-lib only to configure the driver and then operate directly on memory mapped buffers, having a really efficient way to determine which position the hardware processes at the very moment. ALSA5 interface was very good for it (OK, I had to patch alsa-lib to allow write access to the capture buffer). A question: will I get now 2 sound interfaces with 2 voices each, or rather a single interface with 4 voices? Please let me know if this discussion is getting off-topic on alsa-devel list. Best regards, -- Tomasz Motylewski _______________________________________________________________ Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm _______________________________________________ Alsa-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/alsa-devel
