[alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
how does this work?
an explicit explanation would be great, but i'd even settle for a vague handwavy one at this point.
At Tue, 27 Mar 2007 17:55:56 -0700, John Utz wrote:
how does this work?
an explicit explanation would be great, but i'd even settle for a vague handwavy one at this point.
Please elaborate the question...
Takashi
man, it sux to have to type in all the little '>' myself, outlook web access's non rfc compliant behavior diminishes my productivity...
-----Original Message----- From: alsa-devel-bounces@alsa-project.org on behalf of Takashi Iwai Sent: Wed 3/28/2007 3:11 AM To: John Utz Cc: alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
At Tue, 27 Mar 2007 17:55:56 -0700, John Utz wrote:
how does this work?
an explicit explanation would be great, but i'd even settle for a vague handwavy one at this point.
Please elaborate the question...
absolutely! sorry i wasnt clear enuf.
here is the next phase of the problem that i need to solve:
i am working on a space constrained embedded system. it will always have the same:
hardware (via8237rplus with a via vt1618 codec)
software (the audio player application that i maintain)
content format (the music get's encoded by my music programmer coworkers to a fixed format)
since i dont need a generic solution, it would be very helpful to me to not use alsa-lib and just use the alsa modules the way that our current version of the player uses /dev/dsp from the alsa oss-compat kernel modules.
i'd like to open and playback to /dev/snd/pcmC0D0p and /dev/snd/pcmC0D1p directly if it's a reasonable thing to do.
tnx!
johnu
Takashi
_______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
On 3/29/07, John Utz John.Utz@dmx.com wrote:
since i dont need a generic solution, it would be very helpful to me to not use alsa-lib and just use the alsa modules the way that our current version of the player uses /dev/dsp from the alsa oss-compat kernel modules.
i'd like to open and playback to /dev/snd/pcmC0D0p and /dev/snd/pcmC0D1p directly if it's a reasonable thing to do.
The alsa-lib <-> kernel interface is intentionally undocumented. The only officially supported interface to ALSA is via alsa-lib.
If you really refuse to use alsa-lib (it's not that big, I've used it in embedded projects), then just enable ALSA's OSS emulation and continue to use /dev/dsp.
Lee
At Thu, 29 Mar 2007 22:10:50 -0400, Lee Revell wrote:
On 3/29/07, John Utz John.Utz@dmx.com wrote:
since i dont need a generic solution, it would be very helpful to me to not use alsa-lib and just use the alsa modules the way that our current version of the player uses /dev/dsp from the alsa oss-compat kernel modules.
i'd like to open and playback to /dev/snd/pcmC0D0p and /dev/snd/pcmC0D1p directly if it's a reasonable thing to do.
The alsa-lib <-> kernel interface is intentionally undocumented. The only officially supported interface to ALSA is via alsa-lib.
Well, not completely intentional. But we _recommend_ alsa-lib as the primary accessor, and at least I have no interest in writing another accessor right now. The documentaion itself is always good ;)
Takashi
-----Original Message----- From: Takashi Iwai [mailto:tiwai@suse.de] Sent: Fri 3/30/2007 5:52 AM To: Lee Revell Cc: John Utz; alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
The alsa-lib <-> kernel interface is intentionally undocumented. The only officially supported interface to ALSA is via alsa-lib.
* Well, not completely intentional. But we _recommend_ alsa-lib as the * primary accessor, and at least I have no interest in writing another * accessor right now. The documentaion itself is always good ;)
which documentation? i thought that i'd looked at it all and i didnt get a sense of what i needed, did i miss something?
tnx!
johnu
* Takashi
At Fri, 30 Mar 2007 10:00:37 -0700, John Utz wrote:
-----Original Message----- From: Takashi Iwai [mailto:tiwai@suse.de] Sent: Fri 3/30/2007 5:52 AM To: Lee Revell Cc: John Utz; alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/ controlC0 and company? is this on the wiki somewhere?
The alsa-lib <-> kernel interface is intentionally undocumented. The only officially supported interface to ALSA is via alsa-lib.
- Well, not completely intentional. But we _recommend_ alsa-lib as the
- primary accessor, and at least I have no interest in writing another
- accessor right now. The documentaion itself is always good ;)
which documentation? i thought that i'd looked at it all and i didnt get a sense of what i needed, did i miss something?
Well, I should have written "the documentation would be good". It's a follow up to Lee's comment.
There is no official documentation for now indeed, but you can guess the necessary syscalls easily from asound.h.
Takashi
-----Original Message----- From: alsa-devel-bounces@alsa-project.org on behalf of Takashi Iwai Sent: Fri 3/30/2007 11:18 AM To: John Utz Cc: alsa-devel@alsa-project.org; Lee Revell Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
At Fri, 30 Mar 2007 10:00:37 -0700, John Utz wrote:
-----Original Message----- From: Takashi Iwai [mailto:tiwai@suse.de] Sent: Fri 3/30/2007 5:52 AM To: Lee Revell Cc: John Utz; alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/ controlC0 and company? is this on the wiki somewhere?
The alsa-lib <-> kernel interface is intentionally undocumented. The only officially supported interface to ALSA is via alsa-lib.
- Well, not completely intentional. But we _recommend_ alsa-lib as the
- primary accessor, and at least I have no interest in writing another
- accessor right now. The documentaion itself is always good ;)
which documentation? i thought that i'd looked at it all and i didnt get a sense of what i needed, did i miss something?
* * Well, I should have written "the documentation would be good". It's a * follow up to Lee's comment.
:-)
dang.
i was *really* hopeful there for a while!
* There is no official documentation for now indeed, but you can guess * the necessary syscalls easily from asound.h.
thankyou, i'll go look at that!
* Takashi _______________________________________________ Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
as usual, plz forgive my poor citing.
-----Original Message----- From: rlrevell@gmail.com on behalf of Lee Revell Sent: Thu 3/29/2007 7:10 PM To: John Utz Cc: Takashi Iwai; alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
Hello Lee;
Thankyou for taking the time to respond, i appreciate it!
On 3/29/07, John Utz John.Utz@dmx.com wrote:
since i dont need a generic solution, it would be very helpful to me to not use alsa-lib and just use the alsa modules the way that our current version of the player uses /dev/dsp from the alsa oss-compat kernel modules.
i'd like to open and playback to /dev/snd/pcmC0D0p and /dev/snd/pcmC0D1p directly if it's a reasonable thing to do.
* The alsa-lib <-> kernel interface is intentionally undocumented. The * only officially supported interface to ALSA is via alsa-lib.
you can still document things even if you dont support them!
most app writers will use the lib because it does a very good job of dealing with all the inscrutable mandatory details of getting sound out of an arbitrary device.
but it's not the best fit for everybody!
what if you have a very teensy tiny filesystem and you dont want to fill your limited space with text configuration files and executable code that reads the conf files and executes test calls that you as a developer already know the answer to because you are the guy that also wrote the driver for the only hardware that you will ever use?
The lib code is in front of me and i have a compiler and a debugger and i will figure it out in the fullness of time.
Make no mistake, it will be kind of a fun adventure!
But having fun is supposed to be just a side effect of doing a good job for my employer and to do a good job for my employer i need to meet a rather short deadline!
Thus i am executing my responsibility as an effective employee by asking for help on this matter.
* If you really refuse to use alsa-lib (it's not that big, I've used it * in embedded projects),
'refuse' is much too strong of a term!
i seek to minimize:
1. the filesystem space required 2. the build complexity involved in building our product.
i have to cross-compile *everything!* and currently i am the *only* guy in the whole company that's doing this job, i wish to provide as simple a build process as possible for my future self or my windows centric coworkers.
* then just enable ALSA's OSS emulation and continue to use /dev/dsp.
does /dev/dsp support use of 4 channel output? how would one create 4 individual mono streams with /dev/dsp?
i dont think this is possible. am i mistaken?
* Lee
At Fri, 30 Mar 2007 11:22:27 -0700, John Utz wrote:
as usual, plz forgive my poor citing.
-----Original Message----- From: rlrevell@gmail.com on behalf of Lee Revell Sent: Thu 3/29/2007 7:10 PM To: John Utz Cc: Takashi Iwai; alsa-devel@alsa-project.org Subject: Re: [alsa-devel] how does hw:0, 1 eventually get mapped to /dev/snd/controlC0 and company? is this on the wiki somewhere?
Hello Lee;
Thankyou for taking the time to respond, i appreciate it!
On 3/29/07, John Utz John.Utz@dmx.com wrote:
since i dont need a generic solution, it would be very helpful to me to not use
alsa-lib and just use the alsa modules the way that our current version of the player uses /dev/dsp from the alsa oss-compat kernel modules.
i'd like to open and playback to /dev/snd/pcmC0D0p and /dev/snd/pcmC0D1p directly if
it's a reasonable thing to do.
- The alsa-lib <-> kernel interface is intentionally undocumented. The
- only officially supported interface to ALSA is via alsa-lib.
you can still document things even if you dont support them!
most app writers will use the lib because it does a very good job of dealing with all the inscrutable mandatory details of getting sound out of an arbitrary device.
but it's not the best fit for everybody!
what if you have a very teensy tiny filesystem and you dont want to fill your limited space with text configuration files and executable code that reads the conf files and executes test calls that you as a developer already know the answer to because you are the guy that also wrote the driver for the only hardware that you will ever use?
The lib code is in front of me and i have a compiler and a debugger and i will figure it out in the fullness of time.
I know of demands on recent many embedded systems. Now alsa-lib can be built in fairly small size by choosing appropriate components via configure options. Of course, it would be more smaller if the interface is rewritten somehow, though.
Make no mistake, it will be kind of a fun adventure!
But having fun is supposed to be just a side effect of doing a good job for my employer and to do a good job for my employer i need to meet a rather short deadline!
Thus i am executing my responsibility as an effective employee by asking for help on this matter.
- If you really refuse to use alsa-lib (it's not that big, I've used it
- in embedded projects),
'refuse' is much too strong of a term!
i seek to minimize:
- the filesystem space required
- the build complexity involved in building our product.
i have to cross-compile *everything!* and currently i am the *only* guy in the whole company that's doing this job, i wish to provide as simple a build process as possible for my future self or my windows centric coworkers.
- then just enable ALSA's OSS emulation and continue to use /dev/dsp.
does /dev/dsp support use of 4 channel output? how would one create 4 individual mono streams with /dev/dsp?
i dont think this is possible. am i mistaken?
Not impossible but much harder if you cope with ALSA drivers... At least, the current OSS emulation won't work easily for such a purpose as it is.
Takashi
participants (3)
-
John Utz -
Lee Revell -
Takashi Iwai