I know OSS will be replaced with ALSA soon, but i've got a couple of OSS
cleanup patches lined up (module usage count, power management patches
for two cards) for both 2.4.x and 2.5.x, should i continue with them or is
it not worthwhile?
Comments appreciated.
Regards,
Zwane Mwaikambo
> I know OSS will be replaced with ALSA soon, but i've got a couple of OSS
> cleanup patches lined up (module usage count, power management patches
> for two cards) for both 2.4.x and 2.5.x, should i continue with them or is
> it not worthwhile?
Well if you've done the work why not - people will be running 2.4 for a long
time. The PM changes may also be relevant to ALSA anyway
Hi
Out of curiosity, will ALSA also be available on the 24..xx kernels??
Will there be a choice of useing OSS or ALSA??
Will the ALSA drivers be the 0.9 series or the old 0.5 series??? AFAIK they
are very very different in architecture.
I welcome the use of ALSA, as it appears to be a more flexibile solution and
can be used with OSS compatibility from an application point of view. My only
concern is that there is a potential for looseing support for some sound
cards.
John
On December 3, 2001 04:11 am, Zwane Mwaikambo wrote:
> I know OSS will be replaced with ALSA soon, but i've got a couple of OSS
> cleanup patches lined up (module usage count, power management patches
> for two cards) for both 2.4.x and 2.5.x, should i continue with them or is
> it not worthwhile?
>
> Comments appreciated.
>
> Regards,
> Zwane Mwaikambo
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
Zwane Mwaikambo wrote:
> I know OSS will be replaced with ALSA soon, but i've got a couple of OSS
> cleanup patches lined up (module usage count, power management patches
> for two cards) for both 2.4.x and 2.5.x, should i continue with them or is
> it not worthwhile?
IMHO they are useful for 2.5 series, perhaps backported to 2.4 later.
I'm sure people will continue using the OSS drivers even after they
become the "old" sound drivers... for a while at least.
Eventually we will have a pleasant future where we have a sane API that
is scalable for expensive sound hardware, with kernel changes mitigated
by a central userspace kind of libc-for-sound. a.k.a. alsa and
alsa-lib.
Also FWIW, I support getting ALSA into the 2.5 kernel sooner rather than
later. I had hoped to review alsa before it got in, but that kept
getting put off.
Getting ALSA into the kernel in early 2.5 will allow for a longer period
of public review, and a longer period to smooth out the rough spots and
finalize the kernel<->userspace interface.
I have style quibbles with ALSA but the ALSA guys much to their credit
are very responsive and tackle technical issues as soon as they are made
aware. And having alsa-lib sit between kernel and userspace is the
correct and right idea, opening up all sorts of new possibilities for
the future of Linux audio.
Jeff
--
Jeff Garzik | Only so many songs can be sung
Building 1024 | with two lips, two lungs, and one tongue.
MandrakeSoft | - nomeansno
On Mon, 3 Dec 2001, John Gluck wrote:
> Hi
>
> Out of curiosity, will ALSA also be available on the 24..xx kernels??
> Will there be a choice of useing OSS or ALSA??
> Will the ALSA drivers be the 0.9 series or the old 0.5 series??? AFAIK they
> are very very different in architecture.
>
> I welcome the use of ALSA, as it appears to be a more flexibile solution and
> can be used with OSS compatibility from an application point of view. My only
> concern is that there is a potential for looseing support for some sound
> cards.
I doubt ALSA will get into 2.4 since its maintenance only, but i'm not the
final authority on this ;) Also when ALSA starts getting incorporated
into the kernel, they will use their more upto date tree, i would
presume, so they wouldn't start at 0.5. But as everyone else says, i'm
sure we'll be backporting fixes and perhaps even additional card support
into 2.4-OSS as they appear.
Cheers,
Zwane Mwaikambo
Zwane Mwaikambo wrote:
>
> On Mon, 3 Dec 2001, John Gluck wrote:
>
> > Hi
> >
> > Out of curiosity, will ALSA also be available on the 24..xx kernels??
> > Will there be a choice of useing OSS or ALSA??
> > Will the ALSA drivers be the 0.9 series or the old 0.5 series??? AFAIK they
> > are very very different in architecture.
> >
> > I welcome the use of ALSA, as it appears to be a more flexibile solution and
> > can be used with OSS compatibility from an application point of view. My only
> > concern is that there is a potential for looseing support for some sound
> > cards.
>
> I doubt ALSA will get into 2.4 since its maintenance only, but i'm not the
> final authority on this ;) Also when ALSA starts getting incorporated
> into the kernel, they will use their more upto date tree, i would
> presume, so they wouldn't start at 0.5. But as everyone else says, i'm
> sure we'll be backporting fixes and perhaps even additional card support
> into 2.4-OSS as they appear.
IMHO ALSA should -never- go into 2.4. It's fine as a patch but 2.5 is
the time for big merges, and since it's already available for 2.4
outside the kernel there shouldn't be any need for backporting
Jeff
--
Jeff Garzik | Only so many songs can be sung
Building 1024 | with two lips, two lungs, and one tongue.
MandrakeSoft | - nomeansno
Jeff Garzik wrote:
> I'm sure people will continue using the OSS drivers even after they
> become the "old" sound drivers... for a while at least.
Not least because I have reports from my housemate that ALSA drivers are
a b*tch to set up. To be done only if there isn't an OSS driver for
your card. Whereis with OSS you just load a module and its done.
-- Jamie
Jamie Lokier wrote:
>
> Jeff Garzik wrote:
> > I'm sure people will continue using the OSS drivers even after they
> > become the "old" sound drivers... for a while at least.
>
> Not least because I have reports from my housemate that ALSA drivers are
> a b*tch to set up. To be done only if there isn't an OSS driver for
> your card. Whereis with OSS you just load a module and its done.
If "modprobe snd-card-via686" doesn't just-work that is a regression
from 2.4 and thus a bug. There should be no alsa-conf or anything else
required for these drivers to work.
Jeff
--
Jeff Garzik | Only so many songs can be sung
Building 1024 | with two lips, two lungs, and one tongue.
MandrakeSoft | - nomeansno
> Not least because I have reports from my housemate that ALSA drivers are
> a b*tch to set up. To be done only if there isn't an OSS driver for
> your card. Whereis with OSS you just load a module and its done.
Thats already a "must fix"
At Tue, 04 Dec 2001 11:55:49 -0500,
Jeff Garzik wrote:
>
> Jamie Lokier wrote:
> >
> > Jeff Garzik wrote:
> > > I'm sure people will continue using the OSS drivers even after they
> > > become the "old" sound drivers... for a while at least.
> >
> > Not least because I have reports from my housemate that ALSA drivers are
> > a b*tch to set up. To be done only if there isn't an OSS driver for
> > your card. Whereis with OSS you just load a module and its done.
>
> If "modprobe snd-card-via686" doesn't just-work that is a regression
> from 2.4 and thus a bug. There should be no alsa-conf or anything else
> required for these drivers to work.
As long as using only ALSA native applications, "modprobe xxx" would
be enough. The subtle configuration is required for OSS emulation,
kmod support and card-specific module options.
Well, but it's surely better to set up module aliases and options
properly for comfortable life :)
There is another issue, that alsa mutes all volumes as default.
I know that's what sometime annoying people. IMO, it's same on OSS
anyway, since everyone needs to set up the mixer as he likes after
start up.
Takashi
Alan Cox wrote:
>>Not least because I have reports from my housemate that ALSA drivers are
>>a b*tch to set up. To be done only if there isn't an OSS driver for
>>your card. Whereis with OSS you just load a module and its done.
>>
>
> Thats already a "must fix"
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
>
Once in 2.5.X will the ALSA drivers still be modular only or will they
be able to be linked statically into the kernel as the current OSS
drivers are able to be?
> As long as using only ALSA native applications, "modprobe xxx" would
> be enough. The subtle configuration is required for OSS emulation,
> kmod support and card-specific module options.
That should be fixable. Trying to open oss emulation devices can load the
OSS emulation.
Alan
Jordan Breeding wrote:
> Once in 2.5.X will the ALSA drivers still be modular only or will they
> be able to be linked statically into the kernel as the current OSS
> drivers are able to be?
They will need to use module_init/exit, which allows static or modular
build at your choice.
Jeff
--
Jeff Garzik | Only so many songs can be sung
Building 1024 | with two lips, two lungs, and one tongue.
MandrakeSoft | - nomeansno
Jeff Garzik wrote:
>>>Out of curiosity, will ALSA also be available on the 24..xx kernels??
>>>Will there be a choice of useing OSS or ALSA??
>>>Will the ALSA drivers be the 0.9 series or the old 0.5 series??? AFAIK they
>
> IMHO ALSA should -never- go into 2.4. It's fine as a patch but 2.5 is
> the time for big merges, and since it's already available for 2.4
> outside the kernel there shouldn't be any need for backporting
>
> Jeff
I agree with Jeff. The better choice is wait and make big changes only
in 2.5. If anyone **needs** alsa, is possible download and configure then
outside kernel 2.4. "Stable" series isn't a good place to make experiences...
Bye,
Piter PUNK
--
____________
/ Piter PUNK \_____________________________________________________
| |
| | E-Mail: [email protected] (personal) |
| .|. [email protected] (professional) |
| /V\ |
| // \\ UIN: 116043354 Homepage: http://www.piterpunk.hpg.com.br |
| /( )\ |
| ^`~'^ |
| #105432 |
`-------------------------------------------------------------------'
On Tue, 4 Dec 2001, Jordan Breeding wrote:
> Once in 2.5.X will the ALSA drivers still be modular only or will they
> be able to be linked statically into the kernel as the current OSS
> drivers are able to be?
IIRC the work already has been done so you can build yourself a giant
oversized monolithic kernel with statically linked ALSA drivers.
-Dan
--
[-] Omae no subete no kichi wa ore no mono da. [-]
> There is another issue, that alsa mutes all volumes as default.
> I know that's what sometime annoying people. IMO, it's same on OSS
> anyway, since everyone needs to set up the mixer as he likes after
> start up.
That's not a bug that's a feature!
The drivers included in the 2.2.x and early 2.4.x series kernels(havn't
tested lately) cause a feedback loop on some laptops that continues until
the mixers are set. (usually 1 to 2 seconds during bootup) I'd *really*
prefer mute.
Gerhard
Alan Cox wrote:
> > Not least because I have reports from my housemate that ALSA drivers are
> > a b*tch to set up. To be done only if there isn't an OSS driver for
> > your card. Whereis with OSS you just load a module and its done.
>
> Thats already a "must fix"
Is the really bad latency reported on this list
as coming from the alsa drivers also "must fix"?
cu
jjs
At Tue, 04 Dec 2001 22:23:11 -0800,
J Sloan wrote:
>
> Alan Cox wrote:
>
> > > Not least because I have reports from my housemate that ALSA drivers are
> > > a b*tch to set up. To be done only if there isn't an OSS driver for
> > > your card. Whereis with OSS you just load a module and its done.
> >
> > Thats already a "must fix"
>
> Is the really bad latency reported on this list
> as coming from the alsa drivers also "must fix"?
At the last time I've tried, the bad latency happened only on OSS
emulation. There must be some small difference of behavior between
OSS and emulation. Yes, this is "must fix", too.
On ALSA native mode you get even lower latency than OSS.
IIRC the chip reported on lkml is ymfpci - it cannot achieve latency
lower than 256/48000 = 5.33 msec, anyway, due to h/w limit..
Takashi