Re: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

Re: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

[Jeff Garzik]

On Thu, Jun 24, 2004 at 04:44:59PM -0700, Andrew Chew wrote:
> This patch adds a new driver under linux/drivers/sound/nvaudio.  The new
> driver is heavily derived from the i810_audio driver, but includes a lot
> of new work in adding multichannel and spdif support.

Well, ICH5 and ICH6 (and ICH4?) support this new stuff too.  I'm open
to a new driver, but maybe rename it to something more vendor-neutral?

And, does it have the ~11 critical bug fixes that went into i810_audio,
to bring it up to version 1.00?

	Jeff

Re: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

[Jeff Garzik]

On Thu, Jun 24, 2004 at 04:44:59PM -0700, Andrew Chew wrote:
> This patch adds a new driver under linux/drivers/sound/nvaudio.  The new
> driver is heavily derived from the i810_audio driver, but includes a lot
> of new work in adding multichannel and spdif support.
> 
> The patch removes the following PCI device IDs from the i810_audio
> driver:

also, if you're gonna do a driver for the newer ich-ish audio,
it's a good time to make the driver MMIO-only...

	Jeff

RE: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

[Andrew Chew]

> On Thu, Jun 24, 2004 at 04:44:59PM -0700, Andrew Chew wrote:
> > This patch adds a new driver under 
> linux/drivers/sound/nvaudio.  The 
> > new driver is heavily derived from the i810_audio driver, 
> but includes 
> > a lot of new work in adding multichannel and spdif support.

> From: Jeff Garzik [mailto:jgarzik@pobox.com] 
> Well, ICH5 and ICH6 (and ICH4?) support this new stuff too.  
> I'm open to a new driver, but maybe rename it to something 
> more vendor-neutral?

I don't know anything about the ICH5 and ICH6.  I'm checking with
another engineer to see if the hardware interface is the same (or
similar) between the ICH5/6 and nForce audio.  I'm definitely in support
of making it more vendor-neutral if another vendor can use this
commonality.

> And, does it have the ~11 critical bug fixes that went into 
> i810_audio, to bring it up to version 1.00?

I believe the nvaudio driver was forked off from the i810_audio driver
at kernel version 2.4.19, so I think the answer is probably no.  I
apologize for not knowing about the bug fixes.  I'll look into that now.

RE: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

[Andrew Chew]

>From diffs between the forkpoint (2.4.19) and 2.4.27-rc1, and
examination of the nvaudio driver, it doesn't look like the nvaudio
driver inherited any recently fixed bugs from the i810_audio driver.

> -----Original Message-----
> From: Jeff Garzik [mailto:jgarzik@pobox.com] 
> Sent: Thursday, June 24, 2004 5:18 PM
> To: Andrew Chew
> Cc: Alan Cox; linux-kernel@vger.kernel.org
> Subject: Re: [PATCH] 2.4.27-rc1, nvaudio, i810_audio
> 
> 
> On Thu, Jun 24, 2004 at 04:44:59PM -0700, Andrew Chew wrote:
> > This patch adds a new driver under 
> linux/drivers/sound/nvaudio.  The 
> > new driver is heavily derived from the i810_audio driver, 
> but includes 
> > a lot of new work in adding multichannel and spdif support.
> 
> Well, ICH5 and ICH6 (and ICH4?) support this new stuff too.  
> I'm open to a new driver, but maybe rename it to something 
> more vendor-neutral?
> 
> And, does it have the ~11 critical bug fixes that went into 
> i810_audio, to bring it up to version 1.00?
> 
> 	Jeff
> 
> 
> 
> 

RE: [PATCH] 2.4.27-rc1, nvaudio, i810_audio

[Andrew Chew]

> From: Jeff Garzik [mailto:jgarzik@pobox.com] 
> Well, ICH5 and ICH6 (and ICH4?) support this new stuff too.  
> I'm open to a new driver, but maybe rename it to something 
> more vendor-neutral?
It looks like the ICH5/6's register map for spdif is different, so it
would require some work to get them working with the nvaudio driver that
I submitted.

> And, does it have the ~11 critical bug fixes that went into 
> i810_audio, to bring it up to version 1.00?
I just realized that you were referring to the DRIVER_VERSION macro in
the driver.  Yes, since this is a first submission of the driver, let's
call it 0.1 instead.  1.0 is definitely not appropriate.