Re: [Openvortex-dev] Status of Manuel's patches and openvortex development status

[Manuel Jander]

Hi,

If you (or anyone else) could help do some revisions regarding what
patches are really applied to current ALSA and which not, that would be
really great.

> >>vortex_adbdma_resetup from au88x0_core.c and its invocation in 
> >>snd_vortex_pcm_trigger from au88x0_pcm.c and declaration in au88x0.h 
> >>seem to be missing in the 1.0.6a.
> >>
> >This one, contributed by someone here on the list (cant remember right
> >now). Its related to PCM streams being paused ans resumed. While pausing
> >the DMA context seems to get lost, so this extra call reestablishes the
> >DMA context.
> >  
> >
> Ok. Is it related to "irq fifo bug" on [some] VIA chipsets? If it is, 
> this is my case ;-)

No. There where other patches involved in automatic latency fixes on VIA
PCI bridges. AFAIK they are included in current ALSA. The "resetup"
stuff is solely for start-pause-resume situations of PCM streams with
some apps, mostly while using OSS emulation.

> BTW, it seems installation of 1.0.6a fixed my previous strange equalizer 
> channel mapping (message-id <address@hidden>, 
> reported earlier in this list).

I don't remember anything about that, but...  :D
Regarding the hardware EQ meters, there are some alsa-mixer flaws
preventing them being visualized correctly; it doesn comply with its own
specs :P.

> >>Also it's unclear for me if keeping vortex "QUAD" status in a member 
> >>variable "isquad", which initialized at startup (as it's done in 
> >>aureal-quadfix.patch), is more preferrable than runtime calculation in 
> >>VORTEX_IS_QUAD macro (upstream variant).
> >>
> >
> >There where some segfault problems, because at some time the ac97 struct
> >could be null and the macro being used. Using a variable that is
> >initialized at startup when the ac97 struct is not null for sure, solved
> >the problem.
> >  
> >
> Hmm, so this means, I should re-apply this patch if I want to get 
> reliable behaviour

I would recommend it.

> >>And what about other patches published in this list?

AFAIK the 2.6.6 patch is the latest patch regarding functionality. The
other 2 or 3 patches made by Jeff Muizelaar, are not functionality
related, but valuable because they clarify the EQ code significantly; i
would not leave them out.

> >All current patches on my website should be considered obsolete... while
> >there could be some things that where silently omitted by the ALSA
> >team :(
> >  
> >
> Well, so far I applied two aureal patches mentioned above and things 
> "work for me". Also I noticed some "cleanup" patches during last two 
> months, some of them seem to be applied, some aren't [yet] (as of 
> 1.0.6a) but none of them seem, e-e-ehm, not "life-critical" for me 
> (perhaps I'm wrong).

Best Regards,

Manuel