[alsa-devel] firewire-kernel-streaming branch into main kernel tree?
Hello Clemens,
it's been a long time since I last inquired about the snd_dice module provided by your firewire-kernel-streaming branch (http://git.alsa-project.org/?p=alsa-kprivate.git;a=shortlog;h=refs/heads/fir...), which is a good sign I guess. In the meantime we have been making heavy use of the snd_dice module for over a year with all of our Weiss DICE based products (obviously playback only), and it has proven to be extremely stable (currently 3.4.18-rt29, patched with firewire-kernel-streaming). Therefore I have been meaning to ask you for some time now whether you have any interest in merging this firewire-kernel-streaming branch into the main kernel tree (at least as an experimental option, mind you), or are there any major objections? Given that you don't have unlimited time at hand to further develop the snd_dice module, let alone provide capture support in the near future, AND the fact that DICE is likely to see its EOL soon (and replaced by a sophisticated successor by TCAT), I realize that this is probably not a high priority. However from our perspective it would be sad to see these ALSA Firewire proceedings be lost and not be appreciated by a larger audience. With respect to this: * what would be the customary approach to requesting such a merge? * what are your reservations, resp. are there any existing shortcomings that you are yet unhappy with? * have you received any other (negative/positive) feedback from snd_dice users? I know of the Vortexbox community, which has been offering an optional patch to support Firewire plackback devices; as far as I know, they're happy with it. * if you need any backing or assistance, we'd be happy to help out
Keen to hear your feedback on this suggestion. Best regards,
Rolf Anderegg
(Weiss Engineering Ltd.)
Rolf Anderegg wrote:
I have been meaning to ask you for some time now whether you have any interest in merging this firewire-kernel-streaming branch into the main kernel tree (at least as an experimental option, mind you), or are there any major objections?
I planned to add capture support before the merge, and it's only the lack of time that has prevented progress.
- have you received any other (negative/positive) feedback from
snd_dice users?
Quite positive, especially from users of playback-only devices where FFADO does not work.
For those devices, merging would make sense now. So would you be happy if I merged the driver in its current form, except being restricted to all/some(?) Weiss devices?
Regards, Clemens
For those devices, merging would make sense now. So would you be happy if I merged the driver in its current form, except being restricted to all/some(?) Weiss devices?
Our findings have shown that snd-dice works with all our DICE based Weiss devices, as far as playback (with sync-source restricted to CLOCK_SOURCE_ARX1) is concerned. So yes, we would be very happy with a merge of the driver's current form.
Regards, Rolf
participants (2)
-
Clemens Ladisch -
Rolf Anderegg