Hi Takashi
On Sat, Jul 11, 2015 at 11:12:11PM +0900, Takashi Sakamoto wrote:
This patchset is for three new drivers and related modification to firewire-lib module. :
- Patch 29-37: for firewire-motu, a new driver for MOTU FireWire series
:
- firewire-motu:
- experimental, for preview to developers
A few general comments about the MOTU driver.
1) Unlike the FFADO driver (where I imagine much of the detail has been taken) there are a lot of magic numbers in this driver both in terms of the registers being written to and the values written to those registers. I appreciate that ALSA may have different coding conventions, but I think it would make sense to use defines to document what these magic values are, especially since in a vast majority of cases we do know what the relevant bits and registers are doing.
2) As has been mentioned before, the streaming drivers for firewire devices in ALSA are a successor to those in the FFADO project. In the case of some of the other firewire drivers (DICE etc) I understand that the ALSA code is predominantly a reimplementation based on published standards. However, given the structure and terminology used within the the experimental MOTU driver, it seems to be based fairly heavily on my FFADO MOTU driver which itself has come about due to a significant time investment on my part over many years to decode and analyse the protocol used by these interfaces. I have no problem with this reuse of knowledge since this is precisely how open source should work. However, it would be appreciated if this prior work by myself within FFADO was at least acknowledged in some way in the ALSA MOTU driver files.
Similarly, knowledge of the requirements of many MOTU devices have been contributed by others as a result of directed experimentation. Those individuals are acknowledged in FFADO source code and in our repo. Where this information has been utilised in the ALSA code I think it is fair that their contribution is noted somehow. I don't know what the ALSA convention is with regard to such things - such as single-line comments to this effect within the code.
As mentioned earlier I will continue to work through the patchset and provide comments and suggestions as time allows over the coming few days.
Regards jonathan