On 2015-04-02 00:24, Takashi Sakamoto wrote: > Hi David, > > On May 31 2015 21:38, David Henningsson wrote: >> About the first patch: >> >> I tested and refined a patch previously submitted by Takashi >> Sakamoto. It seems like nothing has happened on the udev side so I >> suggest we take this one for now. In case udev should improve and get a >> model database for the Firewire bus, then we could consider >> reverting this. > > Thanks for the revised version of my patch (how stupid mistake to use > 'IO_' instead of 'ID_' in my comment!) and I'm sorry to have little time > for this issue... > > Well, I want to test this patchset with my devices in this weekend. In > my understanding, these patches are for devices with different PCM > channels for playback/capture. In this case, your patch allows > pulseaudio clients to send/receive the first 4 channels > (left,right,rear-left,rear-right) to/from the PCM substreams instead of > using whole channels. > > Is my understanding is correct? The second patch works around a bug that could cause multichannel devices with different number of playback and record channels (e g, a device that can only be opened with 10 out and 6 in and nothing else) not to be opened correctly. -- David Henningsson, Canonical Ltd. https://launchpad.net/~diwic
