Hi Yun-hao, > I think it is more likely to be the PTS's limitations. I realized the > problem is in the media channel not in the signaling channel. PTS > wants to use streaming mode, but bluez rejects it, and then PTS aborts > the test. According to the MPS 1.0 Spec 6.2, L2CAP streaming mode is > the recommendation for optimization when using MPS media control > channels, so I think bluez doesn't do anything wrong during this test. do you have the btmon trace for this. I think we should accept ERTM on the media channel. The unfortunate part is that control and media channels both use PSM 25 and so we can’t be really selective when we are the acceptor. If MPS wants us to accept ERTM on the media channel we should allow. For all I care we can even accept it on the control channel, but it is just a waste of overhead on the L2CAP headers. Actually the new Enhanced Flow Control mode on BR/EDR would be more efficient for A2DP. Regards Marcel
