Hello again Clemens.
Il 22 marzo 2019 alle 15.06 Clemens Ladisch clemens@ladisch.de ha scritto:
Guido Trentalancia wrote:
Also, the hardware is the same: Hercules P32 DJ (with exactly the same firmware version 1.46 which is the latest).
Are the USB controllers the same (2.0/3.x, OHCI/UHCI/EHCI/xHCI)?
You wrote that "amidi -p hw:2,0,0 -d" causes the problem. As long as no MIDI messages are actually received, the kernel or driver cannot affect what happens on the USB bus; the polling is done entirely by the host controller.
I have double-checked this too and I can now confirm: "amidi -p hw:2,0,0 -d" does not increase the RX/TX (I/O) bytes count in /proc/asound/card2/midi0.
So, what kind of operation does amidi trigger in the kernel which causes the problem ? What does it actually do ?
I saw you wrote the ALSA USB MIDI driver (sound/usb/midi.c). But the MIDI seems to run fine, I can receive (RX) data from the controller fine.
It is the audio part (card.c, pcm.c, clock.c and so on) that is being disturbed by simultaneous MIDI operation !
What do you say ?
Regards, Clemens