[alsa-devel] [PATCH 1/1] ALSA: usb-audio: add support for Akai MPD16
The decoding routine is based on own reverse-engineering. It seems to recognize all the messages that MPD16 sends during normal operation (that is, via standard MIDI messages).
Configuration (changing pad sensitivity, slider controller and MIDI notes) is not supported in this version.
Signed-off-by: Krzysztof Foltman wdev@foltman.com
diff --git a/sound/usb/midi.c b/sound/usb/midi.c index 2c1558c..cce5dbe 100644 --- a/sound/usb/midi.c +++ b/sound/usb/midi.c @@ -645,6 +645,33 @@ static struct usb_protocol_ops snd_usbmidi_cme_ops = { };
/* + * AKAI MPD16 protocol: one or more chunks consisting of first byte of + * (0x20 | msg_len) and then a MIDI message (msg_len bytes long) + * + * Messages sent: + * 21 FE (active sense) + * 23 90 xx xx (note on) + * 23 Ax xx xx (polyphonic pressure) + * 23 Bx xx xx (control change) + */ +static void snd_usbmidi_akai_input(struct snd_usb_midi_in_endpoint *ep, + uint8_t *buffer, int buffer_length) +{ + unsigned int pos = 0; + while (pos < (unsigned)buffer_length && (buffer[pos] & 0xF8) == 0x20) { + int msg_len = buffer[pos] & 0x0f; + snd_usbmidi_input_data(ep, 0, &buffer[pos + 1], msg_len); + pos += 1 + msg_len; + } +} + +static struct usb_protocol_ops snd_usbmidi_akai_ops = { + .input = snd_usbmidi_akai_input, + .output = snd_usbmidi_standard_output, + .output_packet = snd_usbmidi_output_standard_packet, +}; + +/* * Novation USB MIDI protocol: number of data bytes is in the first byte * (when receiving) (+1!) or in the second byte (when sending); data begins * at the third byte. @@ -1434,6 +1461,8 @@ static struct port_info { EXTERNAL_PORT(0x086a, 0x0001, 8, "%s Broadcast"), EXTERNAL_PORT(0x086a, 0x0002, 8, "%s Broadcast"), EXTERNAL_PORT(0x086a, 0x0003, 4, "%s Broadcast"), + /* Akai MPD16 */ + EXTERNAL_PORT(0x09e8, 0x0062, 0, "%s MIDI"), /* Access Music Virus TI */ EXTERNAL_PORT(0x133e, 0x0815, 0, "%s MIDI"), PORT_INFO(0x133e, 0x0815, 1, "%s Synth", 0, @@ -2035,6 +2064,13 @@ int snd_usbmidi_create(struct snd_card *card, umidi->usb_protocol_ops = &snd_usbmidi_cme_ops; err = snd_usbmidi_detect_per_port_endpoints(umidi, endpoints); break; + case QUIRK_MIDI_AKAI: + umidi->usb_protocol_ops = &snd_usbmidi_akai_ops; + err = snd_usbmidi_detect_per_port_endpoints(umidi, endpoints); + /* endpoint 0 does not carry MIDI data */ + endpoints[0].out_cables = 0; + endpoints[0].in_cables = 0; + break; default: snd_printd(KERN_ERR "invalid quirk type %d\n", quirk->type); err = -ENXIO; diff --git a/sound/usb/midi.h b/sound/usb/midi.h index 2089ec9..2fca80b 100644 --- a/sound/usb/midi.h +++ b/sound/usb/midi.h @@ -37,6 +37,8 @@ struct snd_usb_midi_endpoint_info {
/* for QUIRK_MIDI_CME, data is NULL */
+/* for QUIRK_MIDI_AKAI, data is NULL */ + int snd_usbmidi_create(struct snd_card *card, struct usb_interface *iface, struct list_head *midi_list, diff --git a/sound/usb/quirks-table.h b/sound/usb/quirks-table.h index 91ddef3..7ecad05 100644 --- a/sound/usb/quirks-table.h +++ b/sound/usb/quirks-table.h @@ -1973,6 +1973,17 @@ YAMAHA_DEVICE(0x7010, "UB99"), } },
+/* AKAI devices */ +{ + USB_DEVICE(0x09e8, 0x0062), + .driver_info = (unsigned long) &(const struct snd_usb_audio_quirk) { + .vendor_name = "AKAI", + .product_name = "MPD16", + .ifnum = 0, + .type = QUIRK_MIDI_AKAI, + } +}, + /* TerraTec devices */ { USB_DEVICE_VENDOR_SPEC(0x0ccd, 0x0012), diff --git a/sound/usb/quirks.c b/sound/usb/quirks.c index 136e5b4..b45e54c 100644 --- a/sound/usb/quirks.c +++ b/sound/usb/quirks.c @@ -289,6 +289,7 @@ int snd_usb_create_quirk(struct snd_usb_audio *chip, [QUIRK_MIDI_FASTLANE] = create_any_midi_quirk, [QUIRK_MIDI_EMAGIC] = create_any_midi_quirk, [QUIRK_MIDI_CME] = create_any_midi_quirk, + [QUIRK_MIDI_AKAI] = create_any_midi_quirk, [QUIRK_AUDIO_STANDARD_INTERFACE] = create_standard_audio_quirk, [QUIRK_AUDIO_FIXED_ENDPOINT] = create_fixed_stream_quirk, [QUIRK_AUDIO_EDIROL_UAXX] = create_uaxx_quirk, diff --git a/sound/usb/usbaudio.h b/sound/usb/usbaudio.h index d679e72..06ebf24 100644 --- a/sound/usb/usbaudio.h +++ b/sound/usb/usbaudio.h @@ -74,6 +74,7 @@ enum quirk_type { QUIRK_MIDI_FASTLANE, QUIRK_MIDI_EMAGIC, QUIRK_MIDI_CME, + QUIRK_MIDI_AKAI, QUIRK_MIDI_US122L, QUIRK_AUDIO_STANDARD_INTERFACE, QUIRK_AUDIO_FIXED_ENDPOINT,
Krzysztof Foltman wrote:
- AKAI MPD16 protocol: one or more chunks consisting of first byte of
- (0x20 | msg_len) and then a MIDI message (msg_len bytes long)
- Messages sent:
- 21 FE (active sense)
- 23 90 xx xx (note on)
- 23 Ax xx xx (polyphonic pressure)
- 23 Bx xx xx (control change)
- */
+static void snd_usbmidi_akai_input(struct snd_usb_midi_in_endpoint *ep,
uint8_t *buffer, int buffer_length)+{
- unsigned int pos = 0;
- while (pos < (unsigned)buffer_length && (buffer[pos] & 0xF8) == 0x20) {
int msg_len = buffer[pos] & 0x0f;snd_usbmidi_input_data(ep, 0, &buffer[pos + 1], msg_len);
This might overflow if the buffer ends with a 2x byte.
+static struct usb_protocol_ops snd_usbmidi_akai_ops = {
- .input = snd_usbmidi_akai_input,
- .output = snd_usbmidi_standard_output,
- .output_packet = snd_usbmidi_output_standard_packet,
Assuming that this device doesn't have any output ports, please add a comment that this isn't the actual output protocol.
checkpatch.pl also complains about the use of space after the ampersand in quirks-table.h, where it seems to be a common practice.
I'd value consistency over some stupid script, but this issue isn't important enough to worry over.
Overall, the patch looks fine.
Regards, Clemens
On 05/19/2010 07:59 AM, Clemens Ladisch wrote:
This might overflow if the buffer ends with a 2x byte.
True. I rewrote that part now, using better sanity checking (hopefully). Also, the control endpoint uses 0x1x instead of 0x2x and the length is 9, so the mask 0xF8 was wrong too.
Assuming that this device doesn't have any output ports, please add a comment that this isn't the actual output protocol.
In the most recent version, I've added output support (for control port only, as the device does not seem to handle input on its data port endpoint). It's a little bit hairy, but seems to do the job. I've checked it with amidi, both input and output - even when I deliberately added junk before actual sysex data, or when I put several SysEx messages in one hex string. The device seems to require that individual messages are entirely contained within a single USB packet, which complicates things a little bit.
Thanks!
K.
participants (2)
-
Clemens Ladisch -
Krzysztof Foltman