[alsa-devel] Patch for Digidesign Mbox 2 (set sample rate)
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate was hard-set to 48000Hz by the original author even though the device supports it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch which appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it will not work without the "magic" init sequence that I have removed because it does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it merged. This is my first attempt at submitting a patch for any Linux device driver so please let me know if I need to do anything else :)
Cheers, Vik
On Fri, 28 Apr 2017 22:36:09 +0200, Viktor Radnai wrote:
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate was hard-set to 48000Hz by the original author even though the device supports it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch which appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it will not work without the "magic" init sequence that I have removed because it does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it merged. This is my first attempt at submitting a patch for any Linux device driver so please let me know if I need to do anything else :)
At least it'd be better to the original author in the loop. Damien, could you take a look if you still have the device?
thanks,
Takashi
Hi Takashi,
Thanks, I have also reached out to Damien and he just got back to me. Unfortunately he no longer has the device and is quite busy with other things.
I am going to buy an Mbox 2 and test this myself. You can expect results within two weeks, hopefully sooner.
In the meantime if anyone on the list has this device I would really appreciate if you could test it.
Cheers, Vik
On 3 May 2017 at 11:16, Takashi Iwai tiwai@suse.de wrote:
On Fri, 28 Apr 2017 22:36:09 +0200, Viktor Radnai wrote:
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate
was
hard-set to 48000Hz by the original author even though the device
supports
it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch
which
appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it will not work without the "magic" init sequence that I have removed because it does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it
merged.
This is my first attempt at submitting a patch for any Linux device
driver
so please let me know if I need to do anything else :)
At least it'd be better to the original author in the loop. Damien, could you take a look if you still have the device?
thanks,
Takashi
Hi again,
I have bought an Mbox 2 and tested playback at both 44.1 and 48kHz using Jackd, and 44.1kHz using PulseAudio. It appeared to work without problems (it played without any dropouts and at the correct speed / pitch). I haven't tested recording or done more rigorous testing on the Mbox yet as it only arrived today (I did test recording with the Mini though).
Have you got a proper list of things to test before merging the patch or should I try to make it up myself?
Cheers, Vik
On 6 May 2017 at 09:03, Viktor Radnai viktor.radnai@gmail.com wrote:
Hi Takashi,
Thanks, I have also reached out to Damien and he just got back to me. Unfortunately he no longer has the device and is quite busy with other things.
I am going to buy an Mbox 2 and test this myself. You can expect results within two weeks, hopefully sooner.
In the meantime if anyone on the list has this device I would really appreciate if you could test it.
Cheers, Vik
On 3 May 2017 at 11:16, Takashi Iwai tiwai@suse.de wrote:
On Fri, 28 Apr 2017 22:36:09 +0200, Viktor Radnai wrote:
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate
was
hard-set to 48000Hz by the original author even though the device
supports
it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch
which
appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it
will
not work without the "magic" init sequence that I have removed because
it
does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it
merged.
This is my first attempt at submitting a patch for any Linux device
driver
so please let me know if I need to do anything else :)
At least it'd be better to the original author in the loop. Damien, could you take a look if you still have the device?
thanks,
Takashi
-- My other sig is hilarious
Hi all,
I have tested Viktor's patch on "Mbox 2" and it works very well. Looks like my 'magic' boot sequence was not needed. Rate switching works between 44.1 and 48k. Good job.
Damien
On 03/05/17 20:16, Takashi Iwai wrote:
On Fri, 28 Apr 2017 22:36:09 +0200, Viktor Radnai wrote:
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate was hard-set to 48000Hz by the original author even though the device supports it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch which appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it will not work without the "magic" init sequence that I have removed because it does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it merged. This is my first attempt at submitting a patch for any Linux device driver so please let me know if I need to do anything else :)
At least it'd be better to the original author in the loop. Damien, could you take a look if you still have the device?
thanks,
Takashi
Hi Damien,
Thanks for the test report.
I have now successfully recorded with both the Mbox 2 and Mbox 2 Mini, however I had one occurrence where the Mbox 2 would not initialize even when I repeatedly unplugged and plugged it back. I have added your "magic" boot sequence back and since then I haven't seen this problem. I have been using that version ever since, so I cannot tell if simply unloading and reloading the driver would have fixed it. In any case I will re-submit the patch with the magic boot sequence for inclusion.
Cheers, Vik
On 9 July 2017 at 03:14, Damien Zammit damien@zamaudio.com wrote:
Hi all,
I have tested Viktor's patch on "Mbox 2" and it works very well. Looks like my 'magic' boot sequence was not needed. Rate switching works between 44.1 and 48k. Good job.
Damien
On 03/05/17 20:16, Takashi Iwai wrote:
On Fri, 28 Apr 2017 22:36:09 +0200, Viktor Radnai wrote:
Hi there,
I have recently bought a Digidesign Mbox 2 Mini and found the lack of support for 44100Hz sample rate in the driver limiting (the sample rate
was
hard-set to 48000Hz by the original author even though the device
supports
it, because he couldn't get the frame rate switching to work reliably).
I had a go at hacking the driver and came up with the following patch
which
appears to work on the MBox 2 Mini: https://pastebin.com/kzzTDj9T
I have built and tested this with Linux 4.8.15 without any realtime patches. I am planning to pick up a "full size" Mbox 2 as well, in the meantime if anyone on the list owns one I would appreciate if you could test this and let me know the result. There is a possibility that it
will
not work without the "magic" init sequence that I have removed because
it
does not look like it's needed on the Mini. Although I would be a bit surprised because the two devices are apparently identical to the point that the Mini allows the use of it's physically nonexistent features.
Once someone could confirm that this works, I would like to get it
merged.
This is my first attempt at submitting a patch for any Linux device
driver
so please let me know if I need to do anything else :)
At least it'd be better to the original author in the loop. Damien, could you take a look if you still have the device?
thanks,
Takashi
Hi again,
I think this patch is ready to be applied. I have kept the "magic" init sequence but I have substituted the hexadecimal numbers with preprocessor macros to make the code easier to read.
I will try to run without the init magic for a while and if I don't see any issues I will submit a follow-up patch to remove the entire mbox2_setup_48_24_magic function.
Cheers, Vik
diff -ur linux-source-4.11-orig/sound/usb/quirks.c linux-source-4.11/sound/usb/quirks.c --- linux-source-4.11-orig/sound/usb/quirks.c 2017-06-17 05:47:27.000000000 +0100 +++ linux-source-4.11/sound/usb/quirks.c 2017-07-18 22:37:58.522858262 +0100 @@ -768,23 +768,32 @@
static void mbox2_setup_48_24_magic(struct usb_device *dev) { - u8 srate[3]; - u8 temp[12]; - /* Choose 48000Hz permanently */ - srate[0] = 0x80; - srate[1] = 0xbb; - srate[2] = 0x00; + u8 srate[3] = { 0x80, 0xbb, 0x00 }; + u8 temp[3]; + int err;
/* Send the magic! */ - snd_usb_ctl_msg(dev, usb_rcvctrlpipe(dev, 0), - 0x01, 0x22, 0x0100, 0x0085, &temp, 0x0003); - snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), - 0x81, 0xa2, 0x0100, 0x0085, &srate, 0x0003); - snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), - 0x81, 0xa2, 0x0100, 0x0086, &srate, 0x0003); - snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), - 0x81, 0xa2, 0x0100, 0x0003, &srate, 0x0003); + if ((err = snd_usb_ctl_msg(dev, usb_rcvctrlpipe(dev, 0), UAC_SET_CUR, + USB_TYPE_CLASS | USB_RECIP_ENDPOINT | USB_DIR_OUT, + UAC_EP_CS_ATTR_SAMPLE_RATE << 8, 0x0085, &temp, sizeof(temp))) < 0) + dev_err(&dev->dev, "mbox2: magic setup step #1 failed\n"); + + if ((err = snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), UAC_GET_CUR, + USB_TYPE_CLASS | USB_RECIP_ENDPOINT | USB_DIR_IN, + UAC_EP_CS_ATTR_SAMPLE_RATE << 8, 0x0085, &srate, sizeof(srate))) < 0) + dev_err(&dev->dev, "mbox2: magic setup step #2 failed\n"); + + if ((err = snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), UAC_GET_CUR, + USB_TYPE_CLASS | USB_RECIP_ENDPOINT | USB_DIR_IN, + UAC_EP_CS_ATTR_SAMPLE_RATE << 8, 0x0086, &srate, sizeof(srate))) < 0) + dev_err(&dev->dev, "mbox2: magic setup step #3 failed\n"); + + if ((err = snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), UAC_GET_CUR, + USB_TYPE_CLASS | USB_RECIP_ENDPOINT | USB_DIR_IN, + UAC_EP_CS_ATTR_SAMPLE_RATE << 8, 0x0003, &srate, sizeof(srate))) < 0) + dev_err(&dev->dev, "mbox2: magic setup step #4 failed\n"); + return; }
@@ -815,16 +824,16 @@
count = 0; bootresponse[0] = MBOX2_BOOT_LOADING; - while ((bootresponse[0] == MBOX2_BOOT_LOADING) && (count < 10)) { - msleep(500); /* 0.5 second delay */ + do { snd_usb_ctl_msg(dev, usb_rcvctrlpipe(dev, 0), /* Control magic - load onboard firmware */ - 0x85, 0xc0, 0x0001, 0x0000, &bootresponse, 0x0012); + UAC_GET_MEM, USB_TYPE_VENDOR | USB_RECIP_DEVICE | USB_DIR_IN, 0x0001, 0x0000, &bootresponse, sizeof(bootresponse)); if (bootresponse[0] == MBOX2_BOOT_READY) break; + msleep(500); /* 0.5 second delay */ dev_dbg(&dev->dev, "device not ready, resending boot sequence...\n"); count++; - } + } while ((bootresponse[0] == MBOX2_BOOT_LOADING) && (count < 10));
if (bootresponse[0] != MBOX2_BOOT_READY) { dev_err(&dev->dev, "Unknown bootresponse=%d, or timed out, ignoring device.\n", bootresponse[0]); @@ -847,7 +856,7 @@
mbox2_setup_48_24_magic(dev);
- dev_info(&dev->dev, "Digidesign Mbox 2: 24bit 48kHz"); + dev_info(&dev->dev, "Digidesign Mbox 2 configured");
return 0; /* Successful boot */ } @@ -1111,6 +1120,26 @@ subs->pkt_offset_adj = (emu_samplerate_id >= EMU_QUIRK_SR_176400HZ) ? 4 : 0; }
+static void set_format_mbox2_quirk(struct snd_usb_substream *subs, + struct audioformat *fmt) +{ + struct usb_device *dev = subs->dev; + u8 srate[3]; + int err; + + srate[0] = subs->cur_rate; + srate[1] = subs->cur_rate >> 8; + srate[2] = subs->cur_rate >> 16; + + /* Set the sample rate for endpoint IN 5 (0x0085) until we succeed */ + while ((err = snd_usb_ctl_msg(dev, usb_sndctrlpipe(dev, 0), UAC_SET_CUR, + USB_TYPE_CLASS | USB_RECIP_ENDPOINT | USB_DIR_OUT, + UAC_EP_CS_ATTR_SAMPLE_RATE << 8, 0x0085, srate, sizeof(srate))) < 0) + dev_err(&dev->dev, "mbox2: set sample rate failed, retrying\n"); + + dev_info(&subs->dev->dev, "mbox2: set sample rate to %dHz\n", subs->cur_rate); +} + void snd_usb_set_format_quirk(struct snd_usb_substream *subs, struct audioformat *fmt) { @@ -1121,6 +1150,9 @@ case USB_ID(0x041e, 0x3f19): /* E-Mu 0204 USB */ set_format_emu_quirk(subs, fmt); break; + case USB_ID(0x0dba, 0x3000): /* Digidesign Mbox 2 */ + set_format_mbox2_quirk(subs, fmt); + break; } }
diff -ur linux-source-4.11-orig/sound/usb/quirks-table.h linux-source-4.11/sound/usb/quirks-table.h --- linux-source-4.11-orig/sound/usb/quirks-table.h 2017-06-17 05:47:27.000000000 +0100 +++ linux-source-4.11/sound/usb/quirks-table.h 2017-07-18 17:33:57.620424012 +0100 @@ -3020,12 +3020,13 @@ .attributes = 0x00, .endpoint = 0x03, .ep_attr = USB_ENDPOINT_SYNC_ASYNC, - .rates = SNDRV_PCM_RATE_48000, - .rate_min = 48000, + .rates = SNDRV_PCM_RATE_44100| + SNDRV_PCM_RATE_48000, + .rate_min = 44100, .rate_max = 48000, - .nr_rates = 1, + .nr_rates = 2, .rate_table = (unsigned int[]) { - 48000 + 44100, 48000 } } }, @@ -3045,12 +3046,13 @@ .attributes = UAC_EP_CS_ATTR_SAMPLE_RATE, .endpoint = 0x85, .ep_attr = USB_ENDPOINT_SYNC_SYNC, - .rates = SNDRV_PCM_RATE_48000, - .rate_min = 48000, + .rates = SNDRV_PCM_RATE_44100| + SNDRV_PCM_RATE_48000, + .rate_min = 44100, .rate_max = 48000, - .nr_rates = 1, + .nr_rates = 2, .rate_table = (unsigned int[]) { - 48000 + 44100, 48000 } } },
participants (3)
-
Damien Zammit -
Takashi Iwai -
Viktor Radnai