Skip to contentSkip to conversation

USB Audio Device Class 3.0 support

PATCH Superseded Ruslan Bilovol 4 participants 10 messages Mar 19, 2018

Subsystems (2)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

v2 cover USB Audio Device Class 3.0 support

Hi Takashi,

This patch adds initial USB Audio Device Class 3.0 [1] support to the ALSA that we discussed at ELCE.

The patch was tested with UAC3 gadget [2] that I posted to USB mailing list before. It is good for working with BADD device which implements such topologies like BAIF (Basic Audio Input Functions), BAOF (Basic Audio Output Functions) or compbination of both.

UAC3 spec has changed descriptors laout and many other codes comparing to UAC2 spec, thus making reuse of existing sources impossible or quite complex.

There are still many areas of improvement, as this patch doesn't have UAC3 Mixer Unit support nor some new features like Power Management

I tested this with BeagleBone Black as UAC3 gadget device.

[1] http://www.usb.org/developers/docs/devclass_docs/USB_Audio_v3.0.zip
[2] http://www.spinics.net/lists/linux-usb/msg162482.html

v2 changes:
- moved audioformat quirks to quirks.c
- addressed many comments from Pierre-Louis Bossart
- reorganized audio-v3.h
- added more UAC3 channel relationship definitions

Ruslan Bilovol (2):
ALSA: usb-audio: move audioformat quirks to quirks.c
ALSA: usb: initial USB Audio Device Class 3.0 support

 include/linux/usb/audio-v2.h   |   4 +-
 include/linux/usb/audio-v3.h   | 395 +++++++++++++++++++++++++++++++++++++++++
 include/uapi/linux/usb/audio.h |   1 +
 sound/usb/card.c               |   7 +-
 sound/usb/card.h               |   2 +-
 sound/usb/clock.c              | 228 +++++++++++++++++++++---
 sound/usb/clock.h              |   4 +-
 sound/usb/format.c             |  91 ++++++++--
 sound/usb/format.h             |   6 +-
 sound/usb/mixer.c              | 337 +++++++++++++++++++++++------------
 sound/usb/quirks.c             |  34 ++++
 sound/usb/quirks.h             |   4 +
 sound/usb/stream.c             | 393 +++++++++++++++++++++++++++++++++-------
 13 files changed, 1284 insertions(+), 222 deletions(-)
 create mode 100644 include/linux/usb/audio-v3.h
Signature · 1 line
1.9.1
#1 lore raw

PATCH v2 1/2 ALSA: usb-audio: move audioformat quirks to quirks.c

Offload USB audio interface parsing function by moving quirks to a specially designed location (quirks.c)

---
 sound/usb/quirks.c | 34 ++++++++++++++++++++++++++++++++++
 sound/usb/quirks.h |  4 ++++
 sound/usb/stream.c | 30 +-----------------------------
 3 files changed, 39 insertions(+), 29 deletions(-)
sound/usb/quirks.c +34 -0
index ea8f3de..eeea8e1 100644
@@ -1403,3 +1403,37 @@ u64 snd_usb_interface_dsd_format_quirks(struct snd_usb_audio *chip,
 
 	return 0;
 }
+
+void snd_usb_audioformat_attributes_quirk(struct snd_usb_audio *chip,
+					  struct audioformat *fp,
+					  int stream)
+{
+	switch (chip->usb_id) {
+	case USB_ID(0x0a92, 0x0053): /* AudioTrak Optoplay */
+		/* Optoplay sets the sample rate attribute although
+		 * it seems not supporting it in fact.
+		 */
+		fp->attributes &= ~UAC_EP_CS_ATTR_SAMPLE_RATE;
+		break;
+	case USB_ID(0x041e, 0x3020): /* Creative SB Audigy 2 NX */
+	case USB_ID(0x0763, 0x2003): /* M-Audio Audiophile USB */
+		/* doesn't set the sample rate attribute, but supports it */
+		fp->attributes |= UAC_EP_CS_ATTR_SAMPLE_RATE;
+		break;
+	case USB_ID(0x0763, 0x2001):  /* M-Audio Quattro USB */
+	case USB_ID(0x0763, 0x2012):  /* M-Audio Fast Track Pro USB */
+	case USB_ID(0x047f, 0x0ca1): /* plantronics headset */
+	case USB_ID(0x077d, 0x07af): /* Griffin iMic (note that there is
+					an older model 77d:223) */
+	/*
+	 * plantronics headset and Griffin iMic have set adaptive-in
+	 * although it's really not...
+	 */
+		fp->ep_attr &= ~USB_ENDPOINT_SYNCTYPE;
+		if (stream == SNDRV_PCM_STREAM_PLAYBACK)
+			fp->ep_attr |= USB_ENDPOINT_SYNC_ADAPTIVE;
+		else
+			fp->ep_attr |= USB_ENDPOINT_SYNC_SYNC;
+		break;
+	}
+}
sound/usb/quirks.h +4 -0
index b90c8b7..a80e0dd 100644
@@ -42,4 +42,8 @@ u64 snd_usb_interface_dsd_format_quirks(struct snd_usb_audio *chip,
 					struct audioformat *fp,
 					unsigned int sample_bytes);
 
+void snd_usb_audioformat_attributes_quirk(struct snd_usb_audio *chip,
+					  struct audioformat *fp,
+					  int stream);
+
 #endif /* __USBAUDIO_QUIRKS_H */
sound/usb/stream.c +1 -29
index d1776e5..dbbe854 100644
@@ -678,35 +678,7 @@ int snd_usb_parse_audio_interface(struct snd_usb_audio *chip, int iface_no)
 		INIT_LIST_HEAD(&fp->list);
 
 		/* some quirks for attributes here */
-
-		switch (chip->usb_id) {
-		case USB_ID(0x0a92, 0x0053): /* AudioTrak Optoplay */
-			/* Optoplay sets the sample rate attribute although
-			 * it seems not supporting it in fact.
-			 */
-			fp->attributes &= ~UAC_EP_CS_ATTR_SAMPLE_RATE;
-			break;
-		case USB_ID(0x041e, 0x3020): /* Creative SB Audigy 2 NX */
-		case USB_ID(0x0763, 0x2003): /* M-Audio Audiophile USB */
-			/* doesn't set the sample rate attribute, but supports it */
-			fp->attributes |= UAC_EP_CS_ATTR_SAMPLE_RATE;
-			break;
-		case USB_ID(0x0763, 0x2001):  /* M-Audio Quattro USB */
-		case USB_ID(0x0763, 0x2012):  /* M-Audio Fast Track Pro USB */
-		case USB_ID(0x047f, 0x0ca1): /* plantronics headset */
-		case USB_ID(0x077d, 0x07af): /* Griffin iMic (note that there is
-						an older model 77d:223) */
-		/*
-		 * plantronics headset and Griffin iMic have set adaptive-in
-		 * although it's really not...
-		 */
-			fp->ep_attr &= ~USB_ENDPOINT_SYNCTYPE;
-			if (stream == SNDRV_PCM_STREAM_PLAYBACK)
-				fp->ep_attr |= USB_ENDPOINT_SYNC_ADAPTIVE;
-			else
-				fp->ep_attr |= USB_ENDPOINT_SYNC_SYNC;
-			break;
-		}
+		snd_usb_audioformat_attributes_quirk(chip, fp, stream);
 
 		/* ok, let's parse further... */
 		if (snd_usb_parse_audio_format(chip, fp, format, fmt, stream) < 0) {
Signature · 1 line
1.9.1
#2 lore raw

Re: [PATCH v2 1/2] ALSA: usb-audio: move audioformat quirks to quirks.c

Quoted 5 lines · Ruslan Bilovol

Offload USB audio interface parsing function by moving quirks to a specially designed location (quirks.c)

This one is applied now to for-next branch.

thanks,

Takashi

#3 lore raw

Re: [PATCH v2 1/2] ALSA: usb-audio: move audioformat quirks to quirks.c

Quoted 10 lines · Takashi Iwai
Quoted 5 lines · Ruslan Bilovol

Offload USB audio interface parsing function by moving quirks to a specially designed location (quirks.c)

This one is applied now to for-next branch.

Good, I'll drop this one from my series (although don't see it in the for-next branch yet)

Thanks,
Ruslan

#4 lore raw

PATCH v2 2/2 ALSA: usb: initial USB Audio Device Class 3.0 support

In reply to Ruslan Bilovol

Recently released USB Audio Class 3.0 specification introduces many significant changes comparing to previous versions, like
- new Power Domains, support for LPM/L1
- new Cluster descriptor
- changed layout of all class-specific descriptors
- new High Capability descriptors
- New class-specific String descriptors
- new and removed units
- additional sources for interrupts
- removed Type II Audio Data Formats
- ... and many other things (check spec)

It also provides backward compatibility through
multiple configurations, as well as requires
mandatory support for BADD (Basic Audio Device
Definition) on each ADC3.0 compliant device

This patch adds initial support of UAC3 specification that is enough for Generic I/O Profile (BAOF, BAIF) device support from BADD document.

---
 include/linux/usb/audio-v2.h   |   4 +-
 include/linux/usb/audio-v3.h   | 395 +++++++++++++++++++++++++++++++++++++++++
 include/uapi/linux/usb/audio.h |   1 +
 sound/usb/card.c               |   7 +-
 sound/usb/card.h               |   2 +-
 sound/usb/clock.c              | 228 +++++++++++++++++++++---
 sound/usb/clock.h              |   4 +-
 sound/usb/format.c             |  91 ++++++++--
 sound/usb/format.h             |   6 +-
 sound/usb/mixer.c              | 337 +++++++++++++++++++++++------------
 sound/usb/stream.c             | 363 +++++++++++++++++++++++++++++++++----
 11 files changed, 1245 insertions(+), 193 deletions(-)
 create mode 100644 include/linux/usb/audio-v3.h
include/linux/usb/audio-v2.h +2 -2
index 3119d0a..2db83a1 100644
@@ -34,12 +34,12 @@
  *
  */
 
-static inline bool uac2_control_is_readable(u32 bmControls, u8 control)
+static inline bool uac_v2v3_control_is_readable(u32 bmControls, u8 control)
 {
 	return (bmControls >> (control * 2)) & 0x1;
 }
 
-static inline bool uac2_control_is_writeable(u32 bmControls, u8 control)
+static inline bool uac_v2v3_control_is_writeable(u32 bmControls, u8 control)
 {
 	return (bmControls >> (control * 2)) & 0x2;
 }
include/linux/usb/audio-v3.h new file +395 -0

395 lines not on this page · raw message

include/uapi/linux/usb/audio.h +1 -0
index 17a022c..63794a0 100644
@@ -27,6 +27,7 @@
 /* bInterfaceProtocol values to denote the version of the standard used */
 #define UAC_VERSION_1			0x00
 #define UAC_VERSION_2			0x20
+#define UAC_VERSION_3			0x30
 
 /* A.2 Audio Interface Subclass Codes */
 #define USB_SUBCLASS_AUDIOCONTROL	0x01
sound/usb/card.c +5 -2
index 8018d56..4a1c6bb 100644
@@ -7,6 +7,7 @@
  *	    Alan Cox ([email protected])
  *	    Thomas Sailer ([email protected])
  *
+ *   Audio Class 3.0 support by Ruslan Bilovol <[email protected]>
  *
  *   This program is free software; you can redistribute it and/or modify
  *   it under the terms of the GNU General Public License as published by
@@ -44,6 +45,7 @@
 #include <linux/mutex.h>
 #include <linux/usb/audio.h>
 #include <linux/usb/audio-v2.h>
+#include <linux/usb/audio-v3.h>
 #include <linux/module.h>
 
 #include <sound/control.h>
@@ -281,7 +283,8 @@ static int snd_usb_create_streams(struct snd_usb_audio *chip, int ctrlif)
 		break;
 	}
 
-	case UAC_VERSION_2: {
+	case UAC_VERSION_2:
+	case UAC_VERSION_3: {
 		struct usb_interface_assoc_descriptor *assoc =
 			usb_ifnum_to_if(dev, ctrlif)->intf_assoc;
 
@@ -301,7 +304,7 @@ static int snd_usb_create_streams(struct snd_usb_audio *chip, int ctrlif)
 		}
 
 		if (!assoc) {
-			dev_err(&dev->dev, "Audio class v2 interfaces need an interface association\n");
+			dev_err(&dev->dev, "Audio class v2/v3 interfaces need an interface association\n");
 			return -EINVAL;
 		}
 
sound/usb/card.h +1 -1
index ed87cc8..1406292 100644
@@ -22,7 +22,7 @@ struct audioformat {
 	unsigned char endpoint;		/* endpoint */
 	unsigned char ep_attr;		/* endpoint attributes */
 	unsigned char datainterval;	/* log_2 of data packet interval */
-	unsigned char protocol;		/* UAC_VERSION_1/2 */
+	unsigned char protocol;		/* UAC_VERSION_1/2/3 */
 	unsigned int maxpacksize;	/* max. packet size */
 	unsigned int rates;		/* rate bitmasks */
 	unsigned int rate_min, rate_max;	/* min/max rates */
sound/usb/clock.c +203 -25

338 lines not on this page · raw message

sound/usb/clock.h +2 -2
index 87557ca..076e31b 100644
@@ -6,7 +6,7 @@ int snd_usb_init_sample_rate(struct snd_usb_audio *chip, int iface,
 			     struct usb_host_interface *alts,
 			     struct audioformat *fmt, int rate);
 
-int snd_usb_clock_find_source(struct snd_usb_audio *chip, int entity_id,
-			     bool validate);
+int snd_usb_clock_find_source(struct snd_usb_audio *chip, int protocol,
+			     int entity_id, bool validate);
 
 #endif /* __USBAUDIO_CLOCK_H */
sound/usb/format.c +75 -16

186 lines not on this page · raw message

sound/usb/format.h +5 -1
index 8c3ff9c..e701718 100644
@@ -3,8 +3,12 @@
 #define __USBAUDIO_FORMAT_H
 
 int snd_usb_parse_audio_format(struct snd_usb_audio *chip,
-			       struct audioformat *fp, unsigned int format,
+			       struct audioformat *fp, u64 format,
 			       struct uac_format_type_i_continuous_descriptor *fmt,
 			       int stream);
 
+int snd_usb_parse_audio_format_v3(struct snd_usb_audio *chip,
+			       struct audioformat *fp,
+			       struct uac3_as_header_descriptor *as,
+			       int stream);
 #endif /*  __USBAUDIO_FORMAT_H */
sound/usb/mixer.c +228 -109

456 lines not on this page · raw message

sound/usb/stream.c +328 -35

415 lines not on this page · raw message

Signature · 1 line
1.9.1
#5 lore raw

Re: [PATCH v2 2/2] ALSA: usb: initial USB Audio Device Class 3.0 support

Thank you for the patch! Perhaps something to improve:

Hi Ruslan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on linus/master]
[also build test WARNING on v4.16-rc6 next-20180316]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url: https://github.com/0day-ci/linux/commits/Ruslan-Bilovol/ALSA-usb-audio-move-audioformat-quirks-to-quirks-c/20180319-205541
config: i386-randconfig-x016-201811 (attached as .config)
compiler: gcc-7 (Debian 7.3.0-1) 7.3.0 reproduce:

        # save the attached .config to linux build tree
        make ARCH=i386 

Note: it may well be a FALSE warning. FWIW you are at least aware of it now.
http://gcc.gnu.org/wiki/Better_Uninitialized_Warnings

All warnings (new ones prefixed by >>):

sound/usb/stream.c: In function 'snd_usb_parse_audio_interface':

>> sound/usb/stream.c:956:8: warning: 'fmt' may be used uninitialized in this function [-Wmaybe-uninitialized]
       if (snd_usb_parse_audio_format(chip, fp, format,
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
           fmt, stream) < 0) {
           ~~~~~~~~~~~~

vim +/fmt +956 sound/usb/stream.c

628
629 int snd_usb_parse_audio_interface(struct snd_usb_audio *chip, int iface_no)
630 {
631 struct usb_device *dev;
632 struct usb_interface *iface;
633 struct usb_host_interface *alts;
634 struct usb_interface_descriptor *altsd;
635 int i, altno, err, stream;
636 u64 format = 0;
637 unsigned int num_channels = 0;
638 struct audioformat *fp = NULL;
639 int num, protocol, clock = 0;
640 struct uac_format_type_i_continuous_descriptor *fmt;
641 struct snd_pcm_chmap_elem *chmap_v3 = NULL;
642 unsigned int chconfig;
643
644 dev = chip->dev;
645
646 /* parse the interface's altsettings */
647 iface = usb_ifnum_to_if(dev, iface_no);
648
649 num = iface->num_altsetting;
650
651 /*
652 * Dallas DS4201 workaround: It presents 5 altsettings, but the last
653 * one misses syncpipe, and does not produce any sound.
654 */
655 if (chip->usb_id == USB_ID(0x04fa, 0x4201))
656 num = 4;
657
658 for (i = 0; i < num; i++) {
659 alts = &iface->altsetting[i];
660 altsd = get_iface_desc(alts);
661 protocol = altsd->bInterfaceProtocol;
662 /* skip invalid one */
663 if (((altsd->bInterfaceClass != USB_CLASS_AUDIO ||
664 (altsd->bInterfaceSubClass != USB_SUBCLASS_AUDIOSTREAMING &&
665 altsd->bInterfaceSubClass != USB_SUBCLASS_VENDOR_SPEC)) &&
666 altsd->bInterfaceClass != USB_CLASS_VENDOR_SPEC) ||
667 altsd->bNumEndpoints < 1 ||
668 le16_to_cpu(get_endpoint(alts, 0)->wMaxPacketSize) == 0)
669 continue;
670 /* must be isochronous */
671 if ((get_endpoint(alts, 0)->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) !=
672 USB_ENDPOINT_XFER_ISOC)
673 continue;
674 /* check direction */
675 stream = (get_endpoint(alts, 0)->bEndpointAddress & USB_DIR_IN) ?
676 SNDRV_PCM_STREAM_CAPTURE : SNDRV_PCM_STREAM_PLAYBACK;
677 altno = altsd->bAlternateSetting;
678
679 if (snd_usb_apply_interface_quirk(chip, iface_no, altno))
680 continue;
681
682 /*
683 * Roland audio streaming interfaces are marked with protocols
684 * 0/1/2, but are UAC 1 compatible.
685 */
686 if (USB_ID_VENDOR(chip->usb_id) == 0x0582 &&
687 altsd->bInterfaceClass == USB_CLASS_VENDOR_SPEC &&
688 protocol <= 2)
689 protocol = UAC_VERSION_1;
690
691 chconfig = 0;
692 /* get audio formats */
693 switch (protocol) {
694 default:
695 dev_dbg(&dev->dev, "%u:%d: unknown interface protocol %#02x, assuming v1\n",
696 iface_no, altno, protocol);
697 protocol = UAC_VERSION_1;
698 /* fall through */
699
700 case UAC_VERSION_1: {
701 struct uac1_as_header_descriptor *as =
702 snd_usb_find_csint_desc(alts->extra, alts->extralen, NULL, UAC_AS_GENERAL);
703 struct uac_input_terminal_descriptor *iterm;
704
705 if (!as) {
706 dev_err(&dev->dev,
707 "%u:%d : UAC_AS_GENERAL descriptor not found\n",
708 iface_no, altno);
709 continue;
710 }
711
712 if (as->bLength < sizeof(*as)) {
713 dev_err(&dev->dev,
714 "%u:%d : invalid UAC_AS_GENERAL desc\n",
715 iface_no, altno);
716 continue;
717 }
718
719 format = le16_to_cpu(as->wFormatTag); /* remember the format value */
720
721 iterm = snd_usb_find_input_terminal_descriptor(chip->ctrl_intf,
722 as->bTerminalLink);
723 if (iterm) {
724 num_channels = iterm->bNrChannels;
725 chconfig = le16_to_cpu(iterm->wChannelConfig);
726 }
727
728 break;
729 }
730
731 case UAC_VERSION_2: {
732 struct uac2_input_terminal_descriptor *input_term;
733 struct uac2_output_terminal_descriptor *output_term;
734 struct uac2_as_header_descriptor *as =
735 snd_usb_find_csint_desc(alts->extra, alts->extralen, NULL, UAC_AS_GENERAL);
736
737 if (!as) {
738 dev_err(&dev->dev,
739 "%u:%d : UAC_AS_GENERAL descriptor not found\n",
740 iface_no, altno);
741 continue;
742 }
743
744 if (as->bLength < sizeof(*as)) {
745 dev_err(&dev->dev,
746 "%u:%d : invalid UAC_AS_GENERAL desc\n",
747 iface_no, altno);
748 continue;
749 }
750
751 num_channels = as->bNrChannels;
752 format = le32_to_cpu(as->bmFormats);
753 chconfig = le32_to_cpu(as->bmChannelConfig);
754
755 /* lookup the terminal associated to this interface
756 * to extract the clock */
757 input_term = snd_usb_find_input_terminal_descriptor(chip->ctrl_intf,
758 as->bTerminalLink);
759 if (input_term) {
760 clock = input_term->bCSourceID;
761 if (!chconfig && (num_channels == input_term->bNrChannels))
762 chconfig = le32_to_cpu(input_term->bmChannelConfig);
763 break;
764 }
765
766 output_term = snd_usb_find_output_terminal_descriptor(chip->ctrl_intf,
767 as->bTerminalLink);
768 if (output_term) {
769 clock = output_term->bCSourceID;
770 break;
771 }
772
773 dev_err(&dev->dev,
774 "%u:%d : bogus bTerminalLink %d\n",
775 iface_no, altno, as->bTerminalLink);
776 continue;
777 }
778
779 case UAC_VERSION_3: {
780 struct uac3_input_terminal_descriptor *input_term;
781 struct uac3_output_terminal_descriptor *output_term;
782 struct uac3_as_header_descriptor *as;
783 struct uac3_cluster_header_descriptor *cluster;
784 struct uac3_hc_descriptor_header hc_header;
785 u16 cluster_id, wLength;
786
787 as = snd_usb_find_csint_desc(alts->extra,
788 alts->extralen,
789 NULL, UAC_AS_GENERAL);
790
791 if (!as) {
792 dev_err(&dev->dev,
793 "%u:%d : UAC_AS_GENERAL descriptor not found\n",
794 iface_no, altno);
795 continue;
796 }
797
798 if (as->bLength < sizeof(*as)) {
799 dev_err(&dev->dev,
800 "%u:%d : invalid UAC_AS_GENERAL desc\n",
801 iface_no, altno);
802 continue;
803 }
804
805 cluster_id = le16_to_cpu(as->wClusterDescrID);
806 if (!cluster_id) {
807 dev_err(&dev->dev,
808 "%u:%d : no cluster descriptor\n",
809 iface_no, altno);
810 continue;
811 }
812
813 /*
814 * Get number of channels and channel map through
815 * High Capability Cluster Descriptor
816 *
817 * First step: get High Capability header and
818 * read size of Cluster Descriptor
819 */
820 err = snd_usb_ctl_msg(chip->dev,
821 usb_rcvctrlpipe(chip->dev, 0),
822 UAC3_CS_REQ_HIGH_CAPABILITY_DESCRIPTOR,
823 USB_RECIP_INTERFACE | USB_TYPE_CLASS | USB_DIR_IN,
824 cluster_id,
825 snd_usb_ctrl_intf(chip),
826 &hc_header, sizeof(hc_header));
827 if (err < 0)
828 return err;
829 else if (err != sizeof(hc_header)) {
830 dev_err(&dev->dev,
831 "%u:%d : can't get High Capability descriptor\n",
832 iface_no, altno);
833 return -EIO;
834 }
835
836 /*
837 * Second step: allocate needed amount of memory
838 * and request Cluster Descriptor
839 */
840 wLength = le16_to_cpu(hc_header.wLength);
841 cluster = kzalloc(wLength, GFP_KERNEL);
842 if (!cluster)
843 return -ENOMEM;
844 err = snd_usb_ctl_msg(chip->dev,
845 usb_rcvctrlpipe(chip->dev, 0),
846 UAC3_CS_REQ_HIGH_CAPABILITY_DESCRIPTOR,
847 USB_RECIP_INTERFACE | USB_TYPE_CLASS | USB_DIR_IN,
848 cluster_id,
849 snd_usb_ctrl_intf(chip),
850 cluster, wLength);
851 if (err < 0) {
852 kfree(cluster);
853 return err;
854 } else if (err != wLength) {
855 dev_err(&dev->dev,
856 "%u:%d : can't get Cluster Descriptor\n",
857 iface_no, altno);
858 kfree(cluster);
859 return -EIO;
860 }
861
862 num_channels = cluster->bNrChannels;
863 chmap_v3 = convert_chmap_v3(cluster);
864
865 kfree(cluster);
866
867 format = le64_to_cpu(as->bmFormats);
868
869 /* lookup the terminal associated to this interface
870 * to extract the clock */
871 input_term = snd_usb_find_input_terminal_descriptor(
872 chip->ctrl_intf,
873 as->bTerminalLink);
874
875 if (input_term) {
876 clock = input_term->bCSourceID;
877 break;
878 }
879
880 output_term = snd_usb_find_output_terminal_descriptor(chip->ctrl_intf,
881 as->bTerminalLink);
882 if (output_term) {
883 clock = output_term->bCSourceID;
884 break;
885 }
886
887 dev_err(&dev->dev,
888 "%u:%d : bogus bTerminalLink %d\n",
889 iface_no, altno, as->bTerminalLink);
890 continue;
891 }
892 }
893
894 if (protocol == UAC_VERSION_1 || protocol == UAC_VERSION_2) {
895 /* get format type */
896 fmt = snd_usb_find_csint_desc(alts->extra,
897 alts->extralen,
898 NULL, UAC_FORMAT_TYPE);
899 if (!fmt) {
900 dev_err(&dev->dev,
901 "%u:%d : no UAC_FORMAT_TYPE desc\n",
902 iface_no, altno);
903 continue;
904 }
905 if (((protocol == UAC_VERSION_1) && (fmt->bLength < 8))
906 || ((protocol == UAC_VERSION_2) &&
907 (fmt->bLength < 6))) {
908 dev_err(&dev->dev,
909 "%u:%d : invalid UAC_FORMAT_TYPE desc\n",
910 iface_no, altno);
911 continue;
912 }
913
914 /*
915 * Blue Microphones workaround: The last altsetting is
916 * identical with the previous one, except for a larger
917 * packet size, but is actually a mislabeled two-channel
918 * setting; ignore it.
919 */
920 if (fmt->bNrChannels == 1 &&

   921				    fmt->bSubframeSize == 2 &&
   922				    altno == 2 && num == 3 &&
   923				    fp && fp->altsetting == 1 && fp->channels == 1 &&
   924				    fp->formats == SNDRV_PCM_FMTBIT_S16_LE &&
   925				    protocol == UAC_VERSION_1 &&
   926				    le16_to_cpu(get_endpoint(alts, 0)->wMaxPacketSize) ==
   927								fp->maxpacksize * 2)
   928					continue;
   929			}
   930	
   931			fp = kzalloc(sizeof(*fp), GFP_KERNEL);
   932			if (!fp)
   933				return -ENOMEM;
   934	
   935			fp->iface = iface_no;
   936			fp->altsetting = altno;
   937			fp->altset_idx = i;
   938			fp->endpoint = get_endpoint(alts, 0)->bEndpointAddress;
   939			fp->ep_attr = get_endpoint(alts, 0)->bmAttributes;
   940			fp->datainterval = snd_usb_parse_datainterval(chip, alts);
   941			fp->protocol = protocol;
   942			fp->maxpacksize = le16_to_cpu(get_endpoint(alts, 0)->wMaxPacketSize);
   943			fp->channels = num_channels;
   944			if (snd_usb_get_speed(dev) == USB_SPEED_HIGH)
   945				fp->maxpacksize = (((fp->maxpacksize >> 11) & 3) + 1)
   946						* (fp->maxpacksize & 0x7ff);
   947			fp->attributes = parse_uac_endpoint_attributes(chip, alts, protocol, iface_no);
   948			fp->clock = clock;
   949			INIT_LIST_HEAD(&fp->list);
   950	
   951			/* some quirks for attributes here */
   952			snd_usb_audioformat_attributes_quirk(chip, fp, stream);
   953	
   954			/* ok, let's parse further... */
   955			if (protocol == UAC_VERSION_1 || protocol == UAC_VERSION_2) {
 > 956				if (snd_usb_parse_audio_format(chip, fp, format,

957 fmt, stream) < 0) {
958 kfree(fp->rate_table);
959 kfree(fp);
960 fp = NULL;
961 continue;
962 }
963 } else {
964 struct uac3_as_header_descriptor *as;
965
966 as = snd_usb_find_csint_desc(alts->extra,
967 alts->extralen,
968 NULL, UAC_AS_GENERAL);
969
970 if (snd_usb_parse_audio_format_v3(chip, fp, as,
971 stream) < 0) {
972 kfree(fp->rate_table);
973 kfree(fp);
974 fp = NULL;
975 continue;
976 }
977 }
978
979 /* Create chmap */
980 if (fp->channels != num_channels)
981 chconfig = 0;
982
983 if (protocol == UAC_VERSION_3)
984 fp->chmap = chmap_v3;
985 else
986 fp->chmap = convert_chmap(fp->channels, chconfig,
987 protocol);
988
989 dev_dbg(&dev->dev, "%u:%d: add audio endpoint %#x\n", iface_no, altno, fp->endpoint);
990 err = snd_usb_add_audio_stream(chip, stream, fp);
991 if (err < 0) {
992 list_del(&fp->list); /* unlink for avoiding double-free */
993 kfree(fp->rate_table);
994 kfree(fp->chmap);
995 kfree(fp);
996 return err;
997 }
998 /* try to set the interface... */
999 usb_set_interface(chip->dev, iface_no, altno);
1000 snd_usb_init_pitch(chip, iface_no, alts, fp);
1001 snd_usb_init_sample_rate(chip, iface_no, alts, fp, fp->rate_max);
1002 }
1003 return 0;
1004 }
1005

---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation

#6 lore raw

Re: [PATCH v2 2/2] ALSA: usb: initial USB Audio Device Class 3.0 support

Quoted 19 lines · kbuild test robot

Hi Ruslan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on linus/master]
[also build test WARNING on v4.16-rc6 next-20180316]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url: https://github.com/0day-ci/linux/commits/Ruslan-Bilovol/ALSA-usb-audio-move-audioformat-quirks-to-quirks-c/20180319-205541
config: i386-randconfig-x016-201811 (attached as .config)
compiler: gcc-7 (Debian 7.3.0-1) 7.3.0 reproduce:

        # save the attached .config to linux build tree
        make ARCH=i386 

Note: it may well be a FALSE warning. FWIW you are at least aware of it now.
http://gcc.gnu.org/wiki/Better_Uninitialized_Warnings

Quoted 8 lines

All warnings (new ones prefixed by >>):

sound/usb/stream.c: In function 'snd_usb_parse_audio_interface':

Quoted 1 line
Quoted 1 line

sound/usb/stream.c:956:8: warning: 'fmt' may be used uninitialized in this function [-Wmaybe-uninitialized]

       if (snd_usb_parse_audio_format(chip, fp, format,
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
           fmt, stream) < 0) {
           ~~~~~~~~~~~~

Hm, this looks like a false-postive, although I see it on my local build, too.

I'm not always a fan of unneeded initialization, but this time, we can be lazy and add a NULL initialization...

thanks,

Takashi

#7 lore raw

Re: [PATCH v2 2/2] ALSA: usb: initial USB Audio Device Class 3.0 support

In reply to Ruslan Bilovol
Quoted 19 lines · Ruslan Bilovol

Recently released USB Audio Class 3.0 specification introduces many significant changes comparing to previous versions, like
- new Power Domains, support for LPM/L1
- new Cluster descriptor
- changed layout of all class-specific descriptors
- new High Capability descriptors
- New class-specific String descriptors
- new and removed units
- additional sources for interrupts
- removed Type II Audio Data Formats
- ... and many other things (check spec)

It also provides backward compatibility through multiple configurations, as well as requires mandatory support for BADD (Basic Audio Device Definition) on each ADC3.0 compliant device

Quoted 5 lines

This patch adds initial support of UAC3 specification that is enough for Generic I/O Profile (BAOF, BAIF) device support from BADD document.

Most code changes look fairly straightforward, and not breaking the UAC1/UAC2 stuff. So the stuff is good enough through a quick glance.

I suppose it's better to merge via sound git tree.
Greg, could you check include/linux/usb/* (and uapi) stuff and give an ack if it's OK?

In anyway, I'll check the specs and compare with the definitions in this patch.

And, just minor nitpicking regarding the code change:

it's better to drop the parentheses in a line like below

Review

Quoted 1 line
(quoted diff) +1 -0
+		if ((protocol == UAC_VERSION_1) || (protocol == UAC_VERSION_2)) {

that is,

		if (protocol == UAC_VERSION_1 || protocol == UAC_VERSION_2) {

The indentation became so deep in some functions, and it's better to refactor, but this can be done later once after all implementations are finished.

thanks,

Takashi

maintainer Acked-by
#8 lore raw

Re: [PATCH v2 2/2] ALSA: usb: initial USB Audio Device Class 3.0 support

Quoted 25 lines · Takashi Iwai
Quoted 23 lines · Ruslan Bilovol

Recently released USB Audio Class 3.0 specification introduces many significant changes comparing to previous versions, like
- new Power Domains, support for LPM/L1
- new Cluster descriptor
- changed layout of all class-specific descriptors
- new High Capability descriptors
- New class-specific String descriptors
- new and removed units
- additional sources for interrupts
- removed Type II Audio Data Formats
- ... and many other things (check spec)

It also provides backward compatibility through multiple configurations, as well as requires mandatory support for BADD (Basic Audio Device Definition) on each ADC3.0 compliant device

This patch adds initial support of UAC3 specification that is enough for Generic I/O Profile (BAOF, BAIF) device support from BADD document.

Quoted 8 lines
Quoted 1 line

Most code changes look fairly straightforward, and not breaking the UAC1/UAC2 stuff. So the stuff is good enough through a quick glance.

I suppose it's better to merge via sound git tree.
Greg, could you check include/linux/usb/* (and uapi) stuff and give an ack if it's OK?

#9 lore raw

Re: [PATCH v2 2/2] ALSA: usb: initial USB Audio Device Class 3.0 support

In reply to Takashi Iwai
Quoted 30 lines · Takashi Iwai
Quoted 19 lines · Ruslan Bilovol

Recently released USB Audio Class 3.0 specification introduces many significant changes comparing to previous versions, like
- new Power Domains, support for LPM/L1
- new Cluster descriptor
- changed layout of all class-specific descriptors
- new High Capability descriptors
- New class-specific String descriptors
- new and removed units
- additional sources for interrupts
- removed Type II Audio Data Formats
- ... and many other things (check spec)

It also provides backward compatibility through multiple configurations, as well as requires mandatory support for BADD (Basic Audio Device Definition) on each ADC3.0 compliant device

Quoted 5 lines

This patch adds initial support of UAC3 specification that is enough for Generic I/O Profile (BAOF, BAIF) device support from BADD document.

Most code changes look fairly straightforward, and not breaking the UAC1/UAC2 stuff. So the stuff is good enough through a quick glance.

Quoted 6 lines

I suppose it's better to merge via sound git tree.
Greg, could you check include/linux/usb/* (and uapi) stuff and give an ack if it's OK?

In anyway, I'll check the specs and compare with the definitions in this patch.

Please note that I took "A.12 CHANNEL RELATIONSHIP DEFINITIONS" values from previous Pierre's email (they were in a draft UAC3 spec but disappeared in the final version). I've sent an email with my quiestions to [email protected] but didn't get any reply from them yet.

Quoted 6 lines

And, just minor nitpicking regarding the code change:

it's better to drop the parentheses in a line like below

Quoted 8 lines
Quoted 1 line

+ if ((protocol == UAC_VERSION_1) || (protocol == UAC_VERSION_2)) {

that is,

                if (protocol == UAC_VERSION_1 || protocol == UAC_VERSION_2) {

The indentation became so deep in some functions, and it's better to refactor, but this can be done later once after all implementations are finished.

Sure, I fixed these and also kbuild test robot's complaint; if there will no any more comments, I'm going to send v3 tomorrow.

Thanks,
Ruslan