linux-kernel - Re: bad endpoint address, MOTU FastLane

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4A1A54DB.3090902@ladisch.de>
Date:	Mon, 25 May 2009 10:20:43 +0200
From:	Clemens Ladisch <clemens@...isch.de>
To:	David Fries <david@...es.net>
CC:	David Griffith <dgriffi@...csubak.edu>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Alan Stern <stern@...land.harvard.edu>,
	linux-kernel@...r.kernel.org, linux-usb@...r.kernel.org
Subject: Re: bad endpoint address, MOTU FastLane

David Fries wrote:
> On Tue, May 12, 2009 at 09:55:08AM +0200, Clemens Ladisch wrote:
> > The driver could call usb_set_interface() again for the first interface
> > so that the USB core takes notice of the first set of endpoints again.
> > Please try the patch below.
> > 
> > I guess I'll have to write another quirk for this.
> 
> Adding usb_set_interface lets the device work again, ...
> 
> The MOTU Fastlane device is the only one using QUIRK_MIDI_RAW.  How about
> changing the name to QUIRK_MIDI_FASTLANE to imply it isn't completely
> generic anymore and adding a comment?

And while we're at it, the driver should claim interface 1 to prevent
anyone else from using it.

New patch below.

Are you okay with the tested-by tag?

===========

sound: usb-audio: make the MotU Fastlane work again

Kernel 2.6.18 broke the MotU Fastlane, which uses duplicate endpoint
numbers in a manner that is not only illegal but also confuses the
kernel's endpoint descriptor caching mechanism.  To work around this, we
have to add a separate usb_set_interface() call to guide the USB core to
the correct descriptors.

Signed-off-by: Clemens Ladisch <clemens@...isch.de>
Reported-and-tested-by: David Fries <david@...es.net>

--- linux/sound/usb/usbaudio.h
+++ linux/sound/usb/usbaudio.h
@@ -153,7 +153,7 @@ enum quirk_type {
 	QUIRK_MIDI_YAMAHA,
 	QUIRK_MIDI_MIDIMAN,
 	QUIRK_MIDI_NOVATION,
-	QUIRK_MIDI_RAW,
+	QUIRK_MIDI_FASTLANE,
 	QUIRK_MIDI_EMAGIC,
 	QUIRK_MIDI_CME,
 	QUIRK_MIDI_US122L,
--- linux/sound/usb/usbquirks.h
+++ linux/sound/usb/usbquirks.h
@@ -1868,11 +1868,11 @@ YAMAHA_DEVICE(0x7010, "UB99"),
 		.data = & (const struct snd_usb_audio_quirk[]) {
 			{
 				.ifnum = 0,
-				.type = QUIRK_MIDI_RAW
+				.type = QUIRK_MIDI_FASTLANE
 			},
 			{
 				.ifnum = 1,
-				.type = QUIRK_IGNORE_INTERFACE
+				.type = QUIRK_MIDI_FASTLANE
 			},
 			{
 				.ifnum = -1
--- linux/sound/usb/usbaudio.c
+++ linux/sound/usb/usbaudio.c
@@ -3359,7 +3359,7 @@ static int snd_usb_create_quirk(struct s
 		[QUIRK_MIDI_YAMAHA] = snd_usb_create_midi_interface,
 		[QUIRK_MIDI_MIDIMAN] = snd_usb_create_midi_interface,
 		[QUIRK_MIDI_NOVATION] = snd_usb_create_midi_interface,
-		[QUIRK_MIDI_RAW] = snd_usb_create_midi_interface,
+		[QUIRK_MIDI_FASTLANE] = snd_usb_create_midi_interface,
 		[QUIRK_MIDI_EMAGIC] = snd_usb_create_midi_interface,
 		[QUIRK_MIDI_CME] = snd_usb_create_midi_interface,
 		[QUIRK_AUDIO_STANDARD_INTERFACE] = create_standard_audio_quirk,

--- linux/sound/usb/usbmidi.c
+++ linux/sound/usb/usbmidi.c
@@ -1573,6 +1573,33 @@ static int snd_usbmidi_detect_yamaha(str
 }
 
 /*
+ * Detects the endpoints and ports of the MotU Fastlane, and handles the
+ * interfaces' configuration.
+ */
+static int snd_usbmidi_detect_fastlane(struct snd_usb_midi *umidi,
+				       struct snd_usb_midi_endpoint_info *eps)
+{
+	switch (get_iface_desc(umidi->iface->altsetting)->bInterfaceNumber) {
+	case 0:
+		/*
+		 * Interface 1 contains isochronous endpoints, but with the same
+		 * numbers as in interface 0.  Since it is interface 1 that the
+		 * USB core has most recently seen, these descriptors are now
+		 * associated with the endpoint numbers.  This will foul up our
+		 * attempts to submit bulk/interrupt URBs to the endpoints in
+		 * interface 0, so we have to make sure that the USB core looks
+		 * again at interface 0 by calling usb_set_interface() on it.
+		 */
+		usb_set_interface(umidi->chip->dev, 0, 0);
+		return snd_usbmidi_detect_per_port_endpoints(umidi, eps);
+	case 1:
+		return 1; /* no MIDI device for this interface, just claim it */
+	default:
+		return -ENXIO;
+	}
+}
+
+/*
  * Creates the endpoints and their ports for Midiman devices.
  */
 static int snd_usbmidi_create_endpoints_midiman(struct snd_usb_midi* umidi,
@@ -1778,9 +1805,13 @@ int snd_usb_create_midi_interface(struct
 		umidi->usb_protocol_ops = &snd_usbmidi_novation_ops;
 		err = snd_usbmidi_detect_per_port_endpoints(umidi, endpoints);
 		break;
-	case QUIRK_MIDI_RAW:
+	case QUIRK_MIDI_FASTLANE:
 		umidi->usb_protocol_ops = &snd_usbmidi_raw_ops;
-		err = snd_usbmidi_detect_per_port_endpoints(umidi, endpoints);
+		err = snd_usbmidi_detect_fastlane(umidi, endpoints);
+		if (err == 1) {
+			kfree(umidi);
+			return 0;
+		}
 		break;
 	case QUIRK_MIDI_EMAGIC:
 		umidi->usb_protocol_ops = &snd_usbmidi_emagic_ops;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/