drivers/video/udlfb bind framebuffer to interface.

From: Kay Sievers <redacted>

Udlfb has been binding the framebuffer device to its parent, which
isn't correct and causes confusion with operations like udev remove.

Coming plug and play multiseat support is dependent on this fix.

Signed-off-by: Bernie Thompson <bernie@plugable.com>
---
 drivers/video/udlfb.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/video/udlfb.c b/drivers/video/udlfb.c
index 816a4fd..c6584c9 100644
--- a/drivers/video/udlfb.c
+++ b/drivers/video/udlfb.c

@@ -1549,7 +1549,7 @@ static int dlfb_usb_probe(struct usb_interface *interface,
 	/* We don't register a new USB class. Our client interface is fbdev */
 
 	/* allocates framebuffer driver structure, not framebuffer memory */
-	info = framebuffer_alloc(0, &usbdev->dev);
+	info = framebuffer_alloc(0, &interface->dev);
 	if (!info) {
 		retval = -ENOMEM;
 		pr_err("framebuffer_alloc failed\n");

-- 
1.7.4.1

Hi, Bernie Thompson.

-----Original Message-----
From: <redacted> [mailto:linux-fbdev-
owner@vger.kernel.org] On Behalf Of bernie@plugable.com
Sent: Wednesday, July 06, 2011 9:04 AM
To: <redacted>
Cc: <redacted>; Kay Sievers; Bernie Thompson
Subject: [PATCH] drivers/video/udlfb bind framebuffer to interface.

From: Kay Sievers <redacted>

Udlfb has been binding the framebuffer device to its parent, which
isn't correct and causes confusion with operations like udev remove.

Coming plug and play multiseat support is dependent on this fix.

Signed-off-by: Bernie Thompson <bernie@plugable.com>

This patch author seems to be Kay Sievers [off-list ref].
Should you add Signed-off-by: Kay Sievers [off-list ref] ?

ÿôèº{.nÇ+‰·Ÿ®‰­†+%ŠËÿ±éݶ¥Šwÿº{.nÇ+‰·¥Š{±ıöİzÿâ�Ø^n‡r¡ö¦zË�ëh™¨è­Ú&£ûàz¿äz¹Ş—ú+€Ê+zf£¢·hšˆ§~†­†Ûiÿÿï�êÿ‘êçz_è®æj:+v‰¨ş)ߣøm

On Tue, Jul 05, 2011 at 05:04:11PM -0700, bernie@plugable.com wrote:

Udlfb has been binding the framebuffer device to its parent, which
isn't correct and causes confusion with operations like udev remove.

Coming plug and play multiseat support is dependent on this fix.

Signed-off-by: Bernie Thompson <bernie@plugable.com>

On Sun, Jul 10, 2011 at 12:30:00AM -0700, bernie@plugable.com wrote:

Match udlfb only against vendor-specific class (e.g. only DisplayLink
graphics, not composite standard audio class interfaces). This enables
compatibility with composite graphics+audio devices (e.g. HDMI).

Match udlfb only against compatible subclass 0 and protocol 0 chips.
DisplayLink's USB 3.0 generation chips increment these values
to signal that they have a incompatible protocol, preventing udlfb
from erroneously matching to hardware it does not support.

Tested to confirm proper behavior on both USB 2.0 and USB 3.0
generation devices.

Reported-by: Andrew Kephart <redacted>
Signed-off-by: Bernie Thompson <bernie@plugable.com>

Both queued for 3.0, thanks.