Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 2016-03-21 at 11:46 -0400, Alan Stern wrote: > On Mon, 21 Mar 2016, Oliver Neukum wrote: > > > On Fri, 2016-03-18 at 12:36 -0400, Alan Stern wrote: > > > > Almost. In case of reset_resume() it makes no sense to still > > clear a halt or execute a reset that had been ordered. The > >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 21 Mar 2016, Oliver Neukum wrote: > On Fri, 2016-03-18 at 12:36 -0400, Alan Stern wrote: > > > @@ -1584,10 +1581,8 @@ static int hid_resume(struct usb_interfa > > static int hid_reset_resume(struct usb_interface *intf) > > { > > struct hid_device *hid = usb_get_intfdata(intf); > >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 2016-03-18 at 12:36 -0400, Alan Stern wrote: > @@ -1584,10 +1581,8 @@ static int hid_resume(struct usb_interfa > static int hid_reset_resume(struct usb_interface *intf) > { > struct hid_device *hid = usb_get_intfdata(intf); > - struct usbhid_device *usbhid = hid->driver_data;

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 18 Mar 2016 12:36:42 -0400 (EDT) Alan Stern wrote: > All right, I have taken Oliver's suggestion. The patch below refactors > the code to consolidate the common activities in a new function, > hid_restart_io(). > > Daniel, can you please test this patch?

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 14 Mar 2016, Oliver Neukum wrote: > On Mon, 2016-03-14 at 11:03 -0400, Alan Stern wrote: > > On Mon, 14 Mar 2016, Oliver Neukum wrote: > > > > It seems to me that hid_resume_common() needs to be split into three > > > parts > > > > > > a) doing the stuff for pending resets > > > b)

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 2016-03-14 at 11:03 -0400, Alan Stern wrote: > On Mon, 14 Mar 2016, Oliver Neukum wrote: > > It seems to me that hid_resume_common() needs to be split into three > > parts > > > > a) doing the stuff for pending resets > > b) conditionally restarting IO > > c) reset the flag > >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 14 Mar 2016, Oliver Neukum wrote: > On Fri, 2016-03-11 at 11:54 -0500, Alan Stern wrote: > > > Perhaps I didn't do a good job of addressing the real underlying > > problem. That's why I asked if this was the right approach. > > > > In Daniel's case, usbhid_start() never got called.

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 2016-03-11 at 11:54 -0500, Alan Stern wrote: > Perhaps I didn't do a good job of addressing the real underlying > problem. That's why I asked if this was the right approach. > > In Daniel's case, usbhid_start() never got called. Perhaps that's the > real problem? Anyway, this meant

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 11 Mar 2016, Jiri Kosina wrote: > On Wed, 2 Mar 2016, Alan Stern wrote: > > [ ... snip ... ] > > The patch that fixed Daniel's problem is repeated below for your > > convenience. Is this the right approach? If you think it is, I'll > > submit it officially. > > I've been thinking

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Wed, 2 Mar 2016, Alan Stern wrote: [ ... snip ... ] > The patch that fixed Daniel's problem is repeated below for your > convenience. Is this the right approach? If you think it is, I'll > submit it officially. I've been thinking about this, and while I don't really like such bandaids

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016, Daniel Fraga wrote: > On Tue, 1 Mar 2016 17:15:56 -0500 (EST) > Alan Stern wrote: > > > Don't worry about the Elan driver. Instead, let's see if this patch > > fixes the problem. > > Yes, this patch fixed the problem. I can suspend and

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016 17:15:56 -0500 (EST) Alan Stern wrote: > Don't worry about the Elan driver. Instead, let's see if this patch > fixes the problem. Yes, this patch fixed the problem. I can suspend and resume without those repeated "reset" messages ;)

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016, Daniel Fraga wrote: > On Tue, 1 Mar 2016 15:55:00 -0500 (EST) > Alan Stern wrote: > > > No messages about "usbhid_start urb" or "no input endpoint!" or > > "usbhid_start fail urb"? That means usbhid_start() isn't getting > > called. Which means

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016 15:55:00 -0500 (EST) Alan Stern wrote: > No messages about "usbhid_start urb" or "no input endpoint!" or > "usbhid_start fail urb"? That means usbhid_start() isn't getting > called. Which means the device in question probably isn't being used >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016, Daniel Fraga wrote: > On Tue, 1 Mar 2016 10:25:30 -0500 (EST) > Alan Stern wrote: > > > Now we're making progress! That shows a problem right there; we ought > > to have more stuff about 3-1.6 between those two lines. > > > > The next patch adds

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016 10:25:30 -0500 (EST) Alan Stern wrote: > Now we're making progress! That shows a problem right there; we ought > to have more stuff about 3-1.6 between those two lines. > > The next patch adds some more debugging output. For this test you >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 1 Mar 2016 10:25:30 -0500 (EST) Alan Stern wrote: > Now we're making progress! That shows a problem right there; we ought > to have more stuff about 3-1.6 between those two lines. > > The next patch adds some more debugging output. For this test you >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 29 Feb 2016, Daniel Fraga wrote: > On Mon, 29 Feb 2016 16:28:40 -0500 (EST) > Alan Stern wrote: > > > Okay, that's what I had guessed. Somehow usbhid->urbin is getting set > > to NULL. Maybe the patch below will indicate why. When you post the > > log,

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 29 Feb 2016 16:28:40 -0500 (EST) Alan Stern wrote: > Okay, that's what I had guessed. Somehow usbhid->urbin is getting set > to NULL. Maybe the patch below will indicate why. When you post the > log, include everything that mentions the 3-1.6 device --

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Wed, 24 Feb 2016, Daniel Fraga wrote: > On Wed, 24 Feb 2016 14:24:44 -0500 (EST) > Alan Stern wrote: > > > I intended the patch not to cause any call traces, but it did anyway. > > So let's drop the questionable code and try something that will be > > completely

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Wed, 24 Feb 2016 14:24:44 -0500 (EST) Alan Stern wrote: > I intended the patch not to cause any call traces, but it did anyway. > So let's drop the questionable code and try something that will be > completely safe. Ok, here's what I got: Feb 24 19:16:41

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 23 Feb 2016, Daniel Fraga wrote: > On Tue, 23 Feb 2016 11:48:51 -0500 (EST) > Alan Stern wrote: > > > I don't see any crash in that photo. > > Sorry. By "crash" I mean "unable to use the keyboard". Maybe I > used the wrong term. > > > Also, it looks

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 23 Feb 2016 11:48:51 -0500 (EST) Alan Stern wrote: > I don't see any crash in that photo. Sorry. By "crash" I mean "unable to use the keyboard". Maybe I used the wrong term. > Also, it looks like the log results in this run were different from > what

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 23 Feb 2016, Daniel Fraga wrote: > On Tue, 23 Feb 2016 10:19:30 -0500 (EST) > Alan Stern wrote: > > > All right; let's try a slightly different approach that shouldn't cause > > any crashes at all. > > Unfortunately it still crashed with this new

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 23 Feb 2016 10:19:30 -0500 (EST) Alan Stern wrote: > All right; let's try a slightly different approach that shouldn't cause > any crashes at all. Unfortunately it still crashed with this new patch. I took a picture: http://imgur.com/OJCB129 --

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 22 Feb 2016, Daniel Fraga wrote: > On Mon, 22 Feb 2016 16:22:59 -0500 (EST) > Alan Stern wrote: > > > Unfortunately I really need to see the stuff that shows up before the > > first couple of pictures. Is there any way you can use a serial > > console or

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 22 Feb 2016 16:22:59 -0500 (EST) Alan Stern wrote: > Unfortunately I really need to see the stuff that shows up before the > first couple of pictures. Is there any way you can use a serial > console or network console to capture the log data?

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 22 Feb 2016, Daniel Fraga wrote: > On Mon, 22 Feb 2016 14:30:32 -0500 (EST) > Alan Stern wrote: > > > Well, I'm still puzzled. I tried running that patch on my system > > (under 4.5-rc2) and it worked perfectly. > > > > So let's try for a little more detail.

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Mon, 22 Feb 2016 14:30:32 -0500 (EST) Alan Stern wrote: > Well, I'm still puzzled. I tried running that patch on my system > (under 4.5-rc2) and it worked perfectly. > > So let's try for a little more detail. Please apply this patch instead > of the earlier one.

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 19 Feb 2016, Alan Stern wrote: > On Fri, 19 Feb 2016, Daniel Fraga wrote: > > > On Fri, 19 Feb 2016 14:13:25 -0500 (EST) > > Alan Stern wrote: > > > > > -22 is -EINVAL, so we need to figure out which of the many possible > > > EINVAL errors you're getting.

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 19 Feb 2016, Daniel Fraga wrote: > On Fri, 19 Feb 2016 14:13:25 -0500 (EST) > Alan Stern wrote: > > > -22 is -EINVAL, so we need to figure out which of the many possible > > EINVAL errors you're getting. Try the patch below. > > I applied your patch,

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Fri, 19 Feb 2016 14:13:25 -0500 (EST) Alan Stern wrote: > -22 is -EINVAL, so we need to figure out which of the many possible > EINVAL errors you're getting. Try the patch below. I applied your patch, but when it wakes up from S3, the system is stalled:

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Thu, 18 Feb 2016, Daniel Fraga wrote: > On Thu, 18 Feb 2016 15:23:00 -0500 (EST) > Alan Stern wrote: > > > Something like the patch below (untested). > > > + dev_info(>dev, "post reset hid_start_in -> %d\n", status); > > Ok, so I got the following: > >

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Thu, 18 Feb 2016 15:23:00 -0500 (EST) Alan Stern wrote: > Something like the patch below (untested). > + dev_info(>dev, "post reset hid_start_in -> %d\n", status); Ok, so I got the following: Feb 18 19:22:26 tux kernel: [ 258.693120] usb 3-1.6: reset

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Thu, 18 Feb 2016, Daniel Fraga wrote: > On Thu, 18 Feb 2016 14:30:15 -0500 (EST) > Alan Stern wrote: > > > It looks like there's some problem in the usbhid driver. Apparently > > hid_start_in() gets some sort of error when it submits the input URB, > > because

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Thu, 18 Feb 2016 14:30:15 -0500 (EST) Alan Stern wrote: > It looks like there's some problem in the usbhid driver. Apparently > hid_start_in() gets some sort of error when it submits the input URB, > because that URB doesn't show up in the usbmon output. > > Can

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Thu, 18 Feb 2016, Daniel Fraga wrote: > On Thu, 18 Feb 2016 11:18:37 -0500 (EST) > Alan Stern wrote: > > > No idea. Can you collect a usbmon trace for bus 3, starting before the > > suspend and showing the problems? > > Ok Alan, I attached the requested

Re: "reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

On Tue, 9 Feb 2016, Dâniel Fraga wrote: > I'm using Linux 4.4.1 kernel on a Dell Inspiron 15R 5537 and > everytime I wake up from S3 sleep, the syslog is full of the following > message: > > Feb 9 17:19:05 tux kernel: [19196.253206] usb 3-1.6: reset full-speed USB > device number 6 using

"reset full-speed USB device number 6 using ehci-pci" with Dell Inspiron 15R 5537

I'm using Linux 4.4.1 kernel on a Dell Inspiron 15R 5537 and everytime I wake up from S3 sleep, the syslog is full of the following message: Feb 9 17:19:05 tux kernel: [19196.253206] usb 3-1.6: reset full-speed USB device number 6 using ehci-pci Feb 9 17:19:05 tux kernel: