NetBSD Problem Report #53493

From www@NetBSD.org  Thu Aug  2 03:09:55 2018
Return-Path: <www@NetBSD.org>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits))
	(Client CN "mail.NetBSD.org", Issuer "mail.NetBSD.org CA" (not verified))
	by mollari.NetBSD.org (Postfix) with ESMTPS id C560E7A18A
	for <gnats-bugs@gnats.NetBSD.org>; Thu,  2 Aug 2018 03:09:54 +0000 (UTC)
Message-Id: <20180802030946.A08767A19B@mollari.NetBSD.org>
Date: Thu,  2 Aug 2018 03:09:46 +0000 (UTC)
From: scole_mail@gmx.com
Reply-To: scole_mail@gmx.com
To: gnats-bugs@NetBSD.org
Subject: mouse keeps re-attaching using amd64 8.0 boot.iso CD
X-Send-Pr-Version: www-1.0

>Number:         53493
>Category:       kern
>Synopsis:       mouse keeps re-attaching using amd64 8.0 boot.iso CD
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Aug 02 03:10:00 +0000 2018
>Last-Modified:  Fri Aug 03 03:50:01 +0000 2018
>Originator:     scole
>Release:        8.0
>Organization:
none
>Environment:
NetBSD localhost 8.0 NetBSD 8.0 (GENERIC) #0: Tue Jul 17 14:59:51 UTC 2018  mkrepro@mkrepro.NetBSD.org:/usr/src/sys/arch/amd64/compile/GENERIC amd64
>Description:
When using this CD, ftp://ftp.netbsd.org/pub/NetBSD/NetBSD-8.0/amd64/installation/cdrom/boot.iso
and the shell in the utility menu, the mouse keeps reattaching itself.  Messages get displayed repeatedly in the console like this:

wsmouse0: detached
ums0: detached
uhidev0: detached
uhidev0: at uhub1 port 2 (addr 2) disconnected
uhidev0 at uhub1 port 2 configuration 1 interface 0
uhidev0: Genius (0x458) USB Optical Mouse (0xec), rev 1.10/1.00, addr 2, iclass 3/1
ums0 at uhidev0: 3 buttons and Z dir
wsmouse0 at ums0 mux 0

It seems to happen every minute or so, and but doesn't seem fixed periodic.  This happens to me on a couple different machines.  There is a dmesg from one that had the problem at http://gnats.netbsd.org/53492

The machines currently available to me aren't mine so I can't test any installs on them.

Also, the boot loader and kernel wasn't detecting more the 4G of memory on the 2 different machines I tried that had 8G.   I wasn't sure if that's intentional.  Let me know if a separate PR would be useful for that.

Thanks
>How-To-Repeat:
boot amd64 8.0 boot.iso CD, got to Utility Menu -> /bin/sh
>Fix:

>Audit-Trail:
From: Paul Goyette <paul@whooppee.com>
To: gnats-bugs@NetBSD.org
Cc: kern-bug-people@netbsd.org, gnats-admin@netbsd.org, netbsd-bugs@netbsd.org
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso
 CD
Date: Thu, 2 Aug 2018 11:22:46 +0800 (+08)

 > When using this CD,
 > ftp://ftp.netbsd.org/pub/NetBSD/NetBSD-8.0/amd64/installation/cdrom/boot.iso
 > and the shell in the utility menu, the mouse keeps reattaching itself.
 > Messages get displayed repeatedly in the console like this:
 >
 > wsmouse0: detached
 > ums0: detached
 > uhidev0: detached
 > uhidev0: at uhub1 port 2 (addr 2) disconnected
 > uhidev0 at uhub1 port 2 configuration 1 interface 0
 > uhidev0: Genius (0x458) USB Optical Mouse (0xec), rev 1.10/1.00, addr 2, iclass 3/1
 > ums0 at uhidev0: 3 buttons and Z dir
 > wsmouse0 at ums0 mux 0
 >
 > It seems to happen every minute or so, and but doesn't seem fixed
 > periodic.  This happens to me on a couple different machines.  There
 > is a dmesg from one that had the problem at
 > http://gnats.netbsd.org/53492

 With kernel timestamps (under -current) I see the same sequence, over
 and over.  The timing is at _very_ consistent intervals, with the
 reconnect happening at ~2.2 seconds after the disconnect, and the
 disconnect happens almost exactly 60 seconds later.

 ...
 Aug  2 11:13:24 speedy /netbsd: [ 531029.5503589] wsmouse0: detached
 Aug  2 11:13:24 speedy /netbsd: [ 531029.5503589] ums0: detached
 Aug  2 11:13:24 speedy /netbsd: [ 531029.5503589] uhidev2: detached
 Aug  2 11:13:24 speedy /netbsd: [ 531029.5503589] whidev2: at uhub1 port 6 (addr 6) disconnected
 Aug  2 11:13:26 speedy /netbsd: [ 531031.7310457] uhidev2 at uhub1 port 6 configuration 1 interface 0
 Aug  2 11:13:26 speedy /netbsd: [ 531031.7310457] uhidev2: PixArt (0x45e) Microsoft USB Optical Mouse (0xcb), rev 1.10/1.00, addr 7, iclass 3/1
 Aug  2 11:13:26 speedy /netbsd: [ 531031.7310457] ums0 at uhidev2: 3 buttons and Z dir
 Aug  2 11:13:26 speedy /netbsd: [ 531031.7310457] wsmouse0 at ums0 mux 0
 Aug  2 11:14:26 speedy /netbsd: [ 531091.7299334] wsmouse0: detached
 Aug  2 11:14:26 speedy /netbsd: [ 531091.7299334] ums0: detached
 Aug  2 11:14:26 speedy /netbsd: [ 531091.7299334] uhidev2: detached
 Aug  2 11:14:26 speedy /netbsd: [ 531091.7299334] uhidev2: at uhub1 port 6 (addr 7) disconnected
 Aug  2 11:14:29 speedy /netbsd: [ 531093.9106198] uhidev2 at uhub1 port 6 configuration 1 interface 0
 Aug  2 11:14:29 speedy /netbsd: [ 531093.9106198] uhidev2: PixArt (0x45e) Microsoft USB Optical Mouse (0xcb), rev 1.10/1.00, addr 8, iclass 3/1
 Aug  2 11:14:29 speedy /netbsd: [ 531093.9106198] ums0 at uhidev2: 3 buttons and Z dir
 Aug  2 11:14:29 speedy /netbsd: [ 531093.9106198] wsmouse0 at ums0 mux 0
 ...

 This has been happening to me for at least a year or more, and was
 reported several times (although no PR files).  I know that others
 have also seen this.

 Note that this does NOT happen while the Xserver has ownership of the
 console.  Some suspicion is that the Xserver does something(tm) that
 performs some sort of keep-alive on the mouse's USB connection, and
 that without the Xserver doing the keep-alive the connection times
 out.  But no real investigation was done, as far as I know.

 (Thanks for filing the PR - perhaps we can get some traction!)

 > Also, the boot loader and kernel wasn't detecting more the 4G of
 > memory on the 2 different machines I tried that had 8G.   I wasn't
 > sure if that's intentional.  Let me know if a separate PR would be
 > useful for that.

 Yes, that would be a separate issue.  Please file a new PR.


 +------------------+--------------------------+----------------------------+
 | Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:          |
 | (Retired)        | FA29 0E3B 35AF E8AE 6651 | paul at whooppee dot com   |
 | Kernel Developer | 0786 F758 55DE 53BA 7731 | pgoyette at netbsd dot org |
 +------------------+--------------------------+----------------------------+

From: Robert Elz <kre@munnari.OZ.AU>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 02 Aug 2018 11:13:03 +0700

     Date:        Thu,  2 Aug 2018 03:10:00 +0000 (UTC)
     From:        scole_mail@gmx.com
     Message-ID:  <20180802031000.AB76A7A1CF@mollari.NetBSD.org>

   | >Synopsis:       mouse keeps re-attaching using amd64 8.0 boot.iso CD

   | When using this CD, ftp://ftp.netbsd.org/pub/NetBSD/NetBSD-8.0/amd64/installation/cdrom/boot.iso
   | and the shell in the utility menu, the mouse keeps reattaching itself. 

 I have seen this too, but not from the boot CD, just on a regular running
 kernel, with a USB wireless mouse/keyboard dongle.   As long as the
 mouse is powered on, and nothing has the device open, it detaches
 and reattaches like that (or did, I just replaced that mouse/keyboard a
 couple of days ago - not because of this, but because the keyboard was
 starting to develop "sticky keys") - I don't know yet if the new one acts
 the same (normally I boot it, and have X started quickly enough that
 this doesn't happen.)

 It is really no more than a nuisance - the mouse still works fine.   It is
 just spam on the console and in syslog (which is why I never attempted
 to discover whether it was something the mouse was doing, or the kernel.)

 That system is currently running 8.99.1 (ie: just after -8 was branched) but
 I have observed this since it was installed in the 7.99.29 timeframe (May 
 2016).

 kre

From: mlelstv@serpens.de (Michael van Elst)
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 2 Aug 2018 04:39:22 -0000 (UTC)

 paul@whooppee.com (Paul Goyette) writes:

 >With kernel timestamps (under -current) I see the same sequence, over
 >and over.  The timing is at _very_ consistent intervals, with the
 >reconnect happening at ~2.2 seconds after the disconnect, and the
 >disconnect happens almost exactly 60 seconds later.

 That's no bug. The mouse indeed does indeed disconnect and reconnect
 until it is actually accessed, that's a feature of most mice today.


 -- 
 -- 
                                 Michael van Elst
 Internet: mlelstv@serpens.de
                                 "A potential Snark may lurk in every tree."

From: scole_mail <scole_mail@gmx.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Wed, 01 Aug 2018 23:02:09 -0700

 Paul Goyette <paul@whooppee.com> writes:

 >  
 >  (Thanks for filing the PR - perhaps we can get some traction!)
 >  
 >  > Also, the boot loader and kernel wasn't detecting more the 4G of
 >  > memory on the 2 different machines I tried that had 8G.   I wasn't
 >  > sure if that's intentional.  Let me know if a separate PR would be
 >  > useful for that.
 >  
 >  Yes, that would be a separate issue.  Please file a new PR.
 >  

 This RAM detection issue can be disregarded.  Some of these machines
 actually have less then 8G so it looks like they are all being detected
 correctly.

From: scole_mail <scole_mail@gmx.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Wed, 01 Aug 2018 23:11:47 -0700

 mlelstv@serpens.de (Michael van Elst) writes:

 >  
 >  That's no bug. The mouse indeed does indeed disconnect and reconnect
 >  until it is actually accessed, that's a feature of most mice today.
 >  

 Hmm, I've been using the install CD shell for its ssh and it is kind of
 disruptive with all the messages.  Can I disable one of the devices with
 "boot -c" to prevent the them or switch to another terminal, again this
 is with the boot CD?

From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 2 Aug 2018 08:34:12 +0200

 On Thu, Aug 02, 2018 at 04:40:01AM +0000, Michael van Elst wrote:
 >  That's no bug. The mouse indeed does indeed disconnect and reconnect
 >  until it is actually accessed, that's a feature of most mice today.

 What is the downside of the kernel driver always opening the mouse (and
 what does this actually mean in USB terms)?

 Martin

From: mlelstv@serpens.de (Michael van Elst)
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 2 Aug 2018 11:14:45 -0000 (UTC)

 scole_mail@gmx.com (scole_mail) writes:

 > mlelstv@serpens.de (Michael van Elst) writes:
 > 
 > >  
 > >  That's no bug. The mouse indeed does indeed disconnect and reconnect
 > >  until it is actually accessed, that's a feature of most mice today.
 > >  
 > 
 > Hmm, I've been using the install CD shell for its ssh and it is kind of
 > disruptive with all the messages.  Can I disable one of the devices with
 > "boot -c" to prevent the them or switch to another terminal, again this
 > is with the boot CD?

 Disabling the mouse driver would just attach/detach the generic driver (ugen).
 But you can access the mouse. Escape to a shell and run something like

 cat </dev/wsmouse >/dev/null

 in the background. I hope there is a /dev/wsmouse node, otherwise you
 have to create it with mknod too.

 -- 
 -- 
                                 Michael van Elst
 Internet: mlelstv@serpens.de
                                 "A potential Snark may lurk in every tree."

From: mlelstv@serpens.de (Michael van Elst)
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 2 Aug 2018 11:21:34 -0000 (UTC)

 martin@duskware.de (Martin Husemann) writes:

 > On Thu, Aug 02, 2018 at 04:40:01AM +0000, Michael van Elst wrote:
 > >  That's no bug. The mouse indeed does indeed disconnect and reconnect
 > >  until it is actually accessed, that's a feature of most mice today.
 > 
 > What is the downside of the kernel driver always opening the mouse (and
 > what does this actually mean in USB terms)?

 This would be some magic handling of some USB devices. In USB terms it is
 probably an open pipe that prevents the detach/attach-orgy.

 I first thought this were some kind of power-saving mode, but I found some
 comments that it is meant to help detecting the mouse after a suspend or
 hibernation.

 N.B. for sysinst it might be helpful to run a console redirection (TIOCCONS)
 to avoid disruption of the display.

 -- 
 -- 
                                 Michael van Elst
 Internet: mlelstv@serpens.de
                                 "A potential Snark may lurk in every tree."

From: scole_mail <scole_mail@gmx.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/53493: mouse keeps re-attaching using amd64 8.0 boot.iso CD
Date: Thu, 02 Aug 2018 20:46:51 -0700

 mlelstv@serpens.de (Michael van Elst) writes:

 >  
 >  Disabling the mouse driver would just attach/detach the generic driver (ugen).
 >  But you can access the mouse. Escape to a shell and run something like
 >  
 >  cat </dev/wsmouse >/dev/null
 >  
 >  in the background. I hope there is a /dev/wsmouse node, otherwise you
 >  have to create it with mknod too.
 >  

 That trick worked with an added "&", Thanks!

 /dev/wsmouse was already there.

 It would be nice though if it could be made unnecessary.

 Thank you

Home	PR Database Search