Re: usb problem with 5.2-RC1

On Tuesday 23 December 2003 12:26, Jesse Guardiani wrote:
> Kevin Oberman wrote:
> >> From: Jesse Guardiani <jesse_at_wingnet.net>
> >> Date: Tue, 23 Dec 2003 10:07:52 -0500
> >> Sender: owner-freebsd-current_at_freebsd.org
> >>
> >> Howdy list,
> >>
> >> Upon resuming from an APM induced suspended state as
> >> described here:
> >>
> >>    
> >> http://article.gmane.org/gmane.os.freebsd.current/36652/match=tracking+d
> >>own+ata0+problem
> >>
> >> I experience USB problems. Specifically, my mouse
> >> doesn't work and I get these messages from the kernel:
> >>
> >> Dec 23 09:35:50 trevarthan kernel: ums0: at uhub1 port 2 (addr 2)
> >> disconnected Dec 23 09:35:50 trevarthan kernel: ums0: detached
> >> Dec 23 09:35:58 trevarthan moused: unable to open /dev/ums0: No such
> >> file or directory Dec 23 09:37:09 trevarthan kernel: uhub1: device
> >> problem, disabling port 2 Dec 23 09:38:30 trevarthan kernel: uhub1:
> >> device problem, disabling port 2 Dec 23 09:39:52 trevarthan kernel:
> >> uhub1: device problem, disabling port 1 Dec 23 09:39:52 trevarthan
> >> kernel: uhub1: port error, restarting port 2 Dec 23 09:52:50 trevarthan
> >> kernel: uhub1: device problem, disabling port 2 Dec 23 09:55:10
> >> trevarthan kernel: uhub1: device problem, disabling port 2
> >
> > Yes, this is a well known problem with USB. Some fixes for several of the
> > USB drivers were committed last week, but the primary issues with the
> > USB and resume remain unresolved. Nate has posted a request for help in
> > finding this long-standing bug.
>
> How long-standing is it? Always worked great for me under 5.1-RELEASE.
>
> > The normal work-around is to build without USB and only load USB when
> > needed, but a USB mouse requires it all the time, so...
> >
> > USB also blocks the best power-saving mode (C3) on my system, so I
> > really hope that some of this can be tracked down soon. But, with the
> > latest commits to ACPI code I have finally switched to running with ACPI
> > most of the time. Things do continue to get better.
>
> Before I switched from 5.1-RELEASE to 5.2-BETA (and then -RC1 later) I got
> ACPI S3 working on my laptop. I just had to issue an `atacontrol detach 1`
> before suspending. APM worked great under 5.1-RELEASE also.
>
> S3 doesn't work at all under -RC1 for me. It suspends, but I get a blank
> screen on resume and no keyboard response, AFAICT. Have there been
> meaningful ACPI commits since -RC1?
>
setting the sysctl hw.acpi.reset_video to 1 or 0 should fix the problem.  My 
laptop needs 0, or the problem you describe happens, but others need 1 or the 
problem you describe also happens.
> So far my general impression is that 5.2 is a slight regression from 5.1
> and 5.1 was a slight regression in many ways from 4.x. I know a LOT has
> changed since 5.1 and I can appreciate that. After all, my external
> USB->ATAPI 2.5" usb hard disk enclosure works for the first time in
> 5.2-BETA and -RC1 (it didn't work in 4.8, or 5.1). Also, I'm incredibly
> excited about all the kernel threading work going on right now. I run many
> dual CPU machines at work and I can't wait to try out a stable SMPng setup
> on them.
>
> But I'm craving stability at this point. The bleeding edge is getting
> uncomfortable.
>
> :)
There is a reason it's called CURRENT ;)  Hang in there, there is lots of work 
going on and my current system has only a couple of major outstanding 
problems which is down from 4 or 5 just a couple of months ago.  As it's been 
said, things ARE getting better.

-- 
Anish Mistry
amistry_at_am-productions.biz
AM Productions