Date: Fri, 09 Aug 2002 08:22:31 -0700 From: "Kevin Oberman" <oberman@es.net> To: Andy Sparrow <spadger@best.com> Cc: "Michael W. Collette" <metrol@metrol.net>, FreeBSD Mailing Lists <freebsd-mobile@FreeBSD.ORG> Subject: Re: Thinkpad T23 Coming Out of Sleep Message-ID: <20020809152231.289135D04@ptavv.es.net> In-Reply-To: Your message of "Thu, 08 Aug 2002 22:02:21 EDT." <20020809020221.684967C@CRWdog.demon.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
Andy, please wrap your lines at some reasonable length (72-75)! > Date: Thu, 08 Aug 2002 22:02:21 -0400 > From: Andy Sparrow <spadger@best.com> > Sender: owner-freebsd-mobile@FreeBSD.ORG > > > Running into a very intermittent problem here with my otherwise perfectly > > working laptop. When bringing it out of sleep mode, the system locks, hangs > > for about 30-50 seconds, then commences with a reboot all by itself. > > I've seen this about once every 2-3 weeks on my Omnibook 6000 which > generally gets suspended/resumed at least twice a day (and sometimes > as many as 10-15, if the trackpoint is a little recaltriant coming > out of sleep mode - I /hate/ the touchpad).. Have you tried the HOOKRESUME (and, if that does not do it, INITAFTERSUSPEND) flag on the psm device in your kernel. HOOKRESUME (0x2000) did the trick for my IBM T30. > As my Omnibook hangs on a reboot, this is paperclip time... > > > No messages worth noting in the "messages" log either. > > <AOL> > Meee Too. > </AOL> Do you have 'options AUTO_EOI_1'? If so, try removing it. I doubt you do, as it seems to make my T30 hang like this almost every time. On resume I only get console messages from the ATA devices, but never get a message that /kernel resumed. After removing the option, things worked much better. Smells a lot like some sort of race condition in the APM stuff which may be exacerbated by AUTO_EOI. (I really don't understand exactly what AUTO_EOI_1 does, but it certainly seemed to speed up some operations.) In any case, it leaves interrupts blocked or disabled on resume, at least part of the time. LINT provides only: # AUTO_EOI_1 enables the `automatic EOI' feature for the master 8259A # interrupt controller. This saves about 0.7-1.25 usec for each interrupt. # This option breaks suspend/resume on some portables. I hope someone can figure out a work-around of the last line of that entry. The improvement in interrupt performance is nice. R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020809152231.289135D04>