Re: Suspend broken ?

On Sat, 2005-07-09 at 15:17 -0400, John Baldwin wrote:
> On Saturday 09 July 2005 10:36 am, Stefan Ehmann wrote:
> > On Sat, 2005-07-09 at 09:23 -0400, John Baldwin wrote:
> > > On Tuesday 05 July 2005 01:37 pm, Stefan Ehmann wrote:
> > > > On Mon, 2005-07-04 at 20:48 +0200, Fabian Keil wrote:
> > > > > Stefan Ehmann <shoesoft_at_gmx.net> wrote:
> > > > > > On Mon, 2005-07-04 at 21:33 +0800, Huang wen hui wrote:
> > > > > > > Huang wen hui 写é“:
> > > > > > > >Hi,
> > > > > > > >For my T42p, suspend operation could not work from jun 2,
> > > > > > > >It does work under CURRENT using Jun 1 cvsup.
> > > > > > > >/sys/i386/isa/clock.c revert to v1.220.
> > > > > > >
> > > > > > > but /sys/i386/isa/clock.c revert to v1.220 help that.
> > > > > >
> > > > > > Thanks for tracking that down, helps in my case too.
> > > > > >
> > > > > > I'm running a Toshiba M-30X notebook. For me, resume basically
> > > > > > worked but everything was as slow as if it was running on my old
> > > > > > calculator. (e.g. it takes several seconds from typing a character
> > > > > > until it is displayed on the console).
> > > > >
> > > > > I had these symptoms before I put "device pmtimer" in the kernel.
> > > > > I'm still on 5.4, therefore I don't know if this has anything to
> > > > > do with your problem.
> > > >
> > > > Does not seem to be related. I already had pmtimer in my kernel. And
> > > > this problem also hasn't occured before /sys/i386/isa/clock.c rev 221.
> > >
> > > Try this patch:  I'm going to try to get it committed soon.
> >
> > Applies cleanly with latest sources but I get errors during make kernel:
> >
> > /usr/src/sys/i386/isa/clock.c: In function `set_timer_freq':
> > /usr/src/sys/i386/isa/clock.c:541: error: syntax error before "else"
> > /usr/src/sys/i386/isa/clock.c:547: error: `timer0_max_real_count'
> > undeclared (first use in this function)
> > /usr/src/sys/i386/isa/clock.c:547: error: (Each undeclared identifier is
> > reported only once
> > /usr/src/sys/i386/isa/clock.c:547: error: for each function it appears
> > in.)
> > ...
> >
> > and a bunch of more errors and warnings
> 
> I think you just need to remove the extra { after the if statement on line 539 
> at the end of the line.

This and changing timer0_max_real_count to timer0_real_max_count made it
compilable.

But this time the system is immediately very slow (not just after
suspend) and I get this strange error message on startup:

calcru: runtime went backwards from 68378390 usec to 67512958 usec for
pid 11 (idle: cpu0)