Re: current won't boot

This simple patch works around the problem for me - I think this points
to some bad initialization in pmap_bootstrap that somehow (accidently)
is undone on first context switch.

Hauke, does this work for you too?

Martin

Index: uvm_glue.c
===================================================================
RCS file: /cvsroot/src/sys/uvm/uvm_glue.c,v
retrieving revision 1.145
diff -u -p -r1.145 uvm_glue.c
--- uvm_glue.c	16 Apr 2010 03:21:49 -0000	1.145
+++ uvm_glue.c	1 May 2010 21:32:10 -0000
 <at>  <at>  -430,7 +430,7  <at>  <at>  uvm_scheduler(void)
 	lwp_unlock(l);

 	for (;;) {
-		sched_pstats();
 		(void)kpause("uvm", false, hz, NULL);
+		sched_pstats();
 	}
 }

Re: current won't boot

>This simple patch works around the problem for me - I think this points
>to some bad initialization in pmap_bootstrap that somehow (accidently)
>is undone on first context switch.
>
>Hauke, does this work for you too?
>
>Martin
>
>Attachment converted: Quantum18G:patch (TEXT/R*ch) (0003C8FB)

I don't know the code, so I have no right to comment, but this
fix screams "RACE CONDITION".  If it works, it might be more luck
than design......

-dgl-

Re: current won't boot

On Sat, May 01, 2010 at 05:50:47PM -0500, Donald Lee wrote:
> I don't know the code, so I have no right to comment, but this
> fix screams "RACE CONDITION".  If it works, it might be more luck
> than design......

Of course I'm not going to commit that change, it makes no sense - just
collecting data points.

Martin

Re: current won't boot

At 23:34 Uhr +0200 1.5.2010, Martin Husemann wrote:
>This simple patch works around the problem for me - I think this points
>to some bad initialization in pmap_bootstrap that somehow (accidently)
>is undone on first context switch.
>
>Hauke, does this work for you too?

I'll test on Monday - the IIvx is  <at> workplace.

	hauke

--
"It's never straight up and down"     (DEVO)

Re: current won't boot

At 23:34 Uhr +0200 1.5.2010, Martin Husemann wrote:
>This simple patch works around the problem for me - I think this points
>to some bad initialization in pmap_bootstrap that somehow (accidently)
>is undone on first context switch.
>
>Hauke, does this work for you too?

Nope, no change.

	hauke

--
"It's never straight up and down"     (DEVO)

You've received a private message from a friend!

I read your profile today, I thought I would drop you a line and hope to become your friend! Check my personal
page here:
http://ealing.zoomshare.com/files/photos.htm

Mac LCIII - NetBSD 5.0.2

Hi, I am trying to get NetBSD installed on my old LCIII.
It has an external scsi drive and I have successfully installed NetBSD in the past (an ancient 1.4 version
iirc) and it worked very well.

Anyway, I have attempted to use the "sysinst" install method and booted from the install kernel. The kernel
boots but it hangs somewhere between detecting the mouse and (I guess) attempting to access the disks. I
intend to try the traditional method at some stage to see if I can make some headway.

If anyone on the list can give me any pointers I would be grateful.

Dan Whitehouse

Re: Mac LCIII - NetBSD 5.0.2

At 13:25 Uhr +0100 5.5.2010, Dan Whitehouse wrote:
>The kernel boots but it hangs somewhere between detecting the mouse and (I
>guess) attempting to access the disks.

The kernel locking up in adb initialization during boot when you wiggle the
mouse at the wrong moment is a known issue... Re-try the boot, and make
sure you leave the mouse alone after the Booter takes over.

Alternatively, you could just plop a GENERIC(SBC) kernel on the 1.4
installation, and boot that, then untar the tarballs. Some caveats apply,
though, like making sure you have a populated /etc/pam.d directory.

	hauke

--
"It's never straight up and down"     (DEVO)

Re: Mac LCIII - NetBSD 5.0.2

I don't think the problem is due to the adb bug: I don't touch the mouse after hitting command-b to boot the
install kernel.
I've tried booting the latest GENERIC(SBC) kernel from the MacOS filesystem - it gets a bit further but then
it panicks.

Sadly the 1.4 install is long gone. 

I guess I'm going to have to hit the ftp and start working backwards until I find something that works

Dan

On Wed, May 05, 2010 at 09:17:03PM +0200, Hauke Fath wrote:
> At 13:25 Uhr +0100 5.5.2010, Dan Whitehouse wrote:
> >The kernel boots but it hangs somewhere between detecting the mouse and (I
> >guess) attempting to access the disks.
> 
> The kernel locking up in adb initialization during boot when you wiggle the
> mouse at the wrong moment is a known issue... Re-try the boot, and make
> sure you leave the mouse alone after the Booter takes over.
> 
> Alternatively, you could just plop a GENERIC(SBC) kernel on the 1.4
> installation, and boot that, then untar the tarballs. Some caveats apply,
> though, like making sure you have a populated /etc/pam.d directory.
> 
> 	hauke
> 
> 
> --
> "It's never straight up and down"     (DEVO)
> 

Re: Mac LCIII - NetBSD 5.0.2

Hi,

> I don't think the problem is due to the adb bug: I don't touch the mouse 
> after hitting command-b to boot the install kernel. I've tried booting 
> the latest GENERIC(SBC) kernel from the MacOS filesystem - it gets a bit 
> further but then it panicks.

Try unplugging the mouse before booting. You can boot from the keyboard 
with Apple-b.

John