Re: fdisk MBR contains more than one OpenBSD partition!

2018-05-09 Thread Rodney Polkinghorne
> Think of the fdisk partition as a way to mark off a part of the disk for
> OpenBSD.  It should generally be one contiguous block.  The beginning of
> *the* OpenBSD partition holds the disklabel, which is the important part
> for marking off OpenBSD disk (sub?)partitions.

I think it would it be helpful to say that in the "Each entry has a
type. ... This can be edited using disklabel(8)." paragraph of
fdisk(8).

Rodney



Re: Capturing ddb output when "boot reboot" fails

2018-04-18 Thread Rodney Polkinghorne
Thanks Stuart.

> Try "call cpu_reset".

That made the machine reboot cleanly.  Afterwards, dmesg and
dmesg.boot had captured both the 6.3 boot and the 6.3 reboot, but the
ddb session in between was missing.  Is there a ddb command that
flushes the session log to the message buffer?

> Or take photos and please transcribe the most
> important bits - at least the panic / crash string and function names
> from 'trace' - it's a lot quicker to figure out who needs to see it
> if those are available in plaintext in the email.

(==) Using system config directory "/usr/X11R6/share/X11/xorg.conf.d"
uvm_fault(0xd0c10110, 0xd44fa000, 0, 1) -> e
kernel: page fault trap, code=0
Stopped at  _rb_min+0x12:   movl0(%edx),%esi

ddb{0}> trace
_rb_min(d09c33a8,d1127004) at _rb_min+0x12
uvm_pmr_get1page(1,0,f5ea0798,0,0,0) at uvm_pmr_get1page+0x105
uvm_pmr_getpages(1,0,0,1,0,1,2,f5ea0798) at uvm_pmr_getpages+0x1a3
uvm_pagealloc(d610d630,4000,0,0) at uvm_pagealloc+0x155
uvn_get(d610d630,4000,0,f5ea0854,f5ea085c,0,4,0) at uvn_get+0xbc
uvm_fault(d6513310,fa97000,0,4) at uvm_fault+0xaf6
trap() at trap+0x602
--- trap(number -2107169536) ---
end of kernel
0x14:
ddb{0}>

Rodney



Capturing ddb output when "boot reboot" fails

2018-04-17 Thread Rodney Polkinghorne
Dear list

My old Dell laptop ran stably under 6.1.  After I upgraded to 6.2,
the kernel started to crash with "page fault trap, code=0" every
time I started the X server.  Every other time, this left the file
system in a state that fsck could not repair.  I don't have a spare
laptop for debugging, and it took me a while to install OpenBSD on
a USB key where I was happy to reproduce the crash.  (OK, I was
lazy, and I hoped that someone else would encounter this fault who
was in a better position to diagnose it.)

This could be a hardware fault.  However, it occurs consistently,
when booting from both the hard disk and a USB key.  Version 6.1
has always worked.

The fault still happens in 6.3 and in recent snapshots.  I would
like to follow the instructions in crash(8) about capturing the
output of the ps and trace commands, rebooting, and emailing a
dmesg.

The problem is that the machine won't reboot from ddb.  When I run
"boot reboot", it either hangs or panics, and I have to power cycle
and lose the message buffer.

What's the best way forward?  I could take some screen shots with
a camera and post them on the web.  Alternatively, I'm in Melbourne,
and I'd be happy to lend the laptop to a local developer.

Rodney



Re: Installing a snapshot to a USB key using bsd.rd

2018-03-19 Thread Rodney Polkinghorne
It turns out that the snapshot on Aarnet failed to match its checksum
for a blindingly obvious reason.  It was corrupt.

I installed a snapshot from another mirror, started X, and crashed the
kernel.  I'll try to post the details to bugs@ in a day or two, but
some hand holding would be appreciated.

Thanks everyone

Rodney



Re: Installing a snapshot to a USB key using bsd.rd

2018-03-18 Thread Rodney Polkinghorne
Thanks Stuart

> Better to test sooner, if it still fails, if you can get a good report
> written up there's still some chance of a fix before release.

The machine is a cheap Dell laptop that's about 10 years old.  It's
possible that an innocent change in the software triggered a latent
hardware fault.

> Perhaps you caught the mirror mid-update, does it still happen now?

I did think of that.  Minutes after the install failed, I confirmed
that the mtimes on the Aarnet mirror were 10 hours old.  I might have
been mislead by timezones, and I'll try again tonight.

Rodney



Re: Installing a snapshot to a USB key using bsd.rd

2018-03-18 Thread Rodney Polkinghorne
Thanks for the replies.  I suspect this is the answer I needed: don't
try to install a 6.2 snapshot just before version 6.3 is released,
instead wait for the release and install that.

> You need to be sure to use the bsd.rd from the snapshot!

I'm pretty sure that's what I booted.  Does any other bsd.rd have a
default path of pub/OpenBSD/6.3/i386?

> You say install, which is very different from upgrade.

I'm running 6.1 on my laptop hard disk, and I wanted to install a
recent snapshot to a USB key.  The back story involves an upgrade to
6.2 stable, a kernel failure whenever I started X Windows, and a
desire to reproduce the fault while preserving a bootable system on
the hard disk.

I'll wait for the 6.3 release, and see if X still crashes the kernel.

Thanks again

Rodney



Installing a snapshot to a USB key using bsd.rd

2018-03-16 Thread Rodney Polkinghorne
Hi

This is my first post here, I appreciate how much work you all do, please
be gentle. :-)

Could someone please tell me how to install the latest snapshot, or point
me at some instructions that work?  I tried the following:

1. Download bsd.rd and SHA256.sig from
http://mirror.aarnet.edu.au/pub/OpenBSD/snapshots/

2. Fail to verify the 6.3 signatures, because I'm running 6.1.  (It would
be nice if the signify man page had instructions to download and verify
openbsd--base.pub.)

3. Reboot, with boot> boot sd0a:/root/bsd.rd

4. Choose to install to sd1 (a USB key), default options, location of sets
is http, HTTP server is mirror.aarnet.edu.au.

5. The default server directory is pub/OpenBSD/6.3/i386, override that with
snapshots instead of 6.3

6. Leave all sets selected, and say done

At this point, the installer reported that SHA256.sig had downloaded and
verified, and that bsd downloaded but failed its checksum test.

Possibly the answer is to ignore the checksums, but I want to ask first.

Rodney Polkinghorne