headers
Rob Ristroph | 1 Mar 2003 04:22
Picon
Favicon

Re: use installation disk as repair ?

>>>>> "Russ" == Russ Cox <rsc <at> plan9.bell-labs.com> writes:
Russ> 
Russ> It you boot the install floppy you can draw
Russ> a new window and then run
Russ> 
Russ> 	mount /srv/kfs._dev_sdC0_fs /n/kfs

That worked.  Thanks.  While I was waiting for an answer, I decided to
test the machine a bit more by booting from a floppy linux and a
cdrom-based linux, in order to see if my ethernet card was really
working; I ended up having weird kernel paging errors, and replaced
the ram, and then plan9 came up and found the network just fine.

Russ> and do what you need to do.  The DHCP should
Russ> time out after a couple minutes, though.
Russ> Perhaps you didn't wait long enough.

I left it on waiting all night; however, like I said, it appears to
have been bad memory.

Russ> Once you get booting again, what is the 
Russ> date on your kernel (ls -l /386/9pcdisk)?

Sep 22 23:16

It seems to be working now, however.

--Rob
(Continue reading)

Permalink | Reply | 0 comments |
headers
Christopher Nielsen | 1 Mar 2003 04:41
Picon
Favicon

Re: kernel debug - how

> You can debug the kernel with acid -k but there 
> is no way to write to kernel memory.  
> 
> However, you can set that variable by doing
> 
> 	echo debug >/dev/fd0ctl
> 
> (There is no way to change it back to zero
> after that, which is itself a bug.)

Here's a patch that adds a command to disable debugging
for the floppy driver.

121a122
> 	CMnodebug,
129a131
> 	CMnodebug, "nodebug", 1,
475a478,480
> 			break;
> 		case CMnodebug:
> 			floppydebug = 0;

--

-- 
Christopher Nielsen
"They who can give up essential liberty for temporary
safety, deserve neither liberty nor safety." --Benjamin Franklin
Permalink | Reply | 0 comments |
headers
rob pike, esq. | 1 Mar 2003 08:24

Re: plumber crash (vmware)

that's not a crash, that's a traceback. what's the crash?
invalid address, address error, ...?  in other words, what
was the suicide string that got printed, if any?
if possible, can you make a snap of the dead program,
gzip it, and send it to me? thanks.

-rob
Permalink | Reply | 0 comments |
headers
Russ Cox | 1 Mar 2003 17:10
Favicon

Re: plumber crash (vmware)

That stack trace is completely bogus, which means
either the memory in the proc is very messed up, or
you are looking at the wrong binary.  The latter
seems more likely, especially since you just pulled.
It's possible that the crash happened because the
binary you were running got overwritten or removed.
Pull is supposed to move the old binary out of the
way before making a new one, to lessen these effects,
but maybe you hadn't updated in a while?  That change
to pull happened a week or two ago.

Please run

	ls -l /386/bin/*plumber

and then try

	acid 82 /386/bin/_plumber
	stk()

As Rob said, it would be nice to have the snapshot
of the broken program.

	snap 82 | gzip > plumber.snap.gz
	mail -a plumber.snap.gz rsc <at> plan9.bell-labs.com r <at> google.com

Russ
Permalink | Reply | 0 comments |
headers
A.S. Kukhar | 2 Mar 2003 00:46
Picon

Rc shell FAQ

you can get it from http://ask.km.ru/plan9/doc/rcfaq.html, 
the document attempts to answer some frequently asked 
questions about rc, the Plan 9 shell. FAQ based on ``rc 
shell Frequently Asked Questions'' from Alan Watson (15 Dec 
1994).

Please forward any comments, corrections, and suggestions 
regarding this FAQ to plan9 <at> kp.km.ua.

-kyxap
Permalink | Reply | 4 comments |
headers
Russ Cox | 1 Mar 2003 18:05
Favicon

Re: venti arena

I've never understood all the inconsistencies you
just pointed out, so I dug around for a little while.

The numbers given by fmtarenas are the real ones.

The index uses ``virtual'' addresses which get mapped
into (arena-name, offset) pairs.  The base for the virtual
addresses is 1M.  If you had multiple arena partitions 
you'd see that the index entries for the arenas in that
disk start where the previous disk left off.  These virtual
addresses are the numbers that fmtindex prints.

I think checkarenas is just buggy.  There are head and tail
blocks around each arena's data section, and checkarenas
is printing

	[datastart, datastart+datasize+2*blocksize)

instead of

	[datastart-blocksize, datastart+datasize+blocksize)

That is, checkarenas thinks datastart has the blocksize subtracted
from it already, but that's not the case.  The http://server/index
listing uses the same code, and is thus similarly off.
This is why the backup script I posted subtracts one block size
before using the offset in the listing.  (Though that was just
trial and error on my part at the time.)

Russ
(Continue reading)

Permalink | Reply | 2 comments |
headers
Scott Schwartz | 2 Mar 2003 04:17
Picon

aan documentation

The manpage for aan says it uses a unique protocol.  Is
its operation described anywhere (other than the source)?
Permalink | Reply | 2 comments |
headers
Kenji Arisawa | 2 Mar 2003 16:34
Picon

Re: venti arena

 >The numbers given by fmtarenas are the real ones.
Thanks Russ.

Let arenas of fmtarenas be:
	[S0, E0])
	[S1, E1)
	..
	[Sn, En)
and let [Sk, Ek) be active.
My understanding is:
	[0, Sk) is frozen
	[Ek, En)  is unwritten
	[En, M) is frozen
where M is size of the disk space for arenas.
OK?

Kenji Arisawa
Permalink | Reply | 1 comment |
headers
Russ Cox | 2 Mar 2003 16:52
Favicon

Re: venti arena

Yes, that is true.  Note that if you want to
read the arena data you should probably
use venti/rdarena rather than do it yourself.

Russ
Permalink | Reply | 0 comments |
headers
Peter Bosch | 2 Mar 2003 17:54
Favicon

Re: aan documentation

> The manpage for aan says it uses a unique protocol.  Is
> its operation described anywhere (other than the source)?

read the source, luke.
Permalink | Reply | 1 comment |

Gmane