On Friday 30 December 2005 13:26, Juraj Holtak wrote:
> Am Freitag, den 30.12.2005, 12:09 +0100 schrieb Blaisorblade:
> > On Monday 26 December 2005 20:51, Juraj Holtak wrote:
> > > Am Montag, den 26.12.2005, 20:15 +0100 schrieb Blaisorblade:
> > 1) if my binaries don't run there some host problem
On Monday 26 December 2005 20:51, Juraj Holtak wrote:
> Am Montag, den 26.12.2005, 20:15 +0100 schrieb Blaisorblade:
> > > Same problem? Guess not, post the failures.
> >
> > Ok, correction - I see the same result - disable CONFIG_DEBUG_SPINLOCK
> > probably.
> >
> > It's a debug option anyway - di
On Monday 26 December 2005 20:15, Blaisorblade wrote:
> On Monday 26 December 2005 20:03, Blaisorblade wrote:
> > On Monday 26 December 2005 19:53, Juraj Holtak wrote:
> > > I though, that disabling HIGHMEM and running <256MB ram guests helped.
> > > But that was yesterday. Today I had to 20x start
Hi,
Thanks.. I've got it working now your reply and blaisorblade's helped me fix
it. The network problem that I was talking about is also fixed now. Debian
puts the uml_net in /usr/lib/uml/ which is not in the path. Therefor the UML
kernel can't find and execute it.
Now I can continue to e
tell me about your net config.
it`s a 2.6.15-rc7 guest config
crypto section is back again as it has no impact
to run it safely, move /lib/tls out of way or rename it
to /lib/tls.disabled
your inittab should contain this line and all other vty commented out
0:12345:respawn:/sbin/getty 38400
> > Problematic things are in kernel hacking/debugging, and general features
> > (smp, 3-level pages, etc), watchdog? etc.
>
> I still didn`t figure out how the watchdog stuff works. I have HW
> watchdog on the host but a uml guest watchdog would be nice too. Any doc
> or howto?
AFAIK, there isn'
Am Dienstag, den 27.12.2005, 17:48 + schrieb Antoine Martin:
> Problematic things are in kernel hacking/debugging, and general features
> (smp, 3-level pages, etc), watchdog? etc.
I still didn`t figure out how the watchdog stuff works. I have HW
watchdog on the host but a uml guest watchdog w
> I disabled everything in the cryptographic section too, but I`ll enable
> it again (and report if bug comes back)
AFAIK - this shouldn't be a problem.
>
> I have a lot EXPERIMENTAL in the network section because I need it.
Network stuff should be ok too.
Problematic things are in kernel hacking
I seams to work now.
I disabled TT mode (which is broken and crashes on my server) and did a
static compile (skas0 only). Right now in this moment I run 4 Uml`s with
a complete ram usage in sum of 1,5GB. The biggest one has 500MB.
Currently I`m setting up even more guests.
I disabled everything i
> > And why has the maschine been running with 2.6.11.12 6 months fine,
> > until this christmas when some guy (me) had the crazy idea of a new
> > kernel :-)
I had the exact same issue you describe, I guess you also used your
previous .config with a make ARCH=um oldconfig.
> $ make defconf
On Monday 26 December 2005 20:03, Blaisorblade wrote:
> On Monday 26 December 2005 19:53, Juraj Holtak wrote:
> > I though, that disabling HIGHMEM and running <256MB ram guests helped.
> > But that was yesterday. Today I had to 20x start/kill the guest before
> > he went up (2.6.15-rc7 skas0).
> S
On Monday 26 December 2005 19:53, Juraj Holtak wrote:
> I though, that disabling HIGHMEM and running <256MB ram guests helped.
> But that was yesterday. Today I had to 20x start/kill the guest before
> he went up (2.6.15-rc7 skas0).
Same problem? Guess not, post the failures.
> But YES, you can c
I though, that disabling HIGHMEM and running <256MB ram guests helped.
But that was yesterday. Today I had to 20x start/kill the guest before
he went up (2.6.15-rc7 skas0).
But YES, you can call it a "better result" because I didn`t give up
until it started :-)
I don`t understand it at all :-/
On Sunday 25 December 2005 21:33, Juraj Holtak wrote:
> Hello,
>
> it looks like i`ve found a workaround good enough for me.
>
> I discovered, that the problem is only present, when I boot a guest with
> more than 256MB of ram enabled. I spread my services to more guests (7
> instead of 5) with <25
Hello,
it looks like i`ve found a workaround good enough for me.
I discovered, that the problem is only present, when I boot a guest with
more than 256MB of ram enabled. I spread my services to more guests (7
instead of 5) with <256MB ram. I tested all the kernels I have compiled
before and had n
hi
more weird stuff here
i just wanted to get you the output so I booted the 2.6.14-bs3 gust
kernel and it booted! right after that I shut it diwn and booted again,
but the problem was back.
> Just to clarify, you have three working guests and one hung one, and they
> are all the same version?
On Sat, Dec 24, 2005 at 06:26:05PM +0100, Juraj Holtak wrote:
> The last guest, which has 784MB ram (host is highmem enabled with 2GB
> ram) allocated, fails to boot -> soft lockup just after mounting the /
> filesystem on ext3 in data=journal mode (I didn`t try other modes)
Just to clarify, you h
17 matches
Mail list logo
