On Fri, Apr 29, 2011 at 7:05 PM, Riccardo Murri
wrote:
> But I can still redirect stdin from a named FIFO and attach a "silent"
> process to it:
Sure, but you'll still need a wrapper script which creates this FIFO
in the grid node.
I've had enough fun with sge, lsf, ... ;)
--
Thanks,
//richard
On Fri, Apr 29, 2011 at 6:46 PM, richard -rw- weinberger
wrote:
> I fear you cannot use /dev/null (or any other non-tty file) as stdin. :-(
> UML uses poll() to detect whether input is available from stdin.
> As you can always read from /dev/null poll() doesn't block and
> UML gets flooded with in
On Fri, Apr 29, 2011 at 6:10 PM, Riccardo Murri
wrote:
> Hi Richard,
>
> On Fri, Apr 29, 2011 at 5:47 PM, richard -rw- weinberger
> wrote:
>> Anyway, what is the exact command-line which causes UML to hang?
>> Can you provide me a stand alone test case?
>>
>
> Here it is:
>
I fear you cannot use
Hi Richard,
On Fri, Apr 29, 2011 at 5:47 PM, richard -rw- weinberger
wrote:
> Anyway, what is the exact command-line which causes UML to hang?
> Can you provide me a stand alone test case?
>
Here it is:
* download the 64-bit kernel:
http://uml.devloop.org.uk/kernels/kernel64-2.6.35.7.bz2
*
On Fri, Apr 29, 2011 at 5:14 PM, Riccardo Murri
wrote:
> Hello,
>
> I'm trying to use UML to provide a "standard environment" for
> computational jobs in a Grid computing setting.
>
> So, I launch "linux mem=256M ubd0=apppot0.img con=null con0=fd:0,fd:1
> root=/dev/ubda umid=my". The /etc/inittab
Hello,
I'm trying to use UML to provide a "standard environment" for
computational jobs in a Grid computing setting.
So, I launch "linux mem=256M ubd0=apppot0.img con=null con0=fd:0,fd:1
root=/dev/ubda umid=my". The /etc/inittab within the UML
machine is set up to run a certain non-interactive s
Hi,
My company is using UML to simulate the Linux OS on embedded systems to
do off-target testing. However during a special but complex testing a
UML virtual machine hangs for about every 4 times. It took me a few
weeks to debug it but so far I didn't get any breakthrough.
Symptom:
1 virtu
Antoine Martin wrote:
>> Downgrading to skas-v8.2 makes things work fine again.
> Ah, did you get around to generate a v8.2 patch for 2.6.19?
http://www.theshore.net/~caker/uml/patches/skas-2.6.19.1-v8.2.patch
So far so good. I've got a few hosts running it now, but only a few
UMLs on each one.
> Downgrading to skas-v8.2 makes things work fine again.
Ah, did you get around to generate a v8.2 patch for 2.6.19?
> I think the key to reproducing this is a skas-v9 host. It doesn't
> happen for me on every file system, which is somewhat interesting.
Yep, it seems like a nptl/glibc issue (jus
Antoine Martin wrote:
> The kernels from 2.6.17 onwards on uml.nagafix.co.uk do exhibit the
> problem. But it seems that the host setup is part of the problem. No
> luck for me with 2.6.17 onwards (with skas3 v9), booting either with
> skas0 or skas3 noprocmm. I thought the kernels were statically
The kernels from 2.6.17 onwards on uml.nagafix.co.uk do exhibit the
problem. But it seems that the host setup is part of the problem. No
luck for me with 2.6.17 onwards (with skas3 v9), booting either with
skas0 or skas3 noprocmm. I thought the kernels were statically linked
(well apart from the pc
On Thu, Dec 14, 2006 at 12:02:19PM +, Antoine Martin wrote:
> I believe this issue is linked to Fedora's glibc 2.4 or/and gcc 4.1 on
> amd64 platforms only (I have the same issue on Gentoo 2006.1).
> I reported this issue a while back but haven't had time to investigate:
> Blaisorblade suggeste
> Siva
>
> - Original Message
> From: Antoine Martin <[EMAIL PROTECTED]>
> To: Sivakumar Kuppusamy <[EMAIL PROTECTED]>
> Cc: [email protected]
> Sent: Thursday, 14 December, 2006 5:32:19 PM
> Subject: Re: [uml-user] UML hangs after
nt: Thursday, 14 December, 2006 5:32:19 PM
Subject: Re: [uml-user] UML hangs after VFS mount in FC6
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
I believe this issue is linked to Fedora's glibc 2.4 or/and gcc 4.1 on
amd64 platforms only (I have the same issue on Gentoo 2006.1).
I repo
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
I believe this issue is linked to Fedora's glibc 2.4 or/and gcc 4.1 on
amd64 platforms only (I have the same issue on Gentoo 2006.1).
I reported this issue a while back but haven't had time to investigate:
Blaisorblade suggested that I install gli
Hi,
I am trying to setup UML in Fedora Core 6. The steps that I followed are:
- Downloaded 2.6.18.1 vanilla kernel (linux-2.6.18.1.tar.bz2)
- Downloaded blaisorblade's 2.6.18.1 bb2 patch (uml-2.6.18.1-bb2.patch)
- Applied the patch on vanilla kernel
- make defconfig ARCH=um; make linux ARCH=um
-
On Wed, Aug 02, 2006 at 04:23:35PM +0200, Christoph Rudorff wrote:
> Thanks, that fixed the problem!
>
> I rebooted it 10 times, it is working now. Is mode=tt broken? I also tried
> that before but:
Yeah, it'll be removed when SMP support for skas is done.
Jeff
Am Mittwoch, 2. August 2006 15:44 schrieb Christoph Rudorff:
> Am Mittwoch, 2. August 2006 14:17 schrieb Jeff Dike:
> > On Wed, Aug 02, 2006 at 01:07:24PM +0200, Christoph Rudorff wrote:
> > > TLS probs? No. I removed /lib/tls I double checked if /dev is complete.
> > > If I kill the uml and simply
Am Mittwoch, 2. August 2006 14:17 schrieb Jeff Dike:
> On Wed, Aug 02, 2006 at 01:07:24PM +0200, Christoph Rudorff wrote:
> > TLS probs? No. I removed /lib/tls I double checked if /dev is complete.
> > If I kill the uml and simply retry 2 or 3 times, it boots fine. But they
> > usually do not survi
On Wed, Aug 02, 2006 at 01:07:24PM +0200, Christoph Rudorff wrote:
> I'm using sucessfully several uml's for 2 years now.
>
> I recently upgraded kernels from linux.2.6.11.10-bs5 to linux-2.6.17.7 and
> 2.6.17.7-hmd1.smp.p4_skas3_v9_pre9 for the host machine (hmd = homemade).
>
> I have randomly
Hello ppl!
I'm using sucessfully several uml's for 2 years now.
I recently upgraded kernels from linux.2.6.11.10-bs5 to linux-2.6.17.7 and
2.6.17.7-hmd1.smp.p4_skas3_v9_pre9 for the host machine (hmd = homemade).
I have randomly the problem, that the uml machines hangs on boot at
"VFS: Mounted
On Mon, Mar 27, 2006 at 11:33:21PM -0600, Jack Sparrow wrote:
> [EMAIL PROTECTED] linux-2.6.14.6]# ./linux ubd0=root_fs_slack8.1
> Checking PROT_EXEC mmap in /tmp...OK
> Checking for /proc/mm...found
> Checking for the skas3 patch in the host...found
> UML running in SKAS3 mode
Add stderr=1 to the
Hi,
I am trying to create a working version of uml on a 2.6.14.6 kernel. I
added the skas3 patch to the host kernel, compiled it, installed the
uml-utilities and then compiled the guest kernel after addeing the uml
patch. The linux binary is created alright but when I execute the uml,
it hangs (st
> One odd behaviour is that during these slow periods and even during
> the permanent hanging, the host is writing to the disk at ~3MB/sec -
> in the case of the permanent hang it writes like this continuously,
> until I run 'kill ' on the host.
Okay, I tell a lie - it doesn't write continuously w
Hi Jeff,
> How hard does it hang? You can ping it, but can't type?
It depends - the more I see this the more it seems to be I/O blocking
rather than hanging, e.g. I can type, but if I run 'ls' then the shell
hangs, as if the 'ls' command is waiting for the disk. Ctrl+C or
Ctrl+\ won't work, as
On Tue, Feb 07, 2006 at 02:26:48PM +1000, Adam Nielsen wrote:
> Any other ideas?
How hard does it hang? You can ping it, but can't type? It doesn't even
ping? (substitute mconsole version for ping if you're not playing with the
network)
I've seen one other report of this, but haven't seen it m
Hi Blaisorblade,
Thanks for your reply!
> If you enabled CONFIG_SMP in the _UML_ compilation (verify
> with ./vmlinux --showconfig), disable it now.
I don't think I did:
$ ./uml-2.6.14 --showconfig | grep SMP
CONFIG_BROKEN_ON_SMP=y
Not sure what that option is for, but I thought it was for sho
On Monday 06 February 2006 03:21, Adam Nielsen wrote:
> Hi,
>
> I've got a UML running on a server here, but whenever I copy a large
> number of files across via hostfs it seems to lock up the UML. I have
> gkrellmd running and I can see the system load (of the UML) skyrocket
> just before gkrellm
Hi,
I've got a UML running on a server here, but whenever I copy a large
number of files across via hostfs it seems to lock up the UML. I have
gkrellmd running and I can see the system load (of the UML) skyrocket
just before gkrellmd dies and I lose access to the UML. Sometimes I
can still switc
On Wednesday 27 July 2005 11:33, Kyrre M Begnum wrote:
> Hello list,
> I am using uml in a setting where i am only interested in the output of
> its console, and won't give it any console input.
> I have encountered a phenomenon that if i pipe something to the uml
> command the boot freezez at:
If
On Wed, Jul 27, 2005 at 11:33:44AM +0200, Kyrre M Begnum wrote:
>
> Hello list,
>
> I am using uml in a setting where i am only interested in the output of
> its console, and won't give it any console input.
> I have encountered a phenomenon that if i pipe something to the uml
> command the boo
Hello list,
I am using uml in a setting where i am only interested in the output of
its console, and won't give it any console input.
I have encountered a phenomenon that if i pipe something to the uml
command the boot freezez at:
echo something | ./linux ..
.
.
.
kjournald starting. Co
Johannes Formann <[EMAIL PROTECTED]> wrote:
> login:
> Breakpoint 1, panic (fmt=0x811e1c0 "Kernel stack overflow") at panic.c:58
> 58 panic.c: No such file or directory.
> in panic.c
With a 'bit' help from Jeff I killed that error, that panic was
obsolete, so you'll have to delete s
>
> You mentioned debian... If that's your host OS and you've installed
> the debian uml-utilies package, then I've seen this in every
> uml I've built myself, where I try to use uml's switch program...
>
>
>
> I'm not around my uml debian box at the moment and don't remember
> the exact info..
> I built my own uml kernel with the same config as the one that
is
> working, but still cannot bring up the network when using a
kernel I
> build myself. This means that I cannot turn off the CONFIG_UML_REAL_TIME_CLOCK
> and have a network. Everything else seems to be working. I
have tried
> t
On Tuesday 11 January 2005 12:12, Michael West wrote:
> On Mon, Jan 10, 2005 at 03:20:27PM +0100, Blaisorblade wrote:
> > On Monday 10 January 2005 04:32, Michael West wrote:
> > > I have been compiling many uml kernels with different configs to try to
> > > address my networking issue. I am still
On Mon, Jan 10, 2005 at 03:20:27PM +0100, Blaisorblade wrote:
> On Monday 10 January 2005 04:32, Michael West wrote:
> > I have been compiling many uml kernels with different configs to try to
> > address my networking issue. I am still getting "SIOCSIFFLAGS: No such
> > file or directory" when us
On Monday 10 January 2005 04:32, Michael West wrote:
> I have been compiling many uml kernels with different configs to try to
> address my networking issue. I am still getting "SIOCSIFFLAGS: No such
> file or directory" when using kernels I build myself, while the debian
> packaged kernel works f
I have been compiling many uml kernels with different configs to try to
address my networking issue. I am still getting "SIOCSIFFLAGS: No such
file or directory" when using kernels I build myself, while the debian
packaged kernel works fine ( but has the lockup problem )
Can someone provide me a
On Sun, Jan 09, 2005 at 08:29:06PM +0100, Blaisorblade wrote:
> On Sunday 09 January 2005 19:03, Michael West wrote:
> > Thanks, I tried 27-1um but cannot bring up the network
> >
> > ifup eth0
> > SIOCSIFFLAGS: No such file or directory
> > SIOCSIFFLAGS: No such file or directory
> > Failed to br
On Sunday 09 January 2005 19:03, Michael West wrote:
> Thanks, I tried 27-1um but cannot bring up the network
>
> ifup eth0
> SIOCSIFFLAGS: No such file or directory
> SIOCSIFFLAGS: No such file or directory
> Failed to bring up eth0.
The most likely cause is a kernel configuration change while bu
Thanks, I tried 27-1um but cannot bring up the network
ifup eth0
SIOCSIFFLAGS: No such file or directory
SIOCSIFFLAGS: No such file or directory
Failed to bring up eth0.
I will try 24-1um and let you know if that works.
---
The SF.Net email
--On 8. Januar 2005 13:15:41 +0100 Blaisorblade <[EMAIL PROTECTED]>
wrote:
On Thursday 06 January 2005 01:49, Johannes Formann wrote:
Michael West <[EMAIL PROTECTED]> wrote:
> I am running debian testing on both host and uml. Uml kernel is from
> debian package 2.4.26-1um. Host kernel is 2.4.24
On Thursday 06 January 2005 01:49, Johannes Formann wrote:
> Michael West <[EMAIL PROTECTED]> wrote:
> > I am running debian testing on both host and uml. Uml kernel is from
> > debian package 2.4.26-1um. Host kernel is 2.4.24 with SKAS patch.
>
> It looks like a problem I've with every uml-Patch
Michael West <[EMAIL PROTECTED]> wrote:
> I am running debian testing on both host and uml. Uml kernel is from
> debian package 2.4.26-1um. Host kernel is 2.4.24 with SKAS patch.
It looks like a problem I've with every uml-Patch after 24-1um.
After some load/time it freezes/crashes.
regards
I found that I can cause the system to hang by doing a scp of a 1.5 Gb
file to the uml. The file copy will get 50-80% complete and then the
uml hangs. I notice that the host system is doing some swap/paging when
this occurs.
The host is quite small, PII 300Mhz 384Mb RAM Though the host has ve
My uml system hangs after running for 16-48 hours. I cannot connect,
nor can I use mconsole.
I am running debian testing on both host and uml. Uml kernel is from
debian package 2.4.26-1um. Host kernel is 2.4.24 with SKAS patch.
I am new to UML, how do I investigate this?
--
"In the land
47 matches
Mail list logo
