Lei Yang | 1 Oct 01:30 2004

[uml-user] uml 2.6.8.1 doesn't compile

Dear fellows,

I've been trying to get uml kernel 2.6.8.1 compile and work for a while. 
I downloaded linux-2.6.8.1.tar.bz2 and uml-patch-2.6.8.1-1.bz2,  patched 
the kernel and 'make oldconfig ARCH=um'.
Then I did 'make linux ARCH=um' and goodness, a bounch of errors come out:

In file included from include/linux/capability.h:45,
                 from include/linux/sched.h:7,
                 from include/linux/module.h:10,
                 from init/main.c:16:
include/linux/spinlock.h:39:26: asm/spinlock.h: No such file or directory
In file included from include/linux/capability.h:45,
                 from include/linux/sched.h:7,
                 from include/linux/module.h:10,
                 from init/main.c:16:
include/linux/spinlock.h:397: error: parse error before "spinlock_t"
include/linux/spinlock.h:397: warning: function declaration isn't a 
prototype
In file included from include/linux/sched.h:7,
                 from include/linux/module.h:10,
                 from init/main.c:16:
include/linux/capability.h:47: error: parse error before 
"task_capability_lock"
include/linux/capability.h:47: warning: type defaults to `int' in 
declaration of `task_capability_lock'
include/linux/capability.h:47: warning: data definition has no type or 
storage class
In file included from include/linux/time.h:28,
                 from include/linux/timex.h:186,
(Continue reading)

Damien Hull | 1 Oct 02:02 2004
Picon

[uml-user] how do I setup a netwrok

Here's what I have on my host system. (the one hosting the uml)
1. slack 10
2. kernel 2.4.26 ( the one from the cd)
3. I have /dev/net/tun
4. I have tap devices ie tun0, tun1 etc

here's what I have for my uml.
1. uml kernel 2.6.8-1 with networking
2. slack 10 root file system
3. 128mb swap_fs

How do I connect my uml to the net? 
1. I would like to use the switch daemon if I could
2. Would like people to be able to access it from the Internet. 

-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
Jeff Dike | 1 Oct 03:32 2004

Re: [uml-user] uml 2.6.8.1 doesn't compile

lya755 <at> ece.northwestern.edu said:
> the configuration is almost by default.

"almost"  Why so vague?  If CONFIG_SMP is in there, turn it off.

				Jeff

-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
Jeff Dike | 1 Oct 04:30 2004

Re: [uml-user] AMD 64 support

Johan.Verrept <at> advalvas.be said:
>     is there x86_64 support? Even only compiling uml in 32bit mode
> under x86_64 would be nice.
>     how hard would it be to add full x86_64 support?

If you want the incremental patches page, you'll see x86_64 support going
in.

				Jeff

-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
Jeff Dike | 1 Oct 04:32 2004

Re: [uml-user] Intermittent issues executing uml_net

huber <at> alum.wpi.edu said:
> This works fine, most of the time.  Sometimes, eth0 fails to be
> configured on VM startup.  Looking in dmesg, the errno for execl'ing
> uml_net is 2 (ENOENT).

What exactly goes into dmesg?

> When I use mem=128M, uml_net always is found.  If I increase the
> memory size to 256, uml_net is almost never found.  If I boot using
> 128M, then reboot using 256M, it typically works. (with no other
> changes!)
>
> Strange?  I think so... 

Yeah, me too.

				Jeff

-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
Terry | 1 Oct 05:07 2004
Picon

Re: [uml-user] whitebox install - installing via ssh

I am reading more on this error:
cannot set up thread-local storage: cannot set up LDT for thread-local storage

I am coming up with things that relate to NPTL.  I believe I can
disable this on the host system with:
export LD_ASSUME_KERNEL=2.2.5

Then run the installation but it isn't working, gives same behavior. 
Does this spark someone's thinking perhaps?

On Thu, 30 Sep 2004 08:31:52 -0500, Terry <td3201 <at> gmail.com> wrote:
> I am having problems installing it with both 2.4.26 and 2.4.21.  But
> it must not be a kernel issue, just a guess.  The host kernel is 2.6.
> 
> 
> 
> 
> On Thu, 30 Sep 2004 20:10:04 +1200, Peter
> <peter.repliesignored <at> rimuhosting.com> wrote:
> > UML and the new 2.6 threading libraries don't play nice.  Are you getting this error with a 2.6 UML kernel? 
Try installing using a 2.4 kernel.  Then rm /lib/tls.  And run the distro with 2.6 after that (or keep using
2.4 is your safest best).
> >
> > Cheers, Peter
> >
> >
> > ----- Original Message -----
> > From: "Terry" <td3201 <at> gmail.com>
> > To: <user-mode-linux-user <at> lists.sourceforge.net>
> > Sent: Thursday, September 30, 2004 5:12 PM
(Continue reading)

Josh | 1 Oct 05:44 2004

Re: [uml-user] Intermittent issues executing uml_net

Jeff Dike <jdike <at> addtoit.com> writes:

> huber <at> alum.wpi.edu said:
>> This works fine, most of the time.  Sometimes, eth0 fails to be
>> configured on VM startup.  Looking in dmesg, the errno for execl'ing
>> uml_net is 2 (ENOENT).
>
> What exactly goes into dmesg?

Here's what I get in dmesg when it works:

* modprobe tun
modprobe: Can't locate module tun
* ifconfig tap0 192.168.0.5 netmask 255.255.255.255 up
* bash -c echo 1 > /proc/sys/net/ipv4/ip_forward
* route add -host 192.168.0.11 dev tap0
* bash -c echo 1 > /proc/sys/net/ipv4/conf/tap0/proxy_arp
* arp -Ds 192.168.0.11 tap2 pub
* arp -Ds 192.168.0.11 tap1 pub
* arp -Ds 192.168.0.11 tap3 pub
* arp -Ds 192.168.0.11 eth0 pub

The options to the linux binary are:

ubd0=root_fs ubd1=cyrus_fs ubd2=www_fs ubd3=swapfile mem=256M
con=none con0=fd:0,fd:1 eth0=tuntap,,,192.168.0.5

I'm rebooting the VM right now with mem=256M...

(okay, it's working now.)
(Continue reading)

Anthony Brock | 1 Oct 06:11 2004

RE: [uml-user] uml_mconsole and uml_kernel crashes after 5th proc meminfo

Martin,

You can find more information on this in the list archives. Specifically,
there is a work-around in the following article:

http://marc.theaimsgroup.com/?l=user-mode-linux-devel&m=109027118400445&w=2

Tony

> -----Original Message-----
> From: user-mode-linux-user-admin <at> lists.sourceforge.net
> [mailto:user-mode-linux-user-admin <at> lists.sourceforge.net]On Behalf Of
> Martin Weber
> Sent: Thursday, September 30, 2004 12:19 AM
> To: user-mode-linux-user <at> lists.sourceforge.net
> Subject: [uml-user] uml_mconsole and uml_kernel crashes after 5th proc
> meminfo
>
>
> Hi,
>
> When I call "proc meminfo" from the uml_mconsole around the 5th
> time both uml_mconsole and uml_kernel crashes.
> This happens with kernel 2.6.7 and uml-patch-2 and with kernel
> 2.6.8.1 and uml-patch-2.6.8.1-1.
> There seems to be no problem with kernel 2.4.23.
> Also "cat /proc/meminfo" from inside the uml-instance works correct.
>
> Does anybody know how to fix this proplem? Is there a patch or
> workaround?
(Continue reading)

Johan Verrept | 1 Oct 14:20 2004
Picon

Re: [uml-user] AMD 64 support

Jeff Dike wrote:

>Johan.Verrept <at> advalvas.be said:
>  
>
>>    is there x86_64 support? Even only compiling uml in 32bit mode
>>under x86_64 would be nice.
>>    how hard would it be to add full x86_64 support?
>>    
>>
>
>
>If you want the incremental patches page, you'll see x86_64 support going
>in.
>  
>

Thank you.
Can you tell me which of those patches I should apply to get a wrking build?
Since it seems to contain patches that do not build (x86-64-signals), I 
am reluctant to install them all.

regards,

    J.

-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
(Continue reading)

theo29 the29 | 1 Oct 12:53 2004
Picon

[uml-user] More on kernel module debugging in 2.6.*


One other thing concerning debugging kernel modules.

I must know the start address of the module in kernel memory to manually 
point gdb to it. The umlgdb script does this by using the variable 
module_list. This variable is however not in the 2.6.6 kernel. Does anybody 
know how to get the start-address of module in 2.6.6?

Regards, Halldor

-----------------------------------
I am trying to debug a kernel module of mine, by using the umlgdb script. I 
follow all instructions (i.e. start the
kernel with the parameters output from umlgdb, before hitting enter), but I 
get an error when trying to attach:

(gdb) att 1
Attaching to program: 
/home/halldorisak/workspace/UserModeLinux-2.6.6/linux-2.6.6/linux, process 1
ptrace: Operation not permitted.

I am using skas-patch. Anybody knows why gdb won't attach? Otherwise I guess 
I need to just do this manually,
without umlgdb.

There was some post reporting the same problem a while ago, that person was 
also using skas. However I couldn't find
any solution to that problem there.

Regards, Halldor
(Continue reading)


Gmane