[uml-devel] Configuration problem with UML in kernel 2.6.19.1

[uml-devel] Configuration problem with UML in kernel 2.6.19.1

[uml]

Hello!

I had some problems running linux 2.6.19.1 as client UML.  When  
starting 'mapping mmap stub failed, errno = -12' is printed - with an  
additional Kernel panic. In the file arch/um/Kconfig.i386 the  
definition of STUB_CODE and STUB_DATA depend on the configuration  
symbol HOST_2G_2G which has gone sometimes in between 2.6.18.6 and  
2.6.19.1.

The applied patch solves the problem for me.

Kind regards

Andreas Florath


diff -u linux-2.6.19.1_orig/arch/um/Kconfig.i386  
linux-2.6.19.1/arch/um/Kconfig.i386
--- linux-2.6.19.1_orig/arch/um/Kconfig.i386    2006-12-11  
20:32:53.000000000 +0100
+++ linux-2.6.19.1/arch/um/Kconfig.i386 2007-01-10 09:32:01.000000000 +0100
@@ -67,13 +67,13 @@

  config STUB_CODE
         hex
-       default 0xbfffe000 if !HOST_2G_2G
-       default 0x7fffe000 if HOST_2G_2G
+       default 0xbfffe000 if !HOST_VMSPLIT_2G
+       default 0x7fffe000 if HOST_VMSPLIT_2G

  config STUB_DATA
         hex
-       default 0xbffff000 if !HOST_2G_2G
-       default 0x7ffff000 if HOST_2G_2G
+       default 0xbffff000 if !HOST_VMSPLIT_2G
+       default 0x7ffff000 if HOST_VMSPLIT_2G

  config STUB_START
         hex



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Configuration problem with UML in kernel 2.6.19.1

[Jeff Dike]

On Wed, Jan 10, 2007 at 10:28:57AM +0100, uml@flonatel.org wrote:
> Hello!
> 
> I had some problems running linux 2.6.19.1 as client UML.  When  
> starting 'mapping mmap stub failed, errno = -12' is printed - with an  
> additional Kernel panic. In the file arch/um/Kconfig.i386 the  
> definition of STUB_CODE and STUB_DATA depend on the configuration  
> symbol HOST_2G_2G which has gone sometimes in between 2.6.18.6 and  
> 2.6.19.1.
> 
> The applied patch solves the problem for me.

Excellent, thanks for chasing that down.

				Jeff

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Configuration problem with UML in kernel 2.6.19.1

[Blaisorblade]

On Wednesday 10 January 2007 22:46, Jeff Dike wrote:
> On Wed, Jan 10, 2007 at 10:28:57AM +0100, uml@flonatel.org wrote:
> > Hello!
> >
> > I had some problems running linux 2.6.19.1 as client UML.  When
> > starting 'mapping mmap stub failed, errno = -12' is printed - with an
> > additional Kernel panic. In the file arch/um/Kconfig.i386 the
> > definition of STUB_CODE and STUB_DATA depend on the configuration
> > symbol HOST_2G_2G which has gone sometimes in between 2.6.18.6 and
> > 2.6.19.1.
> >
> > The applied patch solves the problem for me.
>
> Excellent, thanks for chasing that down.

This means that running UML under Xen should be retested! Setting HOST_VMSPLIT 
does not fully work without this patch!

Matt, you were reporting a problem about this, can you try this patch (should 
also apply on 2.6.18 kernels)?
After applying the patch, do not forget to give a "make silentoldconfig 
ARCH=um" before recompiling.
If you have lost interest just say that, so we'll know we can't test that for 
now.

Bye!
-- 
Inform me of my mistakes, so I can add them to my list!
Paolo Giarrusso, aka Blaisorblade
http://www.user-mode-linux.org/~blaisorblade
Chiacchiera con i tuoi amici in tempo reale! 
 http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com 


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Configuration problem with UML in kernel 2.6.19.1

[Blaisorblade]

On Wednesday 17 January 2007 18:15, Pravin wrote:
> Thanx  :-)
> Its working now.... :-)
>
> One more thing that I would like to tell here.
> I did used same compiled kernel on other machine which was having
> 2.6.17-1.2187_FC5.
> And that kernel 2.6.19 without your patch did worked fine....
>
> but same kernel 2.6.19 was not working on my machine which is Debian 
> 2.6.18. After applying your patch and changing the configuration to
> HOST_2G_2G its working fine now on my Debian machine 2.6.18.
>
> Can I know what excatly was the problem... ?

Usually the host kernel "gives"* a certain space (3G) to userspace apps. Some 
host kernels give less memory (say 2,75G, 2G, or even 1G) to userspace apps 
(these days, it depends on the CONFIG_VMSPLIT_* settings, before it could be 
tuned via external patches like -ck patchset or probably others).

Most application use the available memory, but UML, for now, needs to know 
that it can use a given amount - say 2G. It will run fine when more memory is 
available, but not when less memory is available.

At last, why different splits? While 3G is usual (normally named 3G/1G), on a 
host with 1G of RAM you either use HIGHMEM (with a little slowdown) or use a 
2,75G/1,25G split.

* This is "virtual address space", not a statical split assigning physical RAM 
to different purposes - but I won't go in the details here.

> My main objective is to run SELinux on UML kernel and test it.

Probably Antoine Martin is the one having more experience on UML+SELinux. This 
is his page on the argument:

http://uml.nagafix.co.uk/SELinux/

> Let me know if there is anything I can do to help here.
>
> On 1/17/07, Blaisorblade <blaisorblade@yahoo.it> wrote:
> > You mentioned trying to use CONFIG_HOST_VMSPLIT to solve a problem, that
> > is of little use without this applied, if you run UML in SKAS0 mode.
> >
> > ----------  Forwarded Message  ----------

-- 
Inform me of my mistakes, so I can add them to my list!
Paolo Giarrusso, aka Blaisorblade
http://www.user-mode-linux.org/~blaisorblade
Chiacchiera con i tuoi amici in tempo reale! 
 http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com 


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel