user-mode-linux-user

[uml-user] Creating Wheezy filesystem from scratch

[Ricardo N. <ric...@gm...>]

I'm trying to create a Debian Wheezy filesystem to use in User Mode Linux.
right now I'm following this
<https://help.ubuntu.com/community/UserModeLinux> tutorial. The commands
after # Remove the other terminal events don't work here and I can't find
the equivalent operations in Wheezy.

When I try to boot all of the system gets frozen and ends up with this
(full log here <http://pastebin.com/mVUH3zxX>):

INIT: Entering runlevel: 2

    [[36minfo[39;49m] Using makefile-style concurrent boot in runlevel 2.
    [   12.690000] Virtual console 5 assigned device '/dev/pts/12'
    [   12.690000] Virtual console 6 assigned device '/dev/pts/19'
    [   12.690000] Virtual console 4 assigned device '/dev/pts/20'
    [   12.690000] Virtual console 3 assigned device '/dev/pts/26'
    [   12.690000] Virtual console 2 assigned device '/dev/pts/27'
    [   12.690000] Virtual console 1 assigned device '/dev/pts/28'

I would use a already prepared filesystem, or even rootstrap, but both of
them forces me to have internet inside the UML machine and, for security
reasons in this environment, I cannot create bridges to my own network.

Do you know how can I pass this until I get a login message?

Thanks

Re: [uml-user] 64-bit UML with 32-bit root

[Richard W. <ric...@gm...>]

On Tue, Sep 1, 2015 at 8:32 AM, Antoine Martin <an...@na...> wrote:
> On 27/08/15 20:14, Mark Morgan Lloyd wrote:
>> The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that
>> it is possible to build UML on a 64-bit system to run a 32-bit guest.
> You can build a 32-bit *kernel* from a 64-bit host.
>> The best that I can manage on e.g. Debian "Jessie" x86-64 is to use
>> make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably
>> run a 32-bit guest. Prerequisites appear to be the multiarch-support
>>   and gcc-multiarch packages.
>>
>> Is it possible to build UML as a 64-bit binary, but to run a 32-bit
>> guest?
> AFAIK, no.
>>  What I'd like to be able to do is to put it on a system which has
>> no multiarch stuff, i.e. to completely sequester the 32-bit libraries
>> etc. within the guest filesystem.
>>
>> [Background: I used UML fairly heavily in the 2.4 era, but I'm a
>> comparative newcomer to x86-64. I'm trying to avoid overuse of
>> multi-arch stuff.]
> AFAIK, there is no multiarch support in in the UML *kernel*.

Yes. 64bit UML has no ia32 emulation like x86_64 has.

-- 
Thanks,
//richard

Re: [uml-user] UML console not coming

[Vijay T. <vi...@in...>]

Dear Sir,
        Did the following and it fixed the problem.
        - Downloaded "linux-3.18.20.tar.xz" from kernel.org
        - compiled it using:
               ARCH=um make mrproper
               ARCH=um make SUBARCH=i386 i386_defconfig
               ARCH=um make SUBARCH=i386

        - Downloaded "Ubuntu-TrustyTahr-i386-root_fs" and 
"Fedora21-x86-root_fs"
        Ran the following command:
         ./linux ubda=Ubuntu-TrustyTahr-i386-root_fs rw mem=512m umid=224
          Ubuntu rootfs is not working and then tried Fedora_root_fs.

         ./linux ubda=Fedora21-x86-root_fs rw mem=512m umid=224
         - It is working properly. Now I am getting the console without 
any error on UML.


      Thanks for your suggestion and help.

Thanks & Regards,
Vijay

On Tuesday 01 September 2015 11:58 AM, Antoine Martin wrote:
> On 01/09/15 11:07, Vijay Tandeker wrote:
>> Dear Sir,
>>           - Downloaded "linux-3.18.20.tar.xz" from kernel.org
>>           - compiled it for UML using "ARCH=um".
>>           - Downloaded "Ubuntu-TrustyTahr-i386-root_fs" and "Fedora21-x86-root_fs".
>>           - Still same issue with Ubuntu as well as Fedora root_fs.
>>
>>           - If needed, I can send the boot logs also.
>>
>>   My system configuration:
>>   - Ubuntu-14.04 32 bit
>>   - Kernel : 3.19
>>   
>>            - Should I download the latest source for kernel v4.0 and above ?
>>
>> Regards,
>> Vijay
>>
>> ________________________________________
>> From: Enjoy Mindful [enj...@gm...]
>> Sent: Tuesday, September 1, 2015 9:00
>> To: Vijay Tandeker
>> Cc: use...@li...
>> Subject: Re: [uml-user] UML console not coming
>>
>> On Sun, Aug 30, 2015 at 6:36 PM, Vijay Tandeker
>> <vi...@in...> wrote:
>>> Dear Sir,
>>>         I am on Ubuntu-14.04 LTS, 32 bit  system and already installed "uml-utilities " package through "apt-get". But still getting the error.
>>>         Can I install the package mentioned by you for Fedora in Ubuntu system?
>> I'm not familiar with Ubuntu. So, I don't know. Could you please try
>> to compile/build uml kernel from upstream source code? It was about
>> one year ago, I had trouble with compiled binary uml kernel downloaded
>> from the internet. I built the uml kernel locally on my PC. And it
>> works for me.
>>
>>
>>
>>> Regards,
>>> Vijay
>>>
>>> ________________________________________
>>> From: Enjoy Mindful [enj...@gm...]
>>> Sent: Sunday, August 30, 2015 4:51
>>> To: Vijay Tandeker
>>> Cc: use...@li...
>>> Subject: Re: [uml-user] UML console not coming
>>>
>>> You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
>>> Fedora). I don't know the package name for Ubuntu.
>>>
>>> On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
>>> <vi...@in...> wrote:
>>>> Hi All,
>>>>      I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>>>>
>>>>      I downloaded the following:
>>>>      - UML Kernel image v3.18 for i386 32 bit
>>>>      - root filesystem for Ubuntu-14 and Fedora21
>>>>
>>>>   My system configuration:
>>>>   - Ubuntu-14.04 32 bit
>>>>   - Kernel : 3.19
>>>>
>>>>   After downloading ran the following command:
>>>>   ./kernel ubda=Ubuntu_root_fs
>>>>
>>>>   After sometime, Got the following error continuously:
>>>>   - getmaster : no usable host pty device
>>>>
>>>>   I am unable to get the console.
>>>>
>>>>   I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
> The web page you downloaded these root_fs images from (judging by the
> filenames used) does mention that some of these images have problems
> with the tty devices.
> Building new kernels is unlikely to fix that.
>
> These images are meant to be generic, for use as chroot, UML or other.
> And they are far from perfect.
> So if you find a fix, please share it.
>
> Cheers
> Antoine
>
>
>
>>>>   Please help.
>>>>
>>>> Regards,
>>>> Vijay
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> _______________________________________________
>>>> User-mode-linux-user mailing list
>>>> Use...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> User-mode-linux-user mailing list
>> Use...@li...
>> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
>
> ------------------------------------------------------------------------------
> _______________________________________________
> User-mode-linux-user mailing list
> Use...@li...
> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML console not coming

[Jon F. <jon...@jf...>]

Vijay Tandeker wrote:
> Hi All,
>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>
>     I downloaded the following:
>     - UML Kernel image v3.18 for i386 32 bit
>     - root filesystem for Ubuntu-14 and Fedora21
>
>  My system configuration:
>  - Ubuntu-14.04 32 bit
>  - Kernel : 3.19
>
>  After downloading ran the following command:
>  ./kernel ubda=Ubuntu_root_fs
>
>  After sometime, Got the following error continuously:
>  - getmaster : no usable host pty device
>
>  I am unable to get the console.
>
>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
>
>  Please help.
>
> Regards,
>
>
> Vijay
Its hard to tell from your description if you are getting the message
from inside the UML or outside. I'm basically guessing it must be from
outside, stuff the kernel is dropping on the screen.

I use and have launched Debian instances, and few other distros, in UML
instances. Since Ubuntu is descended from Debian my experience may
apply. The number one cause for problems that I've had and could give
messages similar to yours has been with the Linux distro inside the UML
trying to open virtual consoles. To cut down on a number of related
issues I usually do something like this:

$ ./kernel con0=null,fd:2 con1=fd:0,fd:1 con=null ssl=null
ubda=Ubuntu_root_fs

This is what the new switches do:

* "con0=null,fd:2": ties "console 0" to stderr, without any input.
   Debian, and probably Ubuntu, use this for boot messages.

* "con1=fd:0,fd:1": ties the stdin and stdout of the current terminal
   session into "console 1", which is usually the first login console.

* "con=null": Turns off all other consoles, so if the distro in the
   UML tries to access any others it will get an error, simply failing
   to launch additional consoles, which shouldn't affect usability.

* "ssl=null": Turns off all serial ports. This for the same reason as
   above.

If you don't use the assorted con and ssl flags to specify which virtual
devices to attach to real devices and which ones to turn off, the Linux
image, inside the UML, will more than likely try to open a dozen or so
consoles and possibly serial ports. That is unless its been customized
specifically for UML use. The default action within the UML kernel
("./kernel") is typically to launch xterms for virtual console and
serial port, which gets to be real messy or to open "ptys" (pseudo
terminals), directly. Xterms need PTYs too... so its the same problem.

I don't know much about the Linux PTY system because in my experience it
just works. Ubuntu does things a bit differently so there might be a
package, or more than likely, a permission problem with your user account.

There is much more to say on this but the revised command above may give
you some instant gratification or at least help to diagnose the problem.
Its also good to read the page documenting those UML CLI (con*, ssl*)
switches: http://user-mode-linux.sourceforge.net/old/input.html.

- Jon

-- 
Sent from my Debian Linux workstation -- http://www.debian.org/intro/about

Jon Foster
JF Possibilities, Inc.
jo...@jf...
541-410-2760
Making computers work for you!

Re: [uml-user] UML console not coming

[Antoine M. <an...@na...>]

On 01/09/15 11:07, Vijay Tandeker wrote:
> Dear Sir,
>          - Downloaded "linux-3.18.20.tar.xz" from kernel.org
>          - compiled it for UML using "ARCH=um".
>          - Downloaded "Ubuntu-TrustyTahr-i386-root_fs" and "Fedora21-x86-root_fs".
>          - Still same issue with Ubuntu as well as Fedora root_fs.
>
>          - If needed, I can send the boot logs also.
>
>  My system configuration:
>  - Ubuntu-14.04 32 bit
>  - Kernel : 3.19
>  
>           - Should I download the latest source for kernel v4.0 and above ?
>
> Regards,
> Vijay
>
> ________________________________________
> From: Enjoy Mindful [enj...@gm...]
> Sent: Tuesday, September 1, 2015 9:00
> To: Vijay Tandeker
> Cc: use...@li...
> Subject: Re: [uml-user] UML console not coming
>
> On Sun, Aug 30, 2015 at 6:36 PM, Vijay Tandeker
> <vi...@in...> wrote:
>> Dear Sir,
>>        I am on Ubuntu-14.04 LTS, 32 bit  system and already installed "uml-utilities " package through "apt-get". But still getting the error.
>>        Can I install the package mentioned by you for Fedora in Ubuntu system?
> I'm not familiar with Ubuntu. So, I don't know. Could you please try
> to compile/build uml kernel from upstream source code? It was about
> one year ago, I had trouble with compiled binary uml kernel downloaded
> from the internet. I built the uml kernel locally on my PC. And it
> works for me.
>
>
>
>> Regards,
>> Vijay
>>
>> ________________________________________
>> From: Enjoy Mindful [enj...@gm...]
>> Sent: Sunday, August 30, 2015 4:51
>> To: Vijay Tandeker
>> Cc: use...@li...
>> Subject: Re: [uml-user] UML console not coming
>>
>> You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
>> Fedora). I don't know the package name for Ubuntu.
>>
>> On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
>> <vi...@in...> wrote:
>>> Hi All,
>>>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>>>
>>>     I downloaded the following:
>>>     - UML Kernel image v3.18 for i386 32 bit
>>>     - root filesystem for Ubuntu-14 and Fedora21
>>>
>>>  My system configuration:
>>>  - Ubuntu-14.04 32 bit
>>>  - Kernel : 3.19
>>>
>>>  After downloading ran the following command:
>>>  ./kernel ubda=Ubuntu_root_fs
>>>
>>>  After sometime, Got the following error continuously:
>>>  - getmaster : no usable host pty device
>>>
>>>  I am unable to get the console.
>>>
>>>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
The web page you downloaded these root_fs images from (judging by the
filenames used) does mention that some of these images have problems
with the tty devices.
Building new kernels is unlikely to fix that.

These images are meant to be generic, for use as chroot, UML or other.
And they are far from perfect.
So if you find a fix, please share it.

Cheers
Antoine



>>>
>>>  Please help.
>>>
>>> Regards,
>>> Vijay
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> _______________________________________________
>>> User-mode-linux-user mailing list
>>> Use...@li...
>>> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
> ------------------------------------------------------------------------------
> _______________________________________________
> User-mode-linux-user mailing list
> Use...@li...
> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] 64-bit UML with 32-bit root

[Antoine M. <an...@na...>]

On 27/08/15 20:14, Mark Morgan Lloyd wrote:
> The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that 
> it is possible to build UML on a 64-bit system to run a 32-bit guest.
You can build a 32-bit *kernel* from a 64-bit host.
> The best that I can manage on e.g. Debian "Jessie" x86-64 is to use 
> make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably 
> run a 32-bit guest. Prerequisites appear to be the multiarch-support 
>   and gcc-multiarch packages.
>
> Is it possible to build UML as a 64-bit binary, but to run a 32-bit 
> guest?
AFAIK, no.
>  What I'd like to be able to do is to put it on a system which has 
> no multiarch stuff, i.e. to completely sequester the 32-bit libraries 
> etc. within the guest filesystem.
>
> [Background: I used UML fairly heavily in the 2.4 era, but I'm a 
> comparative newcomer to x86-64. I'm trying to avoid overuse of 
> multi-arch stuff.]
AFAIK, there is no multiarch support in in the UML *kernel*.

Cheers
Antoine

Re: [uml-user] UML console not coming

[Vijay T. <vi...@in...>]

Dear Sir,
         - Downloaded "linux-3.18.20.tar.xz" from kernel.org
         - compiled it for UML using "ARCH=um".
         - Downloaded "Ubuntu-TrustyTahr-i386-root_fs" and "Fedora21-x86-root_fs".
         - Still same issue with Ubuntu as well as Fedora root_fs.

         - If needed, I can send the boot logs also.

 My system configuration:
 - Ubuntu-14.04 32 bit
 - Kernel : 3.19
 
          - Should I download the latest source for kernel v4.0 and above ?

Regards,
Vijay

________________________________________
From: Enjoy Mindful [enj...@gm...]
Sent: Tuesday, September 1, 2015 9:00
To: Vijay Tandeker
Cc: use...@li...
Subject: Re: [uml-user] UML console not coming

On Sun, Aug 30, 2015 at 6:36 PM, Vijay Tandeker
<vi...@in...> wrote:
> Dear Sir,
>        I am on Ubuntu-14.04 LTS, 32 bit  system and already installed "uml-utilities " package through "apt-get". But still getting the error.
>        Can I install the package mentioned by you for Fedora in Ubuntu system?

I'm not familiar with Ubuntu. So, I don't know. Could you please try
to compile/build uml kernel from upstream source code? It was about
one year ago, I had trouble with compiled binary uml kernel downloaded
from the internet. I built the uml kernel locally on my PC. And it
works for me.



>
> Regards,
> Vijay
>
> ________________________________________
> From: Enjoy Mindful [enj...@gm...]
> Sent: Sunday, August 30, 2015 4:51
> To: Vijay Tandeker
> Cc: use...@li...
> Subject: Re: [uml-user] UML console not coming
>
> You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
> Fedora). I don't know the package name for Ubuntu.
>
> On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
> <vi...@in...> wrote:
>> Hi All,
>>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>>
>>     I downloaded the following:
>>     - UML Kernel image v3.18 for i386 32 bit
>>     - root filesystem for Ubuntu-14 and Fedora21
>>
>>  My system configuration:
>>  - Ubuntu-14.04 32 bit
>>  - Kernel : 3.19
>>
>>  After downloading ran the following command:
>>  ./kernel ubda=Ubuntu_root_fs
>>
>>  After sometime, Got the following error continuously:
>>  - getmaster : no usable host pty device
>>
>>  I am unable to get the console.
>>
>>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
>>
>>  Please help.
>>
>> Regards,
>> Vijay
>>
>>
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> User-mode-linux-user mailing list
>> Use...@li...
>> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML console not coming

[Enjoy M. <enj...@gm...>]

On Sun, Aug 30, 2015 at 6:36 PM, Vijay Tandeker
<vi...@in...> wrote:
> Dear Sir,
>        I am on Ubuntu-14.04 LTS, 32 bit  system and already installed "uml-utilities " package through "apt-get". But still getting the error.
>        Can I install the package mentioned by you for Fedora in Ubuntu system?

I'm not familiar with Ubuntu. So, I don't know. Could you please try
to compile/build uml kernel from upstream source code? It was about
one year ago, I had trouble with compiled binary uml kernel downloaded
from the internet. I built the uml kernel locally on my PC. And it
works for me.



>
> Regards,
> Vijay
>
> ________________________________________
> From: Enjoy Mindful [enj...@gm...]
> Sent: Sunday, August 30, 2015 4:51
> To: Vijay Tandeker
> Cc: use...@li...
> Subject: Re: [uml-user] UML console not coming
>
> You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
> Fedora). I don't know the package name for Ubuntu.
>
> On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
> <vi...@in...> wrote:
>> Hi All,
>>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>>
>>     I downloaded the following:
>>     - UML Kernel image v3.18 for i386 32 bit
>>     - root filesystem for Ubuntu-14 and Fedora21
>>
>>  My system configuration:
>>  - Ubuntu-14.04 32 bit
>>  - Kernel : 3.19
>>
>>  After downloading ran the following command:
>>  ./kernel ubda=Ubuntu_root_fs
>>
>>  After sometime, Got the following error continuously:
>>  - getmaster : no usable host pty device
>>
>>  I am unable to get the console.
>>
>>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
>>
>>  Please help.
>>
>> Regards,
>> Vijay
>>
>>
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> User-mode-linux-user mailing list
>> Use...@li...
>> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML console not coming

[Vijay T. <vi...@in...>]

Dear Sir,
       I am on Ubuntu-14.04 LTS, 32 bit  system and already installed "uml-utilities " package through "apt-get". But still getting the error.
       Can I install the package mentioned by you for Fedora in Ubuntu system?

Regards,
Vijay

________________________________________
From: Enjoy Mindful [enj...@gm...]
Sent: Sunday, August 30, 2015 4:51
To: Vijay Tandeker
Cc: use...@li...
Subject: Re: [uml-user] UML console not coming

You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
Fedora). I don't know the package name for Ubuntu.

On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
<vi...@in...> wrote:
> Hi All,
>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>
>     I downloaded the following:
>     - UML Kernel image v3.18 for i386 32 bit
>     - root filesystem for Ubuntu-14 and Fedora21
>
>  My system configuration:
>  - Ubuntu-14.04 32 bit
>  - Kernel : 3.19
>
>  After downloading ran the following command:
>  ./kernel ubda=Ubuntu_root_fs
>
>  After sometime, Got the following error continuously:
>  - getmaster : no usable host pty device
>
>  I am unable to get the console.
>
>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
>
>  Please help.
>
> Regards,
> Vijay
>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> User-mode-linux-user mailing list
> Use...@li...
> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML console not coming

[Enjoy M. <enj...@gm...>]

You maybe need install uml_utilities-20070815-14.fc21.x86_64 (for
Fedora). I don't know the package name for Ubuntu.

On Sat, Aug 29, 2015 at 9:18 PM, Vijay Tandeker
<vi...@in...> wrote:
> Hi All,
>     I am Vijay, working as software engineer in Tejas Networks Bangalore, India.
>
>     I downloaded the following:
>     - UML Kernel image v3.18 for i386 32 bit
>     - root filesystem for Ubuntu-14 and Fedora21
>
>  My system configuration:
>  - Ubuntu-14.04 32 bit
>  - Kernel : 3.19
>
>  After downloading ran the following command:
>  ./kernel ubda=Ubuntu_root_fs
>
>  After sometime, Got the following error continuously:
>  - getmaster : no usable host pty device
>
>  I am unable to get the console.
>
>  I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.
>
>  Please help.
>
> Regards,
> Vijay
>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> User-mode-linux-user mailing list
> Use...@li...
> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] UML console not coming

[Vijay T. <vi...@in...>]

Hi All,
    I am Vijay, working as software engineer in Tejas Networks Bangalore, India.

    I downloaded the following:
    - UML Kernel image v3.18 for i386 32 bit
    - root filesystem for Ubuntu-14 and Fedora21

 My system configuration:
 - Ubuntu-14.04 32 bit
 - Kernel : 3.19

 After downloading ran the following command:
 ./kernel ubda=Ubuntu_root_fs

 After sometime, Got the following error continuously:
 - getmaster : no usable host pty device

 I am unable to get the console.

 I searched in internet about the solution but no luck. Tried with Fedora root_fs also but no success.

 Please help.

Regards,
Vijay

 

Re: [uml-user] 64-bit UML with 32-bit root

[Mark M. L. <mar...@te...>]

Michael Richardson wrote:
> Mark Morgan Lloyd <mar...@te...> wrote:
>     > The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that
>     > it is possible to build UML on a 64-bit system to run a 32-bit guest.
> 
>     > The best that I can manage on e.g. Debian "Jessie" x86-64 is to use
>     > make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably
>     > run a 32-bit guest. Prerequisites appear to be the multiarch-support
>     > and gcc-multiarch packages.
> 
>     > Is it possible to build UML as a 64-bit binary, but to run a 32-bit
>     > guest? What I'd like to be able to do is to put it on a system which has
>     > no multiarch stuff, i.e. to completely sequester the 32-bit libraries
>     > etc. within the guest filesystem.
> 
> You could try running that 64-bit kernel with a 32-bit binary.
> Put a statically linked 32-bit busybox in the guest file system, or use a
> 32-bit Debian initrd as a test case..
> 
> I suspect that it won't work because the enclosing ("dom0") kernel will have
> set the process to be a 64-bit process to run your kernel, and thus I think
> the entire address space will be 64-bit.

[Nod] Probably fragile at best.

>     > [Background: I used UML fairly heavily in the 2.4 era, but I'm a
>     > comparative newcomer to x86-64. I'm trying to avoid overuse of
>     > multi-arch stuff.]
> 
> I am not sure why you are trying to avoid this, unless you are trying to run
> the results on a system that doesn't have multi-arch.

Trying to keep systems as clean as possible. I had little option but to 
install user-level multi-arch for Acrobat Reader, and there's always 
going to be oddities like Ken Thompson's APL that will never get ported 
to 64-bit, but knowing how robust UML is /if/ it had handled the 
shimming it would have been a very attractive alternative.

> (I'm fighting/putting-off replacing a Fedora10 build system with a Jessie
> system, in great part because the appliance system needs a 64-bit kernel now,
> but my build environment is 32-bit, and not-multiarch capable. Build-root
> is the proper answer)

Jessie's got the expected number of Debianisms in it, even on something 
as ablutions-standard as i386. We're considering looking at Fedora...

Anyway, thanks everybody for the comments, which if nothing else have 
confirmed that while I'd misinterpreted the original text I'd not 
overlooked anything obvious when building kernels etc.

-- 
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk

[Opinions above are the author's, not those of his employers or colleagues]

Re: [uml-user] 64-bit UML with 32-bit root

[Michael R. <mc...@sa...>]

Mark Morgan Lloyd <mar...@te...> wrote:
    > The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that
    > it is possible to build UML on a 64-bit system to run a 32-bit guest.

    > The best that I can manage on e.g. Debian "Jessie" x86-64 is to use
    > make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably
    > run a 32-bit guest. Prerequisites appear to be the multiarch-support
    > and gcc-multiarch packages.

    > Is it possible to build UML as a 64-bit binary, but to run a 32-bit
    > guest? What I'd like to be able to do is to put it on a system which has
    > no multiarch stuff, i.e. to completely sequester the 32-bit libraries
    > etc. within the guest filesystem.

You could try running that 64-bit kernel with a 32-bit binary.
Put a statically linked 32-bit busybox in the guest file system, or use a
32-bit Debian initrd as a test case..

I suspect that it won't work because the enclosing ("dom0") kernel will have
set the process to be a 64-bit process to run your kernel, and thus I think
the entire address space will be 64-bit.

    > [Background: I used UML fairly heavily in the 2.4 era, but I'm a
    > comparative newcomer to x86-64. I'm trying to avoid overuse of
    > multi-arch stuff.]

I am not sure why you are trying to avoid this, unless you are trying to run
the results on a system that doesn't have multi-arch.

(I'm fighting/putting-off replacing a Fedora10 build system with a Jessie
system, in great part because the appliance system needs a 64-bit kernel now,
but my build environment is 32-bit, and not-multiarch capable. Build-root
is the proper answer)

--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        | network architect  [
]     mc...@sa...  http://www.sandelman.ca/        |   ruby on rails    [

Re: [uml-user] 64-bit UML with 32-bit root

[Julien Iguchi-C. <jul...@un...>]

On 08/27/2015 03:14 PM, Mark Morgan Lloyd wrote:
> The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that 
> it is possible to build UML on a 64-bit system to run a 32-bit guest.
> 
> The best that I can manage on e.g. Debian "Jessie" x86-64 is to use 
> make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably 
> run a 32-bit guest. Prerequisites appear to be the multiarch-support 
>   and gcc-multiarch packages.
> 
> Is it possible to build UML as a 64-bit binary, but to run a 32-bit 
> guest? What I'd like to be able to do is to put it on a system which has 
> no multiarch stuff, i.e. to completely sequester the 32-bit libraries 
> etc. within the guest filesystem.
> 
> [Background: I used UML fairly heavily in the 2.4 era, but I'm a 
> comparative newcomer to x86-64. I'm trying to avoid overuse of 
> multi-arch stuff.]
> 

UML 32 bits works on 64 bits because the 32 bits ISA is available. I
don't think it is possible on the other way (albeit pure emulation).

J.

[uml-user] 64-bit UML with 32-bit root

[Mark M. L. <mar...@te...>]

The (unofficial?) FAQ at http://uml.devloop.org.uk/faq.html implies that 
it is possible to build UML on a 64-bit system to run a 32-bit guest.

The best that I can manage on e.g. Debian "Jessie" x86-64 is to use 
make ARCH=um SUBARCH=i386  which results in a 32-bit ELF to presumably 
run a 32-bit guest. Prerequisites appear to be the multiarch-support 
  and gcc-multiarch packages.

Is it possible to build UML as a 64-bit binary, but to run a 32-bit 
guest? What I'd like to be able to do is to put it on a system which has 
no multiarch stuff, i.e. to completely sequester the 32-bit libraries 
etc. within the guest filesystem.

[Background: I used UML fairly heavily in the 2.4 era, but I'm a 
comparative newcomer to x86-64. I'm trying to avoid overuse of 
multi-arch stuff.]

-- 
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk

[Opinions above are the author's, not those of his employers or colleagues]

Re: [uml-user] UML randomly hanging

[Jon F. <jon...@jf...>]

I tried the latest LTS kernel (3.18.20) and I don't see any appreciable
difference. Although maybe not the hard freeze I was seeing, it still
doesn't seem to be able to handle a multi-tasking load like running
Drupal on Apache in the UML. Hit it with 40 simultaneous connections for
a minute and it will never (at least not within a usable timespan)
recover. The symptoms are:

  1. LF (in the UML) of 75 after 15 to 20 mins of no external activity.
It never
      simmered down. Appears that the workers are not actually running.

  2. A bunch of threads running on the host with at most 2% CPU.

  3. Slow response on the pseudo console (con1=fd:0,fd:1) or SSH sessions.
      "top" and "htop" update slowly hanging for several seconds at a time.

I did some digging because the book and other sources mention "ncpus"
controls the number of simultaneous threads it will allow to run on the
host. When I configured the kernel I never found an SMP option in
menuconfig. The resultant binary didn't mention "ncpus" as a CLI
parameter. But I found the note in Kconfig that mentions it only works
in TT mode, not with SKAS, probably does more harm then good and is most
likely broken. Would be nice if the docs in the kernel tree were up to date.

In short it doesn't appear that UML can really support more than one job
at a time and I need to move onto a full virtualization solution. A
shame really I was hopping to go light on this.

THX - Jon


Richard Weinberger wrote:
> On Fri, Aug 14, 2015 at 8:25 PM, Jon Foster
> <jon...@jf...> wrote:
>   
>> I'm using a stock Debian 7 AMD-64 SMP install on a 8 core Intel Xeon
>> server, with 24GB of RAM. I'm using Debian's supplied UML kernel and
>> tools. "uname -a", from within the UML returns: "Linux lamp1 3.2.54 #2
>> Thu Feb 6 22:33:28 UTC 2014 x86_64 GNU/Linux". Using a command like:
>>
>> linux.uml "umid=$MACHNAME" con0=null,fd:2 con1=fd:0,fd:1 \
>>          con=null ssl=null "mem=4096m" "eth0=tuntap,tap0" \
>>          hostfs=/var/local/umlhostfs "ubd0=$ROOT" "ubd1=$SWAP"
>>
>> Obviosuly the environment variables are filled with appropriate values.
>>
>> I'm running a Debian 5 LAMP stack inside the UML. "/tmp" is a regular
>> disk backed filesystem. It runs ok if its not doing anything. But even
>> then from time to time you'll get a second or two pause. If it gets  a
>> modest amount of traffic it will hang for a few minutes, and things just
>> get worse and worse from there.
>>
>> While its idle I've seen "hrtimer" warnings with 9 or 10 digit
>> nanosecond counts. I think this is at the heart of my problem.
>>
>> Anybody have any advice?
>>     
>
> Can you please give a more recent kernel a try?
> 3.2. is really old.
>
>   

-- 
Sent from my Debian Linux workstation -- http://www.debian.org/intro/about

Jon Foster
JF Possibilities, Inc.
jo...@jf...
541-410-2760
Making computers work for you!

Re: [uml-user] per-instance tmpfs backed memory file

[Jon F. <jon...@jf...>]

TEMP worked but not TEMPDIR.

THX - Jon

Richard Weinberger wrote:
> On Fri, Aug 14, 2015 at 8:35 PM, Jon Foster
> <jon...@jf...> wrote:
>   
>> I'm kind of new to the UML thing. I've read that every instance of UML
>> creates a temp file that represents the RAM of that instance. Its
>> obvious that putting that in a "tmpfs" would be very good, but I don't
>> want to mount a tmpfs on "/tmp", its used by too many other things and
>> the size restriction would cause problems.
>>
>> I've read a "per-instance" tmpfs can be setup... but nowhere have I seen
>> how that's done. There doesn't appear to be any CLI switches for telling
>> UML where to put the temp file and $TMPDIR doesn't seem to do anything
>> for it.
>>
>> I'm running stock Debian 7 64bit on Intel Xeons, with their supplied UML
>> packages. The UML kernel version is 3.2.54.
>>
>> How can I tell UML to put its temp file somewhere else?
>>     
>
> UML checks the environment variables TMP, TEMP and TEMPDIR.
> Recent UML kernels use /dev/shm.
>
>   

-- 
Sent from my Debian Linux workstation -- http://www.debian.org/intro/about

Jon Foster
JF Possibilities, Inc.
jo...@jf...
541-410-2760
Making computers work for you!

Re: [uml-user] UML randomly hanging

[Richard W. <ric...@gm...>]

On Fri, Aug 14, 2015 at 8:25 PM, Jon Foster
<jon...@jf...> wrote:
> I'm using a stock Debian 7 AMD-64 SMP install on a 8 core Intel Xeon
> server, with 24GB of RAM. I'm using Debian's supplied UML kernel and
> tools. "uname -a", from within the UML returns: "Linux lamp1 3.2.54 #2
> Thu Feb 6 22:33:28 UTC 2014 x86_64 GNU/Linux". Using a command like:
>
> linux.uml "umid=$MACHNAME" con0=null,fd:2 con1=fd:0,fd:1 \
>          con=null ssl=null "mem=4096m" "eth0=tuntap,tap0" \
>          hostfs=/var/local/umlhostfs "ubd0=$ROOT" "ubd1=$SWAP"
>
> Obviosuly the environment variables are filled with appropriate values.
>
> I'm running a Debian 5 LAMP stack inside the UML. "/tmp" is a regular
> disk backed filesystem. It runs ok if its not doing anything. But even
> then from time to time you'll get a second or two pause. If it gets  a
> modest amount of traffic it will hang for a few minutes, and things just
> get worse and worse from there.
>
> While its idle I've seen "hrtimer" warnings with 9 or 10 digit
> nanosecond counts. I think this is at the heart of my problem.
>
> Anybody have any advice?

Can you please give a more recent kernel a try?
3.2. is really old.

-- 
Thanks,
//richard

Re: [uml-user] per-instance tmpfs backed memory file

[Richard W. <ric...@gm...>]

On Fri, Aug 14, 2015 at 8:35 PM, Jon Foster
<jon...@jf...> wrote:
> I'm kind of new to the UML thing. I've read that every instance of UML
> creates a temp file that represents the RAM of that instance. Its
> obvious that putting that in a "tmpfs" would be very good, but I don't
> want to mount a tmpfs on "/tmp", its used by too many other things and
> the size restriction would cause problems.
>
> I've read a "per-instance" tmpfs can be setup... but nowhere have I seen
> how that's done. There doesn't appear to be any CLI switches for telling
> UML where to put the temp file and $TMPDIR doesn't seem to do anything
> for it.
>
> I'm running stock Debian 7 64bit on Intel Xeons, with their supplied UML
> packages. The UML kernel version is 3.2.54.
>
> How can I tell UML to put its temp file somewhere else?

UML checks the environment variables TMP, TEMP and TEMPDIR.
Recent UML kernels use /dev/shm.

-- 
Thanks,
//richard

[uml-user] UML randomly hanging

[Jon F. <jon...@jf...>]

I'm using a stock Debian 7 AMD-64 SMP install on a 8 core Intel Xeon
server, with 24GB of RAM. I'm using Debian's supplied UML kernel and
tools. "uname -a", from within the UML returns: "Linux lamp1 3.2.54 #2
Thu Feb 6 22:33:28 UTC 2014 x86_64 GNU/Linux". Using a command like:

linux.uml "umid=$MACHNAME" con0=null,fd:2 con1=fd:0,fd:1 \
         con=null ssl=null "mem=4096m" "eth0=tuntap,tap0" \
         hostfs=/var/local/umlhostfs "ubd0=$ROOT" "ubd1=$SWAP"

Obviosuly the environment variables are filled with appropriate values.

I'm running a Debian 5 LAMP stack inside the UML. "/tmp" is a regular
disk backed filesystem. It runs ok if its not doing anything. But even
then from time to time you'll get a second or two pause. If it gets  a
modest amount of traffic it will hang for a few minutes, and things just
get worse and worse from there.

While its idle I've seen "hrtimer" warnings with 9 or 10 digit
nanosecond counts. I think this is at the heart of my problem.

Anybody have any advice?

TIA - Jon

-- 
Sent from my Debian Linux workstation -- http://www.debian.org/intro/about

Jon Foster
JF Possibilities, Inc.
jo...@jf...
541-410-2760
Making computers work for you!

[uml-user] per-instance tmpfs backed memory file

[Jon F. <jon...@jf...>]

I'm kind of new to the UML thing. I've read that every instance of UML
creates a temp file that represents the RAM of that instance. Its
obvious that putting that in a "tmpfs" would be very good, but I don't
want to mount a tmpfs on "/tmp", its used by too many other things and
the size restriction would cause problems.

I've read a "per-instance" tmpfs can be setup... but nowhere have I seen
how that's done. There doesn't appear to be any CLI switches for telling
UML where to put the temp file and $TMPDIR doesn't seem to do anything
for it.

I'm running stock Debian 7 64bit on Intel Xeons, with their supplied UML
packages. The UML kernel version is 3.2.54.

How can I tell UML to put its temp file somewhere else?

THX - Jon

-- 
Sent from my Debian Linux workstation -- http://www.debian.org/intro/about

Jon Foster
JF Possibilities, Inc.
jo...@jf...
541-410-2760
Making computers work for you!

[uml-user] [PATCH 70/74] um/time: Migrate to new 'set-state' interface

[Daniel L. <dan...@li...>]

From: Viresh Kumar <vir...@li...>

Migrate um driver to the new 'set-state' interface provided by
clockevents core, the earlier 'set-mode' interface is marked obsolete
now.

This also enables us to implement callbacks for new states of clockevent
devices, for example: ONESHOT_STOPPED.

Cc: Jeff Dike <jd...@ad...>
Cc: Richard Weinberger <ri...@no...>
Cc: use...@li...
Cc: use...@li...
Signed-off-by: Viresh Kumar <vir...@li...>
Signed-off-by: Daniel Lezcano <dan...@li...>
---
 arch/um/kernel/time.c | 44 ++++++++++++++++++++------------------------
 1 file changed, 20 insertions(+), 24 deletions(-)

diff --git a/arch/um/kernel/time.c b/arch/um/kernel/time.c
index 117568d..5af441e 100644
--- a/arch/um/kernel/time.c
+++ b/arch/um/kernel/time.c
@@ -22,23 +22,16 @@ void timer_handler(int sig, struct siginfo *unused_si, struct uml_pt_regs *regs)
 	local_irq_restore(flags);
 }
 
-static void itimer_set_mode(enum clock_event_mode mode,
-			    struct clock_event_device *evt)
+static int itimer_shutdown(struct clock_event_device *evt)
 {
-	switch (mode) {
-	case CLOCK_EVT_MODE_PERIODIC:
-		set_interval();
-		break;
-
-	case CLOCK_EVT_MODE_SHUTDOWN:
-	case CLOCK_EVT_MODE_UNUSED:
-	case CLOCK_EVT_MODE_ONESHOT:
-		disable_timer();
-		break;
-
-	case CLOCK_EVT_MODE_RESUME:
-		break;
-	}
+	disable_timer();
+	return 0;
+}
+
+static int itimer_set_periodic(struct clock_event_device *evt)
+{
+	set_interval();
+	return 0;
 }
 
 static int itimer_next_event(unsigned long delta,
@@ -48,14 +41,17 @@ static int itimer_next_event(unsigned long delta,
 }
 
 static struct clock_event_device itimer_clockevent = {
-	.name		= "itimer",
-	.rating		= 250,
-	.cpumask	= cpu_all_mask,
-	.features	= CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
-	.set_mode	= itimer_set_mode,
-	.set_next_event = itimer_next_event,
-	.shift		= 32,
-	.irq		= 0,
+	.name			= "itimer",
+	.rating			= 250,
+	.cpumask		= cpu_all_mask,
+	.features		= CLOCK_EVT_FEAT_PERIODIC |
+				  CLOCK_EVT_FEAT_ONESHOT,
+	.set_state_shutdown	= itimer_shutdown,
+	.set_state_periodic	= itimer_set_periodic,
+	.set_state_oneshot	= itimer_shutdown,
+	.set_next_event		= itimer_next_event,
+	.shift			= 32,
+	.irq			= 0,
 };
 
 static irqreturn_t um_timer(int irq, void *dev)
-- 
1.9.1

Re: [uml-user] Segmentation faults when UML is run under gdb

[Sergei A. <sa...@gm...>]

On 6 August 2015 at 14:20, Enjoy Mindful <enj...@gm...> wrote:
> http://copilotco.com/mail-archives/uml.2008/msg01515.html

It helped, thanks.
This valuable information deserves to be here:
http://user-mode-linux.sourceforge.net/hacking.html

Re: [uml-user] Segmentation faults when UML is run under gdb

[Enjoy M. <enj...@gm...>]

http://copilotco.com/mail-archives/uml.2008/msg01515.html

On Thu, Aug 6, 2015 at 7:54 PM, Sergei Antonov <sa...@gm...> wrote:
> Hello!
> I'm trying compile and run UML (version 4.1) on x86_64. It builds and
> runs fine as a normal program. But when run under gdb, it produces
> several segmentation faults.
>
> This is how I build and run it (based on
> http://user-mode-linux.sourceforge.net/hacking.html instructions and
> disk image from http://fs.devloop.org.uk/ ):
> cd linux &&
> make mrproper &&
> make mrproper ARCH=um &&
> git checkout v4.1 &&
> make defconfig ARCH=um &&
> make -j6 ARCH=um &&
> cd .. &&
> gdb --args linux/linux ubda=BusyBox-1.13.2-amd64-root_fs mem=256m
>
>
> And this is a complete gdb output:
>
> ===========================================================
> GNU gdb (GDB) 7.9.1
> Copyright (C) 2015 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "x86_64-unknown-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> <http://www.gnu.org/software/gdb/bugs/>.
> Find the GDB manual and other documentation resources online at:
> <http://www.gnu.org/software/gdb/documentation/>.
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols from linux/linux...done.
> (gdb) r
> Starting program: /root/exfat/linux/linux
> ubda=BusyBox-1.13.2-amd64-root_fs mem=256m
> Core dump limits :
> soft - 0
> hard - NONE
> Core dump limits :
> soft - 0
> hard - NONE
> Core dump limits :
> soft - 0
> hard - NONE
> Core dump limits :
> soft - 0
> hard - NONE
> Checking environment variables for a tempdir...none found
> Checking if /dev/shm is on tmpfs...OK
> Checking PROT_EXEC mmap in /dev/shm...OK
> Core dump limits :
> soft - 0
> hard - NONE
> Checking environment variables for a tempdir...none found
> Checking if /dev/shm is on tmpfs...OK
> Checking PROT_EXEC mmap in /dev/shm...OK
> Initializing cgroup subsys cpuset
> Initializing cgroup subsys cpu
> Initializing cgroup subsys cpuacct
> Linux version 4.1.0 (root@linux64) (gcc version 4.9.2 (GCC) ) #1 Tue
> Aug 4 17:31:01 CEST 2015
> Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 64640
> Kernel command line: ubda=BusyBox-1.13.2-amd64-root_fs mem=256m root=98:0
> PID hash table entries: 1024 (order: 1, 8192 bytes)
> Dentry cache hash table entries: 32768 (order: 6, 262144 bytes)
> Inode-cache hash table entries: 16384 (order: 5, 131072 bytes)
> Memory: 253796K/262144K available (2471K kernel code, 624K rwdata,
> 780K rodata, 109K init, 161K bss, 8348K reserved, 0K cma-reserved)
> NR_IRQS:15
> clocksource itimer: mask: 0xffffffffffffffff max_cycles: 0x1d854df40,
> max_idle_ns: 3526361616960 ns
> Calibrating delay loop... 2649.29 BogoMIPS (lpj=13246464)
> pid_max: default: 32768 minimum: 301
> Mount-cache hash table entries: 512 (order: 0, 4096 bytes)
> Mountpoint-cache hash table entries: 512 (order: 0, 4096 bytes)
> Initializing cgroup subsys blkio
> Initializing cgroup subsys devices
> Initializing cgroup subsys freezer
> Checking that host ptys support output SIGIO...Yes
> Checking that host ptys support SIGIO on close...No, enabling workaround
> devtmpfs: initialized
> Using 2.6 host AIO
> clocksource jiffies: mask: 0xffffffff max_cycles: 0xffffffff,
> max_idle_ns: 19112604462750000 ns
> NET: Registered protocol family 16
> Switched to clocksource itimer
> NET: Registered protocol family 2
> TCP established hash table entries: 2048 (order: 2, 16384 bytes)
> TCP bind hash table entries: 2048 (order: 2, 16384 bytes)
> TCP: Hash tables configured (established 2048 bind 2048)
> UDP hash table entries: 256 (order: 1, 8192 bytes)
> UDP-Lite hash table entries: 256 (order: 1, 8192 bytes)
> NET: Registered protocol family 1
> console [stderr0] disabled
> mconsole (version 2) initialized on /root/.uml/1BxIWh/mconsole
> Checking host MADV_REMOVE support...OK
> futex hash table entries: 256 (order: 0, 6144 bytes)
> VFS: Disk quotas dquot_6.6.0
> VFS: Dquot-cache hash table entries: 512 (order 0, 4096 bytes)
> io scheduler noop registered
> io scheduler deadline registered (default)
> NET: Registered protocol family 17
> Initialized stdio console driver
> Console initialized on /dev/tty0
> console [tty0] enabled
> Initializing software serial port version 1
> console [mc-1] enabled
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x00007ffff78b16c4 in memset () from /lib64/libc.so.6
> (gdb) bt
> #0  0x00007ffff78b16c4 in memset () from /lib64/libc.so.6
> #1  0x00000000601b14f3 in check_partition (hd=<optimized out>,
> bdev=<optimized out>) at block/partitions/check.c:165
> #2  0x00000000601b0abb in rescan_partitions (disk=0x70800000,
> bdev=0x0) at block/partition-generic.c:433
> #3  0x00000000600e7783 in __blkdev_get (bdev=0x6f802880,
> mode=<optimized out>, for_part=0) at fs/block_dev.c:1213
> #4  0x00000000600e7c3b in blkdev_get (bdev=0x0, mode=<optimized out>,
> holder=0x0) at fs/block_dev.c:1317
> #5  0x00000000601ae7e3 in register_disk (disk=<optimized out>) at
> block/genhd.c:556
> #6  add_disk (disk=0x6fd94000) at block/genhd.c:618
> #7  0x0000000060026270 in ubd_disk_register (major=<optimized out>,
> size=<optimized out>, unit=1614123056, disk_out=0x603f6340
> <ubd_gendisk>) at arch/um/drivers/ubd_kern.c:835
> #8  0x0000000060026b1a in ubd_add (n=0, error_out=0x6fc4de90) at
> arch/um/drivers/ubd_kern.c:872
> #9  0x0000000060003c23 in ubd_init () at arch/um/drivers/ubd_kern.c:1073
> #10 0x00000000600184c4 in do_one_initcall (fn=0x60003b5a <ubd_init>)
> at init/main.c:788
> #11 0x0000000060001e38 in do_initcall_level (level=<optimized out>) at
> init/main.c:853
> #12 do_initcalls () at init/main.c:861
> #13 do_basic_setup () at init/main.c:880
> #14 kernel_init_freeable () at init/main.c:1001
> #15 0x000000006027ce2f in kernel_init (unused=<optimized out>) at
> init/main.c:931
> #16 0x0000000060019a7f in new_thread_handler () at arch/um/kernel/process.c:129
> #17 0x0000000000000000 in ?? ()
> (gdb) c
> Continuing.
>
> Program received signal SIGSEGV, Segmentation fault.
> n_tty_open (tty=0x6fcb5c00) at drivers/tty/n_tty.c:1929
> 1929 ldata->overrun_time = jiffies;
> (gdb) bt
> #0  n_tty_open (tty=0x6fcb5c00) at drivers/tty/n_tty.c:1929
> #1  0x00000000601d1f49 in tty_ldisc_open (tty=0x70804288,
> ld=<optimized out>) at drivers/tty/tty_ldisc.c:450
> #2  0x00000000601d289a in tty_ldisc_setup (tty=0x6fcb5c00, o_tty=0x0)
> at drivers/tty/tty_ldisc.c:735
> #3  0x00000000601cc503 in tty_init_dev (driver=0x6fdb3d00, idx=0) at
> drivers/tty/tty_io.c:1548
> #4  0x00000000601cc906 in tty_open (inode=0x6f8001e0, filp=0x6fd2d900)
> at drivers/tty/tty_io.c:2091
> #5  0x00000000600bce52 in chrdev_open (inode=0x6f8001e0,
> filp=0x6fd2d900) at fs/char_dev.c:388
> #6  0x00000000600b74c2 in do_dentry_open (f=0x6fd2d900,
> open=0x600bcd15 <chrdev_open>, cred=<optimized out>) at fs/open.c:734
> #7  0x00000000600b7690 in vfs_open (path=<optimized out>,
> filp=<optimized out>, cred=<optimized out>) at fs/open.c:871
> #8  0x00000000600c42ad in do_last (nd=0x6fc4de10, path=0x6fc4ddb0,
> file=0x6fd2d900, op=<optimized out>, opened=<optimized out>,
> name=<optimized out>) at fs/namei.c:3104
> #9  0x00000000600c62ca in path_openat (dfd=<optimized out>,
> pathname=<optimized out>, nd=0x6fc4de10, op=0x6fc4df0c,
> flags=<optimized out>) at fs/namei.c:3243
> #10 0x00000000600c7443 in do_filp_open (dfd=-100, pathname=0x6fe37000,
> op=0x6fc4df0c) at fs/namei.c:3290
> #11 0x00000000600b860e in do_sys_open (dfd=-100, filename=<optimized
> out>, flags=<optimized out>, mode=<optimized out>) at fs/open.c:1014
> #12 0x00000000600b86e0 in SYSC_open (mode=<optimized out>,
> flags=<optimized out>, filename=<optimized out>) at fs/open.c:1032
> #13 SyS_open (filename=<optimized out>, flags=<optimized out>,
> mode=<optimized out>) at fs/open.c:1027
> #14 0x0000000060001e6b in kernel_init_freeable () at init/main.c:1004
> #15 0x000000006027ce2f in kernel_init (unused=<optimized out>) at
> init/main.c:931
> #16 0x0000000060019a7f in new_thread_handler () at arch/um/kernel/process.c:129
> #17 0x0000000000000000 in ?? ()
> (gdb) c
> Continuing.
> EXT4-fs (ubda): mounting ext3 file system using the ext4 subsystem
> EXT4-fs (ubda): mounted filesystem with ordered data mode. Opts: (null)
> VFS: Mounted root (ext3 filesystem) readonly on device 98:0.
> devtmpfs: mounted
>
> Program received signal SIGSEGV, Segmentation fault.
> n_tty_open (tty=0x6fd16800) at drivers/tty/n_tty.c:1929
> 1929 ldata->overrun_time = jiffies;
> (gdb)
> Continuing.
> Virtual console 5 assigned device '/dev/pts/1'
>
> Program received signal SIGSEGV, Segmentation fault.
> n_tty_open (tty=0x6fd16000) at drivers/tty/n_tty.c:1929
> 1929 ldata->overrun_time = jiffies;
> (gdb)
> Continuing.
> Serial line 0 assigned device '/dev/pts/2'
>
> Please press Enter to activate this console.
> ===========================================================
>
>
> The first fault happens in check_partition(), then several more happen
> in some tty code. I investigated the first fault a little and found
> this.
>
> 1. The memset is called by this line:
>   memset(state->parts, 0, state->limit * sizeof(state->parts[0]));
>
> 2. I printk-ed its parameters:
>   printk("  %s %p %lu\n", __func__, state->parts, state->limit *
> sizeof(state->parts[0]));
> they look normal:
>   check_partition 0000000070800000 2048
>
> 3. Even a 1-byte memset produces segfault too:
>   memset(state->parts, 0, 1);
>
> 4. The memory is allocated by vzalloc() in allocate_partitions().
>
> 5. Writing 1 byte with vwrite() does not fault and returns 0.
>   vwrite(state->parts, &some_variable, 1);
>
>
> I tried changing a kernel version (4.0) and a version of gdb (7.7.1) -
> it was all the same.
>
> How do I fix this problem and debug UML without getting scary and
> disturbing faults?
>
> ------------------------------------------------------------------------------
> _______________________________________________
> User-mode-linux-user mailing list
> Use...@li...
> https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] Segmentation faults when UML is run under gdb

[Sergei A. <sa...@gm...>]

Hello!
I'm trying compile and run UML (version 4.1) on x86_64. It builds and
runs fine as a normal program. But when run under gdb, it produces
several segmentation faults.

This is how I build and run it (based on
http://user-mode-linux.sourceforge.net/hacking.html instructions and
disk image from http://fs.devloop.org.uk/ ):
cd linux &&
make mrproper &&
make mrproper ARCH=um &&
git checkout v4.1 &&
make defconfig ARCH=um &&
make -j6 ARCH=um &&
cd .. &&
gdb --args linux/linux ubda=BusyBox-1.13.2-amd64-root_fs mem=256m


And this is a complete gdb output:

===========================================================
GNU gdb (GDB) 7.9.1
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-unknown-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from linux/linux...done.
(gdb) r
Starting program: /root/exfat/linux/linux
ubda=BusyBox-1.13.2-amd64-root_fs mem=256m
Core dump limits :
soft - 0
hard - NONE
Core dump limits :
soft - 0
hard - NONE
Core dump limits :
soft - 0
hard - NONE
Core dump limits :
soft - 0
hard - NONE
Checking environment variables for a tempdir...none found
Checking if /dev/shm is on tmpfs...OK
Checking PROT_EXEC mmap in /dev/shm...OK
Core dump limits :
soft - 0
hard - NONE
Checking environment variables for a tempdir...none found
Checking if /dev/shm is on tmpfs...OK
Checking PROT_EXEC mmap in /dev/shm...OK
Initializing cgroup subsys cpuset
Initializing cgroup subsys cpu
Initializing cgroup subsys cpuacct
Linux version 4.1.0 (root@linux64) (gcc version 4.9.2 (GCC) ) #1 Tue
Aug 4 17:31:01 CEST 2015
Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 64640
Kernel command line: ubda=BusyBox-1.13.2-amd64-root_fs mem=256m root=98:0
PID hash table entries: 1024 (order: 1, 8192 bytes)
Dentry cache hash table entries: 32768 (order: 6, 262144 bytes)
Inode-cache hash table entries: 16384 (order: 5, 131072 bytes)
Memory: 253796K/262144K available (2471K kernel code, 624K rwdata,
780K rodata, 109K init, 161K bss, 8348K reserved, 0K cma-reserved)
NR_IRQS:15
clocksource itimer: mask: 0xffffffffffffffff max_cycles: 0x1d854df40,
max_idle_ns: 3526361616960 ns
Calibrating delay loop... 2649.29 BogoMIPS (lpj=13246464)
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 512 (order: 0, 4096 bytes)
Mountpoint-cache hash table entries: 512 (order: 0, 4096 bytes)
Initializing cgroup subsys blkio
Initializing cgroup subsys devices
Initializing cgroup subsys freezer
Checking that host ptys support output SIGIO...Yes
Checking that host ptys support SIGIO on close...No, enabling workaround
devtmpfs: initialized
Using 2.6 host AIO
clocksource jiffies: mask: 0xffffffff max_cycles: 0xffffffff,
max_idle_ns: 19112604462750000 ns
NET: Registered protocol family 16
Switched to clocksource itimer
NET: Registered protocol family 2
TCP established hash table entries: 2048 (order: 2, 16384 bytes)
TCP bind hash table entries: 2048 (order: 2, 16384 bytes)
TCP: Hash tables configured (established 2048 bind 2048)
UDP hash table entries: 256 (order: 1, 8192 bytes)
UDP-Lite hash table entries: 256 (order: 1, 8192 bytes)
NET: Registered protocol family 1
console [stderr0] disabled
mconsole (version 2) initialized on /root/.uml/1BxIWh/mconsole
Checking host MADV_REMOVE support...OK
futex hash table entries: 256 (order: 0, 6144 bytes)
VFS: Disk quotas dquot_6.6.0
VFS: Dquot-cache hash table entries: 512 (order 0, 4096 bytes)
io scheduler noop registered
io scheduler deadline registered (default)
NET: Registered protocol family 17
Initialized stdio console driver
Console initialized on /dev/tty0
console [tty0] enabled
Initializing software serial port version 1
console [mc-1] enabled

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff78b16c4 in memset () from /lib64/libc.so.6
(gdb) bt
#0  0x00007ffff78b16c4 in memset () from /lib64/libc.so.6
#1  0x00000000601b14f3 in check_partition (hd=<optimized out>,
bdev=<optimized out>) at block/partitions/check.c:165
#2  0x00000000601b0abb in rescan_partitions (disk=0x70800000,
bdev=0x0) at block/partition-generic.c:433
#3  0x00000000600e7783 in __blkdev_get (bdev=0x6f802880,
mode=<optimized out>, for_part=0) at fs/block_dev.c:1213
#4  0x00000000600e7c3b in blkdev_get (bdev=0x0, mode=<optimized out>,
holder=0x0) at fs/block_dev.c:1317
#5  0x00000000601ae7e3 in register_disk (disk=<optimized out>) at
block/genhd.c:556
#6  add_disk (disk=0x6fd94000) at block/genhd.c:618
#7  0x0000000060026270 in ubd_disk_register (major=<optimized out>,
size=<optimized out>, unit=1614123056, disk_out=0x603f6340
<ubd_gendisk>) at arch/um/drivers/ubd_kern.c:835
#8  0x0000000060026b1a in ubd_add (n=0, error_out=0x6fc4de90) at
arch/um/drivers/ubd_kern.c:872
#9  0x0000000060003c23 in ubd_init () at arch/um/drivers/ubd_kern.c:1073
#10 0x00000000600184c4 in do_one_initcall (fn=0x60003b5a <ubd_init>)
at init/main.c:788
#11 0x0000000060001e38 in do_initcall_level (level=<optimized out>) at
init/main.c:853
#12 do_initcalls () at init/main.c:861
#13 do_basic_setup () at init/main.c:880
#14 kernel_init_freeable () at init/main.c:1001
#15 0x000000006027ce2f in kernel_init (unused=<optimized out>) at
init/main.c:931
#16 0x0000000060019a7f in new_thread_handler () at arch/um/kernel/process.c:129
#17 0x0000000000000000 in ?? ()
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
n_tty_open (tty=0x6fcb5c00) at drivers/tty/n_tty.c:1929
1929 ldata->overrun_time = jiffies;
(gdb) bt
#0  n_tty_open (tty=0x6fcb5c00) at drivers/tty/n_tty.c:1929
#1  0x00000000601d1f49 in tty_ldisc_open (tty=0x70804288,
ld=<optimized out>) at drivers/tty/tty_ldisc.c:450
#2  0x00000000601d289a in tty_ldisc_setup (tty=0x6fcb5c00, o_tty=0x0)
at drivers/tty/tty_ldisc.c:735
#3  0x00000000601cc503 in tty_init_dev (driver=0x6fdb3d00, idx=0) at
drivers/tty/tty_io.c:1548
#4  0x00000000601cc906 in tty_open (inode=0x6f8001e0, filp=0x6fd2d900)
at drivers/tty/tty_io.c:2091
#5  0x00000000600bce52 in chrdev_open (inode=0x6f8001e0,
filp=0x6fd2d900) at fs/char_dev.c:388
#6  0x00000000600b74c2 in do_dentry_open (f=0x6fd2d900,
open=0x600bcd15 <chrdev_open>, cred=<optimized out>) at fs/open.c:734
#7  0x00000000600b7690 in vfs_open (path=<optimized out>,
filp=<optimized out>, cred=<optimized out>) at fs/open.c:871
#8  0x00000000600c42ad in do_last (nd=0x6fc4de10, path=0x6fc4ddb0,
file=0x6fd2d900, op=<optimized out>, opened=<optimized out>,
name=<optimized out>) at fs/namei.c:3104
#9  0x00000000600c62ca in path_openat (dfd=<optimized out>,
pathname=<optimized out>, nd=0x6fc4de10, op=0x6fc4df0c,
flags=<optimized out>) at fs/namei.c:3243
#10 0x00000000600c7443 in do_filp_open (dfd=-100, pathname=0x6fe37000,
op=0x6fc4df0c) at fs/namei.c:3290
#11 0x00000000600b860e in do_sys_open (dfd=-100, filename=<optimized
out>, flags=<optimized out>, mode=<optimized out>) at fs/open.c:1014
#12 0x00000000600b86e0 in SYSC_open (mode=<optimized out>,
flags=<optimized out>, filename=<optimized out>) at fs/open.c:1032
#13 SyS_open (filename=<optimized out>, flags=<optimized out>,
mode=<optimized out>) at fs/open.c:1027
#14 0x0000000060001e6b in kernel_init_freeable () at init/main.c:1004
#15 0x000000006027ce2f in kernel_init (unused=<optimized out>) at
init/main.c:931
#16 0x0000000060019a7f in new_thread_handler () at arch/um/kernel/process.c:129
#17 0x0000000000000000 in ?? ()
(gdb) c
Continuing.
EXT4-fs (ubda): mounting ext3 file system using the ext4 subsystem
EXT4-fs (ubda): mounted filesystem with ordered data mode. Opts: (null)
VFS: Mounted root (ext3 filesystem) readonly on device 98:0.
devtmpfs: mounted

Program received signal SIGSEGV, Segmentation fault.
n_tty_open (tty=0x6fd16800) at drivers/tty/n_tty.c:1929
1929 ldata->overrun_time = jiffies;
(gdb)
Continuing.
Virtual console 5 assigned device '/dev/pts/1'

Program received signal SIGSEGV, Segmentation fault.
n_tty_open (tty=0x6fd16000) at drivers/tty/n_tty.c:1929
1929 ldata->overrun_time = jiffies;
(gdb)
Continuing.
Serial line 0 assigned device '/dev/pts/2'

Please press Enter to activate this console.
===========================================================


The first fault happens in check_partition(), then several more happen
in some tty code. I investigated the first fault a little and found
this.

1. The memset is called by this line:
  memset(state->parts, 0, state->limit * sizeof(state->parts[0]));

2. I printk-ed its parameters:
  printk("  %s %p %lu\n", __func__, state->parts, state->limit *
sizeof(state->parts[0]));
they look normal:
  check_partition 0000000070800000 2048

3. Even a 1-byte memset produces segfault too:
  memset(state->parts, 0, 1);

4. The memory is allocated by vzalloc() in allocate_partitions().

5. Writing 1 byte with vwrite() does not fault and returns 0.
  vwrite(state->parts, &some_variable, 1);


I tried changing a kernel version (4.0) and a version of gdb (7.7.1) -
it was all the same.

How do I fix this problem and debug UML without getting scary and
disturbing faults?