LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [RFC] UML kernel & rootfs bundle with every kernel release ?
@ 2007-03-31 22:58 devzero
  2007-04-01  0:01 ` Parag Warudkar
  2007-04-01 19:08 ` Blaisorblade
  0 siblings, 2 replies; 20+ messages in thread
From: devzero @ 2007-03-31 22:58 UTC (permalink / raw)
  To: linux-kernel; +Cc: blaisorblade, jdike

Hello !

i`m not very much into UML for the last months, but while playing around with dm-loop i just got one idea i`d like to share.

Whenever you want to test some new kernel (feature), you may put you main system at risk, exactly know what you`re doing - or - use UserModeLinux.

The "problem" with UML is:

- you needs to compile an UML kernel first.
- you needs some "basic" knowlege about UML to get things running.
- you need to create an appropriate filesystem image for UML - or find some for download
- you need to copy appropriate kernel modules inside
- you need to put kernel sources inside, have compiler......
- you may need appropriate modle-init-tools,initrd, kernel specific tools (updated dmsetup, updated....whatever)

in short: 
it`s quite some work to be done to have your uml 2.6.21 with root-fs up and running and working cleanly.
whenever i search the net for some appropriate UML fs image, those i find are very often old and outdated...

is there a project/website which is offering such ready to run "UML kernel+rootfs release bundles" for download (i.e. new kernel,generic  root-fs, modules inside, sources inside, compiler inside - in sync with the latest stable vanilla) , or , would it make sense to establish such project ?
i.e. besides releasing the kernel, also releasing sort of a kernel "runtime kit" and/or "devkit" ?

i think this could be very helpful for linux-kernel, because it could be tested by more people more quickly, more easily and thus, more often.
just download, do few steps for setup, start up that virtual machine and there you go testing, hacking into the sources, do all that things you never would do on your main system,  whatever....

it would probably also add benefit to UML itself.

does this sound dumb? i don`t know, so please comment.

regards
roland

PS:
ok, this would be some 500M to 1G download, but there`s lot`s of bandwidth today - and P2P/Bittorrent.....
_______________________________________________________________
SMS schreiben mit WEB.DE FreeMail - einfach, schnell und
kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192


^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-03-31 22:58 [RFC] UML kernel & rootfs bundle with every kernel release ? devzero
@ 2007-04-01  0:01 ` Parag Warudkar
  2007-04-01 19:08 ` Blaisorblade
  1 sibling, 0 replies; 20+ messages in thread
From: Parag Warudkar @ 2007-04-01  0:01 UTC (permalink / raw)
  To: linux-kernel

Hi 
 
<devzero <at> web.de> writes:

> Whenever you want to test some new kernel (feature), you may put you main
system at risk, exactly know what
> you`re doing - or - use UserModeLinux.

Why won't qemu work better in this case? I generally keep a debian testing
installation on disk and when I compile a new kernel I just point qemu to load
it with the debian root fs. It's fast enough (even the kernel mode accelerator
module is GPLed now) and you don't need to mess around with your main system's
kernel. You can even test different arches - like both x86_64 and i386 on one
x64 box for example.

It doesn't work for any particular driver related testing but UML won't either.
And it won't benefit UML development but I don't know if that was your main
objective as opposed to general kernel experimentation and testing.

Parag


^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-03-31 22:58 [RFC] UML kernel & rootfs bundle with every kernel release ? devzero
  2007-04-01  0:01 ` Parag Warudkar
@ 2007-04-01 19:08 ` Blaisorblade
  2007-04-01 19:58   ` [uml-devel] " Antoine Martin
  1 sibling, 1 reply; 20+ messages in thread
From: Blaisorblade @ 2007-04-01 19:08 UTC (permalink / raw)
  To: devzero, user-mode-linux-devel; +Cc: linux-kernel, jdike

On domenica 1 aprile 2007, devzero@web.de wrote:
> Hello !
>
> i`m not very much into UML for the last months, but while playing around
> with dm-loop i just got one idea i`d like to share.
>
> Whenever you want to test some new kernel (feature), you may put you main
> system at risk, exactly know what you`re doing - or - use UserModeLinux.
>
> The "problem" with UML is:
>
> - you needs to compile an UML kernel first.
> - you needs some "basic" knowlege about UML to get things running.
> - you need to create an appropriate filesystem image for UML - or find some
> for download 

> - you need to copy appropriate kernel modules inside 

Well, you just build usually monolithic kernels, so you skip this problem.
And an idea that likely netkit uses is taking modules through an hostfs mount 
(or, it would be an easier setup). But you need modules to have uid 0...

> - you need to put kernel sources inside, have compiler......
> - you may need appropriate modle-init-tools,initrd, kernel specific tools
> (updated dmsetup, updated....whatever)

You found 2.4 images, but 2.6 ones do exist.

> in short:
> it`s quite some work to be done to have your uml 2.6.21 with root-fs up and
> running and working cleanly. whenever i search the net for some appropriate
> UML fs image, those i find are very often old and outdated...

Hmm... I'd think we need a wizard for configuration. Plus some distro-like 
work for some specific issues - if I want to deploy a VM with hostname x, 
network config y, and with Xnest running, I need an easier way to do that.

You can add settings on kernel command line and parse them inside UML - we 
need standard packaged utilities for that (one of the rootfs builder 
installed such stuff).

> is there a project/website which is offering such ready to run "UML
> kernel+rootfs release bundles" for download (i.e. new kernel,generic 
> root-fs, modules inside, sources inside, compiler inside - in sync with the
> latest stable vanilla) , or , would it make sense to establish such project
> ? i.e. besides releasing the kernel, also releasing sort of a kernel
> "runtime kit" and/or "devkit" ?

The runtime kit is there on nagafix.co.uk. The devkit is a main idea - most of 
the work is to put something on the UML wiki and market well the idea - 
creating such an image would be easy. But I haven't clear what you're talking 
about - kernel development (why sources inside) or userspace development?

Also integrating all possible debug stuff would be useful, but I don't know 
what's needed.

> i think this could be very helpful for linux-kernel, because it could be
> tested by more people more quickly, more easily and thus, more often. just
> download, do few steps for setup, start up that virtual machine and there
> you go testing, hacking into the sources, do all that things you never
> would do on your main system,  whatever....
>
> it would probably also add benefit to UML itself.

We need three things:
a) more performance
b) more users

c) more developers

a) leads to c), and b) too.

> does this sound dumb? i don`t know, so please comment.

No, it's not dumb. I'm even wanting to have a "Vmware-like" interface. Or at 
least standard scripts for guest management.

> regards
> roland
>
> PS:
> ok, this would be some 500M to 1G download, but there`s lot`s of bandwidth
> today - and P2P/Bittorrent.....

uml.nagafix.co.uk has some good kernels + images. With compression, they're 
even as little as 50 Mb.

However, making UML easier to use, and marketing it for more application, is 
very important. Various project do exist but they're not integrated, and they 
do not try to (netkit is for network experimentation, but is also better as 
VM management tool).

In short, we'd need somebody helping out really with the website (there is the 
wiki but you must request an account via email, and it's not the main 
website), with uml-utilities, and with new uml-utilities (you know 
dm-snapshot is a faster COW? Something to setup it automatically would be 
good).

There would be more to say on this, but I can't right now (I've other stuff to 
do).

-- 
Inform me of my mistakes, so I can add them to my list!
Paolo Giarrusso, aka Blaisorblade
http://www.user-mode-linux.org/~blaisorblade

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-01 19:08 ` Blaisorblade
@ 2007-04-01 19:58   ` Antoine Martin
  2007-04-02  2:05     ` Jeff Dike
  0 siblings, 1 reply; 20+ messages in thread
From: Antoine Martin @ 2007-04-01 19:58 UTC (permalink / raw)
  To: Blaisorblade; +Cc: devzero, user-mode-linux-devel, jdike, linux-kernel

[...]
>> in short:
>> it`s quite some work to be done to have your uml 2.6.21 with root-fs up and
>> running and working cleanly. whenever i search the net for some appropriate
>> UML fs image, those i find are very often old and outdated...
> 
> Hmm... I'd think we need a wizard for configuration. Plus some distro-like 
> work for some specific issues - if I want to deploy a VM with hostname x, 
> network config y, and with Xnest running, I need an easier way to do that.
I think this can be solved by having the network auto-configure itself 
and make the startup scripts responsible for bringing up the interface 
and starting dhcp server on that interface. This is simple and does not 
require changes to the images. (it takes care of dns, ntp, smb, etc..)

> You can add settings on kernel command line and parse them inside UML - we 
> need standard packaged utilities for that (one of the rootfs builder 
> installed such stuff).
That would certainly complicate things quite a bit.

>> is there a project/website which is offering such ready to run "UML
>> kernel+rootfs release bundles" for download (i.e. new kernel,generic 
>> root-fs, modules inside, sources inside, compiler inside - in sync with the
>> latest stable vanilla) , or , would it make sense to establish such project
>> ? i.e. besides releasing the kernel, also releasing sort of a kernel
>> "runtime kit" and/or "devkit" ?
> 
> The runtime kit is there on nagafix.co.uk. The devkit is a main idea - most of 
> the work is to put something on the UML wiki and market well the idea - 
> creating such an image would be easy. But I haven't clear what you're talking 
> about - kernel development (why sources inside) or userspace development?
> 
> Also integrating all possible debug stuff would be useful, but I don't know 
> what's needed.
> 
>> i think this could be very helpful for linux-kernel, because it could be
>> tested by more people more quickly, more easily and thus, more often. just
>> download, do few steps for setup, start up that virtual machine and there
>> you go testing, hacking into the sources, do all that things you never
>> would do on your main system,  whatever....
>>
>> it would probably also add benefit to UML itself.
> 
> We need three things:
> a) more performance
> b) more users
> 
> c) more developers
> 
> a) leads to c), and b) too.
> 
>> does this sound dumb? i don`t know, so please comment.
> 
> No, it's not dumb. I'm even wanting to have a "Vmware-like" interface. Or at 
> least standard scripts for guest management.
I reckon that one critical thing which could drastically increase the 
user base would be to have a working virtual framebuffer implementation.
my 2p

Cheers
Antoine



> 
>> regards
>> roland
>>
>> PS:
>> ok, this would be some 500M to 1G download, but there`s lot`s of bandwidth
>> today - and P2P/Bittorrent.....
> 
> uml.nagafix.co.uk has some good kernels + images. With compression, they're 
> even as little as 50 Mb.
> 
> However, making UML easier to use, and marketing it for more application, is 
> very important. Various project do exist but they're not integrated, and they 
> do not try to (netkit is for network experimentation, but is also better as 
> VM management tool).
> 
> In short, we'd need somebody helping out really with the website (there is the 
> wiki but you must request an account via email, and it's not the main 
> website), with uml-utilities, and with new uml-utilities (you know 
> dm-snapshot is a faster COW? Something to setup it automatically would be 
> good).
> 
> There would be more to say on this, but I can't right now (I've other stuff to 
> do).
> 


^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-01 19:58   ` [uml-devel] " Antoine Martin
@ 2007-04-02  2:05     ` Jeff Dike
  2007-04-02 10:21       ` Antoine Martin
  0 siblings, 1 reply; 20+ messages in thread
From: Jeff Dike @ 2007-04-02  2:05 UTC (permalink / raw)
  To: Antoine Martin; +Cc: Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
> I reckon that one critical thing which could drastically increase the 
> user base would be to have a working virtual framebuffer implementation.

Why?  I've never understood what a framebuffer gives you that you
don't have now.

				Jeff

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

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02  2:05     ` Jeff Dike
@ 2007-04-02 10:21       ` Antoine Martin
  2007-04-02 11:22         ` Geert Uytterhoeven
                           ` (3 more replies)
  0 siblings, 4 replies; 20+ messages in thread
From: Antoine Martin @ 2007-04-02 10:21 UTC (permalink / raw)
  To: Jeff Dike; +Cc: Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

Jeff Dike wrote:
> On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
>> I reckon that one critical thing which could drastically increase the 
>> user base would be to have a working virtual framebuffer implementation.
> 
> Why?  I've never understood what a framebuffer gives you that you
> don't have now.
Just like the network auto-configuration via dhcp, it would allow users 
to download images+kernel and run them like appliances without 
understanding anything about X or UML, just click and run.
We are all capable of setting up Xvfb here, but most users are not, 
which is why they download ready-made images.
It would also make it a lot easier to focus on writing a management UI, 
hell if there isn't one shortly after, I'll do one myself!
Think of a UML browser image (running IE via wine in a limited image 
with just X + wine + IE - I would much prefer that to having wine+IE 
installed locally), testing framebuffer apps like gtk-fb/cairo-fb 
without risking your dev environment, etc...

Antoine

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 10:21       ` Antoine Martin
@ 2007-04-02 11:22         ` Geert Uytterhoeven
  2007-04-02 12:50           ` Geert Uytterhoeven
  2007-04-02 14:53           ` Jeff Dike
  2007-04-02 14:55         ` Jeff Dike
                           ` (2 subsequent siblings)
  3 siblings, 2 replies; 20+ messages in thread
From: Geert Uytterhoeven @ 2007-04-02 11:22 UTC (permalink / raw)
  To: Antoine Martin
  Cc: Jeff Dike, Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

On Mon, 2 Apr 2007, Antoine Martin wrote:
> Jeff Dike wrote:
> > On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
> >> I reckon that one critical thing which could drastically increase the 
> >> user base would be to have a working virtual framebuffer implementation.
> > 
> > Why?  I've never understood what a framebuffer gives you that you
> > don't have now.
> Just like the network auto-configuration via dhcp, it would allow users 
> to download images+kernel and run them like appliances without 
> understanding anything about X or UML, just click and run.
> We are all capable of setting up Xvfb here, but most users are not, 
> which is why they download ready-made images.
> It would also make it a lot easier to focus on writing a management UI, 
> hell if there isn't one shortly after, I'll do one myself!
> Think of a UML browser image (running IE via wine in a limited image 
> with just X + wine + IE - I would much prefer that to having wine+IE 
> installed locally), testing framebuffer apps like gtk-fb/cairo-fb 
> without risking your dev environment, etc...

There are patches floating around for a UML frame buffer device.
Gerd Kraxel^H^H^H^H^H^HHoffmann did one using plain X11, which worked
great when I gave it a try.

A few years ago I tried doing one myself using gdk/gtk, but it didn't
work reliably on all systems. My main motivation was doing development for
different frame buffer memory organizations (think Amiga and Atari bitplanes)
on UML.

I suggest taking a look at Gerd's patches. IIRC, he posted them to lkml
last year.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 11:22         ` Geert Uytterhoeven
@ 2007-04-02 12:50           ` Geert Uytterhoeven
  2007-04-02 14:53           ` Jeff Dike
  1 sibling, 0 replies; 20+ messages in thread
From: Geert Uytterhoeven @ 2007-04-02 12:50 UTC (permalink / raw)
  To: Antoine Martin
  Cc: Jeff Dike, Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

On Mon, 2 Apr 2007, Geert Uytterhoeven wrote:
> Gerd Kraxel^H^H^H^H^H^HHoffmann did one using plain X11, which worked
       ^^^^^^
Sorry, this should have been `Knorr'.

> great when I gave it a try.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 11:22         ` Geert Uytterhoeven
  2007-04-02 12:50           ` Geert Uytterhoeven
@ 2007-04-02 14:53           ` Jeff Dike
  2007-04-02 15:43             ` Antoine Martin
  1 sibling, 1 reply; 20+ messages in thread
From: Jeff Dike @ 2007-04-02 14:53 UTC (permalink / raw)
  To: Geert Uytterhoeven
  Cc: Antoine Martin, Blaisorblade, devzero, user-mode-linux-devel,
	linux-kernel

On Mon, Apr 02, 2007 at 01:22:00PM +0200, Geert Uytterhoeven wrote:
> There are patches floating around for a UML frame buffer device.
> Gerd Kraxel^H^H^H^H^H^HHoffmann did one using plain X11, which worked
> great when I gave it a try.
> 
> I suggest taking a look at Gerd's patches. IIRC, he posted them to lkml
> last year.

I have Gerd's patch sitting in my tree, but it's never really worked
for me, so I haven't pushed it.

				Jeff

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

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 10:21       ` Antoine Martin
  2007-04-02 11:22         ` Geert Uytterhoeven
@ 2007-04-02 14:55         ` Jeff Dike
  2007-04-02 15:43           ` Antoine Martin
  2007-04-02 20:40         ` roland
  2007-04-03 17:43         ` Blaisorblade
  3 siblings, 1 reply; 20+ messages in thread
From: Jeff Dike @ 2007-04-02 14:55 UTC (permalink / raw)
  To: Antoine Martin; +Cc: Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

On Mon, Apr 02, 2007 at 11:21:43AM +0100, Antoine Martin wrote:
> Just like the network auto-configuration via dhcp, it would allow users 
> to download images+kernel and run them like appliances without 
> understanding anything about X or UML, just click and run.

True, but I don't see that as being revolutionary.  If people are
going to run a VM, presumably they have some purpose in mind, but how
often does that purpose involve having a GUI?

				Jeff

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

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 14:55         ` Jeff Dike
@ 2007-04-02 15:43           ` Antoine Martin
  0 siblings, 0 replies; 20+ messages in thread
From: Antoine Martin @ 2007-04-02 15:43 UTC (permalink / raw)
  To: Jeff Dike; +Cc: Blaisorblade, devzero, user-mode-linux-devel, linux-kernel

Jeff Dike wrote:
> On Mon, Apr 02, 2007 at 11:21:43AM +0100, Antoine Martin wrote:
>> Just like the network auto-configuration via dhcp, it would allow users 
>> to download images+kernel and run them like appliances without 
>> understanding anything about X or UML, just click and run.
> 
> True, but I don't see that as being revolutionary.  If people are
> going to run a VM, presumably they have some purpose in mind, but how
> often does that purpose involve having a GUI?
I think the GUI is more important for new users than anyone else.

Antoine

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 14:53           ` Jeff Dike
@ 2007-04-02 15:43             ` Antoine Martin
  2007-04-03 23:49               ` Antoine Martin
  0 siblings, 1 reply; 20+ messages in thread
From: Antoine Martin @ 2007-04-02 15:43 UTC (permalink / raw)
  To: Jeff Dike
  Cc: Geert Uytterhoeven, Blaisorblade, devzero, user-mode-linux-devel,
	linux-kernel

Jeff Dike wrote:
> On Mon, Apr 02, 2007 at 01:22:00PM +0200, Geert Uytterhoeven wrote:
>> There are patches floating around for a UML frame buffer device.
>> Gerd Kraxel^H^H^H^H^H^HHoffmann did one using plain X11, which worked
>> great when I gave it a try.
>>
>> I suggest taking a look at Gerd's patches. IIRC, he posted them to lkml
>> last year.
> 
> I have Gerd's patch sitting in my tree, but it's never really worked
> for me, so I haven't pushed it.
Yep, I tried them and it didn't work for me either, I'll try again though.

Antoine

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 10:21       ` Antoine Martin
  2007-04-02 11:22         ` Geert Uytterhoeven
  2007-04-02 14:55         ` Jeff Dike
@ 2007-04-02 20:40         ` roland
  2007-04-02 21:44           ` Jeff Dike
  2007-04-03 17:43         ` Blaisorblade
  3 siblings, 1 reply; 20+ messages in thread
From: roland @ 2007-04-02 20:40 UTC (permalink / raw)
  To: Antoine Martin, Jeff Dike
  Cc: Blaisorblade, user-mode-linux-devel, linux-kernel

>> Why?  I've never understood what a framebuffer gives you that you
>> don't have now.
> Just like the network auto-configuration via dhcp, it would allow users to 
> download images+kernel and run them like appliances without understanding 
> anything about X or UML, just click and run.

mhh, i also think framebuffer would be nice, but the main idea behind my 
posting is have some recent linux kernel up and running without any effort - 
with main focus on kernel, not on apps.

sure, it would be nice to have X for the unexperienced users, but if you run 
uml and get networking up and running, you just do ssh -X into your uml and 
then you have your X. those people who run the latest kernel for testing 
special things are not unexperienced, but having a packaged kernel+rootfs 
"ready to run" would also make _their_ life easier.  think of testing 
several kernels in parallel to find/fix regressions.

booting into something graphical with kdm/gdm login, firefox inside etc 
would be really nice, but that would be probably overkill for the purpose it 
should serve. framebuffer and X could be added later to uml, and i`m sure, 
it will exist one day.

btw - i think besides that packaged uml+rootfs, the same thing could be 
distributed in other formats, i.e. qemu, vmware, M$ virtual pc (add your 
favourite v12n solution here)

regards
roland



----- Original Message ----- 
From: "Antoine Martin" <antoine@nagafix.co.uk>
To: "Jeff Dike" <jdike@addtoit.com>
Cc: "Blaisorblade" <blaisorblade@yahoo.it>; <devzero@web.de>; 
<user-mode-linux-devel@lists.sourceforge.net>; 
<linux-kernel@vger.kernel.org>
Sent: Monday, April 02, 2007 12:21 PM
Subject: Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel 
release ?


> Jeff Dike wrote:
>> On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
>>> I reckon that one critical thing which could drastically increase the 
>>> user base would be to have a working virtual framebuffer implementation.
>>
>> Why?  I've never understood what a framebuffer gives you that you
>> don't have now.
> Just like the network auto-configuration via dhcp, it would allow users to 
> download images+kernel and run them like appliances without understanding 
> anything about X or UML, just click and run.
> We are all capable of setting up Xvfb here, but most users are not, which 
> is why they download ready-made images.
> It would also make it a lot easier to focus on writing a management UI, 
> hell if there isn't one shortly after, I'll do one myself!
> Think of a UML browser image (running IE via wine in a limited image with 
> just X + wine + IE - I would much prefer that to having wine+IE installed 
> locally), testing framebuffer apps like gtk-fb/cairo-fb without risking 
> your dev environment, etc...
>
> Antoine 


^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 20:40         ` roland
@ 2007-04-02 21:44           ` Jeff Dike
  2007-04-03  4:44             ` Jason Lunz
  0 siblings, 1 reply; 20+ messages in thread
From: Jeff Dike @ 2007-04-02 21:44 UTC (permalink / raw)
  To: roland; +Cc: Antoine Martin, Blaisorblade, user-mode-linux-devel, linux-kernel

On Mon, Apr 02, 2007 at 10:40:10PM +0200, roland wrote:
> btw - i think besides that packaged uml+rootfs, the same thing could be 
> distributed in other formats, i.e. qemu, vmware, M$ virtual pc (add your 
> favourite v12n solution here)

There are sites (http://uml.nagafix.co.uk/ being the best one I know
of) where, with two downloads, two uncompressions, and one command
line later, you have a booted UML.

The only way I know of to improve on this, aside from inprovements in
the booted distro, is to package the filesystem as a rootfs within the
UML kernel binary.  I've considered this, but haven't done anything
with it.

				Jeff

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

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 21:44           ` Jeff Dike
@ 2007-04-03  4:44             ` Jason Lunz
  0 siblings, 0 replies; 20+ messages in thread
From: Jason Lunz @ 2007-04-03  4:44 UTC (permalink / raw)
  To: Jeff Dike
  Cc: roland, Antoine Martin, Blaisorblade, user-mode-linux-devel,
	linux-kernel

On Mon, Apr 02, 2007 at 05:44:34PM -0400, Jeff Dike wrote:
> There are sites (http://uml.nagafix.co.uk/ being the best one I know
> of) where, with two downloads, two uncompressions, and one command
> line later, you have a booted UML.
> 
> The only way I know of to improve on this, aside from inprovements in
> the booted distro, is to package the filesystem as a rootfs within the
> UML kernel binary.  I've considered this, but haven't done anything
> with it.

I've done the converse: package the uml kernel within the rootfs image,
and use a script that plays the part of bootloader. With ext2 at least,
it's fairly easy to use the debugfs 'cat' command for this.

That way, you simply distribute the fs image with a companion script
that can boot any number of such images.

Jason

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 10:21       ` Antoine Martin
                           ` (2 preceding siblings ...)
  2007-04-02 20:40         ` roland
@ 2007-04-03 17:43         ` Blaisorblade
  2007-04-03 17:53           ` Antoine Martin
  3 siblings, 1 reply; 20+ messages in thread
From: Blaisorblade @ 2007-04-03 17:43 UTC (permalink / raw)
  To: Antoine Martin; +Cc: Jeff Dike, devzero, user-mode-linux-devel, linux-kernel

On lunedì 2 aprile 2007, Antoine Martin wrote:
> Jeff Dike wrote:
> > On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
> >> I reckon that one critical thing which could drastically increase the
> >> user base would be to have a working virtual framebuffer implementation.
> >
> > Why?  I've never understood what a framebuffer gives you that you
> > don't have now.
>
> Just like the network auto-configuration via dhcp,
Hmm... for that to be completely plug-and-play you need to make sure a dhcp 
server on the host exists.

Vmware runs a separate DHCP server exactly for this, even if we should avoid 
that as much as possible.

> it would allow users 
> to download images+kernel and run them like appliances without
> understanding anything about X or UML, just click and run.

> We are all capable of setting up Xvfb here, but most users are not,
> which is why they download ready-made images.

What about installing and pre-configuring Xnest on the image? With a suitable 
script calling xhost on the host, it just works.

This project did it:
http://umlbuilder.sourceforge.net/

although it stopped working for me ages ago (probably for some UML bug). I 
built a Mandrake image (that I now lost) with Xnest configured. With a script 
on the host which passes the host IP and that calls xhost, it should work 
easily. And btw, we need a standard startup script anyway.

> It would also make it a lot easier to focus on writing a management UI,
> hell if there isn't one shortly after, I'll do one myself!

Why not one management UI running from the host, a-la vmware? Possibly, with 
as much code as possible in scripting languages, for better transparency.

> Think of a UML browser image (running IE via wine in a limited image
> with just X + wine + IE - I would much prefer that to having wine+IE
> installed locally), testing framebuffer apps like gtk-fb/cairo-fb
> without risking your dev environment, etc...
>
> Antoine



-- 
Inform me of my mistakes, so I can add them to my list!
Paolo Giarrusso, aka Blaisorblade
http://www.user-mode-linux.org/~blaisorblade

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-03 17:43         ` Blaisorblade
@ 2007-04-03 17:53           ` Antoine Martin
  0 siblings, 0 replies; 20+ messages in thread
From: Antoine Martin @ 2007-04-03 17:53 UTC (permalink / raw)
  To: Blaisorblade; +Cc: Jeff Dike, devzero, user-mode-linux-devel, linux-kernel

Blaisorblade wrote:
> On lunedì 2 aprile 2007, Antoine Martin wrote:
>> Jeff Dike wrote:
>>> On Sun, Apr 01, 2007 at 08:58:45PM +0100, Antoine Martin wrote:
>>>> I reckon that one critical thing which could drastically increase the
>>>> user base would be to have a working virtual framebuffer implementation.
>>> Why?  I've never understood what a framebuffer gives you that you
>>> don't have now.
>> Just like the network auto-configuration via dhcp,
> Hmm... for that to be completely plug-and-play you need to make sure a dhcp 
> server on the host exists.
> 
> Vmware runs a separate DHCP server exactly for this, even if we should avoid 
> that as much as possible.
It is trivial to write a shell script that takes care of setting up the 
interface and checks for the presence of a dhcp server. (the dhcpd.conf 
can be generated easily)

The problem lies with the configuration options, I have got a bunch of 
shell scripts to take care of that, but it is quite hard to make them 
suitable for global consumption by the average user:
* even with just tap networking, I use 4 different ways of hooking them 
up the outside world
* firewalling and mac address filtering
* bridging issues and ebtables
* tmpfs and memory issues
* selinux and chrooting issues...
etc...

[snip]
> although it stopped working for me ages ago (probably for some UML bug). I 
> built a Mandrake image (that I now lost) with Xnest configured. With a script 
> on the host which passes the host IP and that calls xhost, it should work 
> easily. And btw, we need a standard startup script anyway.
It would work, but it's not a pretty solution, it requires 
customizations to the guest and it would not be intuitive to new users.
I would much prefer the ability to just run any distro (even framebuffer 
based ones) without modifications using the virtual framebuffer.

>> It would also make it a lot easier to focus on writing a management UI,
>> hell if there isn't one shortly after, I'll do one myself!
> 
> Why not one management UI running from the host, a-la vmware?
Yes ,that was what I had in mind.
> Possibly, with 
> as much code as possible in scripting languages, for better transparency.
Definitely, (see above)

The management tools I have written export the guests' settings to the 
filesystem in the form of a shell script file, all the utility scripts 
just read those settings and do their stuff. (all in simple shell)
Antoine


>> Think of a UML browser image (running IE via wine in a limited image
>> with just X + wine + IE - I would much prefer that to having wine+IE
>> installed locally), testing framebuffer apps like gtk-fb/cairo-fb
>> without risking your dev environment, etc...
>>
>> Antoine
> 
> 
> 


^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-02 15:43             ` Antoine Martin
@ 2007-04-03 23:49               ` Antoine Martin
  2007-04-04  6:35                 ` Geert Uytterhoeven
  0 siblings, 1 reply; 20+ messages in thread
From: Antoine Martin @ 2007-04-03 23:49 UTC (permalink / raw)
  To: Antoine Martin
  Cc: Jeff Dike, Geert Uytterhoeven, Blaisorblade, devzero,
	user-mode-linux-devel, linux-kernel, kraxel

Antoine Martin wrote:
> Jeff Dike wrote:
>> On Mon, Apr 02, 2007 at 01:22:00PM +0200, Geert Uytterhoeven wrote:
>>> There are patches floating around for a UML frame buffer device.
>>> Gerd Kraxel^H^H^H^H^H^HHoffmann did one using plain X11, which worked
>>> great when I gave it a try.
>>>
>>> I suggest taking a look at Gerd's patches. IIRC, he posted them to lkml
>>> last year.
>>
>> I have Gerd's patch sitting in my tree, but it's never really worked
>> for me, so I haven't pushed it.
> Yep, I tried them and it didn't work for me either, I'll try again though.
I did try again, and I found this patch:
http://dl.bytesex.org/patches/uml-2.6.11/uml-x11-fb
Which applied (2 rejects by hand), but failed with
"Mapping memory: Cannot allocate memory"
in mmap64:
os_map_memory [...]
   at arch/um/os-Linux/process.c:164

and this one:
http://www.suse.de/~kraxel/uml/patches/2.6.18-rc4/uml-x11-fb
which applied cleanly, but is not letting me set the option - Kconfig is 
beyond me:

arch/um/Kconfig:144:warning: 'select' used by config symbol 'X11_FB' 
refer to undefined symbol 'FB_CFB_FILLRECT'
arch/um/Kconfig:145:warning: 'select' used by config symbol 'X11_FB' 
refer to undefined symbol 'FB_CFB_COPYAREA'
arch/um/Kconfig:146:warning: 'select' used by config symbol 'X11_FB' 
refer to undefined symbol 'FB_CFB_IMAGEBLIT'

So that's a non-starter for me...

Antoine

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-03 23:49               ` Antoine Martin
@ 2007-04-04  6:35                 ` Geert Uytterhoeven
  2007-04-04 20:27                   ` Antoine Martin
  0 siblings, 1 reply; 20+ messages in thread
From: Geert Uytterhoeven @ 2007-04-04  6:35 UTC (permalink / raw)
  To: Antoine Martin
  Cc: Jeff Dike, Blaisorblade, devzero,
	User-mode Linux Kernel Development, Linux Kernel Development,
	kraxel

On Wed, 4 Apr 2007, Antoine Martin wrote:
> and this one:
> http://www.suse.de/~kraxel/uml/patches/2.6.18-rc4/uml-x11-fb
> which applied cleanly, but is not letting me set the option - Kconfig is
> beyond me:
> 
> arch/um/Kconfig:144:warning: 'select' used by config symbol 'X11_FB' refer to
> undefined symbol 'FB_CFB_FILLRECT'
> arch/um/Kconfig:145:warning: 'select' used by config symbol 'X11_FB' refer to
> undefined symbol 'FB_CFB_COPYAREA'
> arch/um/Kconfig:146:warning: 'select' used by config symbol 'X11_FB' refer to
> undefined symbol 'FB_CFB_IMAGEBLIT'

Try adding

    source "drivers/video/Kconfig"

to arch/um/Kconfig.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

^ permalink raw reply	[flat|nested] 20+ messages in thread

* Re: [uml-devel] [RFC] UML kernel & rootfs bundle with every kernel release ?
  2007-04-04  6:35                 ` Geert Uytterhoeven
@ 2007-04-04 20:27                   ` Antoine Martin
  0 siblings, 0 replies; 20+ messages in thread
From: Antoine Martin @ 2007-04-04 20:27 UTC (permalink / raw)
  To: Geert Uytterhoeven
  Cc: Jeff Dike, Blaisorblade, devzero,
	User-mode Linux Kernel Development, Linux Kernel Development,
	kraxel

Geert Uytterhoeven wrote:
> On Wed, 4 Apr 2007, Antoine Martin wrote:
>> and this one:
>> http://www.suse.de/~kraxel/uml/patches/2.6.18-rc4/uml-x11-fb
>> which applied cleanly, but is not letting me set the option - Kconfig is
>> beyond me:
>>
>> arch/um/Kconfig:144:warning: 'select' used by config symbol 'X11_FB' refer to
>> undefined symbol 'FB_CFB_FILLRECT'
>> arch/um/Kconfig:145:warning: 'select' used by config symbol 'X11_FB' refer to
>> undefined symbol 'FB_CFB_COPYAREA'
>> arch/um/Kconfig:146:warning: 'select' used by config symbol 'X11_FB' refer to
>> undefined symbol 'FB_CFB_IMAGEBLIT'
> 
> Try adding
> 
>     source "drivers/video/Kconfig"
> 
> to arch/um/Kconfig.
Ahh, thanks for that, it did the trick.

I also needed to add this to include/asm-um/pgtable.h:
#define io_remap_pfn_range(vma, vaddr, pfn, size, prot)         \
                 remap_pfn_range(vma, vaddr, pfn, size, prot)

And replace config.h with autoconf.h in arch/um/drivers/x11_kern.c

Then I hit:
arch/um/drivers/built-in.o:include/linux/input.h:1168: more undefined 
references to `input_event' follow
So I added: "select INPUT" to Kconfig
(starting with a defconfig might have done the trick too)

Then I hit:
arch/um/drivers/built-in.o: In function `x11_probe':
arch/um/drivers/x11_kern.c:442: undefined reference to `init_input_dev'
So I replaced:
         init_input_dev(kd->mouse);
with:
	kd->mouse = input_allocate_device();

That gave me a working kernel, but no x11 window appeared....
(I tried adding x11=640x480 to the command line too - no effect)

I then tried this patch:
http://user-mode-linux.sourceforge.net/work/current/2.6/2.6.21-rc1/patches/x11-fb
(fixing similar issues)

I had to comment out the line containing "soft_cursor" (couldn't figure 
out where to get it from)
It also required a lot of search and replace to change all instances of:
	kd->mouse.XXX
to
	kd->mouse->XXX
(same for kbd and changes to the initialization code)


Looks like the CONFIG_FB_CFB_FILLRECT and friends should not be set:
   LD      .tmp_vmlinux1
drivers/built-in.o: In function `cfb_copyarea':
drivers/video/cfbcopyarea.c:344: multiple definition of `cfb_copyarea'
arch/um/drivers/built-in.o:arch/um/drivers/../../../drivers/video/cfbcopyarea.c:344: 
first defined here
drivers/built-in.o: In function `cfb_imageblit':
drivers/video/cfbimgblt.c:263: multiple definition of `cfb_imageblit'
arch/um/drivers/built-in.o:arch/um/drivers/../../../drivers/video/cfbimgblt.c:263: 
first defined here
drivers/built-in.o: In function `cfb_fillrect':
drivers/video/cfbfillrect.c:345: multiple definition of `cfb_fillrect'
arch/um/drivers/built-in.o:arch/um/drivers/../../../drivers/video/cfbfillrect.c:345: 


Finally, this one pops up an X11 window, but it dies instantly:
[...]
Checking syscall emulation patch for ptrace...missing

Program received signal SIGSEGV, Segmentation fault.
kref_get (kref=0x5a5a5a5a5a5a5aa6) at lib/kref.c:32
32              WARN_ON(!atomic_read(&kref->refcount));
(gdb) bt
#0  kref_get (kref=0x5a5a5a5a5a5a5aa6) at lib/kref.c:32
#1  0x00000000600f4aa2 in kobject_get (kobj=0x5a5a5a5a5a5a5a8a) at 
lib/kobject.c:394
#2  0x0000000060109a95 in class_get (cls=<value optimized out>) at 
include/linux/kobject.h:145
#3  0x000000006010a3e5 in class_device_add (class_dev=0x602c9b38) at 
drivers/base/class.c:594
#4  0x000000006010ff5d in input_register_device (dev=0x602c9488) at 
drivers/input/input.c:987
#5  0x000000006001ed60 in x11_probe () at arch/um/drivers/x11_kern.c:482
#6  0x000000006000444d in x11_init () at arch/um/drivers/x11_kern.c:504
#7  0x00000000600015a6 in do_initcalls () at init/main.c:651
#8  0x0000000060001690 in do_basic_setup () at init/main.c:695
#9  0x000000006000f9ce in init (unused=<value optimized out>) at 
init/main.c:744
#10 0x00000000600241ca in run_kernel_thread (fn=0x6000f985 <init>, 
arg=0x0, jmp_ptr=<value optimized out>)
     at arch/um/os-Linux/process.c:289
#11 0x00000000600149b8 in new_thread_handler () at 
arch/um/kernel/skas/process.c:64
#12 0xa55a5a5a5a5a5a5a in ?? ()
#13 0x00000000602bd100 in ?? ()
#14 0x0000000060205f40 in ident_map ()
#15 0x0000000000000000 in ?? ()

And I am now completely out of my depth!

Antoine

> 
> Gr{oetje,eeting}s,
> 
> 						Geert
> 
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
> 
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
> 							    -- Linus Torvalds
> 


^ permalink raw reply	[flat|nested] 20+ messages in thread

end of thread, other threads:[~2007-04-04 20:27 UTC | newest]

Thread overview: 20+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-03-31 22:58 [RFC] UML kernel & rootfs bundle with every kernel release ? devzero
2007-04-01  0:01 ` Parag Warudkar
2007-04-01 19:08 ` Blaisorblade
2007-04-01 19:58   ` [uml-devel] " Antoine Martin
2007-04-02  2:05     ` Jeff Dike
2007-04-02 10:21       ` Antoine Martin
2007-04-02 11:22         ` Geert Uytterhoeven
2007-04-02 12:50           ` Geert Uytterhoeven
2007-04-02 14:53           ` Jeff Dike
2007-04-02 15:43             ` Antoine Martin
2007-04-03 23:49               ` Antoine Martin
2007-04-04  6:35                 ` Geert Uytterhoeven
2007-04-04 20:27                   ` Antoine Martin
2007-04-02 14:55         ` Jeff Dike
2007-04-02 15:43           ` Antoine Martin
2007-04-02 20:40         ` roland
2007-04-02 21:44           ` Jeff Dike
2007-04-03  4:44             ` Jason Lunz
2007-04-03 17:43         ` Blaisorblade
2007-04-03 17:53           ` Antoine Martin

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).