Installing drivers on USB sticks

[Brian Murray-5]

I've recently discovered hundreds of bug reports regarding people trying
to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
Ubuntu and it failing[1].  Part of the large volume is due to the fact
that the bug reports are automatically reported by apport as package
installation failures.  When looking at these bug reports I had a couple
of questions.

Is installing drivers on a Live USB version of Ubuntu something that is
supported?

It seems to me that there are a lot of people who think it should be and
I believe it would a useful way for people to test the development
release of Ubuntu and the latest drivers.  If it isn't though we should
stop the automatic reporting of this classification of bug reports[2].

There are also some bug reports[3] that have empty
DpkgTerminalLog.txt[4] attachments.

How did this happen and can it be fixed?

Can these be marked as duplicates of bug 557023? (Provided they are on
live media and about installing drivers of course.)


[1] http://launchpad.net/bugs/557023
[2] I've written bug patterns for a couple of packages for this
[3] http://launchpad.net/bugs/696656
[4] http://launchpadlibrarian.net/61549804/DpkgTerminalLog.txt

Thanks,
--
Brian Murray
Ubuntu Bug Master

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[komputes]

On 01/04/2011 05:08 PM, Brian Murray wrote:

> I've recently discovered hundreds of bug reports regarding people trying
> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> Ubuntu and it failing[1].  Part of the large volume is due to the fact
> that the bug reports are automatically reported by apport as package
> installation failures.  When looking at these bug reports I had a couple
> of questions.
>
> Is installing drivers on a Live USB version of Ubuntu something that is
> supported?
>
> It seems to me that there are a lot of people who think it should be and
> I believe it would a useful way for people to test the development
> release of Ubuntu and the latest drivers.  If it isn't though we should
> stop the automatic reporting of this classification of bug reports[2].
>
> There are also some bug reports[3] that have empty
> DpkgTerminalLog.txt[4] attachments.
>
> How did this happen and can it be fixed?
>
> Can these be marked as duplicates of bug 557023? (Provided they are on
> live media and about installing drivers of course.)
>
>
> [1] http://launchpad.net/bugs/557023
> [2] I've written bug patterns for a couple of packages for this
> [3] http://launchpad.net/bugs/696656
> [4] http://launchpadlibrarian.net/61549804/DpkgTerminalLog.txt
>
> Thanks,

Hi Brian,

I often use full installations on USB keys.

My original reason for doing this was because I was going to stores and
looking at different hardware under Ubuntu to see what works. I was
visiting friends and testing hardware compatability on their laptops. I
was submitting checkbox results for machines so that they may feed the
Hardware Database.

Should this use case be supported? I would think so. Since the modules
needed load when the appropriate hardware is detected I think it should
be supported. I did a little research on additional drivers that don't
come with/in the Linux kernel and posted them.
http://ubuntuforums.org/showthread.php?t=1329514


Note that I have had problems in the past with these keys. Once such
problem occured after using a USB stick from a computer using NVIDIA
video to a computer using INTEL video. Once gnome-session starts the
background and gnome-panel items and letters were reversed upside-down.
I would have thought that's X.org would autodetect hardware before
setting up the video server.
http://pad.lv/544813

There are two other issues that come to mind with this setup:
(Regression) This was fixed and re-appeared. When selecting to install
to a USB key in Ubiquity, the installer installs MBR to sda (usb key
being sdb).
http://pad.lv/549756

When ubiquity makes decisions on what it will install based on the
hardware you are running the installer on (e.g. 4GB RAM + Net =
PAE-enabled kernel)
No bug reported for this. Detecting hardware in the installer seems like
a new feature.

I'm able to assist with testing and feedback should you require
assistance with this important effort to support this use case.

Cheers,

-komputes

  (]( -. .- )[)

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[Martin Pitt-4]

Hello Brian,

Brian Murray [2011-01-04 14:08 -0800]:
> I've recently discovered hundreds of bug reports regarding people trying
> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> Ubuntu and it failing[1].

There are two main reasons for this:

 - update-initramfs not working (http://launchpad.net/bugs/557023)

 - There is too little RAM on a live system to download, build,
   install, and run large drivers like nvidia and fglrx. Therefore
   these should be disabled on a live system. This is tracked in
   http://launchpad.net/bugs/685017, and I'll fix it for alpha-2.

It would still be nice to fix 557023 so that this at least works with
smaller drivers like broadcom-wl (this doesn't actually need
an initramfs update, I believe, but it might happen anyway).

> [3] http://launchpad.net/bugs/696656

Duped.

Thanks,

Martin
--
Martin Pitt                        | http://www.piware.de
Ubuntu Developer (www.ubuntu.com)  | Debian Developer  (www.debian.org)

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[Brian Murray-5]

On Tue, Jan 04, 2011 at 11:53:54PM +0100, Martin Pitt wrote:

> Hello Brian,
>
> Brian Murray [2011-01-04 14:08 -0800]:
> > I've recently discovered hundreds of bug reports regarding people trying
> > to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> > Ubuntu and it failing[1].
>
> There are two main reasons for this:
>
>  - update-initramfs not working (http://launchpad.net/bugs/557023)
>
>  - There is too little RAM on a live system to download, build,
>    install, and run large drivers like nvidia and fglrx. Therefore
>    these should be disabled on a live system. This is tracked in
>    http://launchpad.net/bugs/685017, and I'll fix it for alpha-2.
>
> It would still be nice to fix 557023 so that this at least works with
> smaller drivers like broadcom-wl (this doesn't actually need
> an initramfs update, I believe, but it might happen anyway).
>
> > [3] http://launchpad.net/bugs/696656
>
> Duped.

This was an example of a whole class of bugs that have been reported in
Launchpad.  Is it also safe to make those duplicates of bug 685017?

Thanks again,
--
Brian Murray
Ubuntu Bug Master

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[Bryce Harrington-5]

On Tue, Jan 04, 2011 at 02:08:28PM -0800, Brian Murray wrote:
> I've recently discovered hundreds of bug reports regarding people trying
> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> Ubuntu and it failing[1].  Part of the large volume is due to the fact
> that the bug reports are automatically reported by apport as package
> installation failures.  When looking at these bug reports I had a couple
> of questions.

Yeah this is something we've been studying from the X.org side as well.
This is by far the most common X bug reported against natty so far.
 
> Is installing drivers on a Live USB version of Ubuntu something that is
> supported?

I experimented a bunch with it and fglrx before the holidays, and simply
could not get it to work.  It needs to update the initrd for it to work,
but that file lives outside the persistent file system.

So, the way things are designed with USB LiveCD environments, no it is
not supported currently.

People have reported success at installing drivers when they don't
require initrd updates (e.g. old versions of -nvidia).

> It seems to me that there are a lot of people who think it should be and
> I believe it would a useful way for people to test the development
> release of Ubuntu and the latest drivers.  If it isn't though we should
> stop the automatic reporting of this classification of bug reports[2].

Yes, it is clear people want to test Ubuntu this way.  It would be great
if we could enable it.  Although, given the differences in how the USB
live environment is laid out compared with a normal installation, it
might require a good bit of debugging itself(!)

For instance, it's possible with usb-creator to allocate all space to
the persistent file, which causes the new initrd to be written to begin
with.  And other problems...

Anyway, I agree we ought to either stop the automatic reporting, or make
the reports auto-dupe.

Even better would be to get Jockey to not offer to install these drivers
in the first place.  (E.g. bug #685017)  When we feel it is a feature we
definitely do feel comfortable supporting, then we could re-enable it.

> There are also some bug reports[3] that have empty
> DpkgTerminalLog.txt[4] attachments.
>
> How did this happen and can it be fixed?
>
> Can these be marked as duplicates of bug 557023? (Provided they are on
> live media and about installing drivers of course.)

Yes, especially for -nvidia and -fglrx feel free to dupe them up; as it
is the reports are just causing noise for us.

Bryce

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[Scott Kitterman-3]

On Tuesday, January 04, 2011 05:08:28 pm Brian Murray wrote:

> I've recently discovered hundreds of bug reports regarding people trying
> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> Ubuntu and it failing[1].  Part of the large volume is due to the fact
> that the bug reports are automatically reported by apport as package
> installation failures.  When looking at these bug reports I had a couple
> of questions.
>
> Is installing drivers on a Live USB version of Ubuntu something that is
> supported?
>
> It seems to me that there are a lot of people who think it should be and
> I believe it would a useful way for people to test the development
> release of Ubuntu and the latest drivers.  If it isn't though we should
> stop the automatic reporting of this classification of bug reports[2].
>
> There are also some bug reports[3] that have empty
> DpkgTerminalLog.txt[4] attachments.
>
> How did this happen and can it be fixed?
>
> Can these be marked as duplicates of bug 557023? (Provided they are on
> live media and about installing drivers of course.)
>
>
> [1] http://launchpad.net/bugs/557023
> [2] I've written bug patterns for a couple of packages for this
> [3] http://launchpad.net/bugs/696656
> [4] http://launchpadlibrarian.net/61549804/DpkgTerminalLog.txt
>
> Thanks,

I've done this successfully with bcmwl on my Dell Mini 10v a few times with
Lucid and probably close to a dozen times with Maverick (including while
sitting in a UDS plenary when I messed up my system and was trying to get back
to something usable).  I understand that video support might be hard, but I
think that in many cases you need to network to fix things so supporting
wireless drivers is an important use case.

Scott K

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[komputes]

On 01/04/2011 05:44 PM, komputes wrote:

> On 01/04/2011 05:08 PM, Brian Murray wrote:
>> I've recently discovered hundreds of bug reports regarding people trying
>> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
>> Ubuntu and it failing[1].  Part of the large volume is due to the fact
>> that the bug reports are automatically reported by apport as package
>> installation failures.  When looking at these bug reports I had a couple
>> of questions.
>>
>> Is installing drivers on a Live USB version of Ubuntu something that is
>> supported?
>>
>> It seems to me that there are a lot of people who think it should be and
>> I believe it would a useful way for people to test the development
>> release of Ubuntu and the latest drivers.  If it isn't though we should
>> stop the automatic reporting of this classification of bug reports[2].
>>
>> There are also some bug reports[3] that have empty
>> DpkgTerminalLog.txt[4] attachments.
>>
>> How did this happen and can it be fixed?
>>
>> Can these be marked as duplicates of bug 557023? (Provided they are on
>> live media and about installing drivers of course.)
>>
>>
>> [1] http://launchpad.net/bugs/557023
>> [2] I've written bug patterns for a couple of packages for this
>> [3] http://launchpad.net/bugs/696656
>> [4] http://launchpadlibrarian.net/61549804/DpkgTerminalLog.txt
>>
>> Thanks,
> Hi Brian,
>
> I often use full installations on USB keys.
>
> My original reason for doing this was because I was going to stores and
> looking at different hardware under Ubuntu to see what works. I was
> visiting friends and testing hardware compatability on their laptops. I
> was submitting checkbox results for machines so that they may feed the
> Hardware Database.
>
> Should this use case be supported? I would think so. Since the modules
> needed load when the appropriate hardware is detected I think it should
> be supported. I did a little research on additional drivers that don't
> come with/in the Linux kernel and posted them.
> http://ubuntuforums.org/showthread.php?t=1329514
>
>
> Note that I have had problems in the past with these keys. Once such
> problem occured after using a USB stick from a computer using NVIDIA
> video to a computer using INTEL video. Once gnome-session starts the
> background and gnome-panel items and letters were reversed upside-down.
> I would have thought that's X.org would autodetect hardware before
> setting up the video server.
> http://pad.lv/544813
>
> There are two other issues that come to mind with this setup:
> (Regression) This was fixed and re-appeared. When selecting to install
> to a USB key in Ubiquity, the installer installs MBR to sda (usb key
> being sdb).
> http://pad.lv/549756

Sorry, made a mistake there, that one was fixed in Lucid.
This is the correct one for the reappearance of the issue in maverick.
http://pad.lv/684292

> When ubiquity makes decisions on what it will install based on the
> hardware you are running the installer on (e.g. 4GB RAM + Net =
> PAE-enabled kernel)
> No bug reported for this. Detecting hardware in the installer seems like
> a new feature.
>
> I'm able to assist with testing and feedback should you require
> assistance with this important effort to support this use case.
>
> Cheers,
>
> -komputes
>
>   (]( -. .- )[)

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[komputes]

On 01/04/2011 07:04 PM, Brian Murray wrote:

> On Tue, Jan 04, 2011 at 11:53:54PM +0100, Martin Pitt wrote:
>> Hello Brian,
>>
>> Brian Murray [2011-01-04 14:08 -0800]:
>>> I've recently discovered hundreds of bug reports regarding people trying
>>> to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
>>> Ubuntu and it failing[1].
>> There are two main reasons for this:
>>
>>  - update-initramfs not working (http://launchpad.net/bugs/557023)
>>
>>  - There is too little RAM on a live system to download, build,
>>    install, and run large drivers like nvidia and fglrx. Therefore
>>    these should be disabled on a live system. This is tracked in
>>    http://launchpad.net/bugs/685017, and I'll fix it for alpha-2.
>>
>> It would still be nice to fix 557023 so that this at least works with
>> smaller drivers like broadcom-wl (this doesn't actually need
>> an initramfs update, I believe, but it might happen anyway).
>>
>>> [3] http://launchpad.net/bugs/696656
>> Duped.
> This was an example of a whole class of bugs that have been reported in
> Launchpad.  Is it also safe to make those duplicates of bug 685017?
>
> Thanks again,

These issues and limitations of casper stop users from having a
persistent, updated clone of an ubuntu CD image using usb-creator.

I have seen many complains that updating using a key (built with
usb-creator) causes the key to be non-usable afterwards. I use
usb-creator for the purpose of installing to computers via USB, not
running a full installation with the ability to update the kernel.
usb-creator works, it does what it's supposed to and fits the purpose of
making a USB installer. However, since the image contains update-manager
which will inevitably check for updates, we are setting up the user to
mess up their system because of live environment limitations. How can we
correct this or at least minimize the chances of someone updating? How
can we make it clear that there is a difference between a USB CD clone
and a USB installation?

Personally, I do not face any of these issues since I use ubiquity to do
a full installations to a USB drives. I believe that this use case
should also be tested and supported.

Cheers,

-komputes

  (]( -. .- )[)

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

[Brian Murray-5]

On Tue, Jan 04, 2011 at 04:04:40PM -0800, Brian Murray wrote:

> On Tue, Jan 04, 2011 at 11:53:54PM +0100, Martin Pitt wrote:
> > Hello Brian,
> >
> > Brian Murray [2011-01-04 14:08 -0800]:
> > > I've recently discovered hundreds of bug reports regarding people trying
> > > to install drivers (bcmwl, fglrx, nvidia) on their Live USB version of
> > > Ubuntu and it failing[1].
> >
> > There are two main reasons for this:
> >
> >  - update-initramfs not working (http://launchpad.net/bugs/557023)
> >
> >  - There is too little RAM on a live system to download, build,
> >    install, and run large drivers like nvidia and fglrx. Therefore
> >    these should be disabled on a live system. This is tracked in
> >    http://launchpad.net/bugs/685017, and I'll fix it for alpha-2.
> >
> > It would still be nice to fix 557023 so that this at least works with
> > smaller drivers like broadcom-wl (this doesn't actually need
> > an initramfs update, I believe, but it might happen anyway).
> >
> > > [3] http://launchpad.net/bugs/696656
> >
> > Duped.
>
> This was an example of a whole class of bugs that have been reported in
> Launchpad.  Is it also safe to make those duplicates of bug 685017?

Actually thinking about this more.  It is easy to find these
apport-package bugs, reported from LiveMedia and with "empty"
DpkgTerminalLog files, in Launchpad after they are reported.  However,
one can't really write a bug pattern for them as we are not looking for
a specific string but rather the absence of information.  How could we
go about preventing these reports on Maverick from continuing to come
in?  Would an SRU of the fix for bug 557023 do it?

--
Brian Murray
Ubuntu Bug Master

--
ubuntu-devel mailing list
[hidden email]
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel