Ubuntu

{pci,usb}.ids need to go to root partition

Reported by Martin Pitt on 2009-05-05
12
This bug affects 1 person
Affects Status Importance Assigned to Milestone
pciutils (Ubuntu)
Medium
Unassigned
udev (Ubuntu)
Medium
Unassigned
usbutils (Ubuntu)
Medium
Unassigned

Bug Description

When /usr is on a separate partition, /etc/init/udevtrigger.conf potentially runs before /usr is mounted, and thus usb.ids and pci.ids are not available yet to fill the udev database. For this to work, these files need to be moved to e. g. /lib/hwdata/.

Martin Pitt (pitti) wrote :

As discussed further in the channel, we could alternatively replay all events once we have /usr available, to populate the udev db with the names later.

Martin Pitt (pitti) wrote :

This doesn't actually seem to be a problem in practice, closing.

Changed in udev-extras (Ubuntu):
status: New → Invalid
Martin Pitt (pitti) wrote :

Current discussion on #ubuntu-devel turned out that this is a problem after all.

affects: udev-extras (Ubuntu) → udev (Ubuntu)
Changed in udev (Ubuntu):
status: Invalid → Triaged

The upstream recommendation of /lib/hwdata seems sane

Scott James Remnant [2009-09-08 14:27 -0000]:
> The upstream recommendation of /lib/hwdata seems sane

That wouldn't help the actual problem, though, unless we'd copy it to
the initramfs?

On Tue, 2009-09-08 at 15:01 +0000, Martin Pitt wrote:

> Scott James Remnant [2009-09-08 14:27 -0000]:
> > The upstream recommendation of /lib/hwdata seems sane
>
> That wouldn't help the actual problem, though, unless we'd copy it to
> the initramfs?
>
Why wouldn't it?

This surely isn't being used in the initramfs (why would we copy this
rule there?)

Scott
--
Scott James Remnant
<email address hidden>

Martin Pitt (pitti) wrote :

Scott James Remnant [2009-09-08 18:02 -0000]:
> This surely isn't being used in the initramfs (why would we copy this
> rule there?)

I thought the problem was that we don't do an udevadm trigger after we
leave the initramfs, and thus the devices which were already detected
earlier wouldn't get their rules re-run?

If I misunderstood the problem, and it's solely about the time in
between leaving the initramfs and mounting /usr, it's fine, of course.

On Tue, 2009-09-08 at 19:01 +0000, Martin Pitt wrote:

> Scott James Remnant [2009-09-08 18:02 -0000]:
> > This surely isn't being used in the initramfs (why would we copy this
> > rule there?)
>
> I thought the problem was that we don't do an udevadm trigger after we
> leave the initramfs, and thus the devices which were already detected
> earlier wouldn't get their rules re-run?
>
> If I misunderstood the problem, and it's solely about the time in
> between leaving the initramfs and mounting /usr, it's fine, of course.
>
No, we do a trigger in the initramfs, and again before /usr is mounted -
we don't do a third trigger *after* /usr is mounted.

So these just need to be in the root filesystem for the "primary"
trigger

Scott
--
Scott James Remnant
<email address hidden>

Changed in udev (Ubuntu):
importance: Undecided → High
Martin Pitt (pitti) on 2010-08-27
summary: - udev-db needs {pci,usb}.ids early
+ {pci,usb}.ids need to go to root partition
Changed in pciutils (Ubuntu):
status: New → Triaged
importance: Undecided → Medium
Changed in usbutils (Ubuntu):
importance: Undecided → Medium
status: New → Triaged
Changed in udev (Ubuntu):
importance: High → Medium
description: updated
description: updated
Tim Gardner (timg-tpi) wrote :

Martin - is this still an issue ? And are we going to fix this in Ubuntu ?

Martin Pitt (pitti) wrote :

Indeed this has been obsoleted with the move to hwdb in saucy.

Changed in pciutils (Ubuntu):
status: Triaged → Invalid
Changed in udev (Ubuntu):
status: Triaged → Invalid
Changed in usbutils (Ubuntu):
status: Triaged → Invalid
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers