[gentoo-user] udev-197 and /etc/udev/rules.d/

[gentoo-user] udev-197 and /etc/udev/rules.d/

[»Q«]

udev-197-r3 gave me this postinstall warning:

  Upstream has removed the persistent-cd rules generator. If you need
  persistent names for these devices, place udev rules for them
  in /etc/udev/rules.d.

Well, I have had such a rule for a long time, and it worked ok until I
installed udev-197, which seems to be ignoring it.  I'm at a loss to
see what's wrong.

Here's /etc/udev/rules.d/12-opticaldrive.rules, just one line:

KERNEL=="sr0", SUBSYSTEM=="block", NAME="opticaldrive", SYMLINK+="%k", SYMLINK+="cdrom", SYMLINK+="cdrw", SYMLINK+="dvd", SYMLINK+="dvdrom", SYMLINK+="dvdrw"

And here's the first part of udevadm info for the device:

 looking at device '/devices/pci0000:00/0000:00:1f.2/ata2/host1/target1:0:0/1:0:0:0/block/sr0':
   KERNEL=="sr0"
   SUBSYSTEM=="block"
   DRIVER==""
   ATTR{ro}=="0"
   ATTR{size}=="4"
   ATTR{stat}=="       0        0        0        0        0        0        0        0        0        0        0"
   ATTR{range}=="1"
   ATTR{discard_alignment}=="0"
   ATTR{events}=="media_change eject_request"
   ATTR{ext_range}=="1"
   ATTR{events_poll_msecs}=="-1"
   ATTR{alignment_offset}=="0"
   ATTR{inflight}=="       0        0"
   ATTR{removable}=="1"
   ATTR{capability}=="119"
   ATTR{events_async}==""

[gentoo-user] Re: udev-197 and /etc/udev/rules.d/

[Hartmut Figge]

»Q«:

>udev-197-r3 gave me this postinstall warning:
>
>  Upstream has removed the persistent-cd rules generator. If you need
>  persistent names for these devices, place udev rules for them
>  in /etc/udev/rules.d.
>
>Well, I have had such a rule for a long time, and it worked ok until I
>installed udev-197, which seems to be ignoring it.

I was bitten by that too.

>I'm at a loss to see what's wrong.

If that can be solved, i may reconsider going back to udev-171. ;)

Hartmut
-- 
Usenet-ABC-Wiki http://www.usenet-abc.de/wiki/
Von Usern fuer User  :-)

Re: [gentoo-user] udev-197 and /etc/udev/rules.d/

[Marc Joliet]

[-- Attachment #1: Type: text/plain, Size: 2402 bytes --]

Am Mon, 21 Jan 2013 21:56:00 -0600
schrieb »Q« <boxcars@gmx.net>:

> udev-197-r3 gave me this postinstall warning:
> 
>   Upstream has removed the persistent-cd rules generator. If you need
>   persistent names for these devices, place udev rules for them
>   in /etc/udev/rules.d.
> 
> Well, I have had such a rule for a long time, and it worked ok until I
> installed udev-197, which seems to be ignoring it.  I'm at a loss to
> see what's wrong.
> 
> Here's /etc/udev/rules.d/12-opticaldrive.rules, just one line:
> 
> KERNEL=="sr0", SUBSYSTEM=="block", NAME="opticaldrive", SYMLINK+="%k", SYMLINK+="cdrom", SYMLINK+="cdrw", SYMLINK+="dvd", SYMLINK+="dvdrom", SYMLINK+="dvdrw"
> 
> And here's the first part of udevadm info for the device:
> 
>  looking at device '/devices/pci0000:00/0000:00:1f.2/ata2/host1/target1:0:0/1:0:0:0/block/sr0':
>    KERNEL=="sr0"
>    SUBSYSTEM=="block"
>    DRIVER==""
>    ATTR{ro}=="0"
>    ATTR{size}=="4"
>    ATTR{stat}=="       0        0        0        0        0        0        0        0        0        0        0"
>    ATTR{range}=="1"
>    ATTR{discard_alignment}=="0"
>    ATTR{events}=="media_change eject_request"
>    ATTR{ext_range}=="1"
>    ATTR{events_poll_msecs}=="-1"
>    ATTR{alignment_offset}=="0"
>    ATTR{inflight}=="       0        0"
>    ATTR{removable}=="1"
>    ATTR{capability}=="119"
>    ATTR{events_async}==""

Well, I did the udev upgrade, and the following rules in
/etc/udev/rules.d/70-persistent-cd.rules still work:

  KERNEL=="sr0", SYMLINK+="cdrom", ENV{GENERATED}="1"
  KERNEL=="sr0", SYMLINK+="cdrw", ENV{GENERATED}="1"
  KERNEL=="sr0", SYMLINK+="dvd", ENV{GENERATED}="1"
  KERNEL=="sr0", SYMLINK+="dvdrw", ENV{GENERATED}="1"

I don't really think it's that the symlinks are spread out over four lines, but
you can try it anyway. The ENV{GENERATED} bit is left over from the
auto-generated rule; I have my doubts regarding its necessity.

[ Luckily for me, I had no problems upgrading udev, save for a change in
  the naming scheme of the files in /dev/input/by-id/, which screwed up my
  pulseaudio settings (I use the volume keys on my keyboard), which took all of 1
  minute to fix after I noticed pulseaudio wasn't starting. ]

HTH
-- 
Marc Joliet
--
"People who think they know everything really annoy those of us who know we
don't" - Bjarne Stroustrup

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] udev-197 and /etc/udev/rules.d/

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 591 bytes --]

On Mon, 21 Jan 2013 21:56:00 -0600, »Q« wrote:

> Here's /etc/udev/rules.d/12-opticaldrive.rules, just one line:
> 
> KERNEL=="sr0", SUBSYSTEM=="block", NAME="opticaldrive", SYMLINK+="%k",
> SYMLINK+="cdrom", SYMLINK+="cdrw", SYMLINK+="dvd", SYMLINK+="dvdrom",
> SYMLINK+="dvdrw"

ISTR a change in udev that prevented renaming devices. Put it all as
symlinks instead of renaming and trying to symlink back to %k. It seems
that all the replies with working examples do it this way too.


-- 
Neil Bothwick

Not one shred of evidence supports the notion that life is serious.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] udev-197 and /etc/udev/rules.d/

[Marc Joliet]

[-- Attachment #1: Type: text/plain, Size: 1934 bytes --]

Am Tue, 22 Jan 2013 14:23:57 +0000
schrieb Neil Bothwick <neil@digimed.co.uk>:

> On Mon, 21 Jan 2013 21:56:00 -0600, »Q« wrote:
> 
> > Here's /etc/udev/rules.d/12-opticaldrive.rules, just one line:
> > 
> > KERNEL=="sr0", SUBSYSTEM=="block", NAME="opticaldrive", SYMLINK+="%k",
> > SYMLINK+="cdrom", SYMLINK+="cdrw", SYMLINK+="dvd", SYMLINK+="dvdrom",
> > SYMLINK+="dvdrw"
> 
> ISTR a change in udev that prevented renaming devices. Put it all as
> symlinks instead of renaming and trying to symlink back to %k. It seems
> that all the replies with working examples do it this way too.

I was interested enough to look this up. I looked through the git log of
my /etc/udev/rules.d/ and found that in early October 2012 I committed a change
to that effect, so something did change at some point.

However, I can't find any reference to that in the udev changelog. In fact, it
actually looks like it's a kernel change and that udev is really just "obeying"
the kernel [0] [1].

But I did find out that this is in fact documented in the udev(7) man page:

   NAME
     The name to use for a network interface. The name of a device node cannot
     be changed by udev, only additional symlinks can be created.

So currently you can only change network interface names, and nothing else.

[0] I haven't searched extensively, but found a related Email from Greg K-H
(search for "rename") that points out that device node renaming has
problematic/fragile behaviour:
http://lkml.indiana.edu/hypermail/linux/kernel/1010.1/00427.html.

[1] https://www.kernel.org/doc/htmldocs/device-drivers/API-device-rename.html
(documentation generated from linux-3.0-rc7). Relevant quote: "Device nodes are
not renamed at all, there isn't even support for that in the kernel now."

HTH
-- 
Marc Joliet
--
"People who think they know everything really annoy those of us who know we
don't" - Bjarne Stroustrup

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

[gentoo-user] Re: udev-197 and /etc/udev/rules.d/

[»Q«]

On Tue, 22 Jan 2013 18:51:59 +0100
Marc Joliet <marcec@gmx.de> wrote:

> Am Tue, 22 Jan 2013 14:23:57 +0000
> schrieb Neil Bothwick <neil@digimed.co.uk>:
> 
> > ISTR a change in udev that prevented renaming devices. Put it all as
> > symlinks instead of renaming and trying to symlink back to %k. It
> > seems that all the replies with working examples do it this way too.

That works, thanks.

[snip]
 
> So currently you can only change network interface names, and nothing
> else.
> 
> [0] I haven't searched extensively, but found a related Email from
> Greg K-H (search for "rename") that points out that device node
> renaming has problematic/fragile behaviour:
> http://lkml.indiana.edu/hypermail/linux/kernel/1010.1/00427.html.
> 
> [1]
> https://www.kernel.org/doc/htmldocs/device-drivers/API-device-rename.html
> (documentation generated from linux-3.0-rc7). Relevant quote: "Device
> nodes are not renamed at all, there isn't even support for that in
> the kernel now."

Thanks for the discussion (snipped) and the links.  Something Greg K-H
said about kernel names reminded me that I've been meaning to stop
using them in fstab as well.

I've been a Gentoo user for twelve years, and I've learned a lot --
including some bad habits, heh.