[gentoo-user] [HEADSUP] New udev-186 breaks pulseaudio

[gentoo-user] [HEADSUP] New udev-186 breaks pulseaudio

[walt]

udev-186 replaces libudev.so.0 with libudev.so.1, and pulseaudio
won't compile against it.  If you need pulse, avoid udev-186.

I backed down to udev-182-r3, which fixed the problem, but I had
to run revdep-rebuild (again) to fix all the other packages that
did build against libudev.so.1 and now had to be rebuilt a second
time against libudev.so.0.

Also, even udev-182 did some breakage to the udev scripts installed
by hplip, but that's easy to fix.  Apparently the recent udev has
replaced the SYSFS keyword with ATTR.

This simple fix I found with google seems to work for me:

#cd /lib64/udev/rules.d
#sed -i s/SYSFS/ATTR/g *

Doesn't it seem that breakage this big should be obvious to the
devs before the changes go public, even on ~arch?  Hrmph!

Re: [gentoo-user] [HEADSUP] New udev-186 breaks pulseaudio

[Paul Hartman]

On Thu, Jul 5, 2012 at 7:01 AM, walt <w41ter@gmail.com> wrote:
> udev-186 replaces libudev.so.0 with libudev.so.1, and pulseaudio
> won't compile against it.  If you need pulse, avoid udev-186.
>
> I backed down to udev-182-r3, which fixed the problem, but I had
> to run revdep-rebuild (again) to fix all the other packages that
> did build against libudev.so.1 and now had to be rebuilt a second
> time against libudev.so.0.

I ran into the same think. lvm2 also doesn't build against it, which I
think could potentially result in an unbootable system in the right
situation.

> Also, even udev-182 did some breakage to the udev scripts installed
> by hplip, but that's easy to fix.  Apparently the recent udev has
> replaced the SYSFS keyword with ATTR.

I believe that was a configuration problem in the latest hplip ebuild;
the previous version didn't do that. I think udev has been warning
about SYSFS deprecation for 3 years.

> This simple fix I found with google seems to work for me:
>
> #cd /lib64/udev/rules.d
> #sed -i s/SYSFS/ATTR/g *

That also works.

> Doesn't it seem that breakage this big should be obvious to the
> devs before the changes go public, even on ~arch?  Hrmph!

I guess it is called "unstable" for good reason sometimes. :) The bugs
we encounter, report and sometimes actually solve, will help the
stable users down the road so they don't have the same problems.

Re: [gentoo-user] [HEADSUP] New udev-186 breaks pulseaudio

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 700 bytes --]

On Thu, 5 Jul 2012 09:40:15 -0500, Paul Hartman wrote:

> > I backed down to udev-182-r3, which fixed the problem, but I had
> > to run revdep-rebuild (again) to fix all the other packages that
> > did build against libudev.so.1 and now had to be rebuilt a second
> > time against libudev.so.0.  
> 
> I ran into the same think. lvm2 also doesn't build against it, which I
> think could potentially result in an unbootable system in the right
> situation.

Sync again, this has been fixed, at least for lvm2 and
system-config-printer-common, I don't use pulseaudio. 

https://bugs.gentoo.org/show_bug.cgi?id=424810


-- 
Neil Bothwick

Angular Momentum Makes The World Go 'Round

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

Re: [gentoo-user] [HEADSUP] New udev-186 breaks pulseaudio

[Todd Goodman]

* Neil Bothwick <neil@digimed.co.uk> [120705 11:53]:
> On Thu, 5 Jul 2012 09:40:15 -0500, Paul Hartman wrote:
> 
> > > I backed down to udev-182-r3, which fixed the problem, but I had
> > > to run revdep-rebuild (again) to fix all the other packages that
> > > did build against libudev.so.1 and now had to be rebuilt a second
> > > time against libudev.so.0.  
> > 
> > I ran into the same think. lvm2 also doesn't build against it, which I
> > think could potentially result in an unbootable system in the right
> > situation.
> 
> Sync again, this has been fixed, at least for lvm2 and
> system-config-printer-common, I don't use pulseaudio. 
> 
> https://bugs.gentoo.org/show_bug.cgi?id=424810

There's a new pulseaudio release coming along too:

https://bugs.gentoo.org/show_bug.cgi?id=423411

Todd

[gentoo-user] Re: [HEADSUP] New udev-186 breaks pulseaudio[FIXED]

[walt]

On 07/05/2012 08:48 AM, Neil Bothwick wrote:
> On Thu, 5 Jul 2012 09:40:15 -0500, Paul Hartman wrote:
> 
>>> I backed down to udev-182-r3, which fixed the problem, but I had
>>> to run revdep-rebuild (again) to fix all the other packages that
>>> did build against libudev.so.1 and now had to be rebuilt a second
>>> time against libudev.so.0.  
>>
>> I ran into the same think. lvm2 also doesn't build against it, which I
>> think could potentially result in an unbootable system in the right
>> situation.
> 
> Sync again, this has been fixed, at least for lvm2 and
> system-config-printer-common, I don't use pulseaudio. 

All better now, including pulse.  I suppose if I'd updated later
in the same day I wouldn't have hit this bug.  I clearly need to
sleep an extra 8 hours every day so this never happens again :)

Re: [gentoo-user] Re: [HEADSUP] New udev-186 breaks pulseaudio[FIXED]

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 656 bytes --]

On Fri, 06 Jul 2012 11:44:56 -0700, walt wrote:

> > Sync again, this has been fixed, at least for lvm2 and
> > system-config-printer-common, I don't use pulseaudio.   
> 
> All better now, including pulse.  I suppose if I'd updated later
> in the same day I wouldn't have hit this bug.  I clearly need to
> sleep an extra 8 hours every day so this never happens again :)

If everyone took your advice, the bug would simply have waited another 8
hours to be discovered. If you run testing, you are effectively saying to
the devs "Go on ,have a go, see if you can break my system" :)


-- 
Neil Bothwick

If you cannot fix it, feature it.

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