Bug#430707: libio-multiplex-perl: FTBFS: tests (sometimes!) fail during build

[Gunnar Wolf]

gregor herrmann dijo [Fri, Jun 29, 2007 at 07:03:35PM +0200]:
> Ok, some more tests:
> 2.6.22-rc4-686 (without any parameters): works (both on the console
> and under X)
> 2.6.22-rc4-686 (with clocksource=pit): works on the console but
> not under X
> 2.6.22-rc5-686 (without parameters): works on the console but
> not under X
> 2.6.22-rc5-686 (with clocksource=pit): works (both on the console  
> and under X)
> 2.6.22-rc6-686 (without parameters): works (both on the console  
> and under X)
> 2.6.22-rc6-686 (with clocksource=pit): works (both on the console  
> and under X)

Ugh, funny :-/

> (Without setting clocksource the laptop freezes each other minute and
> doesn't reboot at least with some kernels but that's a different
> problem.)
> 
> So the problem is -rc4 and -rc5 with/out the kernel parameter
> "clocksource=pit" in combination with my laptop and X. Or something
> else :-)
> 
> I guess we can close this bug as it's probably not
> libio-multiplex-perl's fault.

And the bug does not impact any official kernel versions. However,
please pull the kernel team's attention towards this bug, as we don't
want it to break later on. It still baffles me how something such as
standard as IO::Multiplex (it's just wrapping IO::Select, right?)
fails because of a (hopefully minor) kernel behavior change!

-- 
Gunnar Wolf - gwolf at gwolf.org - (+52-55)5623-0154 / 1451-2244
PGP key 1024D/8BB527AF 2001-10-23
Fingerprint: 0C79 D2D1 2C4E 9CE4 5973  F800 D80E F35A 8BB5 27AF