[Rpm-devel] rpm-4.4.6 on solaris9-sparc64
n3npq.jbj at gmail.com
Wed Jun 28 15:04:17 EDT 2006
On Jun 28, 2006, at 2:52 PM, Tim Mooney wrote:
> In regard to: Re: [Rpm-devel] rpm-4.4.6 on solaris9-sparc64, Jeff
>> We now return to trying to figger why Tim Mooney is seeing
>> hang on a platform that I have no reasonable access to.
> Just so we're all on the same page: I only see the hang after
> applying the
> reaper patch. Without the patch (stock rpm 4.4.6), there's no
> hang, but
> there's the original problem of the scriptlet being removed before the
> interpreter gets to it.
Scriptlet removed means parent thinks child executed correctly, almost
certainly a closely related problem.
Hmm, perhaps you are on a SMP machine and jmacs is not? That might
explain both results.
> Since we already have negative confirmation from Jonathan and I'm
> the only
> person that's seen this, I would hate for you to spend much more
> time on
> this until someone else reproduces it. Until and unless that happens,
> let's assume it's operator error. I'll look at it some more once my
> Solaris 8 build box is back.
Not a waste, the signal handling in rpm needs as many eyes as possible.
Even if I have to trick people into looking at the code ;-)
Already there's been the beneficial result that reaper = 0 has been
broken for quite some time afaict from cvs.
73 de Jeff
More information about the Rpm-devel