Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 02 Feb 2012 20:10:17 +0200
From:      Andriy Gapon <avg@FreeBSD.org>
To:        =?windows-1251?Q?=CA=EE=ED=FC=EA=EE=E2_=C5=E2=E3=E5=ED=E8=E9?= <kes-kes@yandex.ru>
Cc:        freebsd-bugs@FreeBSD.org, "Eugene M. Zheganin" <emz@norma.perm.ru>, eugene@zhegan.in
Subject:   Re: kern/164705: inability to terminate process in D state
Message-ID:  <4F2AD189.605@FreeBSD.org>
In-Reply-To: <1849394483.20120202200211@yandex.ru>
References:  <201202020923.q129NDxR000887@freefall.freebsd.org> <1259591729.20120202113710@yandex.ru> <4F2A5EA7.9020106@FreeBSD.org> <4F2A6123.8030909@norma.perm.ru> <4F2A89C4.4090907@FreeBSD.org> <1849394483.20120202200211@yandex.ru>

next in thread | previous in thread | raw e-mail | index | archive | help
on 02/02/2012 20:02 Коньков Евгений said the following:
> Здравствуйте, Andriy.
> 
> Вы писали 2 февраля 2012 г., 15:04:04:
> 
> AG> on 02/02/2012 12:10 Eugene M. Zheganin said the following:
>>> On 02.02.2012 16:00, Andriy Gapon wrote:
>>>> on 02/02/2012 11:37 Коньков Евгений said the following:
>>>>> why close? keep the problem open until resolve.
>>>> Resolve exactly what?
>>> Yeah, someone's enormous ego is not exactly a FreeBSD problem.
>>> I can see it's grown up to an unresolvable size.
> 
> AG> Anything on the technical side you might want to add?
> 
> people with sufficiant knowledge looking through opened PRs maybe will
> notice this, take attention of it and fix.
> 
> I do not have any knowledge of kernel design I just report a PR
> and if this is really problem it must be opened until fix. IMHO.
> 
> this PR and PR bin/164526 may be related and this one
> may be closed as 'dublicate'  but not as:
>   'this problem can not be resolved'

PR 164526 reports a concrete problem that can be diagnosed and hopefully fixed.
This PR is equivalent to "when FreeBSD has bugs it sucks", only about a sub-class
of bugs with some specific symptoms (process being stuck and unkillable).  There
could be different actual underlying bugs and bug classes - driver bugs,
deadlocks, etc.  Hence, this PR can not be diagnosed and fixed.
A new (concrete/useful) PR can be opened at any time, it's a low cost operation.

> 2:
> thank you about your explanation of "bad and good news"
> 


-- 
Andriy Gapon



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4F2AD189.605>