pagg
[Top] [All Lists]

Re: [PATCH 0/4] add task handling notifier

To: Jan Beulich <jbeulich@xxxxxxxxxx>
Subject: Re: [PATCH 0/4] add task handling notifier
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Wed, 9 Jan 2008 10:03:44 +0000
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>, hch@xxxxxxxxxxxxx, pagg@xxxxxxxxxxx, erikj@xxxxxxx, pj@xxxxxxx, matthltc@xxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx
In-reply-to: <4784A751.76E4.0078.0@novell.com>
References: <476A780C.76E4.0078.0@novell.com> <20071223122621.GA19310@infradead.org> <20071225140526.547a882f.akpm@linux-foundation.org> <47838ACB.76E4.0078.0@novell.com> <20080108141424.de5d8fba.akpm@linux-foundation.org> <4784A751.76E4.0078.0@novell.com>
Sender: pagg-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.17 (2007-11-01)
On Wed, Jan 09, 2008 at 09:52:01AM +0000, Jan Beulich wrote:
> Yes. And the unidentified feature is NLKD. But as with other notifiers (most
> notably the module unload one), all reasonable kernel debuggers should
> need them (or do explicit patching of the mentioned source files). As I
> explained before, I think that if kernel debuggers aren't allowed into the
> tree, they should at least be allowed to co-exist (since the argument of
> requiring in-tree users and submitting code for mainline inclusion is void
> if political/personal reasons exclude certain code from even being
> considered for inclusion).

We already have a kernel debugger in tree, xmon.  There's anotherone
hopefully getting in soon (kgdb), so they can do the right thing as the
other subsystems.

And as usual, we're not going to provide hooks for out of tree mess.


<Prev in Thread] Current Thread [Next in Thread>