Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 14 Sep 2004 17:36:55 +0100
From:      Christoph Hellwig <hch@infradead.org>
To:        Jeremie Le Hen <jeremie.le-hen@epita.fr>
Cc:        Nikita Danilov <nikita@clusterfs.com>
Subject:   Re: FreeBSD on Xserve?
Message-ID:  <20040914173655.A9479@infradead.org>
In-Reply-To: <20040913171329.GA16110@garak.epita.fr>; from jeremie.le-hen@epita.fr on Mon, Sep 13, 2004 at 07:13:29PM %2B0200
References:  <20040912183437.GF20097@mongers.org> <E1C6aQi-000KyU-00.shmukler-mail-ru@f26.mail.ru> <16708.47393.249768.90818@thebsh.namesys.com> <20040913171329.GA16110@garak.epita.fr>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Sep 13, 2004 at 07:13:29PM +0200, Jeremie Le Hen wrote:
> I would add that the UML patch applied to the hosted kernel source deeply
> modifies the ptrace(2) infrastructure.  All UML processes are in fact
> processes on the host kernel, but the UML kernel ptrace's so that it reroutes
> all invoked syscalls to itself and mmap's processes address space to its
> own one.

That's not ture.  For the ptrace method you describe unmodified host
kernels work.  There's an alternate implementation using multiple address
spaces per process that needs a heavily patched kernel.  (the SKAS patch
you mention below)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040914173655.A9479>