Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 02 Dec 1998 16:58:22 -0800
From:      Mike Smith <mike@smith.net.au>
To:        Robert Watson <robert+freebsd@cyrus.watson.org>
Cc:        Mike Smith <mike@smith.net.au>, Andrzej Bialecki <abial@nask.pl>, Pascal Hofstee <daeron@Wit401305.student.utwente.nl>, Shawn Ramsey <shawn@cpl.net>, oZZ!!! <osa@etrust.ru>, current@FreeBSD.ORG
Subject:   Re: StarOffice-5.0... 
Message-ID:  <199812030058.QAA01111@dingo.cdrom.com>
In-Reply-To: Your message of "Wed, 02 Dec 1998 12:58:44 EST." <Pine.BSF.3.96.981201105221.19304F-100000@fledge.watson.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
> On Sun, 29 Nov 1998, Mike Smith wrote:
> 
> > > Yes, I've got the diffs against relatively fresh current. BTW, I asked
> > > this question on -emulation, but got back a profound silence... Can we/
> > > should we incorporate this patch, and hide it under a kernel option, say
> > > PROCFS_CMDLINE? The life would be soooo easier for people new to our linux
> > > emulation...
> > 
> > A couple of things:
> > 
> >  - If it's part of our emulation support, it should probably be the 
> >    default (cringe).
> >  - Your patch doesn't preserve the remainder of the commandline 
> >    arguments.  Feel like fixing this and resubmitting it?
> 
> I suppose there is no easy way for the procfs code to determine if the
> relevant calls are coming from a linux emulated process or not? 

I don't think there's a uniquifier for the ABI mode in the proc 
structure, no.

> If there
> were, then the command line stuff could be made only to appear for Linux
> processes.  I'm guessing, however, that only the creds and arguments to
> the vfs call make it that far down, and use of curproc seems
> inappropriate?

There's probably a pointer to the current process in the arguments.

> Or, if loadable kernel modules (or whatever they are called today) could
> add hook functions to the procfs module via some symbol or another, then
> new items could be added based on context.  That however, seems a little
> too busy, given that stackable file systems don't work. :)

Yeah.  I was playing with a generic framework for componentised VFS', 
but I haven't had time to do much with it lately, and the proper 
semantics for things like locking were giving me some confusion.  The 
real answer is a linux-procfs, which mounts on /compat/linux/procfs.

-- 
\\  Sometimes you're ahead,       \\  Mike Smith
\\  sometimes you're behind.      \\  mike@smith.net.au
\\  The race is long, and in the  \\  msmith@freebsd.org
\\  end it's only with yourself.  \\  msmith@cdrom.com



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199812030058.QAA01111>