Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 22 May 2012 14:17:03 -0700
From: Roland McGrath <mcgrathr@...gle.com>
To: Al Viro <viro@...iv.linux.org.uk>
Cc: Will Drewry <wad@...omium.org>, "H. Peter Anvin" <hpa@...or.com>, Indan Zupancic <indan@....nu>, 
	Eric Paris <netdev@...isplace.org>, linux-kernel@...r.kernel.org, 
	linux-security-module@...r.kernel.org, kernel-hardening@...ts.openwall.com, 
	mingo@...hat.com, oleg@...hat.com, peterz@...radead.org, rdunlap@...otime.net, 
	tglx@...utronix.de, luto@....edu, eparis@...hat.com, 
	serge.hallyn@...onical.com, pmoore@...hat.com, akpm@...ux-foundation.org, 
	corbet@....net, eric.dumazet@...il.com, markus@...omium.org, 
	coreyb@...ux.vnet.ibm.com, keescook@...omium.org
Subject: Re: seccomp and ptrace. what is the correct order?

On Tue, May 22, 2012 at 2:07 PM, Al Viro <viro@...iv.linux.org.uk> wrote:
> FWIW, I'd prefer to have all that done inside __audit_syscall_entry(),
> with
>        context->arch       = syscall_get_arch(current, regs);
>        context->major      = syscall_get_nr(current, regs);
>        syscall_get_arguments(current, regs, 0, 4, context->argv);
> done instead of initializations from arguments we are doing there now.
> I seriously doubt that it would lead to worse code than what we currently
> have.  If nothing else, we won't be passing that pile of arguments around.

I always felt the same way about the audit code.  (As a bonus, if the audit
folks ever decide they want all six syscall arguments instead of just four,
they wouldn't have to touch every arch.)  But it will certainly produce
drastically worse code for ia64.  (Not that anybody cares about ia64.)


Thanks,
Roland

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.