[klibc] proposed faccessat

[maximilian attems]

On Tue, Apr 20, 2010 at 12:08:02PM +0800, Herbert Xu wrote:
> On Tue, Apr 20, 2010 at 06:04:06AM +0200, maximilian attems wrote:
> > patch is also available in current patch queue:
> > git pull git://git.debian.org/users/maks/klibc.git maks
> > 
> > Run it on latest dash and it could still test readable files as readable.
> > 
> > >From b454cb46858debcd88e27964a94ebd9cbebf35f6 Mon Sep 17 00:00:00 2001
> > From: maximilian attems <max at stro.at>
> > Date: Fri, 16 Apr 2010 05:18:40 +0200
> > Subject: [PATCH] [klibc] add minimal faccessat()
> > MIME-Version: 1.0
> > Content-Type: text/plain; charset=UTF-8
> > Content-Transfer-Encoding: 8bit
> > 
> > Implement as wrapper around sys_faccessat().
> 
> This isn't quite right.  I'd suggest we either tweak dash's build
> to not detect faccessat on klibc, or implement faccessat properly
> in klibc by doing what the glibc version does.

if it is not klibc domain to have it we can merge hunk of my
patch to rename sys_faccessat() to __faccessat,
then dash config wouldn't see it in klibc and this fun would be over :)

as sidenote nor dietlibc nor uclibc seem to have it,
bad that the kernel implementation seems incomplete.

not really my call, waiting for hpa to weight in.
although I'd still like to know what you'd miss Herbert?