This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH v2 6/10] Tilera (and Linux asm-generic) support for glibc
- From: Chris Metcalf <cmetcalf at tilera dot com>
- To: "Joseph S. Myers" <joseph at codesourcery dot com>
- Cc: Roland McGrath <roland at hack dot frob dot com>, <libc-alpha at sourceware dot org>
- Date: Sat, 3 Dec 2011 19:55:11 -0500
- Subject: Re: [PATCH v2 6/10] Tilera (and Linux asm-generic) support for glibc
- References: <201111100054.pAA0sf6u025585@farm-0002.internal.tilera.com> <201111100056.pAA0uGrW025802@farm-0002.internal.tilera.com> <20111111180008.7147E2C0FE@topped-with-meat.com> <4EBD7DFE.9040105@tilera.com> <Pine.LNX.4.64.1111112158270.13833@digraph.polyomino.org.uk> <4EC1DCA1.2040406@tilera.com> <Pine.LNX.4.64.1111151711260.7813@digraph.polyomino.org.uk>
On 11/15/2011 12:13 PM, Joseph S. Myers wrote:
> On Mon, 14 Nov 2011, Chris Metcalf wrote:
>
>>> Yes, FSF glibc doesn't support building glibc with -Os.
>> In practice, though, it works fine for us, once I fixed that one lone
>> callsite in faccessat() to directly call __fxstatat64(). My patch seems to
>> obsolete the patch you passed along from Sandra Loosemore; perhaps other
>> call sites have already been fixed up since then?
> Quite possibly. Or the set of call sites may depend on the target
> architecture (especially with the "generic" syscall interface you could be
> using a significantly different set of call sites to other architectures).
I wanted to follow up on this; in particular, if my change to faccessat
allows glibc to build with -Os for all the "linux-generic" architectures, I
suspect it may be worth it. It also seems to follow the convention of
using the "x" versions of the stat functions internally, and passing
_STAT_VER, to avoid the overhead of checking the version as soon as you
enter the called function.
The patch is here: http://sourceware.org/ml/libc-alpha/2011-11/msg00024.html .
--
Chris Metcalf, Tilera Corp.
http://www.tilera.com