From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1E6ngY-0001WI-Pp
	for qemu-devel@nongnu.org; Sun, 21 Aug 2005 07:08:19 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1E6ngU-0001Uw-HC
	for qemu-devel@nongnu.org; Sun, 21 Aug 2005 07:08:17 -0400
Received: from [199.232.76.173] (helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1E6ngT-0001RI-Ly
	for qemu-devel@nongnu.org; Sun, 21 Aug 2005 07:08:13 -0400
Received: from [65.74.133.11] (helo=mail.codesourcery.com)
	by monty-python.gnu.org with esmtp
	(TLS-1.0:DHE_RSA_3DES_EDE_CBC_SHA:24) (Exim 4.34) id 1E6nbf-0005R7-Nr
	for qemu-devel@nongnu.org; Sun, 21 Aug 2005 07:03:16 -0400
From: Paul Brook <paul@codesourcery.com>
Subject: Re: [Qemu-devel] [patch] NPTL/TLS support
Date: Sun, 21 Aug 2005 12:02:47 +0100
References: <200508141653.25261.paul@codesourcery.com>
	<430850FE.9080306@bellard.org>
In-Reply-To: <430850FE.9080306@bellard.org>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200508211202.48579.paul@codesourcery.com>
Reply-To: qemu-devel@nongnu.org
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: qemu-devel@nongnu.org

> > - The Arm magic kernel code page. This is used on recent Arm kernels to
> > provide efficient access to kernel/CPU features. For example atomic
> > operations, and reading the TLS register (which may or may not be a real
> > hardware register). This is simplified version of the VDSO used on recent
> > x86 and ppc kernels.
>
> On i386 there are AUXENTs to define the address of the VDSO page
> (AT_SYSINFO and AT_SYSINFO_EHDR). If it was supported by the ARM glibc,
> it would avoid some hacks you need to do to intercept calls to it. Can
> you check that ?

Arm does not [yet] use a full VDSO.
As Arm is primarily aimed at embedded targets it was decided that the overhead 
of a VDSO was undesirable.

In particular a fixed page mapping means an application can use shifter tricks 
to generate the address and avoid the PLT indirection.

e.g.:
__aeabi_read_tp:
        mov     r0, #0xffff0fff
        sub     pc, r0, #31

Paul