From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:40098)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1an3Ns-0005db-Bc
	for qemu-devel@nongnu.org; Mon, 04 Apr 2016 08:13:33 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1an3Nm-0005kp-Oj
	for qemu-devel@nongnu.org; Mon, 04 Apr 2016 08:13:32 -0400
Received: from mx1.redhat.com ([209.132.183.28]:35394)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1an3Nm-0005kl-Ji
	for qemu-devel@nongnu.org; Mon, 04 Apr 2016 08:13:26 -0400
References: <87y48tu53d.fsf@linaro.org> <57022D7A.1040507@redhat.com>
	<57024478.2080506@gmail.com> <87wpodty5d.fsf@linaro.org>
	<57025113.7000101@redhat.com> <87vb3xtwg1.fsf@linaro.org>
From: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <57025A5E.3020008@redhat.com>
Date: Mon, 4 Apr 2016 14:13:18 +0200
MIME-Version: 1.0
In-Reply-To: <87vb3xtwg1.fsf@linaro.org>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: [Qemu-devel] Should we introduce a TranslationRegion with its
 own codegen buffer?
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: =?UTF-8?Q?Alex_Benn=c3=a9e?= <alex.bennee@linaro.org>
Cc: MTTCG Devel <mttcg@greensocs.com>, Peter Maydell <peter.maydell@linaro.org>, QEMU Developers <qemu-devel@nongnu.org>, Cota@braap.org, Sergey Fedorov <serge.fdrv@gmail.com>, Richard Henderson <rth@twiddle.net>



On 04/04/2016 14:01, Alex Benn=C3=A9e wrote:
> > >   - is pretty ugly anyway
> >
> > What exactly makes it ugly?
>
> Subjective I guess but the use of tb_lock_recursive() and
> tb_lock_reset() seem not as clean as they could be. Is this stuff Fred
> re-introduced?

I think tb_lock_recursive is mine, while tb_lock_reset is from Fred.

I concede that those functions are a bit ugly, but that is unrelated to
the "tb_find_fast outside tb_lock" idea (which was from Fred and gives a
~20% or so speedup at -smp 2; more as the number of VCPUs increases).

You could make tb_lock recursive I guess if you want to avoid
tb_lock_recursive() but it really just sweeps it under the carpet.
tb_lock_reset() is needed anyway because it is used after longjmp.

Paolo