From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [140.186.70.92] (port=35282 helo=eggs.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1PpcjC-0000Py-Dw for qemu-devel@nongnu.org; Wed, 16 Feb 2011 03:27:15 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1PpcjB-0004Ca-EO for qemu-devel@nongnu.org; Wed, 16 Feb 2011 03:27:14 -0500 Received: from fmmailgate01.web.de ([217.72.192.221]:35839) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1PpcjB-0004Bs-1e for qemu-devel@nongnu.org; Wed, 16 Feb 2011 03:27:13 -0500 Message-ID: <4D5B8A1E.9000404@web.de> Date: Wed, 16 Feb 2011 09:26:06 +0100 From: Jan Kiszka MIME-Version: 1.0 References: <4D58FAF2.6070106@siemens.com> <20110215220330.GA2547@amt.cnet> In-Reply-To: <20110215220330.GA2547@amt.cnet> Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enig5A0EB08E1FDB3C9C5A5C6F0C" Sender: jan.kiszka@web.de Subject: [Qemu-devel] Re: [RFC] Some more io-thread optimizations List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Marcelo Tosatti Cc: Paolo Bonzini , Anthony Liguori , qemu-devel , Aurelien Jarno This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enig5A0EB08E1FDB3C9C5A5C6F0C Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable On 2011-02-15 23:03, Marcelo Tosatti wrote: > On Mon, Feb 14, 2011 at 10:50:42AM +0100, Jan Kiszka wrote: >> Hi, >> >> patch below further reduces the io-thread overhead in tcg mode so that= >> specifically emulating smp boxes gets noticeably faster. Its essence: >> poll the file descriptors until select returns 0, keeping the global >> mutex locked. This reduces ping pong with the vcpu threads, most >> noticeably in tcg mode where we run in lock-step. >> >> Split up in two patches, I'm planning to route those changes via the k= vm >> queue (as they collide with other patches there). >> >> Jan >=20 > Not sure this makes sense for all cases. There could be scenarios where= > a single pass is more efficient (think latency to acquire mutex from > vcpu context in kvm mode, with intensive file IO in progress). Yeah, likely true. Only TCG has these insane long lock-holding times and requires signal-based mutex handover. I will exclude KVM from this. Jan --------------enig5A0EB08E1FDB3C9C5A5C6F0C Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.15 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAk1biiIACgkQitSsb3rl5xT0nACbBwjxncCgWw70RO6OckiVEneP 3IAAoNUiUqolYJC6nr7DJJhrojnQ5RA6 =LMhI -----END PGP SIGNATURE----- --------------enig5A0EB08E1FDB3C9C5A5C6F0C--