From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:58531) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1X2yGD-0001lI-RF for qemu-devel@nongnu.org; Fri, 04 Jul 2014 03:50:30 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1X2yG4-0008Ln-Pb for qemu-devel@nongnu.org; Fri, 04 Jul 2014 03:50:21 -0400 Received: from mail-wg0-x234.google.com ([2a00:1450:400c:c00::234]:62551) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1X2yG4-0008Lj-I4 for qemu-devel@nongnu.org; Fri, 04 Jul 2014 03:50:12 -0400 Received: by mail-wg0-f52.google.com with SMTP id b13so1234425wgh.23 for ; Fri, 04 Jul 2014 00:50:11 -0700 (PDT) Date: Fri, 4 Jul 2014 09:50:08 +0200 From: Stefan Hajnoczi Message-ID: <20140704075008.GA32739@stefanha-thinkpad.redhat.com> References: <1404395521-11158-1-git-send-email-stefanha@redhat.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="uAKRQypu60I7Lcqm" Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [PATCH 0/2] coroutine: dynamically scale pool size List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Ming Lei Cc: Kevin Wolf , Paolo Bonzini , qemu-devel , Stefan Hajnoczi --uAKRQypu60I7Lcqm Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Jul 04, 2014 at 01:10:39PM +0800, Ming Lei wrote: > On Thu, Jul 3, 2014 at 9:51 PM, Stefan Hajnoczi wro= te: > > The coroutine pool reuses exited coroutines to make qemu_coroutine_crea= te() > > cheap. The size of the pool is capped to prevent it from hogging memor= y after > > a period of high coroutine activity. Previously the max size was hardc= oded to > > 64 but this doesn't scale with guest size. > > > > A guest with lots of disks can do more parallel I/O and therefore requi= res a > > larger coroutine pool size. This series tries to solve the problem by = scaling > > pool size according to the number of drives. > > > > Ming: Please let me know if this eliminates the rt_sigprocmask system c= alls you > > are seeing. It should solve part of the performance regression you hav= e seen > > in qemu.git/master virtio-blk dataplane. >=20 > With both the two coroutine patches and the block plug&unplug patches, > performance of qemu.git/master virtio-blk dataplane can recover to level = of > QEMU 2.0. /me does the happy dance Thanks for your great efforts in investigating and fixing the regression! Stefan --uAKRQypu60I7Lcqm Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJTtlywAAoJEJykq7OBq3PIOyIIAL7gzG+1whjrSZvt8bT6DSXk p3uU/0w1CntBNEARgQB4PrhLIC3Gqm30e8hZueEJuWI2pSPJwJMatPo6MbTuQbik S84o7PX87G19W4qALEp0/IsT4rWSzGayvPyPCypIukg2j2Hs4GFuKg3otCTMO+PX 5nz8+fHP9Cs7peQCGh8XRK/rPuiNkhhGZYYLNy9TVqg5/S0KwBsL2eRcSiDEbXXi VfLWlgZdve+IKUt4beqVl8S54R/SSinHF0L3TmN5SDp0eBB+PeGxOQTr41k1ISt/ /rOIMRlqzzNifUAguPY+ojhL4vg28PiyVM2YjEDlnc2q33uI+l0jJrQW2pLo8dU= =zFv5 -----END PGP SIGNATURE----- --uAKRQypu60I7Lcqm--