From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:48007) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dtFuL-0003AY-Oh for qemu-devel@nongnu.org; Sat, 16 Sep 2017 12:25:30 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dtFuK-00066N-Tk for qemu-devel@nongnu.org; Sat, 16 Sep 2017 12:25:29 -0400 References: <20170912112855.24269-1-berrange@redhat.com> <20170912112855.24269-2-berrange@redhat.com> From: Max Reitz Message-ID: Date: Sat, 16 Sep 2017 18:25:19 +0200 MIME-Version: 1.0 In-Reply-To: <20170912112855.24269-2-berrange@redhat.com> Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="frj6fnd4rdbfNpXe5ksRIINTJWj3G7EHS" Subject: Re: [Qemu-devel] [PATCH v3 1/7] block: use 1 MB bounce buffers for crypto instead of 16KB List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Daniel P. Berrange" , qemu-devel@nongnu.org Cc: qemu-block@nongnu.org, Kevin Wolf , Eric Blake , Stefan Hajnoczi This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --frj6fnd4rdbfNpXe5ksRIINTJWj3G7EHS From: Max Reitz To: "Daniel P. Berrange" , qemu-devel@nongnu.org Cc: qemu-block@nongnu.org, Kevin Wolf , Eric Blake , Stefan Hajnoczi Message-ID: Subject: Re: [PATCH v3 1/7] block: use 1 MB bounce buffers for crypto instead of 16KB References: <20170912112855.24269-1-berrange@redhat.com> <20170912112855.24269-2-berrange@redhat.com> In-Reply-To: <20170912112855.24269-2-berrange@redhat.com> Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable On 2017-09-12 13:28, Daniel P. Berrange wrote: > Using 16KB bounce buffers creates a significant performance > penalty for I/O to encrypted volumes on storage which high > I/O latency (rotating rust & network drives), because it > triggers lots of fairly small I/O operations. >=20 > On tests with rotating rust, and cache=3Dnone|directsync, > write speed increased from 2MiB/s to 32MiB/s, on a par > with that achieved by the in-kernel luks driver. With > other cache modes the in-kernel driver is still notably > faster because it is able to report completion of the > I/O request before any encryption is done, while the > in-QEMU driver must encrypt the data before completion. >=20 > Reviewed-by: Eric Blake > Reviewed-by: Stefan Hajnoczi > Signed-off-by: Daniel P. Berrange > --- > block/crypto.c | 12 +++++------- > 1 file changed, 5 insertions(+), 7 deletions(-) Reviewed-by: Max Reitz --frj6fnd4rdbfNpXe5ksRIINTJWj3G7EHS Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- iQFGBAEBCAAwFiEEkb62CjDbPohX0Rgp9AfbAGHVz0AFAlm9UG8SHG1yZWl0ekBy ZWRoYXQuY29tAAoJEPQH2wBh1c9ADncIALQ69xtviQVAvucTMXjnpuZ0YyNqXRC3 gfz8MS5Yis/ZDuTdtmaO7NsUZKBdtCJfljgPlqFrzaZmElqDanFKwu8gkIrTakTy vIwaKIzgA2oIvLOIcGpQzoo4feItJTVSiPaFNFkiXmcIZ4OFASWcm8lNcY9J2xA7 sPAtFWCM4wLXpfxNwUaLQNZ6+4tMn8Ou1ixcy/FQz0Rm0Rljy/ibOKo4COZjFruf THrD16Zk2ZtSPwIeaezSt9y3CLyvl0AsOBlZf5SZhKTpnnkMnKsbv5+nRF5ViJO9 PA3Nz27+NbcNmHtmBiT8RiLwYxulOIixd/Y88Ng8dvGks9WsZjIFa88= =JzQC -----END PGP SIGNATURE----- --frj6fnd4rdbfNpXe5ksRIINTJWj3G7EHS--