From mboxrd@z Thu Jan 1 00:00:00 1970 From: NeilBrown Subject: Re: [PATCH v3 02/14] md: move two macros into md.h Date: Mon, 27 Mar 2017 20:52:24 +1100 Message-ID: <87wpbb6o87.fsf@notabene.neil.brown.name> References: <20170316161235.27110-1-tom.leiming@gmail.com> <20170316161235.27110-3-tom.leiming@gmail.com> <87tw6j8be6.fsf@notabene.neil.brown.name> <20170324165325.nek4kb4yezz2xmow@kernel.org> <20170327091553.GF6879@infradead.org> Mime-Version: 1.0 Content-Type: multipart/signed; boundary="=-=-="; micalg=pgp-sha256; protocol="application/pgp-signature" Return-path: In-Reply-To: <20170327091553.GF6879@infradead.org> Sender: linux-block-owner@vger.kernel.org To: Shaohua Li Cc: Ming Lei , Jens Axboe , linux-raid@vger.kernel.org, linux-block@vger.kernel.org, Christoph Hellwig List-Id: linux-raid.ids --=-=-= Content-Type: text/plain Content-Transfer-Encoding: quoted-printable On Mon, Mar 27 2017, Christoph Hellwig wrote: > On Fri, Mar 24, 2017 at 09:53:25AM -0700, Shaohua Li wrote: >>=20 >> I had the same concern when I looked at this patch firstly. The number f= or >> raid1/10 doesn't need to be the same. But if we don't move the number to= a >> generic header, the third patch will become a little more complicated. I >> eventually ignored this issue. If we really need different number for ra= id1/10, >> lets do it at that time. > > Which brings up my usual queastion: Is is really that benefitical for > us to keep the raid1.c code around instead of making it a special short > cut case in raid10.c? Patches welcome. They would need to handle write-mostly and write-behind. They would also need to avoid the assumption of a chunk size for RAID1. Undoubtedly do-able. Hard to say how beneficial it would be, or how much it would cost. NeilBrown --=-=-= Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEEG8Yp69OQ2HB7X0l6Oeye3VZigbkFAljY4NkACgkQOeye3VZi gbmpVxAAqKNg+kIzJKQ41rcH8t7TAe9n0SXc223FGYgxY+bEpkLmEUrPIE5rhijj CkfmNko7+JuJcmdHxf7p42iWGngfJvULpqZ4g/SWD3G0bUyOwvaSEg3ayJYoyj5H tMY7QGyAgS9xAdfOwRem0Z35paLJEw9AFZcoRR/CBtpZY9D6QfLeiqhHEhkbhedW YHxbbBPbB441mxf57h4Ap7duKDXD6XGmgjKGCpwI2Css8PMsvzA3uitpvTjcX8px x727GdeKwa+uWt6ag9fqFBAHPrz37RVPRhOrD2Ew88ZwQEWbgwjOS9Kas02jw9VY p5npEpybzvjcT3ZzqMTWgIac7ZFojLAeaqDsRBW2MHoJjM/w17qlcmc47/F3vJcb KALHxj7ji0rG4IbrNZa9QoAd+zvXdQs4S/r9AsEajRf+vg9lqCrw9HqJlJPzXgac AwSCOF/B/aKBf7vm5dBnot7xU87wTlRSM0jQ8uyx9uTAHv5hefAFGo86c1r3Z1IQ b3pIyhQiq+QWIMFsiz6eaHNfZrkBR0zreOwQbmn4BqC03KVdxkTdiNOoD3MdLMvR AdvXRmKYAr2BTIaz8WRIVxldbtGnoNbNHUbB/BrXqrR0bkXsYxdZzvPZTdXJSTnZ jqSnaLVyZvbbBYuzsZRTbjB/4PYkcahARe/FmZ2GhkP+vOiTOng= =hQbl -----END PGP SIGNATURE----- --=-=-=-- From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx2.suse.de ([195.135.220.15]:33372 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753102AbdC0JyB (ORCPT ); Mon, 27 Mar 2017 05:54:01 -0400 From: NeilBrown To: Christoph Hellwig , Shaohua Li Date: Mon, 27 Mar 2017 20:52:24 +1100 Cc: Ming Lei , Jens Axboe , linux-raid@vger.kernel.org, linux-block@vger.kernel.org, Christoph Hellwig Subject: Re: [PATCH v3 02/14] md: move two macros into md.h In-Reply-To: <20170327091553.GF6879@infradead.org> References: <20170316161235.27110-1-tom.leiming@gmail.com> <20170316161235.27110-3-tom.leiming@gmail.com> <87tw6j8be6.fsf@notabene.neil.brown.name> <20170324165325.nek4kb4yezz2xmow@kernel.org> <20170327091553.GF6879@infradead.org> Message-ID: <87wpbb6o87.fsf@notabene.neil.brown.name> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="=-=-="; micalg=pgp-sha256; protocol="application/pgp-signature" Sender: linux-block-owner@vger.kernel.org List-Id: linux-block@vger.kernel.org --=-=-= Content-Type: text/plain Content-Transfer-Encoding: quoted-printable On Mon, Mar 27 2017, Christoph Hellwig wrote: > On Fri, Mar 24, 2017 at 09:53:25AM -0700, Shaohua Li wrote: >>=20 >> I had the same concern when I looked at this patch firstly. The number f= or >> raid1/10 doesn't need to be the same. But if we don't move the number to= a >> generic header, the third patch will become a little more complicated. I >> eventually ignored this issue. If we really need different number for ra= id1/10, >> lets do it at that time. > > Which brings up my usual queastion: Is is really that benefitical for > us to keep the raid1.c code around instead of making it a special short > cut case in raid10.c? Patches welcome. They would need to handle write-mostly and write-behind. They would also need to avoid the assumption of a chunk size for RAID1. Undoubtedly do-able. Hard to say how beneficial it would be, or how much it would cost. NeilBrown --=-=-= Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEEG8Yp69OQ2HB7X0l6Oeye3VZigbkFAljY4NkACgkQOeye3VZi gbmpVxAAqKNg+kIzJKQ41rcH8t7TAe9n0SXc223FGYgxY+bEpkLmEUrPIE5rhijj CkfmNko7+JuJcmdHxf7p42iWGngfJvULpqZ4g/SWD3G0bUyOwvaSEg3ayJYoyj5H tMY7QGyAgS9xAdfOwRem0Z35paLJEw9AFZcoRR/CBtpZY9D6QfLeiqhHEhkbhedW YHxbbBPbB441mxf57h4Ap7duKDXD6XGmgjKGCpwI2Css8PMsvzA3uitpvTjcX8px x727GdeKwa+uWt6ag9fqFBAHPrz37RVPRhOrD2Ew88ZwQEWbgwjOS9Kas02jw9VY p5npEpybzvjcT3ZzqMTWgIac7ZFojLAeaqDsRBW2MHoJjM/w17qlcmc47/F3vJcb KALHxj7ji0rG4IbrNZa9QoAd+zvXdQs4S/r9AsEajRf+vg9lqCrw9HqJlJPzXgac AwSCOF/B/aKBf7vm5dBnot7xU87wTlRSM0jQ8uyx9uTAHv5hefAFGo86c1r3Z1IQ b3pIyhQiq+QWIMFsiz6eaHNfZrkBR0zreOwQbmn4BqC03KVdxkTdiNOoD3MdLMvR AdvXRmKYAr2BTIaz8WRIVxldbtGnoNbNHUbB/BrXqrR0bkXsYxdZzvPZTdXJSTnZ jqSnaLVyZvbbBYuzsZRTbjB/4PYkcahARe/FmZ2GhkP+vOiTOng= =hQbl -----END PGP SIGNATURE----- --=-=-=--