From owner-svn-src-vendor@freebsd.org Wed Sep 28 23:46:09 2016 Return-Path: Delivered-To: svn-src-vendor@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id CC13AC01AC6; Wed, 28 Sep 2016 23:46:09 +0000 (UTC) (envelope-from mav@FreeBSD.org) Received: from repo.freebsd.org (repo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:0]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 9EF6EEA3; Wed, 28 Sep 2016 23:46:09 +0000 (UTC) (envelope-from mav@FreeBSD.org) Received: from repo.freebsd.org ([127.0.1.37]) by repo.freebsd.org (8.15.2/8.15.2) with ESMTP id u8SNk8ne000399; Wed, 28 Sep 2016 23:46:08 GMT (envelope-from mav@FreeBSD.org) Received: (from mav@localhost) by repo.freebsd.org (8.15.2/8.15.2/Submit) id u8SNk8LY000398; Wed, 28 Sep 2016 23:46:08 GMT (envelope-from mav@FreeBSD.org) Message-Id: <201609282346.u8SNk8LY000398@repo.freebsd.org> X-Authentication-Warning: repo.freebsd.org: mav set sender to mav@FreeBSD.org using -f From: Alexander Motin Date: Wed, 28 Sep 2016 23:46:08 +0000 (UTC) To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-vendor@freebsd.org Subject: svn commit: r306423 - vendor-sys/illumos/dist/uts/common/fs/zfs X-SVN-Group: vendor-sys MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-BeenThere: svn-src-vendor@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: SVN commit messages for the vendor work area tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 28 Sep 2016 23:46:09 -0000 Author: mav Date: Wed Sep 28 23:46:08 2016 New Revision: 306423 URL: https://svnweb.freebsd.org/changeset/base/306423 Log: 7402 Create tunable to ignore hole_birth feature Until we can resolve the numerous hole_birth bugs that have cropped up recently, and come up with a way going forwards to protect users from corruption, we should disable the hole_birth feature. Using a tunable allows those who are confident that their data is correct to continue to take advantage of the feature. Closes #188 Reviewed by: Matthew Ahrens Author: Paul Dagnelie Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/dmu_traverse.c Modified: vendor-sys/illumos/dist/uts/common/fs/zfs/dmu_traverse.c ============================================================================== --- vendor-sys/illumos/dist/uts/common/fs/zfs/dmu_traverse.c Wed Sep 28 23:44:13 2016 (r306422) +++ vendor-sys/illumos/dist/uts/common/fs/zfs/dmu_traverse.c Wed Sep 28 23:46:08 2016 (r306423) @@ -39,6 +39,7 @@ #include int32_t zfs_pd_bytes_max = 50 * 1024 * 1024; /* 50MB */ +boolean_t send_holes_without_birth_time = B_TRUE; typedef struct prefetch_data { kmutex_t pd_mtx; @@ -253,7 +254,8 @@ traverse_visitbp(traverse_data_t *td, co * * Note that the meta-dnode cannot be reallocated. */ - if ((!td->td_realloc_possible || + if (!send_holes_without_birth_time && + (!td->td_realloc_possible || zb->zb_object == DMU_META_DNODE_OBJECT) && td->td_hole_birth_enabled_txg <= td->td_min_txg) return (0);