From patchwork Fri Jun 17 01:19:08 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Darrick J. Wong" X-Patchwork-Id: 9182045 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 75F7B60760 for ; Fri, 17 Jun 2016 01:19:21 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 66B592839F for ; Fri, 17 Jun 2016 01:19:21 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 5B55C283A1; Fri, 17 Jun 2016 01:19:21 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, UNPARSEABLE_RELAY autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 7675D2839F for ; Fri, 17 Jun 2016 01:19:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754793AbcFQBTT (ORCPT ); Thu, 16 Jun 2016 21:19:19 -0400 Received: from userp1040.oracle.com ([156.151.31.81]:28202 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754091AbcFQBTS (ORCPT ); Thu, 16 Jun 2016 21:19:18 -0400 Received: from userv0022.oracle.com (userv0022.oracle.com [156.151.31.74]) by userp1040.oracle.com (Sentrion-MTA-4.3.2/Sentrion-MTA-4.3.2) with ESMTP id u5H1JDX4005112 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Fri, 17 Jun 2016 01:19:14 GMT Received: from aserv0122.oracle.com (aserv0122.oracle.com [141.146.126.236]) by userv0022.oracle.com (8.14.4/8.13.8) with ESMTP id u5H1JDZJ006678 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Fri, 17 Jun 2016 01:19:13 GMT Received: from abhmp0013.oracle.com (abhmp0013.oracle.com [141.146.116.19]) by aserv0122.oracle.com (8.13.8/8.13.8) with ESMTP id u5H1JArv006008; Fri, 17 Jun 2016 01:19:11 GMT Received: from localhost (/10.145.178.207) by default (Oracle Beehive Gateway v4.0) with ESMTP ; Thu, 16 Jun 2016 18:19:10 -0700 Subject: [PATCH 012/119] xfs: during btree split, save new block key & ptr for future insertion From: "Darrick J. Wong" To: david@fromorbit.com, darrick.wong@oracle.com Cc: linux-fsdevel@vger.kernel.org, vishal.l.verma@intel.com, xfs@oss.sgi.com Date: Thu, 16 Jun 2016 18:19:08 -0700 Message-ID: <146612634887.12839.6210102273046676244.stgit@birch.djwong.org> In-Reply-To: <146612627129.12839.3827886950949809165.stgit@birch.djwong.org> References: <146612627129.12839.3827886950949809165.stgit@birch.djwong.org> User-Agent: StGit/0.17.1-dirty MIME-Version: 1.0 X-Source-IP: userv0022.oracle.com [156.151.31.74] Sender: linux-fsdevel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-fsdevel@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP When a btree block has to be split, we pass the new block's ptr from xfs_btree_split() back to xfs_btree_insert() via a pointer parameter; however, we pass the block's key through the cursor's record. It is a little weird to "initialize" a record from a key since the non-key attributes will have garbage values. When we go to add support for interval queries, we have to be able to pass the lowest and highest keys accessible via a pointer. There's no clean way to pass this back through the cursor's record field. Therefore, pass the key directly back to xfs_btree_insert() the same way that we pass the btree_ptr. As a bonus, we no longer need init_rec_from_key and can drop it from the codebase. Signed-off-by: Darrick J. Wong --- fs/xfs/libxfs/xfs_alloc_btree.c | 12 ---------- fs/xfs/libxfs/xfs_bmap_btree.c | 12 ---------- fs/xfs/libxfs/xfs_btree.c | 44 +++++++++++++++++++------------------- fs/xfs/libxfs/xfs_btree.h | 2 -- fs/xfs/libxfs/xfs_ialloc_btree.c | 10 --------- 5 files changed, 22 insertions(+), 58 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/fs/xfs/libxfs/xfs_alloc_btree.c b/fs/xfs/libxfs/xfs_alloc_btree.c index d9b42425..5ba2dac 100644 --- a/fs/xfs/libxfs/xfs_alloc_btree.c +++ b/fs/xfs/libxfs/xfs_alloc_btree.c @@ -212,17 +212,6 @@ xfs_allocbt_init_key_from_rec( } STATIC void -xfs_allocbt_init_rec_from_key( - union xfs_btree_key *key, - union xfs_btree_rec *rec) -{ - ASSERT(key->alloc.ar_startblock != 0); - - rec->alloc.ar_startblock = key->alloc.ar_startblock; - rec->alloc.ar_blockcount = key->alloc.ar_blockcount; -} - -STATIC void xfs_allocbt_init_rec_from_cur( struct xfs_btree_cur *cur, union xfs_btree_rec *rec) @@ -406,7 +395,6 @@ static const struct xfs_btree_ops xfs_allocbt_ops = { .get_minrecs = xfs_allocbt_get_minrecs, .get_maxrecs = xfs_allocbt_get_maxrecs, .init_key_from_rec = xfs_allocbt_init_key_from_rec, - .init_rec_from_key = xfs_allocbt_init_rec_from_key, .init_rec_from_cur = xfs_allocbt_init_rec_from_cur, .init_ptr_from_cur = xfs_allocbt_init_ptr_from_cur, .key_diff = xfs_allocbt_key_diff, diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c index db0c71e..714b387 100644 --- a/fs/xfs/libxfs/xfs_bmap_btree.c +++ b/fs/xfs/libxfs/xfs_bmap_btree.c @@ -600,17 +600,6 @@ xfs_bmbt_init_key_from_rec( } STATIC void -xfs_bmbt_init_rec_from_key( - union xfs_btree_key *key, - union xfs_btree_rec *rec) -{ - ASSERT(key->bmbt.br_startoff != 0); - - xfs_bmbt_disk_set_allf(&rec->bmbt, be64_to_cpu(key->bmbt.br_startoff), - 0, 0, XFS_EXT_NORM); -} - -STATIC void xfs_bmbt_init_rec_from_cur( struct xfs_btree_cur *cur, union xfs_btree_rec *rec) @@ -760,7 +749,6 @@ static const struct xfs_btree_ops xfs_bmbt_ops = { .get_minrecs = xfs_bmbt_get_minrecs, .get_dmaxrecs = xfs_bmbt_get_dmaxrecs, .init_key_from_rec = xfs_bmbt_init_key_from_rec, - .init_rec_from_key = xfs_bmbt_init_rec_from_key, .init_rec_from_cur = xfs_bmbt_init_rec_from_cur, .init_ptr_from_cur = xfs_bmbt_init_ptr_from_cur, .key_diff = xfs_bmbt_key_diff, diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c index 046fbcf..a096539 100644 --- a/fs/xfs/libxfs/xfs_btree.c +++ b/fs/xfs/libxfs/xfs_btree.c @@ -2862,10 +2862,9 @@ xfs_btree_make_block_unfull( int *index, /* new tree index */ union xfs_btree_ptr *nptr, /* new btree ptr */ struct xfs_btree_cur **ncur, /* new btree cursor */ - union xfs_btree_rec *nrec, /* new record */ + union xfs_btree_key *key, /* key of new block */ int *stat) { - union xfs_btree_key key; /* new btree key value */ int error = 0; if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && @@ -2910,13 +2909,12 @@ xfs_btree_make_block_unfull( * If this works we have to re-set our variables because we * could be in a different block now. */ - error = xfs_btree_split(cur, level, nptr, &key, ncur, stat); + error = xfs_btree_split(cur, level, nptr, key, ncur, stat); if (error || *stat == 0) return error; *index = cur->bc_ptrs[level]; - cur->bc_ops->init_rec_from_key(&key, nrec); return 0; } @@ -2929,16 +2927,16 @@ xfs_btree_insrec( struct xfs_btree_cur *cur, /* btree cursor */ int level, /* level to insert record at */ union xfs_btree_ptr *ptrp, /* i/o: block number inserted */ - union xfs_btree_rec *recp, /* i/o: record data inserted */ + union xfs_btree_key *key, /* i/o: block key for ptrp */ struct xfs_btree_cur **curp, /* output: new cursor replacing cur */ int *stat) /* success/failure */ { struct xfs_btree_block *block; /* btree block */ struct xfs_buf *bp; /* buffer for block */ - union xfs_btree_key key; /* btree key */ union xfs_btree_ptr nptr; /* new block ptr */ struct xfs_btree_cur *ncur; /* new btree cursor */ - union xfs_btree_rec nrec; /* new record count */ + union xfs_btree_key nkey; /* new block key */ + union xfs_btree_rec rec; /* record to insert */ int optr; /* old key/record index */ int ptr; /* key/record index */ int numrecs;/* number of records */ @@ -2947,8 +2945,14 @@ xfs_btree_insrec( int i; #endif + /* Make a key out of the record data to be inserted, and save it. */ + if (level == 0) { + cur->bc_ops->init_rec_from_cur(cur, &rec); + cur->bc_ops->init_key_from_rec(key, &rec); + } + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); - XFS_BTREE_TRACE_ARGIPR(cur, level, *ptrp, recp); + XFS_BTREE_TRACE_ARGIPR(cur, level, *ptrp, &rec); ncur = NULL; @@ -2973,9 +2977,6 @@ xfs_btree_insrec( return 0; } - /* Make a key out of the record data to be inserted, and save it. */ - cur->bc_ops->init_key_from_rec(&key, recp); - optr = ptr; XFS_BTREE_STATS_INC(cur, insrec); @@ -2992,10 +2993,10 @@ xfs_btree_insrec( /* Check that the new entry is being inserted in the right place. */ if (ptr <= numrecs) { if (level == 0) { - ASSERT(cur->bc_ops->recs_inorder(cur, recp, + ASSERT(cur->bc_ops->recs_inorder(cur, &rec, xfs_btree_rec_addr(cur, ptr, block))); } else { - ASSERT(cur->bc_ops->keys_inorder(cur, &key, + ASSERT(cur->bc_ops->keys_inorder(cur, key, xfs_btree_key_addr(cur, ptr, block))); } } @@ -3008,7 +3009,7 @@ xfs_btree_insrec( xfs_btree_set_ptr_null(cur, &nptr); if (numrecs == cur->bc_ops->get_maxrecs(cur, level)) { error = xfs_btree_make_block_unfull(cur, level, numrecs, - &optr, &ptr, &nptr, &ncur, &nrec, stat); + &optr, &ptr, &nptr, &ncur, &nkey, stat); if (error || *stat == 0) goto error0; } @@ -3058,7 +3059,7 @@ xfs_btree_insrec( #endif /* Now put the new data in, bump numrecs and log it. */ - xfs_btree_copy_keys(cur, kp, &key, 1); + xfs_btree_copy_keys(cur, kp, key, 1); xfs_btree_copy_ptrs(cur, pp, ptrp, 1); numrecs++; xfs_btree_set_numrecs(block, numrecs); @@ -3079,7 +3080,7 @@ xfs_btree_insrec( xfs_btree_shift_recs(cur, rp, 1, numrecs - ptr + 1); /* Now put the new data in, bump numrecs and log it. */ - xfs_btree_copy_recs(cur, rp, recp, 1); + xfs_btree_copy_recs(cur, rp, &rec, 1); xfs_btree_set_numrecs(block, ++numrecs); xfs_btree_log_recs(cur, bp, ptr, numrecs); #ifdef DEBUG @@ -3095,7 +3096,7 @@ xfs_btree_insrec( /* If we inserted at the start of a block, update the parents' keys. */ if (optr == 1) { - error = xfs_btree_updkey(cur, &key, level + 1); + error = xfs_btree_updkey(cur, key, level + 1); if (error) goto error0; } @@ -3105,7 +3106,7 @@ xfs_btree_insrec( * we are at the far right edge of the tree, update it. */ if (xfs_btree_is_lastrec(cur, block, level)) { - cur->bc_ops->update_lastrec(cur, block, recp, + cur->bc_ops->update_lastrec(cur, block, &rec, ptr, LASTREC_INSREC); } @@ -3115,7 +3116,7 @@ xfs_btree_insrec( */ *ptrp = nptr; if (!xfs_btree_ptr_is_null(cur, &nptr)) { - *recp = nrec; + *key = nkey; *curp = ncur; } @@ -3146,14 +3147,13 @@ xfs_btree_insert( union xfs_btree_ptr nptr; /* new block number (split result) */ struct xfs_btree_cur *ncur; /* new cursor (split result) */ struct xfs_btree_cur *pcur; /* previous level's cursor */ - union xfs_btree_rec rec; /* record to insert */ + union xfs_btree_key key; /* key of block to insert */ level = 0; ncur = NULL; pcur = cur; xfs_btree_set_ptr_null(cur, &nptr); - cur->bc_ops->init_rec_from_cur(cur, &rec); /* * Loop going up the tree, starting at the leaf level. @@ -3165,7 +3165,7 @@ xfs_btree_insert( * Insert nrec/nptr into this level of the tree. * Note if we fail, nptr will be null. */ - error = xfs_btree_insrec(pcur, level, &nptr, &rec, &ncur, &i); + error = xfs_btree_insrec(pcur, level, &nptr, &key, &ncur, &i); if (error) { if (pcur != cur) xfs_btree_del_cursor(pcur, XFS_BTREE_ERROR); diff --git a/fs/xfs/libxfs/xfs_btree.h b/fs/xfs/libxfs/xfs_btree.h index b955e5d..b99c018 100644 --- a/fs/xfs/libxfs/xfs_btree.h +++ b/fs/xfs/libxfs/xfs_btree.h @@ -158,8 +158,6 @@ struct xfs_btree_ops { /* init values of btree structures */ void (*init_key_from_rec)(union xfs_btree_key *key, union xfs_btree_rec *rec); - void (*init_rec_from_key)(union xfs_btree_key *key, - union xfs_btree_rec *rec); void (*init_rec_from_cur)(struct xfs_btree_cur *cur, union xfs_btree_rec *rec); void (*init_ptr_from_cur)(struct xfs_btree_cur *cur, diff --git a/fs/xfs/libxfs/xfs_ialloc_btree.c b/fs/xfs/libxfs/xfs_ialloc_btree.c index 89c21d7..88da2ad 100644 --- a/fs/xfs/libxfs/xfs_ialloc_btree.c +++ b/fs/xfs/libxfs/xfs_ialloc_btree.c @@ -146,14 +146,6 @@ xfs_inobt_init_key_from_rec( } STATIC void -xfs_inobt_init_rec_from_key( - union xfs_btree_key *key, - union xfs_btree_rec *rec) -{ - rec->inobt.ir_startino = key->inobt.ir_startino; -} - -STATIC void xfs_inobt_init_rec_from_cur( struct xfs_btree_cur *cur, union xfs_btree_rec *rec) @@ -314,7 +306,6 @@ static const struct xfs_btree_ops xfs_inobt_ops = { .get_minrecs = xfs_inobt_get_minrecs, .get_maxrecs = xfs_inobt_get_maxrecs, .init_key_from_rec = xfs_inobt_init_key_from_rec, - .init_rec_from_key = xfs_inobt_init_rec_from_key, .init_rec_from_cur = xfs_inobt_init_rec_from_cur, .init_ptr_from_cur = xfs_inobt_init_ptr_from_cur, .key_diff = xfs_inobt_key_diff, @@ -336,7 +327,6 @@ static const struct xfs_btree_ops xfs_finobt_ops = { .get_minrecs = xfs_inobt_get_minrecs, .get_maxrecs = xfs_inobt_get_maxrecs, .init_key_from_rec = xfs_inobt_init_key_from_rec, - .init_rec_from_key = xfs_inobt_init_rec_from_key, .init_rec_from_cur = xfs_inobt_init_rec_from_cur, .init_ptr_from_cur = xfs_finobt_init_ptr_from_cur, .key_diff = xfs_inobt_key_diff,