From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-9.8 required=3.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id E6A59C2D0BF for ; Fri, 13 Dec 2019 05:53:51 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id AA1E0227BF for ; Fri, 13 Dec 2019 05:53:51 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=samsung.com header.i=@samsung.com header.b="LQHi5QEr" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1732155AbfLMFxu (ORCPT ); Fri, 13 Dec 2019 00:53:50 -0500 Received: from mailout1.samsung.com ([203.254.224.24]:16923 "EHLO mailout1.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727102AbfLMFxq (ORCPT ); Fri, 13 Dec 2019 00:53:46 -0500 Received: from epcas1p3.samsung.com (unknown [182.195.41.47]) by mailout1.samsung.com (KnoxPortal) with ESMTP id 20191213055344epoutp016e43b36b3b3829d15e315c236879b212~f2HyZml2f0557205572epoutp014 for ; Fri, 13 Dec 2019 05:53:44 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 mailout1.samsung.com 20191213055344epoutp016e43b36b3b3829d15e315c236879b212~f2HyZml2f0557205572epoutp014 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=samsung.com; s=mail20170921; t=1576216424; bh=n/rDBhxWckXTuf5F7VnL+Givw9EJ/999mKdt8sYl/aM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=LQHi5QErhy4+T9Em/66ndLUoK4NhOAADlAasPc5JpCtDuQaYFulIf/Z75evwB36oc TNrmzO2tw37lSnDXWDQzU6DcgTKJEB3SZ4e1SWqnOfW7Cx3wIWLtUGjVN3wC8HJwEW KCWmZiEFrp/V/J2IA0u8vFyZqSCsuj4bPffLu1tw= Received: from epsnrtp1.localdomain (unknown [182.195.42.162]) by epcas1p1.samsung.com (KnoxPortal) with ESMTP id 20191213055343epcas1p19d19fd87c94fe788d323bdf979118245~f2HyEXdw60207402074epcas1p11; Fri, 13 Dec 2019 05:53:43 +0000 (GMT) Received: from epsmges1p1.samsung.com (unknown [182.195.40.164]) by epsnrtp1.localdomain (Postfix) with ESMTP id 47Z0GG5bMBzMqYlm; Fri, 13 Dec 2019 05:53:42 +0000 (GMT) Received: from epcas1p4.samsung.com ( [182.195.41.48]) by epsmges1p1.samsung.com (Symantec Messaging Gateway) with SMTP id AA.BA.57028.66723FD5; Fri, 13 Dec 2019 14:53:42 +0900 (KST) Received: from epsmtrp2.samsung.com (unknown [182.195.40.14]) by epcas1p2.samsung.com (KnoxPortal) with ESMTPA id 20191213055342epcas1p2e9f2971a4a7d392173e885b446b3e2c3~f2HwqPtJ62013320133epcas1p2m; Fri, 13 Dec 2019 05:53:42 +0000 (GMT) Received: from epsmgms1p2new.samsung.com (unknown [182.195.42.42]) by epsmtrp2.samsung.com (KnoxPortal) with ESMTP id 20191213055342epsmtrp2dfd7a548ae8f4b198179ae7663082a06~f2Hwpik8o0142901429epsmtrp2D; Fri, 13 Dec 2019 05:53:42 +0000 (GMT) X-AuditID: b6c32a35-50bff7000001dec4-12-5df327667d88 Received: from epsmtip2.samsung.com ( [182.195.34.31]) by epsmgms1p2new.samsung.com (Symantec Messaging Gateway) with SMTP id D9.92.06569.66723FD5; Fri, 13 Dec 2019 14:53:42 +0900 (KST) Received: from localhost.localdomain (unknown [10.88.103.87]) by epsmtip2.samsung.com (KnoxPortal) with ESMTPA id 20191213055342epsmtip2caa212c913027a98d529476e9ccb0286~f2HwfTiP61079710797epsmtip2K; Fri, 13 Dec 2019 05:53:42 +0000 (GMT) From: Namjae Jeon To: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org Cc: gregkh@linuxfoundation.org, valdis.kletnieks@vt.edu, hch@lst.de, sj1557.seo@samsung.com, Namjae Jeon Subject: [PATCH v7 08/13] exfat: add exfat cache Date: Fri, 13 Dec 2019 00:50:23 -0500 Message-Id: <20191213055028.5574-9-namjae.jeon@samsung.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20191213055028.5574-1-namjae.jeon@samsung.com> X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFnrCKsWRmVeSWpSXmKPExsWy7bCmgW6a+udYg84rmhbNi9ezWaxcfZTJ Ys/ekywWl3fNYbP4Mb3eYsu/I6wWl95/YHFg99g/dw27x+6bDWwefVtWMXp83iTncWj7G7YA 1qgcm4zUxJTUIoXUvOT8lMy8dFsl7+B453hTMwNDXUNLC3MlhbzE3FRbJRefAF23zBygK5QU yhJzSoFCAYnFxUr6djZF+aUlqQoZ+cUltkqpBSk5BYYGBXrFibnFpXnpesn5uVaGBgZGpkCV CTkZV85dZiw44Vbxa2pCA+Nbiy5GTg4JAROJpc83MYLYQgI7GCXWTTaGsD8xStx/HtLFyAVk f2OUuLlqNStMw4s1Z1ghEnsZJSZ9WcMC4QB1nFr2iqmLkYODTUBb4s8WUZAGEQF7ic2zD4DV MAu0MEosOP2DGSQhLGAg8f7DSVaQehYBVYmlPxNAwrwC1hLvj31jhFgmL7F6wwGwck4BG4l5 kz+wg8yREFjAJvHt9H82iCIXiWlNT6FsYYlXx7ewQ9hSEp/f7WUDmS8hUC3xcT8zRLiDUeLF d1sI21ji5voNYCcwC2hKrN+lDxFWlNj5ey7YCcwCfBLvvvawQkzhlehoE4IoUZXou3SYCcKW luhq/wC11ENi2dqn0PDsB4bbhKwJjHKzEBYsYGRcxSiWWlCcm55abFhgiBxZmxjBKUzLdAfj lHM+hxgFOBiVeHhXJH6KFWJNLCuuzD3EKMHBrCTCa18DFOJNSaysSi3Kjy8qzUktPsRoCgzG icxSosn5wPSaVxJvaGpkbGxsYWJmbmZqrCTOy/HjYqyQQHpiSWp2ampBahFMHxMHp1QDo4fm tTLtsxq3F+Zxr+/eeW7W235F7fmdBho8fkL7OLbm56Scfe7/tj+0e3Vo/s8i02fe82bcvSLv 6NInImxq5rLl69Kwp03zd3H/K1mzU+LI7ebagPvnV3/wWid6VPDzf54m5gfvAu7YqimWXrC/ x3yG+WjYYpEOp0WLl5aU6JxpS22eKK7UqMRSnJFoqMVcVJwIABPhhZZ3AwAA X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFlrALMWRmVeSWpSXmKPExsWy7bCSvG6a+udYgw1vhSyaF69ns1i5+iiT xZ69J1ksLu+aw2bxY3q9xZZ/R1gtLr3/wOLA7rF/7hp2j903G9g8+rasYvT4vEnO49D2N2wB rFFcNimpOZllqUX6dglcGVfOXWYsOOFW8WtqQgPjW4suRk4OCQETiRdrzrB2MXJxCAnsZpRY t3gaC0RCWuLYiTPMXYwcQLawxOHDxRA1HxglHk7/yg4SZxPQlvizRRSkXETAUaJ312EWkBpm gS5GiUdN35hBEsICBhLvP5xkBalnEVCVWPozASTMK2At8f7YN0aIVfISqzccACvnFLCRmDf5 AzuILQRUc/ftK7YJjHwLGBlWMUqmFhTnpucWGxYY5aWW6xUn5haX5qXrJefnbmIEh5uW1g7G EyfiDzEKcDAq8fCuSPwUK8SaWFZcmXuIUYKDWUmE174GKMSbklhZlVqUH19UmpNafIhRmoNF SZxXPv9YpJBAemJJanZqakFqEUyWiYNTqoGRo3DHtylzbDmDFcSkVx1R+LB0Ude0aU1K5Yyc 91ouh4dcEn379yuPl325xoozy55KS7CcYL3sZXyeP1j5P2+LpqLFhoftwZcFDx0o0dRR3mN4 OFV1WYr+7inBCet4mp/8CP+Ttq1FPddCafmle7f0yld9mvGYS/JI1MN0W/kPbJPqj3hMOJ6u xFKckWioxVxUnAgA0jFd4jMCAAA= X-CMS-MailID: 20191213055342epcas1p2e9f2971a4a7d392173e885b446b3e2c3 X-Msg-Generator: CA Content-Type: text/plain; charset="utf-8" X-Sendblock-Type: SVC_REQ_APPROVE CMS-TYPE: 101P DLP-Filter: Pass X-CFilter-Loop: Reflected X-CMS-RootMailID: 20191213055342epcas1p2e9f2971a4a7d392173e885b446b3e2c3 References: <20191213055028.5574-1-namjae.jeon@samsung.com> Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org This adds the implementation of exfat cache. Signed-off-by: Namjae Jeon Signed-off-by: Sungjong Seo --- fs/exfat/cache.c | 325 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 325 insertions(+) create mode 100644 fs/exfat/cache.c diff --git a/fs/exfat/cache.c b/fs/exfat/cache.c new file mode 100644 index 000000000000..03d0824fc368 --- /dev/null +++ b/fs/exfat/cache.c @@ -0,0 +1,325 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * linux/fs/fat/cache.c + * + * Written 1992,1993 by Werner Almesberger + * + * Mar 1999. AV. Changed cache, so that it uses the starting cluster instead + * of inode number. + * May 1999. AV. Fixed the bogosity with FAT32 (read "FAT28"). Fscking lusers. + * Copyright (C) 2012-2013 Samsung Electronics Co., Ltd. + */ + +#include +#include +#include + +#include "exfat_raw.h" +#include "exfat_fs.h" + +#define EXFAT_CACHE_VALID 0 +#define EXFAT_MAX_CACHE 16 + +struct exfat_cache { + struct list_head cache_list; + unsigned int nr_contig; /* number of contiguous clusters */ + unsigned int fcluster; /* cluster number in the file. */ + unsigned int dcluster; /* cluster number on disk. */ +}; + +struct exfat_cache_id { + unsigned int id; + unsigned int nr_contig; + unsigned int fcluster; + unsigned int dcluster; +}; + +static struct kmem_cache *exfat_cachep; + +static void exfat_cache_init_once(void *c) +{ + struct exfat_cache *cache = (struct exfat_cache *)c; + + INIT_LIST_HEAD(&cache->cache_list); +} + +int exfat_cache_init(void) +{ + exfat_cachep = kmem_cache_create("exfat_cache", + sizeof(struct exfat_cache), + 0, SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD, + exfat_cache_init_once); + if (!exfat_cachep) + return -ENOMEM; + return 0; +} + +void exfat_cache_shutdown(void) +{ + if (!exfat_cachep) + return; + kmem_cache_destroy(exfat_cachep); +} + +void exfat_cache_init_inode(struct inode *inode) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + + spin_lock_init(&ei->cache_lru_lock); + ei->nr_caches = 0; + ei->cache_valid_id = EXFAT_CACHE_VALID + 1; + INIT_LIST_HEAD(&ei->cache_lru); +} + +static inline struct exfat_cache *exfat_cache_alloc(void) +{ + return kmem_cache_alloc(exfat_cachep, GFP_NOFS); +} + +static inline void exfat_cache_free(struct exfat_cache *cache) +{ + WARN_ON(!list_empty(&cache->cache_list)); + kmem_cache_free(exfat_cachep, cache); +} + +static inline void exfat_cache_update_lru(struct inode *inode, + struct exfat_cache *cache) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + + if (ei->cache_lru.next != &cache->cache_list) + list_move(&cache->cache_list, &ei->cache_lru); +} + +static unsigned int exfat_cache_lookup(struct inode *inode, + unsigned int fclus, struct exfat_cache_id *cid, + unsigned int *cached_fclus, unsigned int *cached_dclus) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + static struct exfat_cache nohit = { .fcluster = 0, }; + struct exfat_cache *hit = &nohit, *p; + unsigned int offset = EXFAT_EOF_CLUSTER; + + spin_lock(&ei->cache_lru_lock); + list_for_each_entry(p, &ei->cache_lru, cache_list) { + /* Find the cache of "fclus" or nearest cache. */ + if (p->fcluster <= fclus && hit->fcluster < p->fcluster) { + hit = p; + if (hit->fcluster + hit->nr_contig < fclus) { + offset = hit->nr_contig; + } else { + offset = fclus - hit->fcluster; + break; + } + } + } + if (hit != &nohit) { + exfat_cache_update_lru(inode, hit); + + cid->id = ei->cache_valid_id; + cid->nr_contig = hit->nr_contig; + cid->fcluster = hit->fcluster; + cid->dcluster = hit->dcluster; + *cached_fclus = cid->fcluster + offset; + *cached_dclus = cid->dcluster + offset; + } + spin_unlock(&ei->cache_lru_lock); + + return offset; +} + +static struct exfat_cache *exfat_cache_merge(struct inode *inode, + struct exfat_cache_id *new) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + struct exfat_cache *p; + + list_for_each_entry(p, &ei->cache_lru, cache_list) { + /* Find the same part as "new" in cluster-chain. */ + if (p->fcluster == new->fcluster) { + if (new->nr_contig > p->nr_contig) + p->nr_contig = new->nr_contig; + return p; + } + } + return NULL; +} + +static void exfat_cache_add(struct inode *inode, + struct exfat_cache_id *new) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + struct exfat_cache *cache, *tmp; + + if (new->fcluster == EXFAT_EOF_CLUSTER) /* dummy cache */ + return; + + spin_lock(&ei->cache_lru_lock); + if (new->id != EXFAT_CACHE_VALID && + new->id != ei->cache_valid_id) + goto unlock; /* this cache was invalidated */ + + cache = exfat_cache_merge(inode, new); + if (cache == NULL) { + if (ei->nr_caches < EXFAT_MAX_CACHE) { + ei->nr_caches++; + spin_unlock(&ei->cache_lru_lock); + + tmp = exfat_cache_alloc(); + if (!tmp) { + spin_lock(&ei->cache_lru_lock); + ei->nr_caches--; + spin_unlock(&ei->cache_lru_lock); + return; + } + + spin_lock(&ei->cache_lru_lock); + cache = exfat_cache_merge(inode, new); + if (cache != NULL) { + ei->nr_caches--; + exfat_cache_free(tmp); + goto out_update_lru; + } + cache = tmp; + } else { + struct list_head *p = ei->cache_lru.prev; + + cache = list_entry(p, + struct exfat_cache, cache_list); + } + cache->fcluster = new->fcluster; + cache->dcluster = new->dcluster; + cache->nr_contig = new->nr_contig; + } +out_update_lru: + exfat_cache_update_lru(inode, cache); +unlock: + spin_unlock(&ei->cache_lru_lock); +} + +/* + * Cache invalidation occurs rarely, thus the LRU chain is not updated. It + * fixes itself after a while. + */ +static void __exfat_cache_inval_inode(struct inode *inode) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + struct exfat_cache *cache; + + while (!list_empty(&ei->cache_lru)) { + cache = list_entry(ei->cache_lru.next, + struct exfat_cache, cache_list); + list_del_init(&cache->cache_list); + ei->nr_caches--; + exfat_cache_free(cache); + } + /* Update. The copy of caches before this id is discarded. */ + ei->cache_valid_id++; + if (ei->cache_valid_id == EXFAT_CACHE_VALID) + ei->cache_valid_id++; +} + +void exfat_cache_inval_inode(struct inode *inode) +{ + struct exfat_inode_info *ei = EXFAT_I(inode); + + spin_lock(&ei->cache_lru_lock); + __exfat_cache_inval_inode(inode); + spin_unlock(&ei->cache_lru_lock); +} + +static inline int cache_contiguous(struct exfat_cache_id *cid, + unsigned int dclus) +{ + cid->nr_contig++; + return cid->dcluster + cid->nr_contig == dclus; +} + +static inline void cache_init(struct exfat_cache_id *cid, + unsigned int fclus, unsigned int dclus) +{ + cid->id = EXFAT_CACHE_VALID; + cid->fcluster = fclus; + cid->dcluster = dclus; + cid->nr_contig = 0; +} + +int exfat_get_cluster(struct inode *inode, unsigned int cluster, + unsigned int *fclus, unsigned int *dclus, + unsigned int *last_dclus, int allow_eof) +{ + struct super_block *sb = inode->i_sb; + struct exfat_sb_info *sbi = EXFAT_SB(sb); + unsigned int limit = sbi->num_clusters; + struct exfat_inode_info *ei = EXFAT_I(inode); + struct exfat_cache_id cid; + unsigned int content; + + if (ei->start_clu == EXFAT_FREE_CLUSTER) { + exfat_fs_error(sb, + "invalid access to exfat cache (entry 0x%08x)", + ei->start_clu); + return -EIO; + } + + *fclus = 0; + *dclus = ei->start_clu; + *last_dclus = *dclus; + + /* + * Don`t use exfat_cache if zero offset or non-cluster allocation + */ + if (cluster == 0 || *dclus == EXFAT_EOF_CLUSTER) + return 0; + + cache_init(&cid, EXFAT_EOF_CLUSTER, EXFAT_EOF_CLUSTER); + + if (exfat_cache_lookup(inode, cluster, &cid, fclus, dclus) == + EXFAT_EOF_CLUSTER) { + /* + * dummy, always not contiguous + * This is reinitialized by cache_init(), later. + */ + WARN_ON(cid.id != EXFAT_CACHE_VALID || + cid.fcluster != EXFAT_EOF_CLUSTER || + cid.dcluster != EXFAT_EOF_CLUSTER || + cid.nr_contig != 0); + } + + if (*fclus == cluster) + return 0; + + while (*fclus < cluster) { + /* prevent the infinite loop of cluster chain */ + if (*fclus > limit) { + exfat_fs_error(sb, + "detected the cluster chain loop (i_pos %u)", + (*fclus)); + return -EIO; + } + + if (exfat_ent_get(sb, *dclus, &content)) + return -EIO; + + *last_dclus = *dclus; + *dclus = content; + (*fclus)++; + + if (content == EXFAT_EOF_CLUSTER) { + if (!allow_eof) { + exfat_fs_error(sb, + "invalid cluster chain (i_pos %u, last_clus 0x%08x is EOF)", + *fclus, (*last_dclus)); + return -EIO; + } + + break; + } + + if (!cache_contiguous(&cid, *dclus)) + cache_init(&cid, *fclus, *dclus); + } + + exfat_cache_add(inode, &cid); + return 0; +} -- 2.17.1