From patchwork Wed Aug 31 02:49:42 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960272 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A5FF5ECAAA1 for ; Wed, 31 Aug 2022 02:50:05 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229819AbiHaCuE (ORCPT ); Tue, 30 Aug 2022 22:50:04 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47142 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229846AbiHaCuD (ORCPT ); Tue, 30 Aug 2022 22:50:03 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3975C26103 for ; Tue, 30 Aug 2022 19:50:02 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914200; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=D5eAGpedc53fq+szSrlzLLzrJyC7UV2N4pAN+4VsjY8=; b=W8vhPdL3gEW3YaW3RHa29L1QvaTjWo1OX9fXRxA8XHtNKTJIAzcyuse6VSxALNfYeoXZc1 Fn0nXOo2VQwBKydQ7Ji59uiIH7wBt1fAVue1+yqMhjqizwu1+VaO6CgCzZ1+wc/4d0/uvJ ODfQUANU5geDwZ7qmWGNytYdXw+8jrg= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-605-8T0vk5YsP8uBjUSFQJzl8w-1; Tue, 30 Aug 2022 22:49:59 -0400 X-MC-Unique: 8T0vk5YsP8uBjUSFQJzl8w-1 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.rdu2.redhat.com [10.11.54.4]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 2941B8037AA; Wed, 31 Aug 2022 02:49:59 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id 2280C2026D4C; Wed, 31 Aug 2022 02:49:57 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 1/6] cifs: Make tcon contain a wrapper structure cached_fids instead of cached_fid Date: Wed, 31 Aug 2022 12:49:42 +1000 Message-Id: <20220831024947.3917507-2-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.78 on 10.11.54.4 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org This wrapper structure will later be expanded to contain a list of fids that are cached and not just the root fid. Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 50 ++++++++++++++++++++++++-------------------- fs/cifs/cached_dir.h | 8 +++++-- fs/cifs/cifsglob.h | 2 +- fs/cifs/misc.c | 6 +++--- 4 files changed, 37 insertions(+), 29 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index b401339f6e73..c2f5b71a3c9f 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -52,7 +52,7 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, dentry = cifs_sb->root; - cfid = tcon->cfid; + cfid = &tcon->cfids->cfid; mutex_lock(&cfid->fid_mutex); if (cfid->is_valid) { cifs_dbg(FYI, "found a cached root file handle\n"); @@ -226,7 +226,7 @@ int open_cached_dir_by_dentry(struct cifs_tcon *tcon, { struct cached_fid *cfid; - cfid = tcon->cfid; + cfid = &tcon->cfids->cfid; mutex_lock(&cfid->fid_mutex); if (cfid->dentry == dentry) { @@ -320,7 +320,7 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) tcon = tlink_tcon(tlink); if (IS_ERR(tcon)) continue; - cfid = tcon->cfid; + cfid = &tcon->cfids->cfid; mutex_lock(&cfid->fid_mutex); if (cfid->dentry) { dput(cfid->dentry); @@ -336,12 +336,14 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) */ void invalidate_all_cached_dirs(struct cifs_tcon *tcon) { - mutex_lock(&tcon->cfid->fid_mutex); - tcon->cfid->is_valid = false; + struct cached_fid *cfid = &tcon->cfids->cfid; + + mutex_lock(&cfid->fid_mutex); + cfid->is_valid = false; /* cached handle is not valid, so SMB2_CLOSE won't be sent below */ - close_cached_dir_lease_locked(tcon->cfid); - memset(&tcon->cfid->fid, 0, sizeof(struct cifs_fid)); - mutex_unlock(&tcon->cfid->fid_mutex); + close_cached_dir_lease_locked(cfid); + memset(&cfid->fid, 0, sizeof(struct cifs_fid)); + mutex_unlock(&cfid->fid_mutex); } static void @@ -355,34 +357,36 @@ smb2_cached_lease_break(struct work_struct *work) int cached_dir_lease_break(struct cifs_tcon *tcon, __u8 lease_key[16]) { - if (tcon->cfid->is_valid && + struct cached_fid *cfid = &tcon->cfids->cfid; + + if (cfid->is_valid && !memcmp(lease_key, - tcon->cfid->fid.lease_key, + cfid->fid.lease_key, SMB2_LEASE_KEY_SIZE)) { - tcon->cfid->time = 0; - INIT_WORK(&tcon->cfid->lease_break, + cfid->time = 0; + INIT_WORK(&cfid->lease_break, smb2_cached_lease_break); queue_work(cifsiod_wq, - &tcon->cfid->lease_break); + &cfid->lease_break); return true; } return false; } -struct cached_fid *init_cached_dir(void) +struct cached_fids *init_cached_dirs(void) { - struct cached_fid *cfid; + struct cached_fids *cfids; - cfid = kzalloc(sizeof(*cfid), GFP_KERNEL); - if (!cfid) + cfids = kzalloc(sizeof(*cfids), GFP_KERNEL); + if (!cfids) return NULL; - INIT_LIST_HEAD(&cfid->dirents.entries); - mutex_init(&cfid->dirents.de_mutex); - mutex_init(&cfid->fid_mutex); - return cfid; + INIT_LIST_HEAD(&cfids->cfid.dirents.entries); + mutex_init(&cfids->cfid.dirents.de_mutex); + mutex_init(&cfids->cfid.fid_mutex); + return cfids; } -void free_cached_dir(struct cifs_tcon *tcon) +void free_cached_dirs(struct cached_fids *cfids) { - kfree(tcon->cfid); + kfree(cfids); } diff --git a/fs/cifs/cached_dir.h b/fs/cifs/cached_dir.h index bd262dc8b179..e430e1102296 100644 --- a/fs/cifs/cached_dir.h +++ b/fs/cifs/cached_dir.h @@ -45,8 +45,12 @@ struct cached_fid { struct cached_dirents dirents; }; -extern struct cached_fid *init_cached_dir(void); -extern void free_cached_dir(struct cifs_tcon *tcon); +struct cached_fids { + struct cached_fid cfid; +}; + +extern struct cached_fids *init_cached_dirs(void); +extern void free_cached_dirs(struct cached_fids *cfids); extern int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, const char *path, struct cifs_sb_info *cifs_sb, diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index f15d7b0c123d..768003e52cc9 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -1221,7 +1221,7 @@ struct cifs_tcon { struct fscache_volume *fscache; /* cookie for share */ #endif struct list_head pending_opens; /* list of incomplete opens */ - struct cached_fid *cfid; /* Cached root fid */ + struct cached_fids *cfids; /* BB add field for back pointer to sb struct(s)? */ #ifdef CONFIG_CIFS_DFS_UPCALL struct list_head ulist; /* cache update list */ diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 87f60f736731..5cb7e1ae00a4 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -117,8 +117,8 @@ tconInfoAlloc(void) ret_buf = kzalloc(sizeof(*ret_buf), GFP_KERNEL); if (!ret_buf) return NULL; - ret_buf->cfid = init_cached_dir(); - if (!ret_buf->cfid) { + ret_buf->cfids = init_cached_dirs(); + if (!ret_buf->cfids) { kfree(ret_buf); return NULL; } @@ -144,7 +144,7 @@ tconInfoFree(struct cifs_tcon *tcon) cifs_dbg(FYI, "Null buffer passed to tconInfoFree\n"); return; } - free_cached_dir(tcon); + free_cached_dirs(tcon->cfids); atomic_dec(&tconInfoAllocCount); kfree(tcon->nativeFileSystem); kfree_sensitive(tcon->password); From patchwork Wed Aug 31 02:49:43 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960273 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 04036ECAAA1 for ; Wed, 31 Aug 2022 02:50:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229846AbiHaCuI (ORCPT ); Tue, 30 Aug 2022 22:50:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47362 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230270AbiHaCuH (ORCPT ); Tue, 30 Aug 2022 22:50:07 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3C95B9FD2 for ; Tue, 30 Aug 2022 19:50:04 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914203; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=7z5RpT9TnzSr0iLUMVbbvF/oqT4/optj5m72it3bt+A=; b=ZygRmuNd0ff5qgyBCg3m0POXqLO1wAKEWpiQznQpQuQUn+wG2cVYkLx7a8jak+KpiDvpcm rMw6RODY9IJ6Ll4jOoR43J1HqblTsbmlrp4iTOoRf1IaexM9plGrgXxloIc2E+jfrF1ART uVEzSG3KT9AToLcmldwkb5RgeepAwZ4= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-434-5henQJOrPyS3rgov-ej8kg-1; Tue, 30 Aug 2022 22:50:01 -0400 X-MC-Unique: 5henQJOrPyS3rgov-ej8kg-1 Received: from smtp.corp.redhat.com (int-mx01.intmail.prod.int.rdu2.redhat.com [10.11.54.1]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 7B45A8037AA; Wed, 31 Aug 2022 02:50:01 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id DE9EA400E88E; Wed, 31 Aug 2022 02:50:00 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 2/6] cifs: cifs: handlecache, only track the dentry for the root handle Date: Wed, 31 Aug 2022 12:49:43 +1000 Message-Id: <20220831024947.3917507-3-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.11.54.1 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index c2f5b71a3c9f..4f3d88f3351b 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -47,11 +47,11 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, if (cifs_sb->root == NULL) return -ENOENT; - if (strlen(path)) + if (!path[0]) + dentry = cifs_sb->root; + else return -ENOENT; - dentry = cifs_sb->root; - cfid = &tcon->cfids->cfid; mutex_lock(&cfid->fid_mutex); if (cfid->is_valid) { @@ -177,7 +177,8 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, cfid->tcon = tcon; cfid->is_valid = true; cfid->dentry = dentry; - dget(dentry); + if (dentry) + dget(dentry); kref_init(&cfid->refcount); /* BB TBD check to see if oplock level check can be removed below */ From patchwork Wed Aug 31 02:49:44 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960277 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id BEDDFECAAA1 for ; Wed, 31 Aug 2022 02:50:30 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229601AbiHaCu3 (ORCPT ); Tue, 30 Aug 2022 22:50:29 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:49134 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231496AbiHaCu2 (ORCPT ); Tue, 30 Aug 2022 22:50:28 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DCD0B7B1FD for ; Tue, 30 Aug 2022 19:50:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914226; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=eKaTUYsQM2h5VssGAi4SRp8bg725rn2ZamEr4+OK8zQ=; b=MKcJj7TI5vAIm7vzn4f6Z0YakLYBuJagNLGQq3voHaJMrHkTdDdrKFY7zsB8kKUSbc8snO UQW5KYfUm90WvVv3psk7iCc/LNZ4riCKA7QyleGND+5Yj5VbaKZqNUdcqqUHNYz5yuHMiD momAryOOPC+AZpz+hyMZPZGkOd1a5+4= Received: from mimecast-mx02.redhat.com (mx3-rdu2.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-393-q2mL9GvtPlCBymv7Y5OFuw-1; Tue, 30 Aug 2022 22:50:17 -0400 X-MC-Unique: q2mL9GvtPlCBymv7Y5OFuw-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id D882A3C01DF3; Wed, 31 Aug 2022 02:50:16 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id D1349909FF; Wed, 31 Aug 2022 02:50:15 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 3/6] cifs: store a pointer to a fid in the cfid structure instead of the struct Date: Wed, 31 Aug 2022 12:49:44 +1000 Message-Id: <20220831024947.3917507-4-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.11.54.5 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org also create a constructor that takes a path name and stores it in the fid. Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 63 ++++++++++++++++++++++++++++++++++++++------ fs/cifs/cached_dir.h | 4 ++- 2 files changed, 58 insertions(+), 9 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index 4f3d88f3351b..0036ba6e0ab0 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -11,6 +11,8 @@ #include "smb2proto.h" #include "cached_dir.h" +struct cached_fid *init_cached_dir(const char *path); + /* * Open the and cache a directory handle. * If error then *cfid is not initialized. @@ -52,7 +54,14 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, else return -ENOENT; - cfid = &tcon->cfids->cfid; + cfid = tcon->cfids->cfid; + if (cfid == NULL) { + cfid = init_cached_dir(path); + tcon->cfids->cfid = cfid; + } + if (cfid == NULL) + return -ENOMEM; + mutex_lock(&cfid->fid_mutex); if (cfid->is_valid) { cifs_dbg(FYI, "found a cached root file handle\n"); @@ -227,7 +236,9 @@ int open_cached_dir_by_dentry(struct cifs_tcon *tcon, { struct cached_fid *cfid; - cfid = &tcon->cfids->cfid; + cfid = tcon->cfids->cfid; + if (cfid == NULL) + return -ENOENT; mutex_lock(&cfid->fid_mutex); if (cfid->dentry == dentry) { @@ -321,7 +332,9 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) tcon = tlink_tcon(tlink); if (IS_ERR(tcon)) continue; - cfid = &tcon->cfids->cfid; + cfid = tcon->cfids->cfid; + if (cfid == NULL) + continue; mutex_lock(&cfid->fid_mutex); if (cfid->dentry) { dput(cfid->dentry); @@ -337,7 +350,10 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) */ void invalidate_all_cached_dirs(struct cifs_tcon *tcon) { - struct cached_fid *cfid = &tcon->cfids->cfid; + struct cached_fid *cfid = tcon->cfids->cfid; + + if (cfid == NULL) + return; mutex_lock(&cfid->fid_mutex); cfid->is_valid = false; @@ -358,7 +374,10 @@ smb2_cached_lease_break(struct work_struct *work) int cached_dir_lease_break(struct cifs_tcon *tcon, __u8 lease_key[16]) { - struct cached_fid *cfid = &tcon->cfids->cfid; + struct cached_fid *cfid = tcon->cfids->cfid; + + if (cfid == NULL) + return false; if (cfid->is_valid && !memcmp(lease_key, @@ -374,6 +393,32 @@ int cached_dir_lease_break(struct cifs_tcon *tcon, __u8 lease_key[16]) return false; } +struct cached_fid *init_cached_dir(const char *path) +{ + struct cached_fid *cfid; + + cfid = kzalloc(sizeof(*cfid), GFP_KERNEL); + if (!cfid) + return NULL; + cfid->path = kstrdup(path, GFP_KERNEL); + if (!cfid->path) { + kfree(cfid); + return NULL; + } + + INIT_LIST_HEAD(&cfid->dirents.entries); + mutex_init(&cfid->dirents.de_mutex); + mutex_init(&cfid->fid_mutex); + return cfid; +} + +void free_cached_dir(struct cached_fid *cfid) +{ + kfree(cfid->path); + cfid->path = NULL; + kfree(cfid); +} + struct cached_fids *init_cached_dirs(void) { struct cached_fids *cfids; @@ -381,13 +426,15 @@ struct cached_fids *init_cached_dirs(void) cfids = kzalloc(sizeof(*cfids), GFP_KERNEL); if (!cfids) return NULL; - INIT_LIST_HEAD(&cfids->cfid.dirents.entries); - mutex_init(&cfids->cfid.dirents.de_mutex); - mutex_init(&cfids->cfid.fid_mutex); + mutex_init(&cfids->cfid_list_mutex); return cfids; } void free_cached_dirs(struct cached_fids *cfids) { + if (cfids->cfid) { + free_cached_dir(cfids->cfid); + cfids->cfid = NULL; + } kfree(cfids); } diff --git a/fs/cifs/cached_dir.h b/fs/cifs/cached_dir.h index e430e1102296..bdf6c3866653 100644 --- a/fs/cifs/cached_dir.h +++ b/fs/cifs/cached_dir.h @@ -31,6 +31,7 @@ struct cached_dirents { }; struct cached_fid { + const char *path; bool is_valid:1; /* Do we have a useable root fid */ bool file_all_info_is_valid:1; bool has_lease:1; @@ -46,7 +47,8 @@ struct cached_fid { }; struct cached_fids { - struct cached_fid cfid; + struct mutex cfid_list_mutex; + struct cached_fid *cfid; }; extern struct cached_fids *init_cached_dirs(void); From patchwork Wed Aug 31 02:49:45 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960276 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 6EFC9ECAAA1 for ; Wed, 31 Aug 2022 02:50:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231161AbiHaCuT (ORCPT ); Tue, 30 Aug 2022 22:50:19 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48392 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231376AbiHaCuS (ORCPT ); Tue, 30 Aug 2022 22:50:18 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3397461D65 for ; Tue, 30 Aug 2022 19:50:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914215; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=68rnSs5rg1oEaq6J6eYw6omAPwWeZwDUkokIJKKbP90=; b=A6ZOYIdKsJv5kPIZeiGN9kqTX2Namy0WpeqfTnLLCMu3wF5VNWdx1d+ofas4yEdMkb+K+l nhctcr0uIzgxIA9X7p8NwfDfrJHQjIhCqtihKk4U9XGsu6/ZEIURuSEKgZegWtwaDZSZ4R Hk5c2DT7jH9vQKWiM47pj1/sDo5qdFM= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-47-Rgfo6QXtO2CbPvTRyTCa4g-1; Tue, 30 Aug 2022 22:50:13 -0400 X-MC-Unique: Rgfo6QXtO2CbPvTRyTCa4g-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 8278C185A7B2; Wed, 31 Aug 2022 02:50:13 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id 2C4329458A; Wed, 31 Aug 2022 02:50:11 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 4/6] cifs: enable caching of directories for which a lease is held Date: Wed, 31 Aug 2022 12:49:45 +1000 Message-Id: <20220831024947.3917507-5-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.11.54.5 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org This expands the directory caching to now cache an open handle for all directories (up to a maximum) and not just the root directory. In this patch, locking and refcounting is intended to work as so: The main function to get a reference to a cached handle is find_or_create_cached_dir() called from open_cached_dir() These functions are protected under the cfid_list_lock spin-lock to make sure we do not race creating new references for cached dirs with deletion of expired ones. An successful open_cached_dir() will take out 2 references to the cfid if this was the very first and successful call to open the directory and it acquired a lease from the server. One reference is for the lease and the other is for the cfid that we return. The is lease reference is tracked by cfid->has_lease. If the directory already has a handle with an active lease, then we just take out one new reference for the cfid and return it. It can happen that we have a thread that tries to open a cached directory where we have a cfid already but we do not, yet, have a working lease. In this case we will just return NULL, and this the caller will fall back to the case when no handle was available. In this model the total number of references we have on a cfid is 1 for while the handle is open and we have a lease, and one additional reference for each open instance of a cfid. Once we get a lease break (cached_dir_lease_break()) we remove the cfid from the list under the spinlock. This prevents any new threads to use it, and we also call smb2_cached_lease_break() via the work_queue in order to drop the reference we got for the lease (we drop it outside of the spin-lock.) Anytime a thread calls close_cached_dir() we also drop a reference to the cfid. When the last reference to the cfid is released smb2_close_cached_fid() will be invoked which will drop the reference ot the dentry we held for this cfid and it will also, if we the handle is open/has a lease also call SMB2_close() to close the handle on the server. Two events require special handling: invalidate_all_cached_dirs() this function is called from SMB2_tdis() and cifs_mark_open_files_invalid(). In both cases the tcon is either gone already or will be shortly so we do not need to actually close the handles. They will be dropped server side as part of the tcon dropping. But we have to be careful about a potential race with a concurrent lease break so we need to take out additional refences to avoid the cfid from being freed while we are still referencing it. free_cached_dirs() which is called from tconInfoFree(). This is called quite late in the umount process so there should no longer be any open handles or files and we can just free all the remaining data. Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 437 +++++++++++++++++++++++++------------------ fs/cifs/cached_dir.h | 20 +- fs/cifs/inode.c | 6 +- fs/cifs/smb2ops.c | 2 +- 4 files changed, 267 insertions(+), 198 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index 0036ba6e0ab0..ce6935c3fe06 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -11,7 +11,53 @@ #include "smb2proto.h" #include "cached_dir.h" -struct cached_fid *init_cached_dir(const char *path); +static struct cached_fid *init_cached_dir(const char *path); +static void free_cached_dir(struct cached_fid *cfid); + +static struct cached_fid *find_or_create_cached_dir(struct cached_fids *cfids, + const char *path, + bool lookup_only) +{ + struct cached_fid *cfid; + + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry(cfid, &cfids->entries, entry) { + if (!strcmp(cfid->path, path)) { + /* + * If it doesn't have a lease it is either not yet + * fully cached or it may be in the process of + * being deleted due to a lease break. + */ + if (!cfid->has_lease) { + spin_unlock(&cfids->cfid_list_lock); + return NULL; + } + kref_get(&cfid->refcount); + spin_unlock(&cfids->cfid_list_lock); + return cfid; + } + } + if (lookup_only) { + spin_unlock(&cfids->cfid_list_lock); + return NULL; + } + if (cfids->num_entries >= MAX_CACHED_FIDS) { + spin_unlock(&cfids->cfid_list_lock); + return NULL; + } + cfid = init_cached_dir(path); + if (cfid == NULL) { + spin_unlock(&cfids->cfid_list_lock); + return NULL; + } + cfid->cfids = cfids; + cfids->num_entries++; + list_add(&cfid->entry, &cfids->entries); + cfid->on_list = true; + kref_get(&cfid->refcount); + spin_unlock(&cfids->cfid_list_lock); + return cfid; +} /* * Open the and cache a directory handle. @@ -33,61 +79,65 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, struct kvec open_iov[SMB2_CREATE_IOV_SIZE]; struct kvec qi_iov[1]; int rc, flags = 0; - __le16 utf16_path = 0; /* Null - since an open of top of share */ + __le16 *utf16_path = NULL; u8 oplock = SMB2_OPLOCK_LEVEL_II; struct cifs_fid *pfid; - struct dentry *dentry; + struct dentry *dentry = NULL; struct cached_fid *cfid; + struct cached_fids *cfids; + - if (tcon == NULL || tcon->nohandlecache || + if (tcon == NULL || tcon->cfids == NULL || tcon->nohandlecache || is_smb1_server(tcon->ses->server)) - return -EOPNOTSUPP; + return -ENOTSUPP; ses = tcon->ses; server = ses->server; + cfids = tcon->cfids; + if (!server->ops->new_lease_key) + return -EIO; + if (cifs_sb->root == NULL) return -ENOENT; + /* + * TODO: for better caching we need to find and use the dentry also + * for non-root directories. + */ if (!path[0]) dentry = cifs_sb->root; - else - return -ENOENT; - cfid = tcon->cfids->cfid; - if (cfid == NULL) { - cfid = init_cached_dir(path); - tcon->cfids->cfid = cfid; - } - if (cfid == NULL) + utf16_path = cifs_convert_path_to_utf16(path, cifs_sb); + if (!utf16_path) return -ENOMEM; - mutex_lock(&cfid->fid_mutex); - if (cfid->is_valid) { - cifs_dbg(FYI, "found a cached root file handle\n"); + cfid = find_or_create_cached_dir(cfids, path, lookup_only); + if (cfid == NULL) { + kfree(utf16_path); + return -ENOENT; + } + /* + * At this point we either have a lease already and we can just + * return it. If not we are guaranteed to be the only thread accessing + * this cfid. + */ + if (cfid->has_lease) { *ret_cfid = cfid; - kref_get(&cfid->refcount); - mutex_unlock(&cfid->fid_mutex); + kfree(utf16_path); return 0; } /* * We do not hold the lock for the open because in case - * SMB2_open needs to reconnect, it will end up calling - * cifs_mark_open_files_invalid() which takes the lock again - * thus causing a deadlock + * SMB2_open needs to reconnect. + * This is safe because no other thread will be able to get a ref + * to the cfid until we have finished opening the file and (possibly) + * aquired a lease. */ - mutex_unlock(&cfid->fid_mutex); - - if (lookup_only) - return -ENOENT; - if (smb3_encryption_required(tcon)) flags |= CIFS_TRANSFORM_REQ; - if (!server->ops->new_lease_key) - return -EIO; - pfid = &cfid->fid; server->ops->new_lease_key(pfid); @@ -108,7 +158,7 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, oparms.reconnect = false; rc = SMB2_open_init(tcon, server, - &rqst[0], &oplock, &oparms, &utf16_path); + &rqst[0], &oplock, &oparms, utf16_path); if (rc) goto oshr_free; smb2_set_next_command(tcon, &rqst[0]); @@ -131,47 +181,13 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, rc = compound_send_recv(xid, ses, server, flags, 2, rqst, resp_buftype, rsp_iov); - mutex_lock(&cfid->fid_mutex); - - /* - * Now we need to check again as the cached root might have - * been successfully re-opened from a concurrent process - */ - - if (cfid->is_valid) { - /* work was already done */ - - /* stash fids for close() later */ - struct cifs_fid fid = { - .persistent_fid = pfid->persistent_fid, - .volatile_fid = pfid->volatile_fid, - }; - - /* - * caller expects this func to set the fid in cfid to valid - * cached root, so increment the refcount. - */ - kref_get(&cfid->refcount); - - mutex_unlock(&cfid->fid_mutex); - - if (rc == 0) { - /* close extra handle outside of crit sec */ - SMB2_close(xid, tcon, fid.persistent_fid, fid.volatile_fid); - } - rc = 0; - goto oshr_free; - } - - /* Cached root is still invalid, continue normaly */ - if (rc) { if (rc == -EREMCHG) { tcon->need_reconnect = true; pr_warn_once("server share %s deleted\n", tcon->treeName); } - goto oshr_exit; + goto oshr_free; } atomic_inc(&tcon->num_remote_opens); @@ -184,49 +200,56 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, #endif /* CIFS_DEBUG2 */ cfid->tcon = tcon; - cfid->is_valid = true; - cfid->dentry = dentry; - if (dentry) + if (dentry) { + cfid->dentry = dentry; dget(dentry); - kref_init(&cfid->refcount); - + } /* BB TBD check to see if oplock level check can be removed below */ - if (o_rsp->OplockLevel == SMB2_OPLOCK_LEVEL_LEASE) { - /* - * See commit 2f94a3125b87. Increment the refcount when we - * get a lease for root, release it if lease break occurs - */ - kref_get(&cfid->refcount); - cfid->has_lease = true; - smb2_parse_contexts(server, o_rsp, - &oparms.fid->epoch, - oparms.fid->lease_key, &oplock, - NULL, NULL); - } else - goto oshr_exit; + if (o_rsp->OplockLevel != SMB2_OPLOCK_LEVEL_LEASE) { + goto oshr_free; + } + smb2_parse_contexts(server, o_rsp, + &oparms.fid->epoch, + oparms.fid->lease_key, &oplock, + NULL, NULL); + qi_rsp = (struct smb2_query_info_rsp *)rsp_iov[1].iov_base; if (le32_to_cpu(qi_rsp->OutputBufferLength) < sizeof(struct smb2_file_all_info)) - goto oshr_exit; + goto oshr_free; if (!smb2_validate_and_copy_iov( le16_to_cpu(qi_rsp->OutputBufferOffset), sizeof(struct smb2_file_all_info), &rsp_iov[1], sizeof(struct smb2_file_all_info), (char *)&cfid->file_all_info)) cfid->file_all_info_is_valid = true; - cfid->time = jiffies; + cfid->is_open = true; + cfid->has_lease = true; -oshr_exit: - mutex_unlock(&cfid->fid_mutex); oshr_free: + kfree(utf16_path); SMB2_open_free(&rqst[0]); SMB2_query_info_free(&rqst[1]); free_rsp_buf(resp_buftype[0], rsp_iov[0].iov_base); free_rsp_buf(resp_buftype[1], rsp_iov[1].iov_base); - if (rc == 0) - *ret_cfid = cfid; - + spin_lock(&cfids->cfid_list_lock); + if (!cfid->has_lease) { + if (cfid->on_list) { + list_del(&cfid->entry); + cfid->on_list = false; + cfids->num_entries--; + } + rc = -ENOENT; + } + spin_unlock(&cfids->cfid_list_lock); + if (rc) { + free_cached_dir(cfid); + cfid = NULL; + } + if (rc == 0) { + *ret_cfid = cfid; + } return rc; } @@ -235,20 +258,22 @@ int open_cached_dir_by_dentry(struct cifs_tcon *tcon, struct cached_fid **ret_cfid) { struct cached_fid *cfid; + struct cached_fids *cfids = tcon->cfids; - cfid = tcon->cfids->cfid; - if (cfid == NULL) + if (cfids == NULL) return -ENOENT; - mutex_lock(&cfid->fid_mutex); - if (cfid->dentry == dentry) { - cifs_dbg(FYI, "found a cached root file handle by dentry\n"); - *ret_cfid = cfid; - kref_get(&cfid->refcount); - mutex_unlock(&cfid->fid_mutex); - return 0; + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry(cfid, &cfids->entries, entry) { + if (dentry && cfid->dentry == dentry) { + cifs_dbg(FYI, "found a cached root file handle by dentry\n"); + kref_get(&cfid->refcount); + *ret_cfid = cfid; + spin_unlock(&cfids->cfid_list_lock); + return 0; + } } - mutex_unlock(&cfid->fid_mutex); + spin_unlock(&cfids->cfid_list_lock); return -ENOENT; } @@ -257,63 +282,29 @@ smb2_close_cached_fid(struct kref *ref) { struct cached_fid *cfid = container_of(ref, struct cached_fid, refcount); - struct cached_dirent *dirent, *q; - if (cfid->is_valid) { - cifs_dbg(FYI, "clear cached root file handle\n"); - SMB2_close(0, cfid->tcon, cfid->fid.persistent_fid, - cfid->fid.volatile_fid); + spin_lock(&cfid->cfids->cfid_list_lock); + if (cfid->on_list) { + list_del(&cfid->entry); + cfid->on_list = false; + cfid->cfids->num_entries--; } + spin_unlock(&cfid->cfids->cfid_list_lock); - /* - * We only check validity above to send SMB2_close, - * but we still need to invalidate these entries - * when this function is called - */ - cfid->is_valid = false; - cfid->file_all_info_is_valid = false; - cfid->has_lease = false; - if (cfid->dentry) { - dput(cfid->dentry); - cfid->dentry = NULL; - } - /* - * Delete all cached dirent names - */ - mutex_lock(&cfid->dirents.de_mutex); - list_for_each_entry_safe(dirent, q, &cfid->dirents.entries, entry) { - list_del(&dirent->entry); - kfree(dirent->name); - kfree(dirent); + dput(cfid->dentry); + cfid->dentry = NULL; + + if (cfid->is_open) { + SMB2_close(0, cfid->tcon, cfid->fid.persistent_fid, + cfid->fid.volatile_fid); } - cfid->dirents.is_valid = 0; - cfid->dirents.is_failed = 0; - cfid->dirents.ctx = NULL; - cfid->dirents.pos = 0; - mutex_unlock(&cfid->dirents.de_mutex); + free_cached_dir(cfid); } void close_cached_dir(struct cached_fid *cfid) { - mutex_lock(&cfid->fid_mutex); kref_put(&cfid->refcount, smb2_close_cached_fid); - mutex_unlock(&cfid->fid_mutex); -} - -void close_cached_dir_lease_locked(struct cached_fid *cfid) -{ - if (cfid->has_lease) { - cfid->has_lease = false; - kref_put(&cfid->refcount, smb2_close_cached_fid); - } -} - -void close_cached_dir_lease(struct cached_fid *cfid) -{ - mutex_lock(&cfid->fid_mutex); - close_cached_dir_lease_locked(cfid); - mutex_unlock(&cfid->fid_mutex); } /* @@ -326,41 +317,62 @@ void close_all_cached_dirs(struct cifs_sb_info *cifs_sb) struct cached_fid *cfid; struct cifs_tcon *tcon; struct tcon_link *tlink; + struct cached_fids *cfids; for (node = rb_first(root); node; node = rb_next(node)) { tlink = rb_entry(node, struct tcon_link, tl_rbnode); tcon = tlink_tcon(tlink); if (IS_ERR(tcon)) continue; - cfid = tcon->cfids->cfid; - if (cfid == NULL) + cfids = tcon->cfids; + if (cfids == NULL) continue; - mutex_lock(&cfid->fid_mutex); - if (cfid->dentry) { + list_for_each_entry(cfid, &cfids->entries, entry) { dput(cfid->dentry); cfid->dentry = NULL; } - mutex_unlock(&cfid->fid_mutex); } } /* - * Invalidate and close all cached dirs when a TCON has been reset + * Invalidate all cached dirs when a TCON has been reset * due to a session loss. */ void invalidate_all_cached_dirs(struct cifs_tcon *tcon) { - struct cached_fid *cfid = tcon->cfids->cfid; - - if (cfid == NULL) - return; - - mutex_lock(&cfid->fid_mutex); - cfid->is_valid = false; - /* cached handle is not valid, so SMB2_CLOSE won't be sent below */ - close_cached_dir_lease_locked(cfid); - memset(&cfid->fid, 0, sizeof(struct cifs_fid)); - mutex_unlock(&cfid->fid_mutex); + struct cached_fids *cfids = tcon->cfids; + struct cached_fid *cfid, *q; + struct list_head entry; + + INIT_LIST_HEAD(&entry); + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry_safe(cfid, q, &cfids->entries, entry) { + list_del(&cfid->entry); + list_add(&cfid->entry, &entry); + cfids->num_entries--; + cfid->is_open = false; + /* To prevent race with smb2_cached_lease_break() */ + kref_get(&cfid->refcount); + } + spin_unlock(&cfids->cfid_list_lock); + + list_for_each_entry_safe(cfid, q, &entry, entry) { + cfid->on_list = false; + list_del(&cfid->entry); + cancel_work_sync(&cfid->lease_break); + if (cfid->has_lease) { + /* + * We lease was never cancelled from the server so we + * we need to drop the reference. + */ + spin_lock(&cfids->cfid_list_lock); + cfid->has_lease = false; + spin_unlock(&cfids->cfid_list_lock); + kref_put(&cfid->refcount, smb2_close_cached_fid); + } + /* Drop the extra reference opened above*/ + kref_put(&cfid->refcount, smb2_close_cached_fid); + } } static void @@ -369,51 +381,83 @@ smb2_cached_lease_break(struct work_struct *work) struct cached_fid *cfid = container_of(work, struct cached_fid, lease_break); - close_cached_dir_lease(cfid); + spin_lock(&cfid->cfids->cfid_list_lock); + cfid->has_lease = false; + spin_unlock(&cfid->cfids->cfid_list_lock); + kref_put(&cfid->refcount, smb2_close_cached_fid); } int cached_dir_lease_break(struct cifs_tcon *tcon, __u8 lease_key[16]) { - struct cached_fid *cfid = tcon->cfids->cfid; + struct cached_fids *cfids = tcon->cfids; + struct cached_fid *cfid; - if (cfid == NULL) + if (cfids == NULL) return false; - if (cfid->is_valid && - !memcmp(lease_key, - cfid->fid.lease_key, - SMB2_LEASE_KEY_SIZE)) { - cfid->time = 0; - INIT_WORK(&cfid->lease_break, - smb2_cached_lease_break); - queue_work(cifsiod_wq, - &cfid->lease_break); - return true; + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry(cfid, &cfids->entries, entry) { + if (cfid->has_lease && + !memcmp(lease_key, + cfid->fid.lease_key, + SMB2_LEASE_KEY_SIZE)) { + cfid->time = 0; + /* + * We found a lease remove it from the list + * so no threads threads can access it. + */ + list_del(&cfid->entry); + cfid->on_list = false; + cfids->num_entries--; + + queue_work(cifsiod_wq, + &cfid->lease_break); + spin_unlock(&cfids->cfid_list_lock); + return true; + } } + spin_unlock(&cfids->cfid_list_lock); return false; } -struct cached_fid *init_cached_dir(const char *path) +static struct cached_fid *init_cached_dir(const char *path) { struct cached_fid *cfid; - cfid = kzalloc(sizeof(*cfid), GFP_KERNEL); + cfid = kzalloc(sizeof(*cfid), GFP_ATOMIC); if (!cfid) return NULL; - cfid->path = kstrdup(path, GFP_KERNEL); + cfid->path = kstrdup(path, GFP_ATOMIC); if (!cfid->path) { kfree(cfid); return NULL; } + INIT_WORK(&cfid->lease_break, smb2_cached_lease_break); + INIT_LIST_HEAD(&cfid->entry); INIT_LIST_HEAD(&cfid->dirents.entries); mutex_init(&cfid->dirents.de_mutex); - mutex_init(&cfid->fid_mutex); + spin_lock_init(&cfid->fid_lock); + kref_init(&cfid->refcount); return cfid; } -void free_cached_dir(struct cached_fid *cfid) +static void free_cached_dir(struct cached_fid *cfid) { + struct cached_dirent *dirent, *q; + + dput(cfid->dentry); + cfid->dentry = NULL; + + /* + * Delete all cached dirent names + */ + list_for_each_entry_safe(dirent, q, &cfid->dirents.entries, entry) { + list_del(&dirent->entry); + kfree(dirent->name); + kfree(dirent); + } + kfree(cfid->path); cfid->path = NULL; kfree(cfid); @@ -426,15 +470,34 @@ struct cached_fids *init_cached_dirs(void) cfids = kzalloc(sizeof(*cfids), GFP_KERNEL); if (!cfids) return NULL; - mutex_init(&cfids->cfid_list_mutex); + spin_lock_init(&cfids->cfid_list_lock); + INIT_LIST_HEAD(&cfids->entries); return cfids; } +/* + * Called from tconInfoFree when we are tearing down the tcon. + * There are no active users or open files/directories at this point. + */ void free_cached_dirs(struct cached_fids *cfids) { - if (cfids->cfid) { - free_cached_dir(cfids->cfid); - cfids->cfid = NULL; + struct cached_fid *cfid, *q; + struct list_head entry; + + INIT_LIST_HEAD(&entry); + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry_safe(cfid, q, &cfids->entries, entry) { + cfid->on_list = false; + cfid->is_open = false; + list_del(&cfid->entry); + list_add(&cfid->entry, &entry); } - kfree(cfids); + spin_unlock(&cfids->cfid_list_lock); + + list_for_each_entry_safe(cfid, q, &entry, entry) { + list_del(&cfid->entry); + free_cached_dir(cfid); + } + + kfree(cfids); } diff --git a/fs/cifs/cached_dir.h b/fs/cifs/cached_dir.h index bdf6c3866653..e536304ca2ce 100644 --- a/fs/cifs/cached_dir.h +++ b/fs/cifs/cached_dir.h @@ -31,14 +31,17 @@ struct cached_dirents { }; struct cached_fid { + struct list_head entry; + struct cached_fids *cfids; const char *path; - bool is_valid:1; /* Do we have a useable root fid */ - bool file_all_info_is_valid:1; bool has_lease:1; + bool is_open:1; + bool on_list:1; + bool file_all_info_is_valid:1; unsigned long time; /* jiffies of when lease was taken */ struct kref refcount; struct cifs_fid fid; - struct mutex fid_mutex; + spinlock_t fid_lock; struct cifs_tcon *tcon; struct dentry *dentry; struct work_struct lease_break; @@ -46,9 +49,14 @@ struct cached_fid { struct cached_dirents dirents; }; +#define MAX_CACHED_FIDS 16 struct cached_fids { - struct mutex cfid_list_mutex; - struct cached_fid *cfid; + /* Must be held when: + * - accessing the cfids->entries list + */ + spinlock_t cfid_list_lock; + int num_entries; + struct list_head entries; }; extern struct cached_fids *init_cached_dirs(void); @@ -61,8 +69,6 @@ extern int open_cached_dir_by_dentry(struct cifs_tcon *tcon, struct dentry *dentry, struct cached_fid **cfid); extern void close_cached_dir(struct cached_fid *cfid); -extern void close_cached_dir_lease(struct cached_fid *cfid); -extern void close_cached_dir_lease_locked(struct cached_fid *cfid); extern void close_all_cached_dirs(struct cifs_sb_info *cifs_sb); extern void invalidate_all_cached_dirs(struct cifs_tcon *tcon); extern int cached_dir_lease_break(struct cifs_tcon *tcon, __u8 lease_key[16]); diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index bac08c20f559..a42397b52882 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -2265,13 +2265,13 @@ cifs_dentry_needs_reval(struct dentry *dentry) return true; if (!open_cached_dir_by_dentry(tcon, dentry->d_parent, &cfid)) { - mutex_lock(&cfid->fid_mutex); + spin_lock(&cfid->fid_lock); if (cfid->time && cifs_i->time > cfid->time) { - mutex_unlock(&cfid->fid_mutex); + spin_unlock(&cfid->fid_lock); close_cached_dir(cfid); return false; } - mutex_unlock(&cfid->fid_mutex); + spin_unlock(&cfid->fid_lock); close_cached_dir(cfid); } /* diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index 96f3b0573606..1ed4b4992025 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -787,7 +787,7 @@ smb2_is_path_accessible(const unsigned int xid, struct cifs_tcon *tcon, rc = open_cached_dir(xid, tcon, full_path, cifs_sb, true, &cfid); if (!rc) { - if (cfid->is_valid) { + if (cfid->has_lease) { close_cached_dir(cfid); return 0; } From patchwork Wed Aug 31 02:49:46 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960275 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B6AC0ECAAD5 for ; Wed, 31 Aug 2022 02:50:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231450AbiHaCuQ (ORCPT ); Tue, 30 Aug 2022 22:50:16 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48026 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231161AbiHaCuO (ORCPT ); Tue, 30 Aug 2022 22:50:14 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id EAE5A6525E for ; Tue, 30 Aug 2022 19:50:12 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914212; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=WVsgJBkdBaEUZF0Uub89yn/xzl2UxzqsyTkmrPlSAzE=; b=ah+0lIX2XsawQHoq17IgyCqCMVuMmismeew1DEAzFfS9H1YM1yKyAMu1VKJ1Qm8ZbTR7gK oKgSW7Es6xq3lhUz9fywMhbIfN7PR6kFkaBdkVo6l+NPot9EUdBgeYBfAPonRzfqwHT0LB z2VJej/bGMG1Qc16TKCk77rkpPJtVWY= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-403-VFJ5Mx32OC-lsnqqJkZDiA-1; Tue, 30 Aug 2022 22:50:10 -0400 X-MC-Unique: VFJ5Mx32OC-lsnqqJkZDiA-1 Received: from smtp.corp.redhat.com (int-mx09.intmail.prod.int.rdu2.redhat.com [10.11.54.9]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 2073F80A0C2; Wed, 31 Aug 2022 02:50:10 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id 3268E492C3B; Wed, 31 Aug 2022 02:50:08 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 5/6] cifs: find and use the dentry for cached non-root directories also Date: Wed, 31 Aug 2022 12:49:46 +1000 Message-Id: <20220831024947.3917507-6-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.85 on 10.11.54.9 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org This allows us to use cached attributes for the entries in a cached directory for as long as a lease is held on the directory itself. Previously we have always allowed "used cached attributes for 1 second" but this extends this to the lifetime of the lease as well as making the caching safer. Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 70 +++++++++++++++++++++++++++++++++++--------- 1 file changed, 56 insertions(+), 14 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index ce6935c3fe06..20aba23d8d88 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -5,6 +5,7 @@ * Copyright (c) 2022, Ronnie Sahlberg */ +#include #include "cifsglob.h" #include "cifsproto.h" #include "cifs_debug.h" @@ -59,6 +60,50 @@ static struct cached_fid *find_or_create_cached_dir(struct cached_fids *cfids, return cfid; } +static struct dentry * +path_to_dentry(struct cifs_sb_info *cifs_sb, const char *full_path) +{ + struct dentry *dentry; + char *path = NULL; + char *s, *p; + char sep; + + path = kstrdup(full_path, GFP_ATOMIC); + if (path == NULL) + return ERR_PTR(-ENOMEM); + + sep = CIFS_DIR_SEP(cifs_sb); + dentry = dget(cifs_sb->root); + s = path; + + do { + struct inode *dir = d_inode(dentry); + struct dentry *child; + + if (!S_ISDIR(dir->i_mode)) { + dput(dentry); + dentry = ERR_PTR(-ENOTDIR); + break; + } + + /* skip separators */ + while (*s == sep) + s++; + if (!*s) + break; + p = s++; + /* next separator */ + while (*s && *s != sep) + s++; + + child = lookup_positive_unlocked(p, dentry, s - p); + dput(dentry); + dentry = child; + } while (!IS_ERR(dentry)); + kfree(path); + return dentry; +} + /* * Open the and cache a directory handle. * If error then *cfid is not initialized. @@ -85,7 +130,6 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, struct dentry *dentry = NULL; struct cached_fid *cfid; struct cached_fids *cfids; - if (tcon == NULL || tcon->cfids == NULL || tcon->nohandlecache || is_smb1_server(tcon->ses->server)) @@ -101,13 +145,6 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, if (cifs_sb->root == NULL) return -ENOENT; - /* - * TODO: for better caching we need to find and use the dentry also - * for non-root directories. - */ - if (!path[0]) - dentry = cifs_sb->root; - utf16_path = cifs_convert_path_to_utf16(path, cifs_sb); if (!utf16_path) return -ENOMEM; @@ -199,12 +236,6 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, oparms.fid->mid = le64_to_cpu(o_rsp->hdr.MessageId); #endif /* CIFS_DEBUG2 */ - cfid->tcon = tcon; - if (dentry) { - cfid->dentry = dentry; - dget(dentry); - } - /* BB TBD check to see if oplock level check can be removed below */ if (o_rsp->OplockLevel != SMB2_OPLOCK_LEVEL_LEASE) { goto oshr_free; } @@ -223,6 +254,17 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, &rsp_iov[1], sizeof(struct smb2_file_all_info), (char *)&cfid->file_all_info)) cfid->file_all_info_is_valid = true; + + if (!strlen(path)) { + dentry = dget(cifs_sb->root); + cfid->dentry = dentry; + } else{ + dentry = path_to_dentry(cifs_sb, path); + if (IS_ERR(dentry)) + goto oshr_free; + cfid->dentry = dentry; + } + cfid->tcon = tcon; cfid->time = jiffies; cfid->is_open = true; cfid->has_lease = true; From patchwork Wed Aug 31 02:49:47 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ronnie Sahlberg X-Patchwork-Id: 12960274 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 924DFECAAD5 for ; Wed, 31 Aug 2022 02:50:14 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229565AbiHaCuN (ORCPT ); Tue, 30 Aug 2022 22:50:13 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47882 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231161AbiHaCuM (ORCPT ); Tue, 30 Aug 2022 22:50:12 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 63A454D4C6 for ; Tue, 30 Aug 2022 19:50:11 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1661914210; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=jVY2WMjgGK08eQlqeVLw/a2bBJHT8CTVNDR1qNLLd8E=; b=fRY6ylVoCFelj0izi+3K33Q6QtlTPQMIVgHAwvFPK40PZG9yAKXUenZ75JXZzCuArGetF6 0O/fgFQDd3p0viEPhapgcew+FsbRIYaCwOjK5qW8YppO5KJzjKcKyRUTfD3B7REVjhwmPd bBj5GnVa4FuerSGYKxGM4li8xDjbKpo= Received: from mimecast-mx02.redhat.com (mx3-rdu2.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-187-cTCq6suPNiSoXs2ylFXhiA-1; Tue, 30 Aug 2022 22:50:07 -0400 X-MC-Unique: cTCq6suPNiSoXs2ylFXhiA-1 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.rdu2.redhat.com [10.11.54.2]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id ECDA51C04B42; Wed, 31 Aug 2022 02:50:06 +0000 (UTC) Received: from localhost.localdomain (vpn2-52-20.bne.redhat.com [10.64.52.20]) by smtp.corp.redhat.com (Postfix) with ESMTP id 29A534010D43; Wed, 31 Aug 2022 02:50:05 +0000 (UTC) From: Ronnie Sahlberg To: linux-cifs Cc: Steve French , Ronnie Sahlberg Subject: [PATCH 6/6] cifs: Add a laundromat thread that will timeout any directory leases we have Date: Wed, 31 Aug 2022 12:49:47 +1000 Message-Id: <20220831024947.3917507-7-lsahlber@redhat.com> In-Reply-To: <20220831024947.3917507-1-lsahlber@redhat.com> References: <20220831024947.3917507-1-lsahlber@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.11.54.2 Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org after a timeout of 5 seconds. Later we will add instrumentation to tweak this value. Signed-off-by: Ronnie Sahlberg --- fs/cifs/cached_dir.c | 87 ++++++++++++++++++++++++++++++++++++++------ fs/cifs/cached_dir.h | 1 + 2 files changed, 76 insertions(+), 12 deletions(-) diff --git a/fs/cifs/cached_dir.c b/fs/cifs/cached_dir.c index 20aba23d8d88..16ab0cb9c53c 100644 --- a/fs/cifs/cached_dir.c +++ b/fs/cifs/cached_dir.c @@ -61,17 +61,12 @@ static struct cached_fid *find_or_create_cached_dir(struct cached_fids *cfids, } static struct dentry * -path_to_dentry(struct cifs_sb_info *cifs_sb, const char *full_path) +path_to_dentry(struct cifs_sb_info *cifs_sb, const char *path) { struct dentry *dentry; - char *path = NULL; - char *s, *p; + const char *s, *p; char sep; - path = kstrdup(full_path, GFP_ATOMIC); - if (path == NULL) - return ERR_PTR(-ENOMEM); - sep = CIFS_DIR_SEP(cifs_sb); dentry = dget(cifs_sb->root); s = path; @@ -100,7 +95,7 @@ path_to_dentry(struct cifs_sb_info *cifs_sb, const char *full_path) dput(dentry); dentry = child; } while (!IS_ERR(dentry)); - kfree(path); + return dentry; } @@ -255,15 +250,14 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, (char *)&cfid->file_all_info)) cfid->file_all_info_is_valid = true; - if (!strlen(path)) { + if (!path[0]) dentry = dget(cifs_sb->root); - cfid->dentry = dentry; - } else{ + else { dentry = path_to_dentry(cifs_sb, path); if (IS_ERR(dentry)) goto oshr_free; - cfid->dentry = dentry; } + cfid->dentry = dentry; cfid->tcon = tcon; cfid->time = jiffies; cfid->is_open = true; @@ -505,6 +499,55 @@ static void free_cached_dir(struct cached_fid *cfid) kfree(cfid); } + + +static int +cifs_cfids_laundromat_thread(void *p) +{ + struct cached_fids *cfids = p; + struct cached_fid *cfid, *q; + struct list_head entry; + + while (!kthread_should_stop()) { + ssleep(1); + INIT_LIST_HEAD(&entry); + if (kthread_should_stop()) + return 0; + spin_lock(&cfids->cfid_list_lock); + list_for_each_entry_safe(cfid, q, &cfids->entries, entry) { + if (jiffies > cfid->time + HZ * 5) { + list_del(&cfid->entry); + list_add(&cfid->entry, &entry); + cfids->num_entries--; + } + } + spin_unlock(&cfids->cfid_list_lock); + + list_for_each_entry_safe(cfid, q, &entry, entry) { + cfid->on_list = false; + list_del(&cfid->entry); + /* + * Cancel, and wait for the work to finish in + * case we are racing with it. + */ + cancel_work_sync(&cfid->lease_break); + if (cfid->has_lease) { + /* + * We lease has not yet been cancelled from + * the server so we need to drop the reference. + */ + spin_lock(&cfids->cfid_list_lock); + cfid->has_lease = false; + spin_unlock(&cfids->cfid_list_lock); + kref_put(&cfid->refcount, smb2_close_cached_fid); + } + } + } + + return 0; +} + + struct cached_fids *init_cached_dirs(void) { struct cached_fids *cfids; @@ -514,6 +557,20 @@ struct cached_fids *init_cached_dirs(void) return NULL; spin_lock_init(&cfids->cfid_list_lock); INIT_LIST_HEAD(&cfids->entries); + + /* + * since we're in a cifs function already, we know that + * this will succeed. No need for try_module_get(). + */ + __module_get(THIS_MODULE); + cfids->laundromat = kthread_run(cifs_cfids_laundromat_thread, + cfids, "cifsd-cfid-laundromat"); + if (IS_ERR(cfids->laundromat)) { + cifs_dbg(VFS, "Failed to start cfids laundromat thread.\n"); + kfree(cfids); + module_put(THIS_MODULE); + return NULL; + } return cfids; } @@ -526,6 +583,12 @@ void free_cached_dirs(struct cached_fids *cfids) struct cached_fid *cfid, *q; struct list_head entry; + if (cfids->laundromat) { + kthread_stop(cfids->laundromat); + cfids->laundromat = NULL; + module_put(THIS_MODULE); + } + INIT_LIST_HEAD(&entry); spin_lock(&cfids->cfid_list_lock); list_for_each_entry_safe(cfid, q, &cfids->entries, entry) { diff --git a/fs/cifs/cached_dir.h b/fs/cifs/cached_dir.h index e536304ca2ce..4ab9b10cd098 100644 --- a/fs/cifs/cached_dir.h +++ b/fs/cifs/cached_dir.h @@ -57,6 +57,7 @@ struct cached_fids { spinlock_t cfid_list_lock; int num_entries; struct list_head entries; + struct task_struct *laundromat; }; extern struct cached_fids *init_cached_dirs(void);