1 files changed, 112 insertions, 58 deletions

diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 75d02eaa127..4e0bee240b9 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -8,14 +8,11 @@
 #include <linux/module.h>
 #include <linux/mount.h>
 #include <linux/parser.h>
-#include <linux/rwsem.h>
 #include <linux/sched.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/statfs.h>
 #include <linux/string.h>
-#include <linux/version.h>
-#include <linux/vmalloc.h>
 
 #include "decode.h"
 #include "super.h"
@@ -47,10 +44,20 @@ const char *ceph_file_part(const char *s, int len)
  */
 static void ceph_put_super(struct super_block *s)
 {
-	struct ceph_client *cl = ceph_client(s);
+	struct ceph_client *client = ceph_sb_to_client(s);
 
 	dout("put_super\n");
-	ceph_mdsc_close_sessions(&cl->mdsc);
+	ceph_mdsc_close_sessions(&client->mdsc);
+
+	/*
+	 * ensure we release the bdi before put_anon_super releases
+	 * the device name.
+	 */
+	if (s->s_bdi == &client->backing_dev_info) {
+		bdi_unregister(&client->backing_dev_info);
+		s->s_bdi = NULL;
+	}
+
 	return;
 }
 
@@ -97,12 +104,40 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf)
 static int ceph_syncfs(struct super_block *sb, int wait)
 {
 	dout("sync_fs %d\n", wait);
-	ceph_osdc_sync(&ceph_client(sb)->osdc);
-	ceph_mdsc_sync(&ceph_client(sb)->mdsc);
+	ceph_osdc_sync(&ceph_sb_to_client(sb)->osdc);
+	ceph_mdsc_sync(&ceph_sb_to_client(sb)->mdsc);
 	dout("sync_fs %d done\n", wait);
 	return 0;
 }
 
+static int default_congestion_kb(void)
+{
+	int congestion_kb;
+
+	/*
+	 * Copied from NFS
+	 *
+	 * congestion size, scale with available memory.
+	 *
+	 *  64MB:    8192k
+	 * 128MB:   11585k
+	 * 256MB:   16384k
+	 * 512MB:   23170k
+	 *   1GB:   32768k
+	 *   2GB:   46340k
+	 *   4GB:   65536k
+	 *   8GB:   92681k
+	 *  16GB:  131072k
+	 *
+	 * This allows larger machines to have larger/more transfers.
+	 * Limit the default to 256M
+	 */
+	congestion_kb = (16*int_sqrt(totalram_pages)) << (PAGE_SHIFT-10);
+	if (congestion_kb > 256*1024)
+		congestion_kb = 256*1024;
+
+	return congestion_kb;
+}
 
 /**
  * ceph_show_options - Show mount options in /proc/mounts
@@ -128,6 +163,35 @@ static int ceph_show_options(struct seq_file *m, struct vfsmount *mnt)
 		seq_puts(m, ",nocrc");
 	if (args->flags & CEPH_OPT_NOASYNCREADDIR)
 		seq_puts(m, ",noasyncreaddir");
+
+	if (args->mount_timeout != CEPH_MOUNT_TIMEOUT_DEFAULT)
+		seq_printf(m, ",mount_timeout=%d", args->mount_timeout);
+	if (args->osd_idle_ttl != CEPH_OSD_IDLE_TTL_DEFAULT)
+		seq_printf(m, ",osd_idle_ttl=%d", args->osd_idle_ttl);
+	if (args->osd_timeout != CEPH_OSD_TIMEOUT_DEFAULT)
+		seq_printf(m, ",osdtimeout=%d", args->osd_timeout);
+	if (args->osd_keepalive_timeout != CEPH_OSD_KEEPALIVE_DEFAULT)
+		seq_printf(m, ",osdkeepalivetimeout=%d",
+			 args->osd_keepalive_timeout);
+	if (args->wsize)
+		seq_printf(m, ",wsize=%d", args->wsize);
+	if (args->rsize != CEPH_MOUNT_RSIZE_DEFAULT)
+		seq_printf(m, ",rsize=%d", args->rsize);
+	if (args->congestion_kb != default_congestion_kb())
+		seq_printf(m, ",write_congestion_kb=%d", args->congestion_kb);
+	if (args->caps_wanted_delay_min != CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT)
+		seq_printf(m, ",caps_wanted_delay_min=%d",
+			 args->caps_wanted_delay_min);
+	if (args->caps_wanted_delay_max != CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT)
+		seq_printf(m, ",caps_wanted_delay_max=%d",
+			   args->caps_wanted_delay_max);
+	if (args->cap_release_safety != CEPH_CAP_RELEASE_SAFETY_DEFAULT)
+		seq_printf(m, ",cap_release_safety=%d",
+			   args->cap_release_safety);
+	if (args->max_readdir != CEPH_MAX_READDIR_DEFAULT)
+		seq_printf(m, ",readdir_max_entries=%d", args->max_readdir);
+	if (args->max_readdir_bytes != CEPH_MAX_READDIR_BYTES_DEFAULT)
+		seq_printf(m, ",readdir_max_bytes=%d", args->max_readdir_bytes);
 	if (strcmp(args->snapdir_name, CEPH_SNAPDIRNAME_DEFAULT))
 		seq_printf(m, ",snapdirname=%s", args->snapdir_name);
 	if (args->name)
@@ -151,35 +215,6 @@ static void ceph_inode_init_once(void *foo)
 	inode_init_once(&ci->vfs_inode);
 }
 
-static int default_congestion_kb(void)
-{
-	int congestion_kb;
-
-	/*
-	 * Copied from NFS
-	 *
-	 * congestion size, scale with available memory.
-	 *
-	 *  64MB:    8192k
-	 * 128MB:   11585k
-	 * 256MB:   16384k
-	 * 512MB:   23170k
-	 *   1GB:   32768k
-	 *   2GB:   46340k
-	 *   4GB:   65536k
-	 *   8GB:   92681k
-	 *  16GB:  131072k
-	 *
-	 * This allows larger machines to have larger/more transfers.
-	 * Limit the default to 256M
-	 */
-	congestion_kb = (16*int_sqrt(totalram_pages)) << (PAGE_SHIFT-10);
-	if (congestion_kb > 256*1024)
-		congestion_kb = 256*1024;
-
-	return congestion_kb;
-}
-
 static int __init init_caches(void)
 {
 	ceph_inode_cachep = kmem_cache_create("ceph_inode_info",
@@ -298,7 +333,9 @@ enum {
 	Opt_osd_idle_ttl,
 	Opt_caps_wanted_delay_min,
 	Opt_caps_wanted_delay_max,
+	Opt_cap_release_safety,
 	Opt_readdir_max_entries,
+	Opt_readdir_max_bytes,
 	Opt_congestion_kb,
 	Opt_last_int,
 	/* int args above */
@@ -329,7 +366,9 @@ static match_table_t arg_tokens = {
 	{Opt_osd_idle_ttl, "osd_idle_ttl=%d"},
 	{Opt_caps_wanted_delay_min, "caps_wanted_delay_min=%d"},
 	{Opt_caps_wanted_delay_max, "caps_wanted_delay_max=%d"},
+	{Opt_cap_release_safety, "cap_release_safety=%d"},
 	{Opt_readdir_max_entries, "readdir_max_entries=%d"},
+	{Opt_readdir_max_bytes, "readdir_max_bytes=%d"},
 	{Opt_congestion_kb, "write_congestion_kb=%d"},
 	/* int args above */
 	{Opt_snapdirname, "snapdirname=%s"},
@@ -378,8 +417,9 @@ static struct ceph_mount_args *parse_mount_args(int flags, char *options,
 	args->caps_wanted_delay_max = CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT;
 	args->rsize = CEPH_MOUNT_RSIZE_DEFAULT;
 	args->snapdir_name = kstrdup(CEPH_SNAPDIRNAME_DEFAULT, GFP_KERNEL);
-	args->cap_release_safety = CEPH_CAPS_PER_RELEASE * 4;
-	args->max_readdir = 1024;
+	args->cap_release_safety = CEPH_CAP_RELEASE_SAFETY_DEFAULT;
+	args->max_readdir = CEPH_MAX_READDIR_DEFAULT;
+	args->max_readdir_bytes = CEPH_MAX_READDIR_BYTES_DEFAULT;
 	args->congestion_kb = default_congestion_kb();
 
 	/* ip1[:port1][,ip2[:port2]...]:/subdir/in/fs */
@@ -487,6 +527,9 @@ static struct ceph_mount_args *parse_mount_args(int flags, char *options,
 		case Opt_readdir_max_entries:
 			args->max_readdir = intval;
 			break;
+		case Opt_readdir_max_bytes:
+			args->max_readdir_bytes = intval;
+			break;
 		case Opt_congestion_kb:
 			args->congestion_kb = intval;
 			break;
@@ -626,9 +669,17 @@ static void ceph_destroy_client(struct ceph_client *client)
 
 	/* unmount */
 	ceph_mdsc_stop(&client->mdsc);
-	ceph_monc_stop(&client->monc);
 	ceph_osdc_stop(&client->osdc);
 
+	/*
+	 * make sure mds and osd connections close out before destroying
+	 * the auth module, which is needed to free those connections'
+	 * ceph_authorizers.
+	 */
+	ceph_msgr_flush();
+
+	ceph_monc_stop(&client->monc);
+
 	ceph_adjust_min_caps(-client->min_caps);
 
 	ceph_debugfs_client_cleanup(client);
@@ -636,6 +687,8 @@ static void ceph_destroy_client(struct ceph_client *client)
 	destroy_workqueue(client->pg_inv_wq);
 	destroy_workqueue(client->trunc_wq);
 
+	bdi_destroy(&client->backing_dev_info);
+
 	if (client->msgr)
 		ceph_messenger_destroy(client->msgr);
 	mempool_destroy(client->wb_pagevec_pool);
@@ -670,9 +723,10 @@ int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid)
 /*
  * true if we have the mon map (and have thus joined the cluster)
  */
-static int have_mon_map(struct ceph_client *client)
+static int have_mon_and_osd_map(struct ceph_client *client)
 {
-	return client->monc.monmap && client->monc.monmap->epoch;
+	return client->monc.monmap && client->monc.monmap->epoch &&
+	       client->osdc.osdmap && client->osdc.osdmap->epoch;
 }
 
 /*
@@ -692,7 +746,7 @@ static struct dentry *open_root_dentry(struct ceph_client *client,
 	dout("open_root_inode opening '%s'\n", path);
 	req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, USE_ANY_MDS);
 	if (IS_ERR(req))
-		return ERR_PTR(PTR_ERR(req));
+		return ERR_CAST(req);
 	req->r_path1 = kstrdup(path, GFP_NOFS);
 	req->r_ino1.ino = CEPH_INO_ROOT;
 	req->r_ino1.snap = CEPH_NOSNAP;
@@ -750,7 +804,7 @@ static int ceph_mount(struct ceph_client *client, struct vfsmount *mnt,
 	if (err < 0)
 		goto out;
 
-	while (!have_mon_map(client)) {
+	while (!have_mon_and_osd_map(client)) {
 		err = -EIO;
 		if (timeout && time_after_eq(jiffies, started + timeout))
 			goto out;
@@ -758,8 +812,8 @@ static int ceph_mount(struct ceph_client *client, struct vfsmount *mnt,
 		/* wait */
 		dout("mount waiting for mon_map\n");
 		err = wait_event_interruptible_timeout(client->auth_wq,
-			       have_mon_map(client) || (client->auth_err < 0),
-			       timeout);
+		       have_mon_and_osd_map(client) || (client->auth_err < 0),
+		       timeout);
 		if (err == -EINTR || err == -ERESTARTSYS)
 			goto out;
 		if (client->auth_err < 0) {
@@ -872,18 +926,21 @@ static int ceph_compare_super(struct super_block *sb, void *data)
 /*
  * construct our own bdi so we can control readahead, etc.
  */
+static atomic_long_t bdi_seq = ATOMIC_INIT(0);
+
 static int ceph_register_bdi(struct super_block *sb, struct ceph_client *client)
 {
 	int err;
 
-	sb->s_bdi = &client->backing_dev_info;
-
 	/* set ra_pages based on rsize mount option? */
 	if (client->mount_args->rsize >= PAGE_CACHE_SIZE)
 		client->backing_dev_info.ra_pages =
 			(client->mount_args->rsize + PAGE_CACHE_SIZE - 1)
 			>> PAGE_SHIFT;
-	err = bdi_register_dev(&client->backing_dev_info, sb->s_dev);
+	err = bdi_register(&client->backing_dev_info, NULL, "ceph-%d",
+			   atomic_long_inc_return(&bdi_seq));
+	if (!err)
+		sb->s_bdi = &client->backing_dev_info;
 	return err;
 }
 
@@ -920,9 +977,9 @@ static int ceph_get_sb(struct file_system_type *fs_type,
 		goto out;
 	}
 
-	if (ceph_client(sb) != client) {
+	if (ceph_sb_to_client(sb) != client) {
 		ceph_destroy_client(client);
-		client = ceph_client(sb);
+		client = ceph_sb_to_client(sb);
 		dout("get_sb got existing client %p\n", client);
 	} else {
 		dout("get_sb using new client %p\n", client);
@@ -940,8 +997,7 @@ static int ceph_get_sb(struct file_system_type *fs_type,
 
 out_splat:
 	ceph_mdsc_close_sessions(&client->mdsc);
-	up_write(&sb->s_umount);
-	deactivate_super(sb);
+	deactivate_locked_super(sb);
 	goto out_final;
 
 out:
@@ -957,9 +1013,6 @@ static void ceph_kill_sb(struct super_block *s)
 	dout("kill_sb %p\n", s);
 	ceph_mdsc_pre_umount(&client->mdsc);
 	kill_anon_super(s);    /* will call put_super after sb is r/o */
-	if (s->s_bdi == &client->backing_dev_info)
-		bdi_unregister(&client->backing_dev_info);
-	bdi_destroy(&client->backing_dev_info);
 	ceph_destroy_client(client);
 }
 
@@ -996,9 +1049,10 @@ static int __init init_ceph(void)
 	if (ret)
 		goto out_icache;
 
-	pr_info("loaded %d.%d.%d (mon/mds/osd proto %d/%d/%d)\n",
-		CEPH_VERSION_MAJOR, CEPH_VERSION_MINOR, CEPH_VERSION_PATCH,
-		CEPH_MONC_PROTOCOL, CEPH_MDSC_PROTOCOL, CEPH_OSDC_PROTOCOL);
+	pr_info("loaded (mon/mds/osd proto %d/%d/%d, osdmap %d/%d %d/%d)\n",
+		CEPH_MONC_PROTOCOL, CEPH_MDSC_PROTOCOL, CEPH_OSDC_PROTOCOL,
+		CEPH_OSDMAP_VERSION, CEPH_OSDMAP_VERSION_EXT,
+		CEPH_OSDMAP_INC_VERSION, CEPH_OSDMAP_INC_VERSION_EXT);
 	return 0;
 
 out_icache: