summaryrefslogtreecommitdiffstats
path: root/drivers/block
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/block')
-rw-r--r--drivers/block/aoe/aoe.h11
-rw-r--r--drivers/block/aoe/aoecmd.c156
-rw-r--r--drivers/block/aoe/aoedev.c3
-rw-r--r--drivers/block/aoe/aoenet.c7
-rw-r--r--drivers/block/brd.c4
-rw-r--r--drivers/block/cciss.c32
-rw-r--r--drivers/block/cryptoloop.c2
-rw-r--r--drivers/block/loop.c2
-rw-r--r--drivers/block/loop.h85
-rw-r--r--drivers/block/mtip32xx/mtip32xx.c11
-rw-r--r--drivers/block/nbd.c11
-rw-r--r--drivers/block/nvme-core.c62
-rw-r--r--drivers/block/nvme-scsi.c3
-rw-r--r--drivers/block/pktcdvd.c3
-rw-r--r--drivers/block/rbd.c1124
-rw-r--r--drivers/block/swim.c2
-rw-r--r--drivers/block/virtio_blk.c2
-rw-r--r--drivers/block/xen-blkback/xenbus.c2
-rw-r--r--drivers/block/xsysace.c7
19 files changed, 969 insertions, 560 deletions
diff --git a/drivers/block/aoe/aoe.h b/drivers/block/aoe/aoe.h
index 175649468c9..025c41d3cb3 100644
--- a/drivers/block/aoe/aoe.h
+++ b/drivers/block/aoe/aoe.h
@@ -1,5 +1,5 @@
-/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */
-#define VERSION "81"
+/* Copyright (c) 2013 Coraid, Inc. See COPYING for GPL terms. */
+#define VERSION "83"
#define AOE_MAJOR 152
#define DEVICE_NAME "aoe"
@@ -196,9 +196,11 @@ struct ktstate {
struct completion rendez;
struct task_struct *task;
wait_queue_head_t *waitq;
- int (*fn) (void);
- char *name;
+ int (*fn) (int);
+ char name[12];
spinlock_t *lock;
+ int id;
+ int active;
};
int aoeblk_init(void);
@@ -222,6 +224,7 @@ int aoecmd_init(void);
struct sk_buff *aoecmd_ata_id(struct aoedev *);
void aoe_freetframe(struct frame *);
void aoe_flush_iocq(void);
+void aoe_flush_iocq_by_index(int);
void aoe_end_request(struct aoedev *, struct request *, int);
int aoe_ktstart(struct ktstate *k);
void aoe_ktstop(struct ktstate *k);
diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c
index fc803ecbbce..99cb944a002 100644
--- a/drivers/block/aoe/aoecmd.c
+++ b/drivers/block/aoe/aoecmd.c
@@ -1,4 +1,4 @@
-/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */
+/* Copyright (c) 2013 Coraid, Inc. See COPYING for GPL terms. */
/*
* aoecmd.c
* Filesystem request handling methods
@@ -35,14 +35,27 @@ module_param(aoe_maxout, int, 0644);
MODULE_PARM_DESC(aoe_maxout,
"Only aoe_maxout outstanding packets for every MAC on eX.Y.");
-static wait_queue_head_t ktiowq;
-static struct ktstate kts;
+/* The number of online cpus during module initialization gives us a
+ * convenient heuristic cap on the parallelism used for ktio threads
+ * doing I/O completion. It is not important that the cap equal the
+ * actual number of running CPUs at any given time, but because of CPU
+ * hotplug, we take care to use ncpus instead of using
+ * num_online_cpus() after module initialization.
+ */
+static int ncpus;
+
+/* mutex lock used for synchronization while thread spawning */
+static DEFINE_MUTEX(ktio_spawn_lock);
+
+static wait_queue_head_t *ktiowq;
+static struct ktstate *kts;
/* io completion queue */
-static struct {
+struct iocq_ktio {
struct list_head head;
spinlock_t lock;
-} iocq;
+};
+static struct iocq_ktio *iocq;
static struct page *empty_page;
@@ -1278,23 +1291,36 @@ out:
* Returns true iff responses needing processing remain.
*/
static int
-ktio(void)
+ktio(int id)
{
struct frame *f;
struct list_head *pos;
int i;
+ int actual_id;
for (i = 0; ; ++i) {
if (i == MAXIOC)
return 1;
- if (list_empty(&iocq.head))
+ if (list_empty(&iocq[id].head))
return 0;
- pos = iocq.head.next;
+ pos = iocq[id].head.next;
list_del(pos);
- spin_unlock_irq(&iocq.lock);
f = list_entry(pos, struct frame, head);
+ spin_unlock_irq(&iocq[id].lock);
ktiocomplete(f);
- spin_lock_irq(&iocq.lock);
+
+ /* Figure out if extra threads are required. */
+ actual_id = f->t->d->aoeminor % ncpus;
+
+ if (!kts[actual_id].active) {
+ BUG_ON(id != 0);
+ mutex_lock(&ktio_spawn_lock);
+ if (!kts[actual_id].active
+ && aoe_ktstart(&kts[actual_id]) == 0)
+ kts[actual_id].active = 1;
+ mutex_unlock(&ktio_spawn_lock);
+ }
+ spin_lock_irq(&iocq[id].lock);
}
}
@@ -1311,7 +1337,7 @@ kthread(void *vp)
complete(&k->rendez); /* tell spawner we're running */
do {
spin_lock_irq(k->lock);
- more = k->fn();
+ more = k->fn(k->id);
if (!more) {
add_wait_queue(k->waitq, &wait);
__set_current_state(TASK_INTERRUPTIBLE);
@@ -1340,7 +1366,7 @@ aoe_ktstart(struct ktstate *k)
struct task_struct *task;
init_completion(&k->rendez);
- task = kthread_run(kthread, k, k->name);
+ task = kthread_run(kthread, k, "%s", k->name);
if (task == NULL || IS_ERR(task))
return -ENOMEM;
k->task = task;
@@ -1353,13 +1379,24 @@ aoe_ktstart(struct ktstate *k)
static void
ktcomplete(struct frame *f, struct sk_buff *skb)
{
+ int id;
ulong flags;
f->r_skb = skb;
- spin_lock_irqsave(&iocq.lock, flags);
- list_add_tail(&f->head, &iocq.head);
- spin_unlock_irqrestore(&iocq.lock, flags);
- wake_up(&ktiowq);
+ id = f->t->d->aoeminor % ncpus;
+ spin_lock_irqsave(&iocq[id].lock, flags);
+ if (!kts[id].active) {
+ spin_unlock_irqrestore(&iocq[id].lock, flags);
+ /* The thread with id has not been spawned yet,
+ * so delegate the work to the main thread and
+ * try spawning a new thread.
+ */
+ id = 0;
+ spin_lock_irqsave(&iocq[id].lock, flags);
+ }
+ list_add_tail(&f->head, &iocq[id].head);
+ spin_unlock_irqrestore(&iocq[id].lock, flags);
+ wake_up(&ktiowq[id]);
}
struct sk_buff *
@@ -1706,6 +1743,17 @@ aoe_failbuf(struct aoedev *d, struct buf *buf)
void
aoe_flush_iocq(void)
{
+ int i;
+
+ for (i = 0; i < ncpus; i++) {
+ if (kts[i].active)
+ aoe_flush_iocq_by_index(i);
+ }
+}
+
+void
+aoe_flush_iocq_by_index(int id)
+{
struct frame *f;
struct aoedev *d;
LIST_HEAD(flist);
@@ -1713,9 +1761,9 @@ aoe_flush_iocq(void)
struct sk_buff *skb;
ulong flags;
- spin_lock_irqsave(&iocq.lock, flags);
- list_splice_init(&iocq.head, &flist);
- spin_unlock_irqrestore(&iocq.lock, flags);
+ spin_lock_irqsave(&iocq[id].lock, flags);
+ list_splice_init(&iocq[id].head, &flist);
+ spin_unlock_irqrestore(&iocq[id].lock, flags);
while (!list_empty(&flist)) {
pos = flist.next;
list_del(pos);
@@ -1738,6 +1786,8 @@ int __init
aoecmd_init(void)
{
void *p;
+ int i;
+ int ret;
/* get_zeroed_page returns page with ref count 1 */
p = (void *) get_zeroed_page(GFP_KERNEL | __GFP_REPEAT);
@@ -1745,22 +1795,72 @@ aoecmd_init(void)
return -ENOMEM;
empty_page = virt_to_page(p);
- INIT_LIST_HEAD(&iocq.head);
- spin_lock_init(&iocq.lock);
- init_waitqueue_head(&ktiowq);
- kts.name = "aoe_ktio";
- kts.fn = ktio;
- kts.waitq = &ktiowq;
- kts.lock = &iocq.lock;
- return aoe_ktstart(&kts);
+ ncpus = num_online_cpus();
+
+ iocq = kcalloc(ncpus, sizeof(struct iocq_ktio), GFP_KERNEL);
+ if (!iocq)
+ return -ENOMEM;
+
+ kts = kcalloc(ncpus, sizeof(struct ktstate), GFP_KERNEL);
+ if (!kts) {
+ ret = -ENOMEM;
+ goto kts_fail;
+ }
+
+ ktiowq = kcalloc(ncpus, sizeof(wait_queue_head_t), GFP_KERNEL);
+ if (!ktiowq) {
+ ret = -ENOMEM;
+ goto ktiowq_fail;
+ }
+
+ mutex_init(&ktio_spawn_lock);
+
+ for (i = 0; i < ncpus; i++) {
+ INIT_LIST_HEAD(&iocq[i].head);
+ spin_lock_init(&iocq[i].lock);
+ init_waitqueue_head(&ktiowq[i]);
+ snprintf(kts[i].name, sizeof(kts[i].name), "aoe_ktio%d", i);
+ kts[i].fn = ktio;
+ kts[i].waitq = &ktiowq[i];
+ kts[i].lock = &iocq[i].lock;
+ kts[i].id = i;
+ kts[i].active = 0;
+ }
+ kts[0].active = 1;
+ if (aoe_ktstart(&kts[0])) {
+ ret = -ENOMEM;
+ goto ktstart_fail;
+ }
+ return 0;
+
+ktstart_fail:
+ kfree(ktiowq);
+ktiowq_fail:
+ kfree(kts);
+kts_fail:
+ kfree(iocq);
+
+ return ret;
}
void
aoecmd_exit(void)
{
- aoe_ktstop(&kts);
+ int i;
+
+ for (i = 0; i < ncpus; i++)
+ if (kts[i].active)
+ aoe_ktstop(&kts[i]);
+
aoe_flush_iocq();
+ /* Free up the iocq and thread speicific configuration
+ * allocated during startup.
+ */
+ kfree(iocq);
+ kfree(kts);
+ kfree(ktiowq);
+
free_page((unsigned long) page_address(empty_page));
empty_page = NULL;
}
diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c
index 98f2965778b..784c92e038d 100644
--- a/drivers/block/aoe/aoedev.c
+++ b/drivers/block/aoe/aoedev.c
@@ -1,4 +1,4 @@
-/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */
+/* Copyright (c) 2013 Coraid, Inc. See COPYING for GPL terms. */
/*
* aoedev.c
* AoE device utility functions; maintains device list.
@@ -518,7 +518,6 @@ void
aoedev_exit(void)
{
flush_scheduled_work();
- aoe_flush_iocq();
flush(NULL, 0, EXITING);
}
diff --git a/drivers/block/aoe/aoenet.c b/drivers/block/aoe/aoenet.c
index 71d3ea8d300..63773a90581 100644
--- a/drivers/block/aoe/aoenet.c
+++ b/drivers/block/aoe/aoenet.c
@@ -1,4 +1,4 @@
-/* Copyright (c) 2012 Coraid, Inc. See COPYING for GPL terms. */
+/* Copyright (c) 2013 Coraid, Inc. See COPYING for GPL terms. */
/*
* aoenet.c
* Ethernet portion of AoE driver
@@ -52,7 +52,7 @@ static struct sk_buff_head skbtxq;
/* enters with txlock held */
static int
-tx(void) __must_hold(&txlock)
+tx(int id) __must_hold(&txlock)
{
struct sk_buff *skb;
struct net_device *ifp;
@@ -205,7 +205,8 @@ aoenet_init(void)
kts.lock = &txlock;
kts.fn = tx;
kts.waitq = &txwq;
- kts.name = "aoe_tx";
+ kts.id = 0;
+ snprintf(kts.name, sizeof(kts.name), "aoe_tx%d", kts.id);
if (aoe_ktstart(&kts))
return -EAGAIN;
dev_add_pack(&aoe_pt);
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index f1a29f8e9d3..9bf4371755f 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -117,13 +117,13 @@ static struct page *brd_insert_page(struct brd_device *brd, sector_t sector)
spin_lock(&brd->brd_lock);
idx = sector >> PAGE_SECTORS_SHIFT;
+ page->index = idx;
if (radix_tree_insert(&brd->brd_pages, idx, page)) {
__free_page(page);
page = radix_tree_lookup(&brd->brd_pages, idx);
BUG_ON(!page);
BUG_ON(page->index != idx);
- } else
- page->index = idx;
+ }
spin_unlock(&brd->brd_lock);
radix_tree_preload_end();
diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c
index 6374dc10352..62b6c2cc80b 100644
--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -168,8 +168,6 @@ static irqreturn_t do_cciss_msix_intr(int irq, void *dev_id);
static int cciss_open(struct block_device *bdev, fmode_t mode);
static int cciss_unlocked_open(struct block_device *bdev, fmode_t mode);
static void cciss_release(struct gendisk *disk, fmode_t mode);
-static int do_ioctl(struct block_device *bdev, fmode_t mode,
- unsigned int cmd, unsigned long arg);
static int cciss_ioctl(struct block_device *bdev, fmode_t mode,
unsigned int cmd, unsigned long arg);
static int cciss_getgeo(struct block_device *bdev, struct hd_geometry *geo);
@@ -235,7 +233,7 @@ static const struct block_device_operations cciss_fops = {
.owner = THIS_MODULE,
.open = cciss_unlocked_open,
.release = cciss_release,
- .ioctl = do_ioctl,
+ .ioctl = cciss_ioctl,
.getgeo = cciss_getgeo,
#ifdef CONFIG_COMPAT
.compat_ioctl = cciss_compat_ioctl,
@@ -1143,16 +1141,6 @@ static void cciss_release(struct gendisk *disk, fmode_t mode)
mutex_unlock(&cciss_mutex);
}
-static int do_ioctl(struct block_device *bdev, fmode_t mode,
- unsigned cmd, unsigned long arg)
-{
- int ret;
- mutex_lock(&cciss_mutex);
- ret = cciss_ioctl(bdev, mode, cmd, arg);
- mutex_unlock(&cciss_mutex);
- return ret;
-}
-
#ifdef CONFIG_COMPAT
static int cciss_ioctl32_passthru(struct block_device *bdev, fmode_t mode,
@@ -1179,7 +1167,7 @@ static int cciss_compat_ioctl(struct block_device *bdev, fmode_t mode,
case CCISS_REGNEWD:
case CCISS_RESCANDISK:
case CCISS_GETLUNINFO:
- return do_ioctl(bdev, mode, cmd, arg);
+ return cciss_ioctl(bdev, mode, cmd, arg);
case CCISS_PASSTHRU32:
return cciss_ioctl32_passthru(bdev, mode, cmd, arg);
@@ -1219,7 +1207,7 @@ static int cciss_ioctl32_passthru(struct block_device *bdev, fmode_t mode,
if (err)
return -EFAULT;
- err = do_ioctl(bdev, mode, CCISS_PASSTHRU, (unsigned long)p);
+ err = cciss_ioctl(bdev, mode, CCISS_PASSTHRU, (unsigned long)p);
if (err)
return err;
err |=
@@ -1261,7 +1249,7 @@ static int cciss_ioctl32_big_passthru(struct block_device *bdev, fmode_t mode,
if (err)
return -EFAULT;
- err = do_ioctl(bdev, mode, CCISS_BIG_PASSTHRU, (unsigned long)p);
+ err = cciss_ioctl(bdev, mode, CCISS_BIG_PASSTHRU, (unsigned long)p);
if (err)
return err;
err |=
@@ -1311,11 +1299,14 @@ static int cciss_getpciinfo(ctlr_info_t *h, void __user *argp)
static int cciss_getintinfo(ctlr_info_t *h, void __user *argp)
{
cciss_coalint_struct intinfo;
+ unsigned long flags;
if (!argp)
return -EINVAL;
+ spin_lock_irqsave(&h->lock, flags);
intinfo.delay = readl(&h->cfgtable->HostWrite.CoalIntDelay);
intinfo.count = readl(&h->cfgtable->HostWrite.CoalIntCount);
+ spin_unlock_irqrestore(&h->lock, flags);
if (copy_to_user
(argp, &intinfo, sizeof(cciss_coalint_struct)))
return -EFAULT;
@@ -1356,12 +1347,15 @@ static int cciss_setintinfo(ctlr_info_t *h, void __user *argp)
static int cciss_getnodename(ctlr_info_t *h, void __user *argp)
{
NodeName_type NodeName;
+ unsigned long flags;
int i;
if (!argp)
return -EINVAL;
+ spin_lock_irqsave(&h->lock, flags);
for (i = 0; i < 16; i++)
NodeName[i] = readb(&h->cfgtable->ServerName[i]);
+ spin_unlock_irqrestore(&h->lock, flags);
if (copy_to_user(argp, NodeName, sizeof(NodeName_type)))
return -EFAULT;
return 0;
@@ -1398,10 +1392,13 @@ static int cciss_setnodename(ctlr_info_t *h, void __user *argp)
static int cciss_getheartbeat(ctlr_info_t *h, void __user *argp)
{
Heartbeat_type heartbeat;
+ unsigned long flags;
if (!argp)
return -EINVAL;
+ spin_lock_irqsave(&h->lock, flags);
heartbeat = readl(&h->cfgtable->HeartBeat);
+ spin_unlock_irqrestore(&h->lock, flags);
if (copy_to_user(argp, &heartbeat, sizeof(Heartbeat_type)))
return -EFAULT;
return 0;
@@ -1410,10 +1407,13 @@ static int cciss_getheartbeat(ctlr_info_t *h, void __user *argp)
static int cciss_getbustypes(ctlr_info_t *h, void __user *argp)
{
BusTypes_type BusTypes;
+ unsigned long flags;
if (!argp)
return -EINVAL;
+ spin_lock_irqsave(&h->lock, flags);
BusTypes = readl(&h->cfgtable->BusTypes);
+ spin_unlock_irqrestore(&h->lock, flags);
if (copy_to_user(argp, &BusTypes, sizeof(BusTypes_type)))
return -EFAULT;
return 0;
diff --git a/drivers/block/cryptoloop.c b/drivers/block/cryptoloop.c
index 8b6bb764b0a..99e773cb70d 100644
--- a/drivers/block/cryptoloop.c
+++ b/drivers/block/cryptoloop.c
@@ -25,9 +25,9 @@
#include <linux/string.h>
#include <linux/crypto.h>
#include <linux/blkdev.h>
-#include <linux/loop.h>
#include <linux/scatterlist.h>
#include <asm/uaccess.h>
+#include "loop.h"
MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("loop blockdevice transferfunction adaptor / CryptoAPI");
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index d92d50fd84b..40e715531aa 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -63,7 +63,6 @@
#include <linux/init.h>
#include <linux/swap.h>
#include <linux/slab.h>
-#include <linux/loop.h>
#include <linux/compat.h>
#include <linux/suspend.h>
#include <linux/freezer.h>
@@ -76,6 +75,7 @@
#include <linux/sysfs.h>
#include <linux/miscdevice.h>
#include <linux/falloc.h>
+#include "loop.h"
#include <asm/uaccess.h>
diff --git a/drivers/block/loop.h b/drivers/block/loop.h
new file mode 100644
index 00000000000..90df5d6485b
--- /dev/null
+++ b/drivers/block/loop.h
@@ -0,0 +1,85 @@
+/*
+ * loop.h
+ *
+ * Written by Theodore Ts'o, 3/29/93.
+ *
+ * Copyright 1993 by Theodore Ts'o. Redistribution of this file is
+ * permitted under the GNU General Public License.
+ */
+#ifndef _LINUX_LOOP_H
+#define _LINUX_LOOP_H
+
+#include <linux/bio.h>
+#include <linux/blkdev.h>
+#include <linux/spinlock.h>
+#include <linux/mutex.h>
+#include <uapi/linux/loop.h>
+
+/* Possible states of device */
+enum {
+ Lo_unbound,
+ Lo_bound,
+ Lo_rundown,
+};
+
+struct loop_func_table;
+
+struct loop_device {
+ int lo_number;
+ int lo_refcnt;
+ loff_t lo_offset;
+ loff_t lo_sizelimit;
+ int lo_flags;
+ int (*transfer)(struct loop_device *, int cmd,
+ struct page *raw_page, unsigned raw_off,
+ struct page *loop_page, unsigned loop_off,
+ int size, sector_t real_block);
+ char lo_file_name[LO_NAME_SIZE];
+ char lo_crypt_name[LO_NAME_SIZE];
+ char lo_encrypt_key[LO_KEY_SIZE];
+ int lo_encrypt_key_size;
+ struct loop_func_table *lo_encryption;
+ __u32 lo_init[2];
+ kuid_t lo_key_owner; /* Who set the key */
+ int (*ioctl)(struct loop_device *, int cmd,
+ unsigned long arg);
+
+ struct file * lo_backing_file;
+ struct block_device *lo_device;
+ unsigned lo_blocksize;
+ void *key_data;
+
+ gfp_t old_gfp_mask;
+
+ spinlock_t lo_lock;
+ struct bio_list lo_bio_list;
+ unsigned int lo_bio_count;
+ int lo_state;
+ struct mutex lo_ctl_mutex;
+ struct task_struct *lo_thread;
+ wait_queue_head_t lo_event;
+ /* wait queue for incoming requests */
+ wait_queue_head_t lo_req_wait;
+
+ struct request_queue *lo_queue;
+ struct gendisk *lo_disk;
+};
+
+/* Support for loadable transfer modules */
+struct loop_func_table {
+ int number; /* filter type */
+ int (*transfer)(struct loop_device *lo, int cmd,
+ struct page *raw_page, unsigned raw_off,
+ struct page *loop_page, unsigned loop_off,
+ int size, sector_t real_block);
+ int (*init)(struct loop_device *, const struct loop_info64 *);
+ /* release is called from loop_unregister_transfer or clr_fd */
+ int (*release)(struct loop_device *);
+ int (*ioctl)(struct loop_device *, int cmd, unsigned long arg);
+ struct module *owner;
+};
+
+int loop_register_transfer(struct loop_func_table *funcs);
+int loop_unregister_transfer(int number);
+
+#endif
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index 847107ef0cc..952dbfe2212 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -3002,7 +3002,8 @@ static int mtip_hw_debugfs_init(struct driver_data *dd)
static void mtip_hw_debugfs_exit(struct driver_data *dd)
{
- debugfs_remove_recursive(dd->dfs_node);
+ if (dd->dfs_node)
+ debugfs_remove_recursive(dd->dfs_node);
}
@@ -3863,7 +3864,7 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio)
struct driver_data *dd = queue->queuedata;
struct scatterlist *sg;
struct bio_vec *bvec;
- int nents = 0;
+ int i, nents = 0;
int tag = 0, unaligned = 0;
if (unlikely(dd->dd_flag & MTIP_DDF_STOP_IO)) {
@@ -3921,11 +3922,12 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio)
}
/* Create the scatter list for this bio. */
- bio_for_each_segment(bvec, bio, nents) {
+ bio_for_each_segment(bvec, bio, i) {
sg_set_page(&sg[nents],
bvec->bv_page,
bvec->bv_len,
bvec->bv_offset);
+ nents++;
}
/* Issue the read/write. */
@@ -4085,7 +4087,8 @@ skip_create_disk:
start_service_thread:
sprintf(thd_name, "mtip_svc_thd_%02d", index);
dd->mtip_svc_handler = kthread_create_on_node(mtip_service_thread,
- dd, dd->numa_node, thd_name);
+ dd, dd->numa_node, "%s",
+ thd_name);
if (IS_ERR(dd->mtip_svc_handler)) {
dev_err(&dd->pdev->dev, "service thread failed to start\n");
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 037288e7874..2dc3b5153f0 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -623,8 +623,10 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
if (!nbd->sock)
return -EINVAL;
+ nbd->disconnect = 1;
+
nbd_send_req(nbd, &sreq);
- return 0;
+ return 0;
}
case NBD_CLEAR_SOCK: {
@@ -654,6 +656,7 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
nbd->sock = SOCKET_I(inode);
if (max_part > 0)
bdev->bd_invalidated = 1;
+ nbd->disconnect = 0; /* we're connected now */
return 0;
} else {
fput(file);
@@ -714,7 +717,8 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
else
blk_queue_flush(nbd->disk->queue, 0);
- thread = kthread_create(nbd_thread, nbd, nbd->disk->disk_name);
+ thread = kthread_create(nbd_thread, nbd, "%s",
+ nbd->disk->disk_name);
if (IS_ERR(thread)) {
mutex_lock(&nbd->tx_lock);
return PTR_ERR(thread);
@@ -742,6 +746,8 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
set_capacity(nbd->disk, 0);
if (max_part > 0)
ioctl_by_bdev(bdev, BLKRRPART, 0);
+ if (nbd->disconnect) /* user requested, ignore socket errors */
+ return 0;
return nbd->harderror;
}
@@ -750,7 +756,6 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
* This is for compatibility only. The queue is always cleared
* by NBD_DO_IT or NBD_CLEAR_SOCK.
*/
- BUG_ON(!nbd->sock && !list_empty(&nbd->queue_head));
return 0;
case NBD_PRINT_DEBUG:
diff --git a/drivers/block/nvme-core.c b/drivers/block/nvme-core.c
index 8efdfaa44a5..ce79a590b45 100644
--- a/drivers/block/nvme-core.c
+++ b/drivers/block/nvme-core.c
@@ -629,7 +629,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns,
struct nvme_command *cmnd;
struct nvme_iod *iod;
enum dma_data_direction dma_dir;
- int cmdid, length, result = -ENOMEM;
+ int cmdid, length, result;
u16 control;
u32 dsmgmt;
int psegs = bio_phys_segments(ns->queue, bio);
@@ -640,6 +640,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns,
return result;
}
+ result = -ENOMEM;
iod = nvme_alloc_iod(psegs, bio->bi_size, GFP_ATOMIC);
if (!iod)
goto nomem;
@@ -977,6 +978,8 @@ static void nvme_cancel_ios(struct nvme_queue *nvmeq, bool timeout)
if (timeout && !time_after(now, info[cmdid].timeout))
continue;
+ if (info[cmdid].ctx == CMD_CTX_CANCELLED)
+ continue;
dev_warn(nvmeq->q_dmadev, "Cancelling I/O %d\n", cmdid);
ctx = cancel_cmdid(nvmeq, cmdid, &fn);
fn(nvmeq->dev, ctx, &cqe);
@@ -1206,7 +1209,7 @@ struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write,
if (addr & 3)
return ERR_PTR(-EINVAL);
- if (!length)
+ if (!length || length > INT_MAX - PAGE_SIZE)
return ERR_PTR(-EINVAL);
offset = offset_in_page(addr);
@@ -1227,7 +1230,8 @@ struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write,
sg_init_table(sg, count);
for (i = 0; i < count; i++) {
sg_set_page(&sg[i], pages[i],
- min_t(int, length, PAGE_SIZE - offset), offset);
+ min_t(unsigned, length, PAGE_SIZE - offset),
+ offset);
length -= (PAGE_SIZE - offset);
offset = 0;
}
@@ -1435,7 +1439,7 @@ static int nvme_user_admin_cmd(struct nvme_dev *dev,
nvme_free_iod(dev, iod);
}
- if (!status && copy_to_user(&ucmd->result, &cmd.result,
+ if ((status >= 0) && copy_to_user(&ucmd->result, &cmd.result,
sizeof(cmd.result)))
status = -EFAULT;
@@ -1633,7 +1637,8 @@ static int set_queue_count(struct nvme_dev *dev, int count)
static int nvme_setup_io_queues(struct nvme_dev *dev)
{
- int result, cpu, i, nr_io_queues, db_bar_size, q_depth;
+ struct pci_dev *pdev = dev->pci_dev;
+ int result, cpu, i, nr_io_queues, db_bar_size, q_depth, q_count;
nr_io_queues = num_online_cpus();
result = set_queue_count(dev, nr_io_queues);
@@ -1642,14 +1647,14 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
if (result < nr_io_queues)
nr_io_queues = result;
+ q_count = nr_io_queues;
/* Deregister the admin queue's interrupt */
free_irq(dev->entry[0].vector, dev->queues[0]);
db_bar_size = 4096 + ((nr_io_queues + 1) << (dev->db_stride + 3));
if (db_bar_size > 8192) {
iounmap(dev->bar);
- dev->bar = ioremap(pci_resource_start(dev->pci_dev, 0),
- db_bar_size);
+ dev->bar = ioremap(pci_resource_start(pdev, 0), db_bar_size);
dev->dbs = ((void __iomem *)dev->bar) + 4096;
dev->queues[0]->q_db = dev->dbs;
}
@@ -1657,19 +1662,36 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
for (i = 0; i < nr_io_queues; i++)
dev->entry[i].entry = i;
for (;;) {
- result = pci_enable_msix(dev->pci_dev, dev->entry,
- nr_io_queues);
+ result = pci_enable_msix(pdev, dev->entry, nr_io_queues);
if (result == 0) {
break;
} else if (result > 0) {
nr_io_queues = result;
continue;
} else {
- nr_io_queues = 1;
+ nr_io_queues = 0;
break;
}
}
+ if (nr_io_queues == 0) {
+ nr_io_queues = q_count;
+ for (;;) {
+ result = pci_enable_msi_block(pdev, nr_io_queues);
+ if (result == 0) {
+ for (i = 0; i < nr_io_queues; i++)
+ dev->entry[i].vector = i + pdev->irq;
+ break;
+ } else if (result > 0) {
+ nr_io_queues = result;
+ continue;
+ } else {
+ nr_io_queues = 1;
+ break;
+ }
+ }
+ }
+
result = queue_request_irq(dev, dev->queues[0], "nvme admin");
/* XXX: handle failure here */
@@ -1850,7 +1872,10 @@ static void nvme_free_dev(struct kref *kref)
{
struct nvme_dev *dev = container_of(kref, struct nvme_dev, kref);
nvme_dev_remove(dev);
- pci_disable_msix(dev->pci_dev);
+ if (dev->pci_dev->msi_enabled)
+ pci_disable_msi(dev->pci_dev);
+ else if (dev->pci_dev->msix_enabled)
+ pci_disable_msix(dev->pci_dev);
iounmap(dev->bar);
nvme_release_instance(dev);
nvme_release_prp_pools(dev);
@@ -1923,8 +1948,14 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
INIT_LIST_HEAD(&dev->namespaces);
dev->pci_dev = pdev;
pci_set_drvdata(pdev, dev);
- dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
- dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
+
+ if (!dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)))
+ dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
+ else if (!dma_set_mask(&pdev->dev, DMA_BIT_MASK(32)))
+ dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
+ else
+ goto disable;
+
result = nvme_set_instance(dev);
if (result)
goto disable;
@@ -1977,7 +2008,10 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id)
unmap:
iounmap(dev->bar);
disable_msix:
- pci_disable_msix(pdev);
+ if (dev->pci_dev->msi_enabled)
+ pci_disable_msi(dev->pci_dev);
+ else if (dev->pci_dev->msix_enabled)
+ pci_disable_msix(dev->pci_dev);
nvme_release_instance(dev);
nvme_release_prp_pools(dev);
disable:
diff --git a/drivers/block/nvme-scsi.c b/drivers/block/nvme-scsi.c
index fed54b03989..102de2f52b5 100644
--- a/drivers/block/nvme-scsi.c
+++ b/drivers/block/nvme-scsi.c
@@ -44,7 +44,6 @@
#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/types.h>
-#include <linux/version.h>
#include <scsi/sg.h>
#include <scsi/scsi.h>
@@ -1654,7 +1653,7 @@ static void nvme_trans_modesel_save_bd(struct nvme_ns *ns, u8 *parm_list,
}
}
-static u16 nvme_trans_modesel_get_mp(struct nvme_ns *ns, struct sg_io_hdr *hdr,
+static int nvme_trans_modesel_get_mp(struct nvme_ns *ns, struct sg_io_hdr *hdr,
u8 *mode_page, u8 page_code)
{
int res = SNTI_TRANSLATION_SUCCESS;
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index 3c08983e600..f5d0ea11d9f 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -83,7 +83,8 @@
#define MAX_SPEED 0xffff
-#define ZONE(sector, pd) (((sector) + (pd)->offset) & ~((pd)->settings.size - 1))
+#define ZONE(sector, pd) (((sector) + (pd)->offset) & \
+ ~(sector_t)((pd)->settings.size - 1))
static DEFINE_MUTEX(pktcdvd_mutex);
static struct pktcdvd_device *pkt_devs[MAX_WRITERS];
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index ca63104136e..4ad2ad9a5bb 100644
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -55,6 +55,39 @@
#define SECTOR_SHIFT 9
#define SECTOR_SIZE (1ULL << SECTOR_SHIFT)
+/*
+ * Increment the given counter and return its updated value.
+ * If the counter is already 0 it will not be incremented.
+ * If the counter is already at its maximum value returns
+ * -EINVAL without updating it.
+ */
+static int atomic_inc_return_safe(atomic_t *v)
+{
+ unsigned int counter;
+
+ counter = (unsigned int)__atomic_add_unless(v, 1, 0);
+ if (counter <= (unsigned int)INT_MAX)
+ return (int)counter;
+
+ atomic_dec(v);
+
+ return -EINVAL;
+}
+
+/* Decrement the counter. Return the resulting value, or -EINVAL */
+static int atomic_dec_return_safe(atomic_t *v)
+{
+ int counter;
+
+ counter = atomic_dec_return(v);
+ if (counter >= 0)
+ return counter;
+
+ atomic_inc(v);
+
+ return -EINVAL;
+}
+
#define RBD_DRV_NAME "rbd"
#define RBD_DRV_NAME_LONG "rbd (rados block device)"
@@ -100,21 +133,20 @@
* block device image metadata (in-memory version)
*/
struct rbd_image_header {
- /* These four fields never change for a given rbd image */
+ /* These six fields never change for a given rbd image */
char *object_prefix;
- u64 features;
__u8 obj_order;
__u8 crypt_type;
__u8 comp_type;
+ u64 stripe_unit;
+ u64 stripe_count;
+ u64 features; /* Might be changeable someday? */
/* The remaining fields need to be updated occasionally */
u64 image_size;
struct ceph_snap_context *snapc;
- char *snap_names;
- u64 *snap_sizes;
-
- u64 stripe_unit;
- u64 stripe_count;
+ char *snap_names; /* format 1 only */
+ u64 *snap_sizes; /* format 1 only */
};
/*
@@ -225,6 +257,7 @@ struct rbd_obj_request {
};
};
struct page **copyup_pages;
+ u32 copyup_page_count;
struct ceph_osd_request *osd_req;
@@ -257,6 +290,7 @@ struct rbd_img_request {
struct rbd_obj_request *obj_request; /* obj req initiator */
};
struct page **copyup_pages;
+ u32 copyup_page_count;
spinlock_t completion_lock;/* protects next_completion */
u32 next_completion;
rbd_img_callback_t callback;
@@ -311,6 +345,7 @@ struct rbd_device {
struct rbd_spec *parent_spec;
u64 parent_overlap;
+ atomic_t parent_ref;
struct rbd_device *parent;
/* protects updating the header */
@@ -337,7 +372,7 @@ enum rbd_dev_flags {
RBD_DEV_FLAG_REMOVING, /* this mapping is being removed */
};
-static DEFINE_MUTEX(ctl_mutex); /* Serialize open/close/setup/teardown */
+static DEFINE_MUTEX(client_mutex); /* Serialize client creation */
static LIST_HEAD(rbd_dev_list); /* devices */
static DEFINE_SPINLOCK(rbd_dev_list_lock);
@@ -359,7 +394,8 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf,
size_t count);
static ssize_t rbd_remove(struct bus_type *bus, const char *buf,
size_t count);
-static int rbd_dev_image_probe(struct rbd_device *rbd_dev);
+static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping);
+static void rbd_spec_put(struct rbd_spec *spec);
static struct bus_attribute rbd_bus_attrs[] = {
__ATTR(add, S_IWUSR, NULL, rbd_add),
@@ -426,7 +462,8 @@ static void rbd_img_parent_read(struct rbd_obj_request *obj_request);
static void rbd_dev_remove_parent(struct rbd_device *rbd_dev);
static int rbd_dev_refresh(struct rbd_device *rbd_dev);
-static int rbd_dev_v2_refresh(struct rbd_device *rbd_dev);
+static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev);
+static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev);
static const char *rbd_dev_v2_snap_name(struct rbd_device *rbd_dev,
u64 snap_id);
static int _rbd_dev_v2_snap_size(struct rbd_device *rbd_dev, u64 snap_id,
@@ -452,10 +489,8 @@ static int rbd_open(struct block_device *bdev, fmode_t mode)
if (removing)
return -ENOENT;
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
(void) get_device(&rbd_dev->dev);
set_device_ro(bdev, rbd_dev->mapping.read_only);
- mutex_unlock(&ctl_mutex);
return 0;
}
@@ -470,9 +505,7 @@ static void rbd_release(struct gendisk *disk, fmode_t mode)
spin_unlock_irq(&rbd_dev->lock);
rbd_assert(open_count_before > 0);
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
put_device(&rbd_dev->dev);
- mutex_unlock(&ctl_mutex);
}
static const struct block_device_operations rbd_bd_ops = {
@@ -482,8 +515,8 @@ static const struct block_device_operations rbd_bd_ops = {
};
/*
- * Initialize an rbd client instance.
- * We own *ceph_opts.
+ * Initialize an rbd client instance. Success or not, this function
+ * consumes ceph_opts. Caller holds client_mutex.
*/
static struct rbd_client *rbd_client_create(struct ceph_options *ceph_opts)
{
@@ -498,30 +531,25 @@ static struct rbd_client *rbd_client_create(struct ceph_options *ceph_opts)
kref_init(&rbdc->kref);
INIT_LIST_HEAD(&rbdc->node);
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-
rbdc->client = ceph_create_client(ceph_opts, rbdc, 0, 0);
if (IS_ERR(rbdc->client))
- goto out_mutex;
+ goto out_rbdc;
ceph_opts = NULL; /* Now rbdc->client is responsible for ceph_opts */
ret = ceph_open_session(rbdc->client);
if (ret < 0)
- goto out_err;
+ goto out_client;
spin_lock(&rbd_client_list_lock);
list_add_tail(&rbdc->node, &rbd_client_list);
spin_unlock(&rbd_client_list_lock);
- mutex_unlock(&ctl_mutex);
dout("%s: rbdc %p\n", __func__, rbdc);
return rbdc;
-
-out_err:
+out_client:
ceph_destroy_client(rbdc->client);
-out_mutex:
- mutex_unlock(&ctl_mutex);
+out_rbdc:
kfree(rbdc);
out_opt:
if (ceph_opts)
@@ -638,17 +666,20 @@ static int parse_rbd_opts_token(char *c, void *private)
/*
* Get a ceph client with specific addr and configuration, if one does
- * not exist create it.
+ * not exist create it. Either way, ceph_opts is consumed by this
+ * function.
*/
static struct rbd_client *rbd_get_client(struct ceph_options *ceph_opts)
{
struct rbd_client *rbdc;
+ mutex_lock_nested(&client_mutex, SINGLE_DEPTH_NESTING);
rbdc = rbd_client_find(ceph_opts);
if (rbdc) /* using an existing client */
ceph_destroy_options(ceph_opts);
else
rbdc = rbd_client_create(ceph_opts);
+ mutex_unlock(&client_mutex);
return rbdc;
}
@@ -726,88 +757,120 @@ static bool rbd_dev_ondisk_valid(struct rbd_image_header_ondisk *ondisk)
}
/*
- * Create a new header structure, translate header format from the on-disk
- * header.
+ * Fill an rbd image header with information from the given format 1
+ * on-disk header.
*/
-static int rbd_header_from_disk(struct rbd_image_header *header,
+static int rbd_header_from_disk(struct rbd_device *rbd_dev,
struct rbd_image_header_ondisk *ondisk)
{
+ struct rbd_image_header *header = &rbd_dev->header;
+ bool first_time = header->object_prefix == NULL;
+ struct ceph_snap_context *snapc;
+ char *object_prefix = NULL;
+ char *snap_names = NULL;
+ u64 *snap_sizes = NULL;
u32 snap_count;
- size_t len;
size_t size;
+ int ret = -ENOMEM;
u32 i;
- memset(header, 0, sizeof (*header));
+ /* Allocate this now to avoid having to handle failure below */
- snap_count = le32_to_cpu(ondisk->snap_count);
+ if (first_time) {
+ size_t len;
- len = strnlen(ondisk->object_prefix, sizeof (ondisk->object_prefix));
- header->object_prefix = kmalloc(len + 1, GFP_KERNEL);
- if (!header->object_prefix)
- return -ENOMEM;
- memcpy(header->object_prefix, ondisk->object_prefix, len);
- header->object_prefix[len] = '\0';
+ len = strnlen(ondisk->object_prefix,
+ sizeof (ondisk->object_prefix));
+ object_prefix = kmalloc(len + 1, GFP_KERNEL);
+ if (!object_prefix)
+ return -ENOMEM;
+ memcpy(object_prefix, ondisk->object_prefix, len);
+ object_prefix[len] = '\0';
+ }
+ /* Allocate the snapshot context and fill it in */
+
+ snap_count = le32_to_cpu(ondisk->snap_count);
+ snapc = ceph_create_snap_context(snap_count, GFP_KERNEL);
+ if (!snapc)
+ goto out_err;
+ snapc->seq = le64_to_cpu(ondisk->snap_seq);
if (snap_count) {
+ struct rbd_image_snap_ondisk *snaps;
u64 snap_names_len = le64_to_cpu(ondisk->snap_names_len);
- /* Save a copy of the snapshot names */
+ /* We'll keep a copy of the snapshot names... */
- if (snap_names_len > (u64) SIZE_MAX)
- return -EIO;
- header->snap_names = kmalloc(snap_names_len, GFP_KERNEL);
- if (!header->snap_names)
+ if (snap_names_len > (u64)SIZE_MAX)
+ goto out_2big;
+ snap_names = kmalloc(snap_names_len, GFP_KERNEL);
+ if (!snap_names)
+ goto out_err;
+
+ /* ...as well as the array of their sizes. */
+
+ size = snap_count * sizeof (*header->snap_sizes);
+ snap_sizes = kmalloc(size, GFP_KERNEL);
+ if (!snap_sizes)
goto out_err;
+
/*
- * Note that rbd_dev_v1_header_read() guarantees
- * the ondisk buffer we're working with has
+ * Copy the names, and fill in each snapshot's id
+ * and size.
+ *
+ * Note that rbd_dev_v1_header_info() guarantees the
+ * ondisk buffer we're working with has
* snap_names_len bytes beyond the end of the
* snapshot id array, this memcpy() is safe.
*/
- memcpy(header->snap_names, &ondisk->snaps[snap_count],
- snap_names_len);
-
- /* Record each snapshot's size */
+ memcpy(snap_names, &ondisk->snaps[snap_count], snap_names_len);
+ snaps = ondisk->snaps;
+ for (i = 0; i < snap_count; i++) {
+ snapc->snaps[i] = le64_to_cpu(snaps[i].id);
+ snap_sizes[i] = le64_to_cpu(snaps[i].image_size);
+ }
+ }
- size = snap_count * sizeof (*header->snap_sizes);
- header->snap_sizes = kmalloc(size, GFP_KERNEL);
- if (!header->snap_sizes)
- goto out_err;
- for (i = 0; i < snap_count; i++)
- header->snap_sizes[i] =
- le64_to_cpu(ondisk->snaps[i].image_size);
+ /* We won't fail any more, fill in the header */
+
+ if (first_time) {
+ header->object_prefix = object_prefix;
+ header->obj_order = ondisk->options.order;
+ header->crypt_type = ondisk->options.crypt_type;
+ header->comp_type = ondisk->options.comp_type;
+ /* The rest aren't used for format 1 images */
+ header->stripe_unit = 0;
+ header->stripe_count = 0;
+ header->features = 0;
} else {
- header->snap_names = NULL;
- header->snap_sizes = NULL;
+ ceph_put_snap_context(header->snapc);
+ kfree(header->snap_names);
+ kfree(header->snap_sizes);
}
- header->features = 0; /* No features support in v1 images */
- header->obj_order = ondisk->options.order;
- header->crypt_type = ondisk->options.crypt_type;
- header->comp_type = ondisk->options.comp_type;
-
- /* Allocate and fill in the snapshot context */
+ /* The remaining fields always get updated (when we refresh) */
header->image_size = le64_to_cpu(ondisk->image_size);
+ header->snapc = snapc;
+ header->snap_names = snap_names;
+ header->snap_sizes = snap_sizes;
- header->snapc = ceph_create_snap_context(snap_count, GFP_KERNEL);
- if (!header->snapc)
- goto out_err;
- header->snapc->seq = le64_to_cpu(ondisk->snap_seq);
- for (i = 0; i < snap_count; i++)
- header->snapc->snaps[i] = le64_to_cpu(ondisk->snaps[i].id);
+ /* Make sure mapping size is consistent with header info */
- return 0;
+ if (rbd_dev->spec->snap_id == CEPH_NOSNAP || first_time)
+ if (rbd_dev->mapping.size != header->image_size)
+ rbd_dev->mapping.size = header->image_size;
+ return 0;
+out_2big:
+ ret = -EIO;
out_err:
- kfree(header->snap_sizes);
- header->snap_sizes = NULL;
- kfree(header->snap_names);
- header->snap_names = NULL;
- kfree(header->object_prefix);
- header->object_prefix = NULL;
+ kfree(snap_sizes);
+ kfree(snap_names);
+ ceph_put_snap_context(snapc);
+ kfree(object_prefix);
- return -ENOMEM;
+ return ret;
}
static const char *_rbd_dev_v1_snap_name(struct rbd_device *rbd_dev, u32 which)
@@ -934,20 +997,11 @@ static int rbd_snap_features(struct rbd_device *rbd_dev, u64 snap_id,
static int rbd_dev_mapping_set(struct rbd_device *rbd_dev)
{
- const char *snap_name = rbd_dev->spec->snap_name;
- u64 snap_id;
+ u64 snap_id = rbd_dev->spec->snap_id;
u64 size = 0;
u64 features = 0;
int ret;
- if (strcmp(snap_name, RBD_SNAP_HEAD_NAME)) {
- snap_id = rbd_snap_id_by_name(rbd_dev, snap_name);
- if (snap_id == CEPH_NOSNAP)
- return -ENOENT;
- } else {
- snap_id = CEPH_NOSNAP;
- }
-
ret = rbd_snap_size(rbd_dev, snap_id, &size);
if (ret)
return ret;
@@ -958,11 +1012,6 @@ static int rbd_dev_mapping_set(struct rbd_device *rbd_dev)
rbd_dev->mapping.size = size;
rbd_dev->mapping.features = features;
- /* If we are mapping a snapshot it must be marked read-only */
-
- if (snap_id != CEPH_NOSNAP)
- rbd_dev->mapping.read_only = true;
-
return 0;
}
@@ -970,14 +1019,6 @@ static void rbd_dev_mapping_clear(struct rbd_device *rbd_dev)
{
rbd_dev->mapping.size = 0;
rbd_dev->mapping.features = 0;
- rbd_dev->mapping.read_only = true;
-}
-
-static void rbd_dev_clear_mapping(struct rbd_device *rbd_dev)
-{
- rbd_dev->mapping.size = 0;
- rbd_dev->mapping.features = 0;
- rbd_dev->mapping.read_only = true;
}
static const char *rbd_segment_name(struct rbd_device *rbd_dev, u64 offset)
@@ -985,12 +1026,16 @@ static const char *rbd_segment_name(struct rbd_device *rbd_dev, u64 offset)
char *name;
u64 segment;
int ret;
+ char *name_format;
name = kmem_cache_alloc(rbd_segment_name_cache, GFP_NOIO);
if (!name)
return NULL;
segment = offset >> rbd_dev->header.obj_order;
- ret = snprintf(name, MAX_OBJ_NAME_SIZE + 1, "%s.%012llx",
+ name_format = "%s.%012llx";
+ if (rbd_dev->image_format == 2)
+ name_format = "%s.%016llx";
+ ret = snprintf(name, MAX_OBJ_NAME_SIZE + 1, name_format,
rbd_dev->header.object_prefix, segment);
if (ret < 0 || ret > MAX_OBJ_NAME_SIZE) {
pr_err("error formatting segment name for #%llu (%d)\n",
@@ -1071,6 +1116,7 @@ static void zero_bio_chain(struct bio *chain, int start_ofs)
buf = bvec_kmap_irq(bv, &flags);
memset(buf + remainder, 0,
bv->bv_len - remainder);
+ flush_dcache_page(bv->bv_page);
bvec_kunmap_irq(buf, &flags);
}
pos += bv->bv_len;
@@ -1098,11 +1144,12 @@ static void zero_pages(struct page **pages, u64 offset, u64 end)
unsigned long flags;
void *kaddr;
- page_offset = (size_t)(offset & ~PAGE_MASK);
- length = min(PAGE_SIZE - page_offset, (size_t)(end - offset));
+ page_offset = offset & ~PAGE_MASK;
+ length = min_t(size_t, PAGE_SIZE - page_offset, end - offset);
local_irq_save(flags);
kaddr = kmap_atomic(*page);
memset(kaddr + page_offset, 0, length);
+ flush_dcache_page(*page);
kunmap_atomic(kaddr);
local_irq_restore(flags);
@@ -1342,20 +1389,18 @@ static void rbd_obj_request_put(struct rbd_obj_request *obj_request)
kref_put(&obj_request->kref, rbd_obj_request_destroy);
}
-static void rbd_img_request_get(struct rbd_img_request *img_request)
-{
- dout("%s: img %p (was %d)\n", __func__, img_request,
- atomic_read(&img_request->kref.refcount));
- kref_get(&img_request->kref);
-}
-
+static bool img_request_child_test(struct rbd_img_request *img_request);
+static void rbd_parent_request_destroy(struct kref *kref);
static void rbd_img_request_destroy(struct kref *kref);
static void rbd_img_request_put(struct rbd_img_request *img_request)
{
rbd_assert(img_request != NULL);
dout("%s: img %p (was %d)\n", __func__, img_request,
atomic_read(&img_request->kref.refcount));
- kref_put(&img_request->kref, rbd_img_request_destroy);
+ if (img_request_child_test(img_request))
+ kref_put(&img_request->kref, rbd_parent_request_destroy);
+ else
+ kref_put(&img_request->kref, rbd_img_request_destroy);
}
static inline void rbd_img_obj_request_add(struct rbd_img_request *img_request,
@@ -1472,6 +1517,12 @@ static void img_request_child_set(struct rbd_img_request *img_request)
smp_mb();
}
+static void img_request_child_clear(struct rbd_img_request *img_request)
+{
+ clear_bit(IMG_REQ_CHILD, &img_request->flags);
+ smp_mb();
+}
+
static bool img_request_child_test(struct rbd_img_request *img_request)
{
smp_mb();
@@ -1484,6 +1535,12 @@ static void img_request_layered_set(struct rbd_img_request *img_request)
smp_mb();
}
+static void img_request_layered_clear(struct rbd_img_request *img_request)
+{
+ clear_bit(IMG_REQ_LAYERED, &img_request->flags);
+ smp_mb();
+}
+
static bool img_request_layered_test(struct rbd_img_request *img_request)
{
smp_mb();
@@ -1827,6 +1884,74 @@ static void rbd_obj_request_destroy(struct kref *kref)
kmem_cache_free(rbd_obj_request_cache, obj_request);
}
+/* It's OK to call this for a device with no parent */
+
+static void rbd_spec_put(struct rbd_spec *spec);
+static void rbd_dev_unparent(struct rbd_device *rbd_dev)
+{
+ rbd_dev_remove_parent(rbd_dev);
+ rbd_spec_put(rbd_dev->parent_spec);
+ rbd_dev->parent_spec = NULL;
+ rbd_dev->parent_overlap = 0;
+}
+
+/*
+ * Parent image reference counting is used to determine when an
+ * image's parent fields can be safely torn down--after there are no
+ * more in-flight requests to the parent image. When the last
+ * reference is dropped, cleaning them up is safe.
+ */
+static void rbd_dev_parent_put(struct rbd_device *rbd_dev)
+{
+ int counter;
+
+ if (!rbd_dev->parent_spec)
+ return;
+
+ counter = atomic_dec_return_safe(&rbd_dev->parent_ref);
+ if (counter > 0)
+ return;
+
+ /* Last reference; clean up parent data structures */
+
+ if (!counter)
+ rbd_dev_unparent(rbd_dev);
+ else
+ rbd_warn(rbd_dev, "parent reference underflow\n");
+}
+
+/*
+ * If an image has a non-zero parent overlap, get a reference to its
+ * parent.
+ *
+ * We must get the reference before checking for the overlap to
+ * coordinate properly with zeroing the parent overlap in
+ * rbd_dev_v2_parent_info() when an image gets flattened. We
+ * drop it again if there is no overlap.
+ *
+ * Returns true if the rbd device has a parent with a non-zero
+ * overlap and a reference for it was successfully taken, or
+ * false otherwise.
+ */
+static bool rbd_dev_parent_get(struct rbd_device *rbd_dev)
+{
+ int counter;
+
+ if (!rbd_dev->parent_spec)
+ return false;
+
+ counter = atomic_inc_return_safe(&rbd_dev->parent_ref);
+ if (counter > 0 && rbd_dev->parent_overlap)
+ return true;
+
+ /* Image was flattened, but parent is not yet torn down */
+
+ if (counter < 0)
+ rbd_warn(rbd_dev, "parent reference overflow\n");
+
+ return false;
+}
+
/*
* Caller is responsible for filling in the list of object requests
* that comprises the image request, and the Linux request pointer
@@ -1835,8 +1960,7 @@ static void rbd_obj_request_destroy(struct kref *kref)
static struct rbd_img_request *rbd_img_request_create(
struct rbd_device *rbd_dev,
u64 offset, u64 length,
- bool write_request,
- bool child_request)
+ bool write_request)
{
struct rbd_img_request *img_request;
@@ -1861,9 +1985,7 @@ static struct rbd_img_request *rbd_img_request_create(
} else {
img_request->snap_id = rbd_dev->spec->snap_id;
}
- if (child_request)
- img_request_child_set(img_request);
- if (rbd_dev->parent_spec)
+ if (rbd_dev_parent_get(rbd_dev))
img_request_layered_set(img_request);
spin_lock_init(&img_request->completion_lock);
img_request->next_completion = 0;
@@ -1873,9 +1995,6 @@ static struct rbd_img_request *rbd_img_request_create(
INIT_LIST_HEAD(&img_request->obj_requests);
kref_init(&img_request->kref);
- rbd_img_request_get(img_request); /* Avoid a warning */
- rbd_img_request_put(img_request); /* TEMPORARY */
-
dout("%s: rbd_dev %p %s %llu/%llu -> img %p\n", __func__, rbd_dev,
write_request ? "write" : "read", offset, length,
img_request);
@@ -1897,15 +2016,54 @@ static void rbd_img_request_destroy(struct kref *kref)
rbd_img_obj_request_del(img_request, obj_request);
rbd_assert(img_request->obj_request_count == 0);
+ if (img_request_layered_test(img_request)) {
+ img_request_layered_clear(img_request);
+ rbd_dev_parent_put(img_request->rbd_dev);
+ }
+
if (img_request_write_test(img_request))
ceph_put_snap_context(img_request->snapc);
- if (img_request_child_test(img_request))
- rbd_obj_request_put(img_request->obj_request);
-
kmem_cache_free(rbd_img_request_cache, img_request);
}
+static struct rbd_img_request *rbd_parent_request_create(
+ struct rbd_obj_request *obj_request,
+ u64 img_offset, u64 length)
+{
+ struct rbd_img_request *parent_request;
+ struct rbd_device *rbd_dev;
+
+ rbd_assert(obj_request->img_request);
+ rbd_dev = obj_request->img_request->rbd_dev;
+
+ parent_request = rbd_img_request_create(rbd_dev->parent,
+ img_offset, length, false);
+ if (!parent_request)
+ return NULL;
+
+ img_request_child_set(parent_request);
+ rbd_obj_request_get(obj_request);
+ parent_request->obj_request = obj_request;
+
+ return parent_request;
+}
+
+static void rbd_parent_request_destroy(struct kref *kref)
+{
+ struct rbd_img_request *parent_request;
+ struct rbd_obj_request *orig_request;
+
+ parent_request = container_of(kref, struct rbd_img_request, kref);
+ orig_request = parent_request->obj_request;
+
+ parent_request->obj_request = NULL;
+ rbd_obj_request_put(orig_request);
+ img_request_child_clear(parent_request);
+
+ rbd_img_request_destroy(kref);
+}
+
static bool rbd_img_obj_end_request(struct rbd_obj_request *obj_request)
{
struct rbd_img_request *img_request;
@@ -2005,9 +2163,9 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
struct rbd_obj_request *obj_request = NULL;
struct rbd_obj_request *next_obj_request;
bool write_request = img_request_write_test(img_request);
- struct bio *bio_list;
+ struct bio *bio_list = 0;
unsigned int bio_offset = 0;
- struct page **pages;
+ struct page **pages = 0;
u64 img_offset;
u64 resid;
u16 opcode;
@@ -2086,13 +2244,17 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request,
obj_request->pages, length,
offset & ~PAGE_MASK, false, false);
+ /*
+ * set obj_request->img_request before formatting
+ * the osd_request so that it gets the right snapc
+ */
+ rbd_img_obj_request_add(img_request, obj_request);
if (write_request)
rbd_osd_req_format_write(obj_request);
else
rbd_osd_req_format_read(obj_request);
obj_request->img_offset = img_offset;
- rbd_img_obj_request_add(img_request, obj_request);
img_offset += length;
resid -= length;
@@ -2114,7 +2276,7 @@ rbd_img_obj_copyup_callback(struct rbd_obj_request *obj_request)
{
struct rbd_img_request *img_request;
struct rbd_device *rbd_dev;
- u64 length;
+ struct page **pages;
u32 page_count;
rbd_assert(obj_request->type == OBJ_REQUEST_BIO);
@@ -2124,12 +2286,14 @@ rbd_img_obj_copyup_callback(struct rbd_obj_request *obj_request)
rbd_dev = img_request->rbd_dev;
rbd_assert(rbd_dev);
- length = (u64)1 << rbd_dev->header.obj_order;
- page_count = (u32)calc_pages_for(0, length);
- rbd_assert(obj_request->copyup_pages);
- ceph_release_page_vector(obj_request->copyup_pages, page_count);
+ pages = obj_request->copyup_pages;
+ rbd_assert(pages != NULL);
obj_request->copyup_pages = NULL;
+ page_count = obj_request->copyup_page_count;
+ rbd_assert(page_count);
+ obj_request->copyup_page_count = 0;
+ ceph_release_page_vector(pages, page_count);
/*
* We want the transfer count to reflect the size of the
@@ -2153,9 +2317,11 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request)
struct ceph_osd_client *osdc;
struct rbd_device *rbd_dev;
struct page **pages;
- int result;
- u64 obj_size;
- u64 xferred;
+ u32 page_count;
+ int img_result;
+ u64 parent_length;
+ u64 offset;
+ u64 length;
rbd_assert(img_request_child_test(img_request));
@@ -2164,46 +2330,74 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request)
pages = img_request->copyup_pages;
rbd_assert(pages != NULL);
img_request->copyup_pages = NULL;
+ page_count = img_request->copyup_page_count;
+ rbd_assert(page_count);
+ img_request->copyup_page_count = 0;
orig_request = img_request->obj_request;
rbd_assert(orig_request != NULL);
- rbd_assert(orig_request->type == OBJ_REQUEST_BIO);
- result = img_request->result;
- obj_size = img_request->length;
- xferred = img_request->xferred;
+ rbd_assert(obj_request_type_valid(orig_request->type));
+ img_result = img_request->result;
+ parent_length = img_request->length;
+ rbd_assert(parent_length == img_request->xferred);
+ rbd_img_request_put(img_request);
- rbd_dev = img_request->rbd_dev;
+ rbd_assert(orig_request->img_request);
+ rbd_dev = orig_request->img_request->rbd_dev;
rbd_assert(rbd_dev);
- rbd_assert(obj_size == (u64)1 << rbd_dev->header.obj_order);
- rbd_img_request_put(img_request);
+ /*
+ * If the overlap has become 0 (most likely because the
+ * image has been flattened) we need to free the pages
+ * and re-submit the original write request.
+ */
+ if (!rbd_dev->parent_overlap) {
+ struct ceph_osd_client *osdc;
- if (result)
- goto out_err;
+ ceph_release_page_vector(pages, page_count);
+ osdc = &rbd_dev->rbd_client->client->osdc;
+ img_result = rbd_obj_request_submit(osdc, orig_request);
+ if (!img_result)
+ return;
+ }
- /* Allocate the new copyup osd request for the original request */
+ if (img_result)
+ goto out_err;
- result = -ENOMEM;
- rbd_assert(!orig_request->osd_req);
+ /*
+ * The original osd request is of no use to use any more.
+ * We need a new one that can hold the two ops in a copyup
+ * request. Allocate the new copyup osd request for the
+ * original request, and release the old one.
+ */
+ img_result = -ENOMEM;
osd_req = rbd_osd_req_create_copyup(orig_request);
if (!osd_req)
goto out_err;
+ rbd_osd_req_destroy(orig_request->osd_req);
orig_request->osd_req = osd_req;
orig_request->copyup_pages = pages;
+ orig_request->copyup_page_count = page_count;
/* Initialize the copyup op */
osd_req_op_cls_init(osd_req, 0, CEPH_OSD_OP_CALL, "rbd", "copyup");
- osd_req_op_cls_request_data_pages(osd_req, 0, pages, obj_size, 0,
+ osd_req_op_cls_request_data_pages(osd_req, 0, pages, parent_length, 0,
false, false);
/* Then the original write request op */
+ offset = orig_request->offset;
+ length = orig_request->length;
osd_req_op_extent_init(osd_req, 1, CEPH_OSD_OP_WRITE,
- orig_request->offset,
- orig_request->length, 0, 0);
- osd_req_op_extent_osd_data_bio(osd_req, 1, orig_request->bio_list,
- orig_request->length);
+ offset, length, 0, 0);
+ if (orig_request->type == OBJ_REQUEST_BIO)
+ osd_req_op_extent_osd_data_bio(osd_req, 1,
+ orig_request->bio_list, length);
+ else
+ osd_req_op_extent_osd_data_pages(osd_req, 1,
+ orig_request->pages, length,
+ offset & ~PAGE_MASK, false, false);
rbd_osd_req_format_write(orig_request);
@@ -2211,13 +2405,13 @@ rbd_img_obj_parent_read_full_callback(struct rbd_img_request *img_request)
orig_request->callback = rbd_img_obj_copyup_callback;
osdc = &rbd_dev->rbd_client->client->osdc;
- result = rbd_obj_request_submit(osdc, orig_request);
- if (!result)
+ img_result = rbd_obj_request_submit(osdc, orig_request);
+ if (!img_result)
return;
out_err:
/* Record the error code and complete the request */
- orig_request->result = result;
+ orig_request->result = img_result;
orig_request->xferred = 0;
obj_request_done_set(orig_request);
rbd_obj_request_complete(orig_request);
@@ -2249,7 +2443,7 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request)
int result;
rbd_assert(obj_request_img_data_test(obj_request));
- rbd_assert(obj_request->type == OBJ_REQUEST_BIO);
+ rbd_assert(obj_request_type_valid(obj_request->type));
img_request = obj_request->img_request;
rbd_assert(img_request != NULL);
@@ -2257,15 +2451,6 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request)
rbd_assert(rbd_dev->parent != NULL);
/*
- * First things first. The original osd request is of no
- * use to use any more, we'll need a new one that can hold
- * the two ops in a copyup request. We'll get that later,
- * but for now we can release the old one.
- */
- rbd_osd_req_destroy(obj_request->osd_req);
- obj_request->osd_req = NULL;
-
- /*
* Determine the byte range covered by the object in the
* child image to which the original request was to be sent.
*/
@@ -2295,18 +2480,16 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request)
}
result = -ENOMEM;
- parent_request = rbd_img_request_create(rbd_dev->parent,
- img_offset, length,
- false, true);
+ parent_request = rbd_parent_request_create(obj_request,
+ img_offset, length);
if (!parent_request)
goto out_err;
- rbd_obj_request_get(obj_request);
- parent_request->obj_request = obj_request;
result = rbd_img_request_fill(parent_request, OBJ_REQUEST_PAGES, pages);
if (result)
goto out_err;
parent_request->copyup_pages = pages;
+ parent_request->copyup_page_count = page_count;
parent_request->callback = rbd_img_obj_parent_read_full_callback;
result = rbd_img_request_submit(parent_request);
@@ -2314,6 +2497,7 @@ static int rbd_img_obj_parent_read_full(struct rbd_obj_request *obj_request)
return 0;
parent_request->copyup_pages = NULL;
+ parent_request->copyup_page_count = 0;
parent_request->obj_request = NULL;
rbd_obj_request_put(obj_request);
out_err:
@@ -2331,6 +2515,7 @@ out_err:
static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request)
{
struct rbd_obj_request *orig_request;
+ struct rbd_device *rbd_dev;
int result;
rbd_assert(!obj_request_img_data_test(obj_request));
@@ -2342,6 +2527,7 @@ static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request)
*/
orig_request = obj_request->obj_request;
obj_request->obj_request = NULL;
+ rbd_obj_request_put(orig_request);
rbd_assert(orig_request);
rbd_assert(orig_request->img_request);
@@ -2353,8 +2539,20 @@ static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request)
obj_request->xferred, obj_request->length);
rbd_obj_request_put(obj_request);
- rbd_assert(orig_request);
- rbd_assert(orig_request->img_request);
+ /*
+ * If the overlap has become 0 (most likely because the
+ * image has been flattened) we need to free the pages
+ * and re-submit the original write request.
+ */
+ rbd_dev = orig_request->img_request->rbd_dev;
+ if (!rbd_dev->parent_overlap) {
+ struct ceph_osd_client *osdc;
+
+ osdc = &rbd_dev->rbd_client->client->osdc;
+ result = rbd_obj_request_submit(osdc, orig_request);
+ if (!result)
+ return;
+ }
/*
* Our only purpose here is to determine whether the object
@@ -2379,7 +2577,6 @@ static void rbd_img_obj_exists_callback(struct rbd_obj_request *obj_request)
out:
if (orig_request->result)
rbd_obj_request_complete(orig_request);
- rbd_obj_request_put(orig_request);
}
static int rbd_img_obj_exists_submit(struct rbd_obj_request *obj_request)
@@ -2512,14 +2709,36 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request)
struct rbd_obj_request *obj_request;
struct rbd_device *rbd_dev;
u64 obj_end;
+ u64 img_xferred;
+ int img_result;
rbd_assert(img_request_child_test(img_request));
+ /* First get what we need from the image request and release it */
+
obj_request = img_request->obj_request;
+ img_xferred = img_request->xferred;
+ img_result = img_request->result;
+ rbd_img_request_put(img_request);
+
+ /*
+ * If the overlap has become 0 (most likely because the
+ * image has been flattened) we need to re-submit the
+ * original request.
+ */
rbd_assert(obj_request);
rbd_assert(obj_request->img_request);
+ rbd_dev = obj_request->img_request->rbd_dev;
+ if (!rbd_dev->parent_overlap) {
+ struct ceph_osd_client *osdc;
- obj_request->result = img_request->result;
+ osdc = &rbd_dev->rbd_client->client->osdc;
+ img_result = rbd_obj_request_submit(osdc, obj_request);
+ if (!img_result)
+ return;
+ }
+
+ obj_request->result = img_result;
if (obj_request->result)
goto out;
@@ -2532,7 +2751,6 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request)
*/
rbd_assert(obj_request->img_offset < U64_MAX - obj_request->length);
obj_end = obj_request->img_offset + obj_request->length;
- rbd_dev = obj_request->img_request->rbd_dev;
if (obj_end > rbd_dev->parent_overlap) {
u64 xferred = 0;
@@ -2540,43 +2758,39 @@ static void rbd_img_parent_read_callback(struct rbd_img_request *img_request)
xferred = rbd_dev->parent_overlap -
obj_request->img_offset;
- obj_request->xferred = min(img_request->xferred, xferred);
+ obj_request->xferred = min(img_xferred, xferred);
} else {
- obj_request->xferred = img_request->xferred;
+ obj_request->xferred = img_xferred;
}
out:
- rbd_img_request_put(img_request);
rbd_img_obj_request_read_callback(obj_request);
rbd_obj_request_complete(obj_request);
}
static void rbd_img_parent_read(struct rbd_obj_request *obj_request)
{
- struct rbd_device *rbd_dev;
struct rbd_img_request *img_request;
int result;
rbd_assert(obj_request_img_data_test(obj_request));
rbd_assert(obj_request->img_request != NULL);
rbd_assert(obj_request->result == (s32) -ENOENT);
- rbd_assert(obj_request->type == OBJ_REQUEST_BIO);
+ rbd_assert(obj_request_type_valid(obj_request->type));
- rbd_dev = obj_request->img_request->rbd_dev;
- rbd_assert(rbd_dev->parent != NULL);
/* rbd_read_finish(obj_request, obj_request->length); */
- img_request = rbd_img_request_create(rbd_dev->parent,
+ img_request = rbd_parent_request_create(obj_request,
obj_request->img_offset,
- obj_request->length,
- false, true);
+ obj_request->length);
result = -ENOMEM;
if (!img_request)
goto out_err;
- rbd_obj_request_get(obj_request);
- img_request->obj_request = obj_request;
-
- result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO,
- obj_request->bio_list);
+ if (obj_request->type == OBJ_REQUEST_BIO)
+ result = rbd_img_request_fill(img_request, OBJ_REQUEST_BIO,
+ obj_request->bio_list);
+ else
+ result = rbd_img_request_fill(img_request, OBJ_REQUEST_PAGES,
+ obj_request->pages);
if (result)
goto out_err;
@@ -2626,6 +2840,7 @@ out:
static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data)
{
struct rbd_device *rbd_dev = (struct rbd_device *)data;
+ int ret;
if (!rbd_dev)
return;
@@ -2633,7 +2848,9 @@ static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data)
dout("%s: \"%s\" notify_id %llu opcode %u\n", __func__,
rbd_dev->header_name, (unsigned long long)notify_id,
(unsigned int)opcode);
- (void)rbd_dev_refresh(rbd_dev);
+ ret = rbd_dev_refresh(rbd_dev);
+ if (ret)
+ rbd_warn(rbd_dev, "header refresh error (%d)\n", ret);
rbd_obj_notify_ack(rbd_dev, notify_id);
}
@@ -2642,7 +2859,7 @@ static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data)
* Request sync osd watch/unwatch. The value of "start" determines
* whether a watch request is being initiated or torn down.
*/
-static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, int start)
+static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, bool start)
{
struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc;
struct rbd_obj_request *obj_request;
@@ -2676,7 +2893,7 @@ static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, int start)
rbd_dev->watch_request->osd_req);
osd_req_op_watch_init(obj_request->osd_req, 0, CEPH_OSD_OP_WATCH,
- rbd_dev->watch_event->cookie, 0, start);
+ rbd_dev->watch_event->cookie, 0, start ? 1 : 0);
rbd_osd_req_format_write(obj_request);
ret = rbd_obj_request_submit(osdc, obj_request);
@@ -2869,9 +3086,16 @@ static void rbd_request_fn(struct request_queue *q)
goto end_request; /* Shouldn't happen */
}
+ result = -EIO;
+ if (offset + length > rbd_dev->mapping.size) {
+ rbd_warn(rbd_dev, "beyond EOD (%llu~%llu > %llu)\n",
+ offset, length, rbd_dev->mapping.size);
+ goto end_request;
+ }
+
result = -ENOMEM;
img_request = rbd_img_request_create(rbd_dev, offset, length,
- write_request, false);
+ write_request);
if (!img_request)
goto end_request;
@@ -3022,17 +3246,11 @@ out:
}
/*
- * Read the complete header for the given rbd device.
- *
- * Returns a pointer to a dynamically-allocated buffer containing
- * the complete and validated header. Caller can pass the address
- * of a variable that will be filled in with the version of the
- * header object at the time it was read.
- *
- * Returns a pointer-coded errno if a failure occurs.
+ * Read the complete header for the given rbd device. On successful
+ * return, the rbd_dev->header field will contain up-to-date
+ * information about the image.
*/
-static struct rbd_image_header_ondisk *
-rbd_dev_v1_header_read(struct rbd_device *rbd_dev)
+static int rbd_dev_v1_header_info(struct rbd_device *rbd_dev)
{
struct rbd_image_header_ondisk *ondisk = NULL;
u32 snap_count = 0;
@@ -3057,22 +3275,22 @@ rbd_dev_v1_header_read(struct rbd_device *rbd_dev)
size += names_size;
ondisk = kmalloc(size, GFP_KERNEL);
if (!ondisk)
- return ERR_PTR(-ENOMEM);
+ return -ENOMEM;
ret = rbd_obj_read_sync(rbd_dev, rbd_dev->header_name,
0, size, ondisk);
if (ret < 0)
- goto out_err;
+ goto out;
if ((size_t)ret < size) {
ret = -ENXIO;
rbd_warn(rbd_dev, "short header read (want %zd got %d)",
size, ret);
- goto out_err;
+ goto out;
}
if (!rbd_dev_ondisk_valid(ondisk)) {
ret = -ENXIO;
rbd_warn(rbd_dev, "invalid header");
- goto out_err;
+ goto out;
}
names_size = le64_to_cpu(ondisk->snap_names_len);
@@ -3080,85 +3298,13 @@ rbd_dev_v1_header_read(struct rbd_device *rbd_dev)
snap_count = le32_to_cpu(ondisk->snap_count);
} while (snap_count != want_count);
- return ondisk;
-
-out_err:
- kfree(ondisk);
-
- return ERR_PTR(ret);
-}
-
-/*
- * reload the ondisk the header
- */
-static int rbd_read_header(struct rbd_device *rbd_dev,
- struct rbd_image_header *header)
-{
- struct rbd_image_header_ondisk *ondisk;
- int ret;
-
- ondisk = rbd_dev_v1_header_read(rbd_dev);
- if (IS_ERR(ondisk))
- return PTR_ERR(ondisk);
- ret = rbd_header_from_disk(header, ondisk);
+ ret = rbd_header_from_disk(rbd_dev, ondisk);
+out:
kfree(ondisk);
return ret;
}
-static void rbd_update_mapping_size(struct rbd_device *rbd_dev)
-{
- if (rbd_dev->spec->snap_id != CEPH_NOSNAP)
- return;
-
- if (rbd_dev->mapping.size != rbd_dev->header.image_size) {
- sector_t size;
-
- rbd_dev->mapping.size = rbd_dev->header.image_size;
- size = (sector_t)rbd_dev->mapping.size / SECTOR_SIZE;
- dout("setting size to %llu sectors", (unsigned long long)size);
- set_capacity(rbd_dev->disk, size);
- }
-}
-
-/*
- * only read the first part of the ondisk header, without the snaps info
- */
-static int rbd_dev_v1_refresh(struct rbd_device *rbd_dev)
-{
- int ret;
- struct rbd_image_header h;
-
- ret = rbd_read_header(rbd_dev, &h);
- if (ret < 0)
- return ret;
-
- down_write(&rbd_dev->header_rwsem);
-
- /* Update image size, and check for resize of mapped image */
- rbd_dev->header.image_size = h.image_size;
- rbd_update_mapping_size(rbd_dev);
-
- /* rbd_dev->header.object_prefix shouldn't change */
- kfree(rbd_dev->header.snap_sizes);
- kfree(rbd_dev->header.snap_names);
- /* osd requests may still refer to snapc */
- ceph_put_snap_context(rbd_dev->header.snapc);
-
- rbd_dev->header.image_size = h.image_size;
- rbd_dev->header.snapc = h.snapc;
- rbd_dev->header.snap_names = h.snap_names;
- rbd_dev->header.snap_sizes = h.snap_sizes;
- /* Free the extra copy of the object prefix */
- if (strcmp(rbd_dev->header.object_prefix, h.object_prefix))
- rbd_warn(rbd_dev, "object prefix changed (ignoring)");
- kfree(h.object_prefix);
-
- up_write(&rbd_dev->header_rwsem);
-
- return ret;
-}
-
/*
* Clear the rbd device's EXISTS flag if the snapshot it's mapped to
* has disappeared from the (just updated) snapshot context.
@@ -3180,26 +3326,30 @@ static void rbd_exists_validate(struct rbd_device *rbd_dev)
static int rbd_dev_refresh(struct rbd_device *rbd_dev)
{
- u64 image_size;
+ u64 mapping_size;
int ret;
rbd_assert(rbd_image_format_valid(rbd_dev->image_format));
- image_size = rbd_dev->header.image_size;
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+ down_write(&rbd_dev->header_rwsem);
+ mapping_size = rbd_dev->mapping.size;
if (rbd_dev->image_format == 1)
- ret = rbd_dev_v1_refresh(rbd_dev);
+ ret = rbd_dev_v1_header_info(rbd_dev);
else
- ret = rbd_dev_v2_refresh(rbd_dev);
+ ret = rbd_dev_v2_header_info(rbd_dev);
/* If it's a mapped snapshot, validate its EXISTS flag */
rbd_exists_validate(rbd_dev);
- mutex_unlock(&ctl_mutex);
- if (ret)
- rbd_warn(rbd_dev, "got notification but failed to "
- " update snaps: %d\n", ret);
- if (image_size != rbd_dev->header.image_size)
+ up_write(&rbd_dev->header_rwsem);
+
+ if (mapping_size != rbd_dev->mapping.size) {
+ sector_t size;
+
+ size = (sector_t)rbd_dev->mapping.size / SECTOR_SIZE;
+ dout("setting size to %llu sectors", (unsigned long long)size);
+ set_capacity(rbd_dev->disk, size);
revalidate_disk(rbd_dev->disk);
+ }
return ret;
}
@@ -3403,6 +3553,8 @@ static ssize_t rbd_image_refresh(struct device *dev,
int ret;
ret = rbd_dev_refresh(rbd_dev);
+ if (ret)
+ rbd_warn(rbd_dev, ": manual header refresh error (%d)\n", ret);
return ret < 0 ? ret : size;
}
@@ -3501,6 +3653,7 @@ static struct rbd_device *rbd_dev_create(struct rbd_client *rbdc,
spin_lock_init(&rbd_dev->lock);
rbd_dev->flags = 0;
+ atomic_set(&rbd_dev->parent_ref, 0);
INIT_LIST_HEAD(&rbd_dev->node);
init_rwsem(&rbd_dev->header_rwsem);
@@ -3650,7 +3803,9 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev)
__le64 snapid;
void *p;
void *end;
+ u64 pool_id;
char *image_id;
+ u64 snap_id;
u64 overlap;
int ret;
@@ -3680,16 +3835,34 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev)
p = reply_buf;
end = reply_buf + ret;
ret = -ERANGE;
- ceph_decode_64_safe(&p, end, parent_spec->pool_id, out_err);
- if (parent_spec->pool_id == CEPH_NOPOOL)
+ ceph_decode_64_safe(&p, end, pool_id, out_err);
+ if (pool_id == CEPH_NOPOOL) {
+ /*
+ * Either the parent never existed, or we have
+ * record of it but the image got flattened so it no
+ * longer has a parent. When the parent of a
+ * layered image disappears we immediately set the
+ * overlap to 0. The effect of this is that all new
+ * requests will be treated as if the image had no
+ * parent.
+ */
+ if (rbd_dev->parent_overlap) {
+ rbd_dev->parent_overlap = 0;
+ smp_mb();
+ rbd_dev_parent_put(rbd_dev);
+ pr_info("%s: clone image has been flattened\n",
+ rbd_dev->disk->disk_name);
+ }
+
goto out; /* No parent? No problem. */
+ }
/* The ceph file layout needs to fit pool id in 32 bits */
ret = -EIO;
- if (parent_spec->pool_id > (u64)U32_MAX) {
+ if (pool_id > (u64)U32_MAX) {
rbd_warn(NULL, "parent pool id too large (%llu > %u)\n",
- (unsigned long long)parent_spec->pool_id, U32_MAX);
+ (unsigned long long)pool_id, U32_MAX);
goto out_err;
}
@@ -3698,13 +3871,51 @@ static int rbd_dev_v2_parent_info(struct rbd_device *rbd_dev)
ret = PTR_ERR(image_id);
goto out_err;
}
- parent_spec->image_id = image_id;
- ceph_decode_64_safe(&p, end, parent_spec->snap_id, out_err);
+ ceph_decode_64_safe(&p, end, snap_id, out_err);
ceph_decode_64_safe(&p, end, overlap, out_err);
+ /*
+ * The parent won't change (except when the clone is
+ * flattened, already handled that). So we only need to
+ * record the parent spec we have not already done so.
+ */
+ if (!rbd_dev->parent_spec) {
+ parent_spec->pool_id = pool_id;
+ parent_spec->image_id = image_id;
+ parent_spec->snap_id = snap_id;
+ rbd_dev->parent_spec = parent_spec;
+ parent_spec = NULL; /* rbd_dev now owns this */
+ }
+
+ /*
+ * We always update the parent overlap. If it's zero we
+ * treat it specially.
+ */
rbd_dev->parent_overlap = overlap;
- rbd_dev->parent_spec = parent_spec;
- parent_spec = NULL; /* rbd_dev now owns this */
+ smp_mb();
+ if (!overlap) {
+
+ /* A null parent_spec indicates it's the initial probe */
+
+ if (parent_spec) {
+ /*
+ * The overlap has become zero, so the clone
+ * must have been resized down to 0 at some
+ * point. Treat this the same as a flatten.
+ */
+ rbd_dev_parent_put(rbd_dev);
+ pr_info("%s: clone image now standalone\n",
+ rbd_dev->disk->disk_name);
+ } else {
+ /*
+ * For the initial probe, if we find the
+ * overlap is zero we just pretend there was
+ * no parent image.
+ */
+ rbd_warn(rbd_dev, "ignoring parent of "
+ "clone with overlap 0\n");
+ }
+ }
out:
ret = 0;
out_err:
@@ -4002,6 +4213,7 @@ static int rbd_dev_v2_snap_context(struct rbd_device *rbd_dev)
for (i = 0; i < snap_count; i++)
snapc->snaps[i] = ceph_decode_64(&p);
+ ceph_put_snap_context(rbd_dev->header.snapc);
rbd_dev->header.snapc = snapc;
dout(" snap context seq = %llu, snap_count = %u\n",
@@ -4053,23 +4265,54 @@ out:
return snap_name;
}
-static int rbd_dev_v2_refresh(struct rbd_device *rbd_dev)
+static int rbd_dev_v2_header_info(struct rbd_device *rbd_dev)
{
+ bool first_time = rbd_dev->header.object_prefix == NULL;
int ret;
- down_write(&rbd_dev->header_rwsem);
-
ret = rbd_dev_v2_image_size(rbd_dev);
if (ret)
- goto out;
- rbd_update_mapping_size(rbd_dev);
+ return ret;
+
+ if (first_time) {
+ ret = rbd_dev_v2_header_onetime(rbd_dev);
+ if (ret)
+ return ret;
+ }
+
+ /*
+ * If the image supports layering, get the parent info. We
+ * need to probe the first time regardless. Thereafter we
+ * only need to if there's a parent, to see if it has
+ * disappeared due to the mapped image getting flattened.
+ */
+ if (rbd_dev->header.features & RBD_FEATURE_LAYERING &&
+ (first_time || rbd_dev->parent_spec)) {
+ bool warn;
+
+ ret = rbd_dev_v2_parent_info(rbd_dev);
+ if (ret)
+ return ret;
+
+ /*
+ * Print a warning if this is the initial probe and
+ * the image has a parent. Don't print it if the
+ * image now being probed is itself a parent. We
+ * can tell at this point because we won't know its
+ * pool name yet (just its pool id).
+ */
+ warn = rbd_dev->parent_spec && rbd_dev->spec->pool_name;
+ if (first_time && warn)
+ rbd_warn(rbd_dev, "WARNING: kernel layering "
+ "is EXPERIMENTAL!");
+ }
+
+ if (rbd_dev->spec->snap_id == CEPH_NOSNAP)
+ if (rbd_dev->mapping.size != rbd_dev->header.image_size)
+ rbd_dev->mapping.size = rbd_dev->header.image_size;
ret = rbd_dev_v2_snap_context(rbd_dev);
dout("rbd_dev_v2_snap_context returned %d\n", ret);
- if (ret)
- goto out;
-out:
- up_write(&rbd_dev->header_rwsem);
return ret;
}
@@ -4079,8 +4322,6 @@ static int rbd_bus_add_dev(struct rbd_device *rbd_dev)
struct device *dev;
int ret;
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-
dev = &rbd_dev->dev;
dev->bus = &rbd_bus_type;
dev->type = &rbd_device_type;
@@ -4089,8 +4330,6 @@ static int rbd_bus_add_dev(struct rbd_device *rbd_dev)
dev_set_name(dev, "%d", rbd_dev->dev_id);
ret = device_register(dev);
- mutex_unlock(&ctl_mutex);
-
return ret;
}
@@ -4484,16 +4723,18 @@ out:
return ret;
}
-/* Undo whatever state changes are made by v1 or v2 image probe */
-
+/*
+ * Undo whatever state changes are made by v1 or v2 header info
+ * call.
+ */
static void rbd_dev_unprobe(struct rbd_device *rbd_dev)
{
struct rbd_image_header *header;
- rbd_dev_remove_parent(rbd_dev);
- rbd_spec_put(rbd_dev->parent_spec);
- rbd_dev->parent_spec = NULL;
- rbd_dev->parent_overlap = 0;
+ /* Drop parent reference unless it's already been done (or none) */
+
+ if (rbd_dev->parent_overlap)
+ rbd_dev_parent_put(rbd_dev);
/* Free dynamic fields from the header, then zero it out */
@@ -4505,72 +4746,22 @@ static void rbd_dev_unprobe(struct rbd_device *rbd_dev)
memset(header, 0, sizeof (*header));
}
-static int rbd_dev_v1_probe(struct rbd_device *rbd_dev)
+static int rbd_dev_v2_header_onetime(struct rbd_device *rbd_dev)
{
int ret;
- /* Populate rbd image metadata */
-
- ret = rbd_read_header(rbd_dev, &rbd_dev->header);
- if (ret < 0)
- goto out_err;
-
- /* Version 1 images have no parent (no layering) */
-
- rbd_dev->parent_spec = NULL;
- rbd_dev->parent_overlap = 0;
-
- dout("discovered version 1 image, header name is %s\n",
- rbd_dev->header_name);
-
- return 0;
-
-out_err:
- kfree(rbd_dev->header_name);
- rbd_dev->header_name = NULL;
- kfree(rbd_dev->spec->image_id);
- rbd_dev->spec->image_id = NULL;
-
- return ret;
-}
-
-static int rbd_dev_v2_probe(struct rbd_device *rbd_dev)
-{
- int ret;
-
- ret = rbd_dev_v2_image_size(rbd_dev);
- if (ret)
- goto out_err;
-
- /* Get the object prefix (a.k.a. block_name) for the image */
-
ret = rbd_dev_v2_object_prefix(rbd_dev);
if (ret)
goto out_err;
- /* Get the and check features for the image */
-
+ /*
+ * Get the and check features for the image. Currently the
+ * features are assumed to never change.
+ */
ret = rbd_dev_v2_features(rbd_dev);
if (ret)
goto out_err;
- /* If the image supports layering, get the parent info */
-
- if (rbd_dev->header.features & RBD_FEATURE_LAYERING) {
- ret = rbd_dev_v2_parent_info(rbd_dev);
- if (ret)
- goto out_err;
-
- /*
- * Don't print a warning for parent images. We can
- * tell this point because we won't know its pool
- * name yet (just its pool id).
- */
- if (rbd_dev->spec->pool_name)
- rbd_warn(rbd_dev, "WARNING: kernel layering "
- "is EXPERIMENTAL!");
- }
-
/* If the image supports fancy striping, get its parameters */
if (rbd_dev->header.features & RBD_FEATURE_STRIPINGV2) {
@@ -4578,28 +4769,11 @@ static int rbd_dev_v2_probe(struct rbd_device *rbd_dev)
if (ret < 0)
goto out_err;
}
-
- /* crypto and compression type aren't (yet) supported for v2 images */
-
- rbd_dev->header.crypt_type = 0;
- rbd_dev->header.comp_type = 0;
-
- /* Get the snapshot context, plus the header version */
-
- ret = rbd_dev_v2_snap_context(rbd_dev);
- if (ret)
- goto out_err;
-
- dout("discovered version 2 image, header name is %s\n",
- rbd_dev->header_name);
+ /* No support for crypto and compression type format 2 images */
return 0;
out_err:
- rbd_dev->parent_overlap = 0;
- rbd_spec_put(rbd_dev->parent_spec);
- rbd_dev->parent_spec = NULL;
- kfree(rbd_dev->header_name);
- rbd_dev->header_name = NULL;
+ rbd_dev->header.features = 0;
kfree(rbd_dev->header.object_prefix);
rbd_dev->header.object_prefix = NULL;
@@ -4628,15 +4802,16 @@ static int rbd_dev_probe_parent(struct rbd_device *rbd_dev)
if (!parent)
goto out_err;
- ret = rbd_dev_image_probe(parent);
+ ret = rbd_dev_image_probe(parent, false);
if (ret < 0)
goto out_err;
rbd_dev->parent = parent;
+ atomic_set(&rbd_dev->parent_ref, 1);
return 0;
out_err:
if (parent) {
- rbd_spec_put(rbd_dev->parent_spec);
+ rbd_dev_unparent(rbd_dev);
kfree(rbd_dev->header_name);
rbd_dev_destroy(parent);
} else {
@@ -4651,10 +4826,6 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
{
int ret;
- ret = rbd_dev_mapping_set(rbd_dev);
- if (ret)
- return ret;
-
/* generate unique id: find highest unique id, add one */
rbd_dev_id_get(rbd_dev);
@@ -4676,13 +4847,17 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
if (ret)
goto err_out_blkdev;
- ret = rbd_bus_add_dev(rbd_dev);
+ ret = rbd_dev_mapping_set(rbd_dev);
if (ret)
goto err_out_disk;
+ set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE);
+
+ ret = rbd_bus_add_dev(rbd_dev);
+ if (ret)
+ goto err_out_mapping;
/* Everything's ready. Announce the disk to the world. */
- set_capacity(rbd_dev->disk, rbd_dev->mapping.size / SECTOR_SIZE);
set_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags);
add_disk(rbd_dev->disk);
@@ -4691,6 +4866,8 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev)
return ret;
+err_out_mapping:
+ rbd_dev_mapping_clear(rbd_dev);
err_out_disk:
rbd_free_disk(rbd_dev);
err_out_blkdev:
@@ -4731,12 +4908,7 @@ static int rbd_dev_header_name(struct rbd_device *rbd_dev)
static void rbd_dev_image_release(struct rbd_device *rbd_dev)
{
- int ret;
-
rbd_dev_unprobe(rbd_dev);
- ret = rbd_dev_header_watch_sync(rbd_dev, 0);
- if (ret)
- rbd_warn(rbd_dev, "failed to cancel watch event (%d)\n", ret);
kfree(rbd_dev->header_name);
rbd_dev->header_name = NULL;
rbd_dev->image_format = 0;
@@ -4748,18 +4920,20 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev)
/*
* Probe for the existence of the header object for the given rbd
- * device. For format 2 images this includes determining the image
- * id.
+ * device. If this image is the one being mapped (i.e., not a
+ * parent), initiate a watch on its header object before using that
+ * object to get detailed information about the rbd image.
*/
-static int rbd_dev_image_probe(struct rbd_device *rbd_dev)
+static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping)
{
int ret;
int tmp;
/*
- * Get the id from the image id object. If it's not a
- * format 2 image, we'll get ENOENT back, and we'll assume
- * it's a format 1 image.
+ * Get the id from the image id object. Unless there's an
+ * error, rbd_dev->spec->image_id will be filled in with
+ * a dynamically-allocated string, and rbd_dev->image_format
+ * will be set to either 1 or 2.
*/
ret = rbd_dev_image_id(rbd_dev);
if (ret)
@@ -4771,14 +4945,16 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev)
if (ret)
goto err_out_format;
- ret = rbd_dev_header_watch_sync(rbd_dev, 1);
- if (ret)
- goto out_header_name;
+ if (mapping) {
+ ret = rbd_dev_header_watch_sync(rbd_dev, true);
+ if (ret)
+ goto out_header_name;
+ }
if (rbd_dev->image_format == 1)
- ret = rbd_dev_v1_probe(rbd_dev);
+ ret = rbd_dev_v1_header_info(rbd_dev);
else
- ret = rbd_dev_v2_probe(rbd_dev);
+ ret = rbd_dev_v2_header_info(rbd_dev);
if (ret)
goto err_out_watch;
@@ -4787,15 +4963,22 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev)
goto err_out_probe;
ret = rbd_dev_probe_parent(rbd_dev);
- if (!ret)
- return 0;
+ if (ret)
+ goto err_out_probe;
+ dout("discovered format %u image, header name is %s\n",
+ rbd_dev->image_format, rbd_dev->header_name);
+
+ return 0;
err_out_probe:
rbd_dev_unprobe(rbd_dev);
err_out_watch:
- tmp = rbd_dev_header_watch_sync(rbd_dev, 0);
- if (tmp)
- rbd_warn(rbd_dev, "unable to tear down watch request\n");
+ if (mapping) {
+ tmp = rbd_dev_header_watch_sync(rbd_dev, false);
+ if (tmp)
+ rbd_warn(rbd_dev, "unable to tear down "
+ "watch request (%d)\n", tmp);
+ }
out_header_name:
kfree(rbd_dev->header_name);
rbd_dev->header_name = NULL;
@@ -4819,6 +5002,7 @@ static ssize_t rbd_add(struct bus_type *bus,
struct rbd_spec *spec = NULL;
struct rbd_client *rbdc;
struct ceph_osd_client *osdc;
+ bool read_only;
int rc = -ENOMEM;
if (!try_module_get(THIS_MODULE))
@@ -4828,13 +5012,15 @@ static ssize_t rbd_add(struct bus_type *bus,
rc = rbd_add_parse_args(buf, &ceph_opts, &rbd_opts, &spec);
if (rc < 0)
goto err_out_module;
+ read_only = rbd_opts->read_only;
+ kfree(rbd_opts);
+ rbd_opts = NULL; /* done with this */
rbdc = rbd_get_client(ceph_opts);
if (IS_ERR(rbdc)) {
rc = PTR_ERR(rbdc);
goto err_out_args;
}
- ceph_opts = NULL; /* rbd_dev client now owns this */
/* pick the pool */
osdc = &rbdc->client->osdc;
@@ -4858,27 +5044,29 @@ static ssize_t rbd_add(struct bus_type *bus,
rbdc = NULL; /* rbd_dev now owns this */
spec = NULL; /* rbd_dev now owns this */
- rbd_dev->mapping.read_only = rbd_opts->read_only;
- kfree(rbd_opts);
- rbd_opts = NULL; /* done with this */
-
- rc = rbd_dev_image_probe(rbd_dev);
+ rc = rbd_dev_image_probe(rbd_dev, true);
if (rc < 0)
goto err_out_rbd_dev;
+ /* If we are mapping a snapshot it must be marked read-only */
+
+ if (rbd_dev->spec->snap_id != CEPH_NOSNAP)
+ read_only = true;
+ rbd_dev->mapping.read_only = read_only;
+
rc = rbd_dev_device_setup(rbd_dev);
- if (!rc)
- return count;
+ if (rc) {
+ rbd_dev_image_release(rbd_dev);
+ goto err_out_module;
+ }
+
+ return count;
- rbd_dev_image_release(rbd_dev);
err_out_rbd_dev:
rbd_dev_destroy(rbd_dev);
err_out_client:
rbd_put_client(rbdc);
err_out_args:
- if (ceph_opts)
- ceph_destroy_options(ceph_opts);
- kfree(rbd_opts);
rbd_spec_put(spec);
err_out_module:
module_put(THIS_MODULE);
@@ -4888,30 +5076,13 @@ err_out_module:
return (ssize_t)rc;
}
-static struct rbd_device *__rbd_get_dev(unsigned long dev_id)
-{
- struct list_head *tmp;
- struct rbd_device *rbd_dev;
-
- spin_lock(&rbd_dev_list_lock);
- list_for_each(tmp, &rbd_dev_list) {
- rbd_dev = list_entry(tmp, struct rbd_device, node);
- if (rbd_dev->dev_id == dev_id) {
- spin_unlock(&rbd_dev_list_lock);
- return rbd_dev;
- }
- }
- spin_unlock(&rbd_dev_list_lock);
- return NULL;
-}
-
static void rbd_dev_device_release(struct device *dev)
{
struct rbd_device *rbd_dev = dev_to_rbd_dev(dev);
rbd_free_disk(rbd_dev);
clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags);
- rbd_dev_clear_mapping(rbd_dev);
+ rbd_dev_mapping_clear(rbd_dev);
unregister_blkdev(rbd_dev->major, rbd_dev->name);
rbd_dev->major = 0;
rbd_dev_id_put(rbd_dev);
@@ -4949,8 +5120,10 @@ static ssize_t rbd_remove(struct bus_type *bus,
size_t count)
{
struct rbd_device *rbd_dev = NULL;
- int target_id;
+ struct list_head *tmp;
+ int dev_id;
unsigned long ul;
+ bool already = false;
int ret;
ret = strict_strtoul(buf, 10, &ul);
@@ -4958,34 +5131,40 @@ static ssize_t rbd_remove(struct bus_type *bus,
return ret;
/* convert to int; abort if we lost anything in the conversion */
- target_id = (int) ul;
- if (target_id != ul)
+ dev_id = (int)ul;
+ if (dev_id != ul)
return -EINVAL;
- mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-
- rbd_dev = __rbd_get_dev(target_id);
- if (!rbd_dev) {
- ret = -ENOENT;
- goto done;
+ ret = -ENOENT;
+ spin_lock(&rbd_dev_list_lock);
+ list_for_each(tmp, &rbd_dev_list) {
+ rbd_dev = list_entry(tmp, struct rbd_device, node);
+ if (rbd_dev->dev_id == dev_id) {
+ ret = 0;
+ break;
+ }
}
+ if (!ret) {
+ spin_lock_irq(&rbd_dev->lock);
+ if (rbd_dev->open_count)
+ ret = -EBUSY;
+ else
+ already = test_and_set_bit(RBD_DEV_FLAG_REMOVING,
+ &rbd_dev->flags);
+ spin_unlock_irq(&rbd_dev->lock);
+ }
+ spin_unlock(&rbd_dev_list_lock);
+ if (ret < 0 || already)
+ return ret;
- spin_lock_irq(&rbd_dev->lock);
- if (rbd_dev->open_count)
- ret = -EBUSY;
- else
- set_bit(RBD_DEV_FLAG_REMOVING, &rbd_dev->flags);
- spin_unlock_irq(&rbd_dev->lock);
- if (ret < 0)
- goto done;
- ret = count;
rbd_bus_del_dev(rbd_dev);
+ ret = rbd_dev_header_watch_sync(rbd_dev, false);
+ if (ret)
+ rbd_warn(rbd_dev, "failed to cancel watch event (%d)\n", ret);
rbd_dev_image_release(rbd_dev);
module_put(THIS_MODULE);
-done:
- mutex_unlock(&ctl_mutex);
- return ret;
+ return count;
}
/*
@@ -5093,6 +5272,7 @@ static void __exit rbd_exit(void)
module_init(rbd_init);
module_exit(rbd_exit);
+MODULE_AUTHOR("Alex Elder <elder@inktank.com>");
MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
MODULE_DESCRIPTION("rados block device");
diff --git a/drivers/block/swim.c b/drivers/block/swim.c
index 2f445b7a174..8ed6ccb748c 100644
--- a/drivers/block/swim.c
+++ b/drivers/block/swim.c
@@ -893,7 +893,7 @@ static int swim_probe(struct platform_device *dev)
swim_base = ioremap(res->start, resource_size(res));
if (!swim_base) {
- return -ENOMEM;
+ ret = -ENOMEM;
goto out_release_io;
}
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 64723953e1c..5cdf88b7ad9 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -20,7 +20,7 @@ module_param(use_bio, bool, S_IRUGO);
static int major;
static DEFINE_IDA(vd_index_ida);
-struct workqueue_struct *virtblk_wq;
+static struct workqueue_struct *virtblk_wq;
struct virtio_blk
{
diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c
index 8bfd1bcf95e..04608a6502d 100644
--- a/drivers/block/xen-blkback/xenbus.c
+++ b/drivers/block/xen-blkback/xenbus.c
@@ -93,7 +93,7 @@ static void xen_update_blkif_status(struct xen_blkif *blkif)
}
invalidate_inode_pages2(blkif->vbd.bdev->bd_inode->i_mapping);
- blkif->xenblkd = kthread_run(xen_blkif_schedule, blkif, name);
+ blkif->xenblkd = kthread_run(xen_blkif_schedule, blkif, "%s", name);
if (IS_ERR(blkif->xenblkd)) {
err = PTR_ERR(blkif->xenblkd);
blkif->xenblkd = NULL;
diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c
index f8ef15f37c5..1393b8871a2 100644
--- a/drivers/block/xsysace.c
+++ b/drivers/block/xsysace.c
@@ -407,7 +407,7 @@ static void ace_dump_regs(struct ace_device *ace)
ace_in32(ace, ACE_CFGLBA), ace_in(ace, ACE_FATSTAT));
}
-void ace_fix_driveid(u16 *id)
+static void ace_fix_driveid(u16 *id)
{
#if defined(__BIG_ENDIAN)
int i;
@@ -463,7 +463,7 @@ static inline void ace_fsm_yieldirq(struct ace_device *ace)
}
/* Get the next read/write request; ending requests that we don't handle */
-struct request *ace_get_next_request(struct request_queue * q)
+static struct request *ace_get_next_request(struct request_queue *q)
{
struct request *req;
@@ -1160,8 +1160,7 @@ static int ace_probe(struct platform_device *dev)
dev_dbg(&dev->dev, "ace_probe(%p)\n", dev);
/* device id and bus width */
- of_property_read_u32(dev->dev.of_node, "port-number", &id);
- if (id < 0)
+ if (of_property_read_u32(dev->dev.of_node, "port-number", &id))
id = 0;
if (of_find_property(dev->dev.of_node, "8-bit", NULL))
bus_width = ACE_BUS_WIDTH_8;
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-01-09 11:52:49 +0000