summaryrefslogtreecommitdiff
path: root/drivers/block
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/block')
-rw-r--r--drivers/block/DAC960.c14
-rw-r--r--drivers/block/Kconfig17
-rw-r--r--drivers/block/Makefile1
-rw-r--r--drivers/block/amiflop.c19
-rw-r--r--drivers/block/aoe/aoeblk.c9
-rw-r--r--drivers/block/aoe/aoechr.c10
-rw-r--r--drivers/block/ataflop.c15
-rw-r--r--drivers/block/brd.c7
-rw-r--r--drivers/block/cciss.c27
-rw-r--r--drivers/block/cpqarray.c15
-rw-r--r--drivers/block/drbd/drbd_main.c11
-rw-r--r--drivers/block/drbd/drbd_receiver.c1
-rw-r--r--drivers/block/floppy.c16
-rw-r--r--drivers/block/loop.c13
-rw-r--r--drivers/block/mg_disk.c3
-rw-r--r--drivers/block/nbd.c7
-rw-r--r--drivers/block/paride/pcd.c15
-rw-r--r--drivers/block/paride/pd.c15
-rw-r--r--drivers/block/paride/pf.c17
-rw-r--r--drivers/block/paride/pg.c8
-rw-r--r--drivers/block/paride/pt.c20
-rw-r--r--drivers/block/pktcdvd.c19
-rw-r--r--drivers/block/ps3disk.c2
-rw-r--r--drivers/block/rbd.c1841
-rw-r--r--drivers/block/rbd_types.h73
-rw-r--r--drivers/block/swim.c15
-rw-r--r--drivers/block/swim3.c15
-rw-r--r--drivers/block/ub.c15
-rw-r--r--drivers/block/viodasd.c11
-rw-r--r--drivers/block/virtio_blk.c23
-rw-r--r--drivers/block/xd.c7
-rw-r--r--drivers/block/xen-blkfront.c13
-rw-r--r--drivers/block/xsysace.c11
-rw-r--r--drivers/block/z2ram.c15
34 files changed, 2140 insertions, 180 deletions
diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c
index 4e2c367fec11..1f286ab461d3 100644
--- a/drivers/block/DAC960.c
+++ b/drivers/block/DAC960.c
@@ -36,7 +36,7 @@
#include <linux/ioport.h>
#include <linux/mm.h>
#include <linux/slab.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/reboot.h>
@@ -54,6 +54,7 @@
#define DAC960_GAM_MINOR 252
+static DEFINE_MUTEX(DAC960_mutex);
static DAC960_Controller_T *DAC960_Controllers[DAC960_MaxControllers];
static int DAC960_ControllerCount;
static struct proc_dir_entry *DAC960_ProcDirectoryEntry;
@@ -81,7 +82,7 @@ static int DAC960_open(struct block_device *bdev, fmode_t mode)
int drive_nr = (long)disk->private_data;
int ret = -ENXIO;
- lock_kernel();
+ mutex_lock(&DAC960_mutex);
if (p->FirmwareType == DAC960_V1_Controller) {
if (p->V1.LogicalDriveInformation[drive_nr].
LogicalDriveState == DAC960_V1_LogicalDrive_Offline)
@@ -99,7 +100,7 @@ static int DAC960_open(struct block_device *bdev, fmode_t mode)
goto out;
ret = 0;
out:
- unlock_kernel();
+ mutex_unlock(&DAC960_mutex);
return ret;
}
@@ -6625,7 +6626,7 @@ static long DAC960_gam_ioctl(struct file *file, unsigned int Request,
long ErrorCode = 0;
if (!capable(CAP_SYS_ADMIN)) return -EACCES;
- lock_kernel();
+ mutex_lock(&DAC960_mutex);
switch (Request)
{
case DAC960_IOCTL_GET_CONTROLLER_COUNT:
@@ -7056,13 +7057,14 @@ static long DAC960_gam_ioctl(struct file *file, unsigned int Request,
default:
ErrorCode = -ENOTTY;
}
- unlock_kernel();
+ mutex_unlock(&DAC960_mutex);
return ErrorCode;
}
static const struct file_operations DAC960_gam_fops = {
.owner = THIS_MODULE,
- .unlocked_ioctl = DAC960_gam_ioctl
+ .unlocked_ioctl = DAC960_gam_ioctl,
+ .llseek = noop_llseek,
};
static struct miscdevice DAC960_gam_dev = {
diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index de277689da61..4b9359a6f6ca 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -488,4 +488,21 @@ config BLK_DEV_HD
If unsure, say N.
+config BLK_DEV_RBD
+ tristate "Rados block device (RBD)"
+ depends on INET && EXPERIMENTAL && BLOCK
+ select CEPH_LIB
+ select LIBCRC32C
+ select CRYPTO_AES
+ select CRYPTO
+ default n
+ help
+ Say Y here if you want include the Rados block device, which stripes
+ a block device over objects stored in the Ceph distributed object
+ store.
+
+ More information at http://ceph.newdream.net/.
+
+ If unsure, say N.
+
endif # BLK_DEV
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index aff5ac925c34..d7f463d6312d 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -37,5 +37,6 @@ obj-$(CONFIG_BLK_DEV_HD) += hd.o
obj-$(CONFIG_XEN_BLKDEV_FRONTEND) += xen-blkfront.o
obj-$(CONFIG_BLK_DEV_DRBD) += drbd/
+obj-$(CONFIG_BLK_DEV_RBD) += rbd.o
swim_mod-objs := swim.o swim_asm.o
diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c
index 327ed27dfe1a..a1725e6488d3 100644
--- a/drivers/block/amiflop.c
+++ b/drivers/block/amiflop.c
@@ -60,7 +60,7 @@
#include <linux/hdreg.h>
#include <linux/delay.h>
#include <linux/init.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/amifdreg.h>
#include <linux/amifd.h>
#include <linux/buffer_head.h>
@@ -109,6 +109,7 @@
#define FD_HD_3 0x55555555 /* high-density 3.5" (1760K) drive */
#define FD_DD_5 0xaaaaaaaa /* double-density 5.25" (440K) drive */
+static DEFINE_MUTEX(amiflop_mutex);
static unsigned long int fd_def_df0 = FD_DD_3; /* default for df0 if it doesn't identify */
module_param(fd_def_df0, ulong, 0);
@@ -1541,9 +1542,9 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&amiflop_mutex);
ret = fd_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
+ mutex_unlock(&amiflop_mutex);
return ret;
}
@@ -1590,11 +1591,11 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
int old_dev;
unsigned long flags;
- lock_kernel();
+ mutex_lock(&amiflop_mutex);
old_dev = fd_device[drive];
if (fd_ref[drive] && old_dev != system) {
- unlock_kernel();
+ mutex_unlock(&amiflop_mutex);
return -EBUSY;
}
@@ -1610,7 +1611,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
rel_fdc();
if (wrprot) {
- unlock_kernel();
+ mutex_unlock(&amiflop_mutex);
return -EROFS;
}
}
@@ -1629,7 +1630,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
printk(KERN_INFO "fd%d: accessing %s-disk with %s-layout\n",drive,
unit[drive].type->name, data_types[system].name);
- unlock_kernel();
+ mutex_unlock(&amiflop_mutex);
return 0;
}
@@ -1638,7 +1639,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
struct amiga_floppy_struct *p = disk->private_data;
int drive = p - unit;
- lock_kernel();
+ mutex_lock(&amiflop_mutex);
if (unit[drive].dirty == 1) {
del_timer (flush_track_timer + drive);
non_int_flush_track (drive);
@@ -1652,7 +1653,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
/* the mod_use counter is handled this way */
floppy_off (drive | 0x40000000);
#endif
- unlock_kernel();
+ mutex_unlock(&amiflop_mutex);
return 0;
}
diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c
index a946929735a5..f21c237a9e5e 100644
--- a/drivers/block/aoe/aoeblk.c
+++ b/drivers/block/aoe/aoeblk.c
@@ -12,9 +12,10 @@
#include <linux/slab.h>
#include <linux/genhd.h>
#include <linux/netdevice.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include "aoe.h"
+static DEFINE_MUTEX(aoeblk_mutex);
static struct kmem_cache *buf_pool_cache;
static ssize_t aoedisk_show_state(struct device *dev,
@@ -125,16 +126,16 @@ aoeblk_open(struct block_device *bdev, fmode_t mode)
struct aoedev *d = bdev->bd_disk->private_data;
ulong flags;
- lock_kernel();
+ mutex_lock(&aoeblk_mutex);
spin_lock_irqsave(&d->lock, flags);
if (d->flags & DEVFL_UP) {
d->nopen++;
spin_unlock_irqrestore(&d->lock, flags);
- unlock_kernel();
+ mutex_unlock(&aoeblk_mutex);
return 0;
}
spin_unlock_irqrestore(&d->lock, flags);
- unlock_kernel();
+ mutex_unlock(&aoeblk_mutex);
return -ENODEV;
}
diff --git a/drivers/block/aoe/aoechr.c b/drivers/block/aoe/aoechr.c
index 4a1b9e7464aa..146296ca4965 100644
--- a/drivers/block/aoe/aoechr.c
+++ b/drivers/block/aoe/aoechr.c
@@ -9,7 +9,7 @@
#include <linux/completion.h>
#include <linux/delay.h>
#include <linux/slab.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/skbuff.h>
#include "aoe.h"
@@ -37,6 +37,7 @@ struct ErrMsg {
char *msg;
};
+static DEFINE_MUTEX(aoechr_mutex);
static struct ErrMsg emsgs[NMSG];
static int emsgs_head_idx, emsgs_tail_idx;
static struct completion emsgs_comp;
@@ -183,16 +184,16 @@ aoechr_open(struct inode *inode, struct file *filp)
{
int n, i;
- lock_kernel();
+ mutex_lock(&aoechr_mutex);
n = iminor(inode);
filp->private_data = (void *) (unsigned long) n;
for (i = 0; i < ARRAY_SIZE(chardevs); ++i)
if (chardevs[i].minor == n) {
- unlock_kernel();
+ mutex_unlock(&aoechr_mutex);
return 0;
}
- unlock_kernel();
+ mutex_unlock(&aoechr_mutex);
return -EINVAL;
}
@@ -265,6 +266,7 @@ static const struct file_operations aoe_fops = {
.open = aoechr_open,
.release = aoechr_rel,
.owner = THIS_MODULE,
+ .llseek = noop_llseek,
};
static char *aoe_devnode(struct device *dev, mode_t *mode)
diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c
index 0f4eec442e5d..4e4cc6c828cb 100644
--- a/drivers/block/ataflop.c
+++ b/drivers/block/ataflop.c
@@ -67,7 +67,7 @@
#include <linux/delay.h>
#include <linux/init.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <asm/atafd.h>
#include <asm/atafdreg.h>
@@ -79,6 +79,7 @@
#undef DEBUG
+static DEFINE_MUTEX(ataflop_mutex);
static struct request *fd_request;
static int fdc_queue;
@@ -1694,9 +1695,9 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&ataflop_mutex);
ret = fd_locked_ioctl(bdev, mode, cmd, arg);
- unlock_kernel();
+ mutex_unlock(&ataflop_mutex);
return ret;
}
@@ -1877,9 +1878,9 @@ static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&ataflop_mutex);
ret = floppy_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&ataflop_mutex);
return ret;
}
@@ -1887,14 +1888,14 @@ static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode)
static int floppy_release(struct gendisk *disk, fmode_t mode)
{
struct atari_floppy_struct *p = disk->private_data;
- lock_kernel();
+ mutex_lock(&ataflop_mutex);
if (p->ref < 0)
p->ref = 0;
else if (!p->ref--) {
printk(KERN_ERR "floppy_release with fd_ref == 0");
p->ref = 0;
}
- unlock_kernel();
+ mutex_unlock(&ataflop_mutex);
return 0;
}
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index 1c7f63792ff8..82bfd5bb4a97 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -15,7 +15,7 @@
#include <linux/blkdev.h>
#include <linux/bio.h>
#include <linux/highmem.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/radix-tree.h>
#include <linux/buffer_head.h> /* invalidate_bh_lrus() */
#include <linux/slab.h>
@@ -55,6 +55,7 @@ struct brd_device {
/*
* Look up and return a brd's page for a given sector.
*/
+static DEFINE_MUTEX(brd_mutex);
static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector)
{
pgoff_t idx;
@@ -402,7 +403,7 @@ static int brd_ioctl(struct block_device *bdev, fmode_t mode,
* ram device BLKFLSBUF has special semantics, we want to actually
* release and destroy the ramdisk data.
*/
- lock_kernel();
+ mutex_lock(&brd_mutex);
mutex_lock(&bdev->bd_mutex);
error = -EBUSY;
if (bdev->bd_openers <= 1) {
@@ -419,7 +420,7 @@ static int brd_ioctl(struct block_device *bdev, fmode_t mode,
error = 0;
}
mutex_unlock(&bdev->bd_mutex);
- unlock_kernel();
+ mutex_unlock(&brd_mutex);
return error;
}
diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c
index 41f9acb8ecd7..f09e6df15aa7 100644
--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -26,7 +26,6 @@
#include <linux/pci.h>
#include <linux/kernel.h>
#include <linux/slab.h>
-#include <linux/smp_lock.h>
#include <linux/delay.h>
#include <linux/major.h>
#include <linux/fs.h>
@@ -66,6 +65,7 @@ MODULE_SUPPORTED_DEVICE("HP Smart Array Controllers");
MODULE_VERSION("3.6.26");
MODULE_LICENSE("GPL");
+static DEFINE_MUTEX(cciss_mutex);
static int cciss_allow_hpsa;
module_param(cciss_allow_hpsa, int, S_IRUGO|S_IWUSR);
MODULE_PARM_DESC(cciss_allow_hpsa,
@@ -299,6 +299,8 @@ static void enqueue_cmd_and_start_io(ctlr_info_t *h,
spin_lock_irqsave(&h->lock, flags);
addQ(&h->reqQ, c);
h->Qdepth++;
+ if (h->Qdepth > h->maxQsinceinit)
+ h->maxQsinceinit = h->Qdepth;
start_io(h);
spin_unlock_irqrestore(&h->lock, flags);
}
@@ -1059,9 +1061,9 @@ static int cciss_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&cciss_mutex);
ret = cciss_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&cciss_mutex);
return ret;
}
@@ -1074,13 +1076,13 @@ static int cciss_release(struct gendisk *disk, fmode_t mode)
ctlr_info_t *h;
drive_info_struct *drv;
- lock_kernel();
+ mutex_lock(&cciss_mutex);
h = get_host(disk);
drv = get_drv(disk);
dev_dbg(&h->pdev->dev, "cciss_release %s\n", disk->disk_name);
drv->usage_count--;
h->usage_count--;
- unlock_kernel();
+ mutex_unlock(&cciss_mutex);
return 0;
}
@@ -1088,9 +1090,9 @@ static int do_ioctl(struct block_device *bdev, fmode_t mode,
unsigned cmd, unsigned long arg)
{
int ret;
- lock_kernel();
+ mutex_lock(&cciss_mutex);
ret = cciss_ioctl(bdev, mode, cmd, arg);
- unlock_kernel();
+ mutex_unlock(&cciss_mutex);
return ret;
}
@@ -4503,6 +4505,12 @@ static __devinit int cciss_kdump_hard_reset_controller(struct pci_dev *pdev)
misc_fw_support = readl(&cfgtable->misc_fw_support);
use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET;
+ /* The doorbell reset seems to cause lockups on some Smart
+ * Arrays (e.g. P410, P410i, maybe others). Until this is
+ * fixed or at least isolated, avoid the doorbell reset.
+ */
+ use_doorbell = 0;
+
rc = cciss_controller_hard_reset(pdev, vaddr, use_doorbell);
if (rc)
goto unmap_cfgtable;
@@ -4696,6 +4704,9 @@ static int __devinit cciss_init_one(struct pci_dev *pdev,
h->scatter_list = kmalloc(h->max_commands *
sizeof(struct scatterlist *),
GFP_KERNEL);
+ if (!h->scatter_list)
+ goto clean4;
+
for (k = 0; k < h->nr_cmds; k++) {
h->scatter_list[k] = kmalloc(sizeof(struct scatterlist) *
h->maxsgentries,
@@ -4765,7 +4776,7 @@ static int __devinit cciss_init_one(struct pci_dev *pdev,
clean4:
kfree(h->cmd_pool_bits);
/* Free up sg elements */
- for (k = 0; k < h->nr_cmds; k++)
+ for (k-- ; k >= 0; k--)
kfree(h->scatter_list[k]);
kfree(h->scatter_list);
cciss_free_sg_chain_blocks(h->cmd_sg_list, h->nr_cmds);
diff --git a/drivers/block/cpqarray.c b/drivers/block/cpqarray.c
index d53b0291c44b..946dad4caef3 100644
--- a/drivers/block/cpqarray.c
+++ b/drivers/block/cpqarray.c
@@ -35,7 +35,7 @@
#include <linux/seq_file.h>
#include <linux/init.h>
#include <linux/hdreg.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/spinlock.h>
#include <linux/blkdev.h>
#include <linux/genhd.h>
@@ -68,6 +68,7 @@ MODULE_LICENSE("GPL");
#define CPQARRAY_DMA_MASK 0xFFFFFFFF /* 32 bit DMA */
+static DEFINE_MUTEX(cpqarray_mutex);
static int nr_ctlr;
static ctlr_info_t *hba[MAX_CTLR];
@@ -845,9 +846,9 @@ static int ida_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&cpqarray_mutex);
ret = ida_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&cpqarray_mutex);
return ret;
}
@@ -859,10 +860,10 @@ static int ida_release(struct gendisk *disk, fmode_t mode)
{
ctlr_info_t *host;
- lock_kernel();
+ mutex_lock(&cpqarray_mutex);
host = get_host(disk);
host->usage_count--;
- unlock_kernel();
+ mutex_unlock(&cpqarray_mutex);
return 0;
}
@@ -1217,9 +1218,9 @@ static int ida_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&cpqarray_mutex);
ret = ida_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
+ mutex_unlock(&cpqarray_mutex);
return ret;
}
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index 8bfedc7164fa..c5dfe6486cf3 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -32,7 +32,7 @@
#include <asm/types.h>
#include <net/sock.h>
#include <linux/ctype.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/proc_fs.h>
@@ -64,6 +64,7 @@ struct after_state_chg_work {
struct completion *done;
};
+static DEFINE_MUTEX(drbd_main_mutex);
int drbdd_init(struct drbd_thread *);
int drbd_worker(struct drbd_thread *);
int drbd_asender(struct drbd_thread *);
@@ -2672,7 +2673,7 @@ static int drbd_open(struct block_device *bdev, fmode_t mode)
unsigned long flags;
int rv = 0;
- lock_kernel();
+ mutex_lock(&drbd_main_mutex);
spin_lock_irqsave(&mdev->req_lock, flags);
/* to have a stable mdev->state.role
* and no race with updating open_cnt */
@@ -2687,7 +2688,7 @@ static int drbd_open(struct block_device *bdev, fmode_t mode)
if (!rv)
mdev->open_cnt++;
spin_unlock_irqrestore(&mdev->req_lock, flags);
- unlock_kernel();
+ mutex_unlock(&drbd_main_mutex);
return rv;
}
@@ -2695,9 +2696,9 @@ static int drbd_open(struct block_device *bdev, fmode_t mode)
static int drbd_release(struct gendisk *gd, fmode_t mode)
{
struct drbd_conf *mdev = gd->private_data;
- lock_kernel();
+ mutex_lock(&drbd_main_mutex);
mdev->open_cnt--;
- unlock_kernel();
+ mutex_unlock(&drbd_main_mutex);
return 0;
}
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index 6ec922c623a1..760ae0df9251 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -3071,7 +3071,6 @@ static int receive_sizes(struct drbd_conf *mdev, enum drbd_packets cmd, unsigned
* we still need to figure out whether we accept that. */
mdev->p_size = p_size;
-#define min_not_zero(l, r) (l == 0) ? r : ((r == 0) ? l : min(l, r))
if (get_ldev(mdev)) {
warn_if_differ_considerably(mdev, "lower level device sizes",
p_size, drbd_get_max_capacity(mdev->ldev));
diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index aa42e7766c6a..767107cce982 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -178,7 +178,6 @@ static int print_unex = 1;
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/bio.h>
-#include <linux/smp_lock.h>
#include <linux/string.h>
#include <linux/jiffies.h>
#include <linux/fcntl.h>
@@ -199,6 +198,7 @@ static int print_unex = 1;
* It's been recommended that take about 1/4 of the default speed
* in some more extreme cases.
*/
+static DEFINE_MUTEX(floppy_mutex);
static int slow_floppy;
#include <asm/dma.h>
@@ -3578,9 +3578,9 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&floppy_mutex);
ret = fd_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
+ mutex_unlock(&floppy_mutex);
return ret;
}
@@ -3641,7 +3641,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
{
int drive = (long)disk->private_data;
- lock_kernel();
+ mutex_lock(&floppy_mutex);
mutex_lock(&open_lock);
if (UDRS->fd_ref < 0)
UDRS->fd_ref = 0;
@@ -3652,7 +3652,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
if (!UDRS->fd_ref)
opened_bdev[drive] = NULL;
mutex_unlock(&open_lock);
- unlock_kernel();
+ mutex_unlock(&floppy_mutex);
return 0;
}
@@ -3670,7 +3670,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
int res = -EBUSY;
char *tmp;
- lock_kernel();
+ mutex_lock(&floppy_mutex);
mutex_lock(&open_lock);
old_dev = UDRS->fd_device;
if (opened_bdev[drive] && opened_bdev[drive] != bdev)
@@ -3747,7 +3747,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
goto out;
}
mutex_unlock(&open_lock);
- unlock_kernel();
+ mutex_unlock(&floppy_mutex);
return 0;
out:
if (UDRS->fd_ref < 0)
@@ -3758,7 +3758,7 @@ out:
opened_bdev[drive] = NULL;
out2:
mutex_unlock(&open_lock);
- unlock_kernel();
+ mutex_unlock(&floppy_mutex);
return res;
}
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index dc552308668e..de3083b0a4f5 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -67,7 +67,7 @@
#include <linux/compat.h>
#include <linux/suspend.h>
#include <linux/freezer.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/writeback.h>
#include <linux/buffer_head.h> /* for invalidate_bdev() */
#include <linux/completion.h>
@@ -78,6 +78,7 @@
#include <asm/uaccess.h>
+static DEFINE_MUTEX(loop_mutex);
static LIST_HEAD(loop_devices);
static DEFINE_MUTEX(loop_devices_mutex);
@@ -478,7 +479,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio)
pos = ((loff_t) bio->bi_sector << 9) + lo->lo_offset;
if (bio_rw(bio) == WRITE) {
- bool barrier = (bio->bi_rw & REQ_HARDBARRIER);
+ bool barrier = !!(bio->bi_rw & REQ_HARDBARRIER);
struct file *file = lo->lo_backing_file;
if (barrier) {
@@ -1510,11 +1511,11 @@ static int lo_open(struct block_device *bdev, fmode_t mode)
{
struct loop_device *lo = bdev->bd_disk->private_data;
- lock_kernel();
+ mutex_lock(&loop_mutex);
mutex_lock(&lo->lo_ctl_mutex);
lo->lo_refcnt++;
mutex_unlock(&lo->lo_ctl_mutex);
- unlock_kernel();
+ mutex_unlock(&loop_mutex);
return 0;
}
@@ -1524,7 +1525,7 @@ static int lo_release(struct gendisk *disk, fmode_t mode)
struct loop_device *lo = disk->private_data;
int err;
- lock_kernel();
+ mutex_lock(&loop_mutex);
mutex_lock(&lo->lo_ctl_mutex);
if (--lo->lo_refcnt)
@@ -1549,7 +1550,7 @@ static int lo_release(struct gendisk *disk, fmode_t mode)
out:
mutex_unlock(&lo->lo_ctl_mutex);
out_unlocked:
- lock_kernel();
+ mutex_unlock(&loop_mutex);
return 0;
}
diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c
index b82c5ce5e9df..76fa3deaee84 100644
--- a/drivers/block/mg_disk.c
+++ b/drivers/block/mg_disk.c
@@ -974,8 +974,7 @@ static int mg_probe(struct platform_device *plat_dev)
host->breq->queuedata = host;
/* mflash is random device, thanx for the noop */
- elevator_exit(host->breq->elevator);
- err = elevator_init(host->breq, "noop");
+ err = elevator_change(host->breq, "noop");
if (err) {
printk(KERN_ERR "%s:%d (elevator_init) fail\n",
__func__, __LINE__);
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 0daa422aa281..a32fb41246f8 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -24,7 +24,7 @@
#include <linux/errno.h>
#include <linux/file.h>
#include <linux/ioctl.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/compiler.h>
#include <linux/err.h>
#include <linux/kernel.h>
@@ -53,6 +53,7 @@
#define DBG_BLKDEV 0x0100
#define DBG_RX 0x0200
#define DBG_TX 0x0400
+static DEFINE_MUTEX(nbd_mutex);
static unsigned int debugflags;
#endif /* NDEBUG */
@@ -717,11 +718,11 @@ static int nbd_ioctl(struct block_device *bdev, fmode_t mode,
dprintk(DBG_IOCTL, "%s: nbd_ioctl cmd=%s(0x%x) arg=%lu\n",
lo->disk->disk_name, ioctl_cmd_to_ascii(cmd), cmd, arg);
- lock_kernel();
+ mutex_lock(&nbd_mutex);
mutex_lock(&lo->tx_lock);
error = __nbd_ioctl(bdev, lo, cmd, arg);
mutex_unlock(&lo->tx_lock);
- unlock_kernel();
+ mutex_unlock(&nbd_mutex);
return error;
}
diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c
index 76f8565e1e8d..62cec6afd7ad 100644
--- a/drivers/block/paride/pcd.c
+++ b/drivers/block/paride/pcd.c
@@ -138,9 +138,10 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY};
#include <linux/cdrom.h>
#include <linux/spinlock.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
+static DEFINE_MUTEX(pcd_mutex);
static DEFINE_SPINLOCK(pcd_lock);
module_param(verbose, bool, 0644);
@@ -227,9 +228,9 @@ static int pcd_block_open(struct block_device *bdev, fmode_t mode)
struct pcd_unit *cd = bdev->bd_disk->private_data;
int ret;
- lock_kernel();
+ mutex_lock(&pcd_mutex);
ret = cdrom_open(&cd->info, bdev, mode);
- unlock_kernel();
+ mutex_unlock(&pcd_mutex);
return ret;
}
@@ -237,9 +238,9 @@ static int pcd_block_open(struct block_device *bdev, fmode_t mode)
static int pcd_block_release(struct gendisk *disk, fmode_t mode)
{
struct pcd_unit *cd = disk->private_data;
- lock_kernel();
+ mutex_lock(&pcd_mutex);
cdrom_release(&cd->info, mode);
- unlock_kernel();
+ mutex_unlock(&pcd_mutex);
return 0;
}
@@ -249,9 +250,9 @@ static int pcd_block_ioctl(struct block_device *bdev, fmode_t mode,
struct pcd_unit *cd = bdev->bd_disk->private_data;
int ret;
- lock_kernel();
+ mutex_lock(&pcd_mutex);
ret = cdrom_ioctl(&cd->info, bdev, mode, cmd, arg);
- unlock_kernel();
+ mutex_unlock(&pcd_mutex);
return ret;
}
diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c
index 985f0d4f1d1e..c0ee1558b9bb 100644
--- a/drivers/block/paride/pd.c
+++ b/drivers/block/paride/pd.c
@@ -153,10 +153,11 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_GEO, D_SBY, D_DLY, D_SLV};
#include <linux/blkdev.h>
#include <linux/blkpg.h>
#include <linux/kernel.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
#include <linux/workqueue.h>
+static DEFINE_MUTEX(pd_mutex);
static DEFINE_SPINLOCK(pd_lock);
module_param(verbose, bool, 0);
@@ -736,14 +737,14 @@ static int pd_open(struct block_device *bdev, fmode_t mode)
{
struct pd_unit *disk = bdev->bd_disk->private_data;
- lock_kernel();
+ mutex_lock(&pd_mutex);
disk->access++;
if (disk->removable) {
pd_special_command(disk, pd_media_check);
pd_special_command(disk, pd_door_lock);
}
- unlock_kernel();
+ mutex_unlock(&pd_mutex);
return 0;
}
@@ -771,10 +772,10 @@ static int pd_ioctl(struct block_device *bdev, fmode_t mode,
switch (cmd) {
case CDROMEJECT:
- lock_kernel();
+ mutex_lock(&pd_mutex);
if (disk->access == 1)
pd_special_command(disk, pd_eject);
- unlock_kernel();
+ mutex_unlock(&pd_mutex);
return 0;
default:
return -EINVAL;
@@ -785,10 +786,10 @@ static int pd_release(struct gendisk *p, fmode_t mode)
{
struct pd_unit *disk = p->private_data;
- lock_kernel();
+ mutex_lock(&pd_mutex);
if (!--disk->access && disk->removable)
pd_special_command(disk, pd_door_unlock);
- unlock_kernel();
+ mutex_unlock(&pd_mutex);
return 0;
}
diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c
index 4457b494882a..635f25dd9e10 100644
--- a/drivers/block/paride/pf.c
+++ b/drivers/block/paride/pf.c
@@ -152,9 +152,10 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_LUN, D_DLY};
#include <linux/spinlock.h>
#include <linux/blkdev.h>
#include <linux/blkpg.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
+static DEFINE_MUTEX(pf_mutex);
static DEFINE_SPINLOCK(pf_spin_lock);
module_param(verbose, bool, 0644);
@@ -302,7 +303,7 @@ static int pf_open(struct block_device *bdev, fmode_t mode)
struct pf_unit *pf = bdev->bd_disk->private_data;
int ret;
- lock_kernel();
+ mutex_lock(&pf_mutex);
pf_identify(pf);
ret = -ENODEV;
@@ -318,7 +319,7 @@ static int pf_open(struct block_device *bdev, fmode_t mode)
if (pf->removable)
pf_lock(pf, 1);
out:
- unlock_kernel();
+ mutex_unlock(&pf_mutex);
return ret;
}
@@ -349,9 +350,9 @@ static int pf_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, u
if (pf->access != 1)
return -EBUSY;
- lock_kernel();
+ mutex_lock(&pf_mutex);
pf_eject(pf);
- unlock_kernel();
+ mutex_unlock(&pf_mutex);
return 0;
}
@@ -360,9 +361,9 @@ static int pf_release(struct gendisk *disk, fmode_t mode)
{
struct pf_unit *pf = disk->private_data;
- lock_kernel();
+ mutex_lock(&pf_mutex);
if (pf->access <= 0) {
- unlock_kernel();
+ mutex_unlock(&pf_mutex);
return -EINVAL;
}
@@ -371,7 +372,7 @@ static int pf_release(struct gendisk *disk, fmode_t mode)
if (!pf->access && pf->removable)
pf_lock(pf, 0);
- unlock_kernel();
+ mutex_unlock(&pf_mutex);
return 0;
}
diff --git a/drivers/block/paride/pg.c b/drivers/block/paride/pg.c
index c397b3ddba9b..6b9a2000d56a 100644
--- a/drivers/block/paride/pg.c
+++ b/drivers/block/paride/pg.c
@@ -162,7 +162,7 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY};
#include <linux/pg.h>
#include <linux/device.h>
#include <linux/sched.h> /* current, TASK_* */
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/jiffies.h>
#include <asm/uaccess.h>
@@ -193,6 +193,7 @@ module_param_array(drive3, int, NULL, 0);
#define ATAPI_IDENTIFY 0x12
+static DEFINE_MUTEX(pg_mutex);
static int pg_open(struct inode *inode, struct file *file);
static int pg_release(struct inode *inode, struct file *file);
static ssize_t pg_read(struct file *filp, char __user *buf,
@@ -234,6 +235,7 @@ static const struct file_operations pg_fops = {
.write = pg_write,
.open = pg_open,
.release = pg_release,
+ .llseek = noop_llseek,
};
static void pg_init_units(void)
@@ -518,7 +520,7 @@ static int pg_open(struct inode *inode, struct file *file)
struct pg *dev = &devices[unit];
int ret = 0;
- lock_kernel();
+ mutex_lock(&pg_mutex);
if ((unit >= PG_UNITS) || (!dev->present)) {
ret = -ENODEV;
goto out;
@@ -547,7 +549,7 @@ static int pg_open(struct inode *inode, struct file *file)
file->private_data = dev;
out:
- unlock_kernel();
+ mutex_unlock(&pg_mutex);
return ret;
}
diff --git a/drivers/block/paride/pt.c b/drivers/block/paride/pt.c
index bc5825fdeaab..7179f79d7468 100644
--- a/drivers/block/paride/pt.c
+++ b/drivers/block/paride/pt.c
@@ -146,7 +146,7 @@ static int (*drives[4])[6] = {&drive0, &drive1, &drive2, &drive3};
#include <linux/mtio.h>
#include <linux/device.h>
#include <linux/sched.h> /* current, TASK_*, schedule_timeout() */
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <asm/uaccess.h>
@@ -189,6 +189,7 @@ module_param_array(drive3, int, NULL, 0);
#define ATAPI_MODE_SENSE 0x1a
#define ATAPI_LOG_SENSE 0x4d
+static DEFINE_MUTEX(pt_mutex);
static int pt_open(struct inode *inode, struct file *file);
static long pt_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
static int pt_release(struct inode *inode, struct file *file);
@@ -239,6 +240,7 @@ static const struct file_operations pt_fops = {
.unlocked_ioctl = pt_ioctl,
.open = pt_open,
.release = pt_release,
+ .llseek = noop_llseek,
};
/* sysfs class support */
@@ -650,9 +652,9 @@ static int pt_open(struct inode *inode, struct file *file)
struct pt_unit *tape = pt + unit;
int err;
- lock_kernel();
+ mutex_lock(&pt_mutex);
if (unit >= PT_UNITS || (!tape->present)) {
- unlock_kernel();
+ mutex_unlock(&pt_mutex);
return -ENODEV;
}
@@ -681,12 +683,12 @@ static int pt_open(struct inode *inode, struct file *file)
}
file->private_data = tape;
- unlock_kernel();
+ mutex_unlock(&pt_mutex);
return 0;
out:
atomic_inc(&tape->available);
- unlock_kernel();
+ mutex_unlock(&pt_mutex);
return err;
}
@@ -704,15 +706,15 @@ static long pt_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
switch (mtop.mt_op) {
case MTREW:
- lock_kernel();
+ mutex_lock(&pt_mutex);
pt_rewind(tape);
- unlock_kernel();
+ mutex_unlock(&pt_mutex);
return 0;
case MTWEOF:
- lock_kernel();
+ mutex_lock(&pt_mutex);
pt_write_fm(tape);
- unlock_kernel();
+ mutex_unlock(&pt_mutex);
return 0;
default:
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index b1cbeb59bb76..ef58fccadad3 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -57,7 +57,6 @@
#include <linux/seq_file.h>
#include <linux/miscdevice.h>
#include <linux/freezer.h>
-#include <linux/smp_lock.h>
#include <linux/mutex.h>
#include <linux/slab.h>
#include <scsi/scsi_cmnd.h>
@@ -86,6 +85,7 @@
#define ZONE(sector, pd) (((sector) + (pd)->offset) & ~((pd)->settings.size - 1))
+static DEFINE_MUTEX(pktcdvd_mutex);
static struct pktcdvd_device *pkt_devs[MAX_WRITERS];
static struct proc_dir_entry *pkt_proc;
static int pktdev_major;
@@ -2369,7 +2369,7 @@ static void pkt_release_dev(struct pktcdvd_device *pd, int flush)
pkt_shrink_pktlist(pd);
}
-static struct pktcdvd_device *pkt_find_dev_from_minor(int dev_minor)
+static struct pktcdvd_device *pkt_find_dev_from_minor(unsigned int dev_minor)
{
if (dev_minor >= MAX_WRITERS)
return NULL;
@@ -2383,7 +2383,7 @@ static int pkt_open(struct block_device *bdev, fmode_t mode)
VPRINTK(DRIVER_NAME": entering open\n");
- lock_kernel();
+ mutex_lock(&pktcdvd_mutex);
mutex_lock(&ctl_mutex);
pd = pkt_find_dev_from_minor(MINOR(bdev->bd_dev));
if (!pd) {
@@ -2411,7 +2411,7 @@ static int pkt_open(struct block_device *bdev, fmode_t mode)
}
mutex_unlock(&ctl_mutex);
- unlock_kernel();
+ mutex_unlock(&pktcdvd_mutex);
return 0;
out_dec:
@@ -2419,7 +2419,7 @@ out_dec:
out:
VPRINTK(DRIVER_NAME": failed open (%d)\n", ret);
mutex_unlock(&ctl_mutex);
- unlock_kernel();
+ mutex_unlock(&pktcdvd_mutex);
return ret;
}
@@ -2428,7 +2428,7 @@ static int pkt_close(struct gendisk *disk, fmode_t mode)
struct pktcdvd_device *pd = disk->private_data;
int ret = 0;
- lock_kernel();
+ mutex_lock(&pktcdvd_mutex);
mutex_lock(&ctl_mutex);
pd->refcnt--;
BUG_ON(pd->refcnt < 0);
@@ -2437,7 +2437,7 @@ static int pkt_close(struct gendisk *disk, fmode_t mode)
pkt_release_dev(pd, flush);
}
mutex_unlock(&ctl_mutex);
- unlock_kernel();
+ mutex_unlock(&pktcdvd_mutex);
return ret;
}
@@ -2773,7 +2773,7 @@ static int pkt_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd,
VPRINTK("pkt_ioctl: cmd %x, dev %d:%d\n", cmd,
MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
- lock_kernel();
+ mutex_lock(&pktcdvd_mutex);
switch (cmd) {
case CDROMEJECT:
/*
@@ -2798,7 +2798,7 @@ static int pkt_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd,
VPRINTK(DRIVER_NAME": Unknown ioctl for %s (%x)\n", pd->name, cmd);
ret = -ENOTTY;
}
- unlock_kernel();
+ mutex_unlock(&pktcdvd_mutex);
return ret;
}
@@ -3046,6 +3046,7 @@ static const struct file_operations pkt_ctl_fops = {
.compat_ioctl = pkt_ctl_compat_ioctl,
#endif
.owner = THIS_MODULE,
+ .llseek = no_llseek,
};
static struct miscdevice pkt_misc = {
diff --git a/drivers/block/ps3disk.c b/drivers/block/ps3disk.c
index e9da874d0419..03688c2da319 100644
--- a/drivers/block/ps3disk.c
+++ b/drivers/block/ps3disk.c
@@ -113,7 +113,7 @@ static void ps3disk_scatter_gather(struct ps3_storage_device *dev,
memcpy(buf, dev->bounce_buf+offset, size);
offset += size;
flush_kernel_dcache_page(bvec->bv_page);
- bvec_kunmap_irq(bvec, &flags);
+ bvec_kunmap_irq(buf, &flags);
i++;
}
}
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
new file mode 100644
index 000000000000..6ec9d53806c5
--- /dev/null
+++ b/drivers/block/rbd.c
@@ -0,0 +1,1841 @@
+/*
+ rbd.c -- Export ceph rados objects as a Linux block device
+
+
+ based on drivers/block/osdblk.c:
+
+ Copyright 2009 Red Hat, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; see the file COPYING. If not, write to
+ the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+
+
+ Instructions for use
+ --------------------
+
+ 1) Map a Linux block device to an existing rbd image.
+
+ Usage: <mon ip addr> <options> <pool name> <rbd image name> [snap name]
+
+ $ echo "192.168.0.1 name=admin rbd foo" > /sys/class/rbd/add
+
+ The snapshot name can be "-" or omitted to map the image read/write.
+
+ 2) List all active blkdev<->object mappings.
+
+ In this example, we have performed step #1 twice, creating two blkdevs,
+ mapped to two separate rados objects in the rados rbd pool
+
+ $ cat /sys/class/rbd/list
+ #id major client_name pool name snap KB
+ 0 254 client4143 rbd foo - 1024000
+
+ The columns, in order, are:
+ - blkdev unique id
+ - blkdev assigned major
+ - rados client id
+ - rados pool name
+ - rados block device name
+ - mapped snapshot ("-" if none)
+ - device size in KB
+
+
+ 3) Create a snapshot.
+
+ Usage: <blkdev id> <snapname>
+
+ $ echo "0 mysnap" > /sys/class/rbd/snap_create
+
+
+ 4) Listing a snapshot.
+
+ $ cat /sys/class/rbd/snaps_list
+ #id snap KB
+ 0 - 1024000 (*)
+ 0 foo 1024000
+
+ The columns, in order, are:
+ - blkdev unique id
+ - snapshot name, '-' means none (active read/write version)
+ - size of device at time of snapshot
+ - the (*) indicates this is the active version
+
+ 5) Rollback to snapshot.
+
+ Usage: <blkdev id> <snapname>
+
+ $ echo "0 mysnap" > /sys/class/rbd/snap_rollback
+
+
+ 6) Mapping an image using snapshot.
+
+ A snapshot mapping is read-only. This is being done by passing
+ snap=<snapname> to the options when adding a device.
+
+ $ echo "192.168.0.1 name=admin,snap=mysnap rbd foo" > /sys/class/rbd/add
+
+
+ 7) Remove an active blkdev<->rbd image mapping.
+
+ In this example, we remove the mapping with blkdev unique id 1.
+
+ $ echo 1 > /sys/class/rbd/remove
+
+
+ NOTE: The actual creation and deletion of rados objects is outside the scope
+ of this driver.
+
+ */
+
+#include <linux/ceph/libceph.h>
+#include <linux/ceph/osd_client.h>
+#include <linux/ceph/mon_client.h>
+#include <linux/ceph/decode.h>
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/blkdev.h>
+
+#include "rbd_types.h"
+
+#define DRV_NAME "rbd"
+#define DRV_NAME_LONG "rbd (rados block device)"
+
+#define RBD_MINORS_PER_MAJOR 256 /* max minors per blkdev */
+
+#define RBD_MAX_MD_NAME_LEN (96 + sizeof(RBD_SUFFIX))
+#define RBD_MAX_POOL_NAME_LEN 64
+#define RBD_MAX_SNAP_NAME_LEN 32
+#define RBD_MAX_OPT_LEN 1024
+
+#define RBD_SNAP_HEAD_NAME "-"
+
+#define DEV_NAME_LEN 32
+
+/*
+ * block device image metadata (in-memory version)
+ */
+struct rbd_image_header {
+ u64 image_size;
+ char block_name[32];
+ __u8 obj_order;
+ __u8 crypt_type;
+ __u8 comp_type;
+ struct rw_semaphore snap_rwsem;
+ struct ceph_snap_context *snapc;
+ size_t snap_names_len;
+ u64 snap_seq;
+ u32 total_snaps;
+
+ char *snap_names;
+ u64 *snap_sizes;
+};
+
+/*
+ * an instance of the client. multiple devices may share a client.
+ */
+struct rbd_client {
+ struct ceph_client *client;
+ struct kref kref;
+ struct list_head node;
+};
+
+/*
+ * a single io request
+ */
+struct rbd_request {
+ struct request *rq; /* blk layer request */
+ struct bio *bio; /* cloned bio */
+ struct page **pages; /* list of used pages */
+ u64 len;
+};
+
+/*
+ * a single device
+ */
+struct rbd_device {
+ int id; /* blkdev unique id */
+
+ int major; /* blkdev assigned major */
+ struct gendisk *disk; /* blkdev's gendisk and rq */
+ struct request_queue *q;
+
+ struct ceph_client *client;
+ struct rbd_client *rbd_client;
+
+ char name[DEV_NAME_LEN]; /* blkdev name, e.g. rbd3 */
+
+ spinlock_t lock; /* queue lock */
+
+ struct rbd_image_header header;
+ char obj[RBD_MAX_OBJ_NAME_LEN]; /* rbd image name */
+ int obj_len;
+ char obj_md_name[RBD_MAX_MD_NAME_LEN]; /* hdr nm. */
+ char pool_name[RBD_MAX_POOL_NAME_LEN];
+ int poolid;
+
+ char snap_name[RBD_MAX_SNAP_NAME_LEN];
+ u32 cur_snap; /* index+1 of current snapshot within snap context
+ 0 - for the head */
+ int read_only;
+
+ struct list_head node;
+};
+
+static spinlock_t node_lock; /* protects client get/put */
+
+static struct class *class_rbd; /* /sys/class/rbd */
+static DEFINE_MUTEX(ctl_mutex); /* Serialize open/close/setup/teardown */
+static LIST_HEAD(rbd_dev_list); /* devices */
+static LIST_HEAD(rbd_client_list); /* clients */
+
+
+static int rbd_open(struct block_device *bdev, fmode_t mode)
+{
+ struct gendisk *disk = bdev->bd_disk;
+ struct rbd_device *rbd_dev = disk->private_data;
+
+ set_device_ro(bdev, rbd_dev->read_only);
+
+ if ((mode & FMODE_WRITE) && rbd_dev->read_only)
+ return -EROFS;
+
+ return 0;
+}
+
+static const struct block_device_operations rbd_bd_ops = {
+ .owner = THIS_MODULE,
+ .open = rbd_open,
+};
+
+/*
+ * Initialize an rbd client instance.
+ * We own *opt.
+ */
+static struct rbd_client *rbd_client_create(struct ceph_options *opt)
+{
+ struct rbd_client *rbdc;
+ int ret = -ENOMEM;
+
+ dout("rbd_client_create\n");
+ rbdc = kmalloc(sizeof(struct rbd_client), GFP_KERNEL);
+ if (!rbdc)
+ goto out_opt;
+
+ kref_init(&rbdc->kref);
+ INIT_LIST_HEAD(&rbdc->node);
+
+ rbdc->client = ceph_create_client(opt, rbdc);
+ if (IS_ERR(rbdc->client))
+ goto out_rbdc;
+ opt = NULL; /* Now rbdc->client is responsible for opt */
+
+ ret = ceph_open_session(rbdc->client);
+ if (ret < 0)
+ goto out_err;
+
+ spin_lock(&node_lock);
+ list_add_tail(&rbdc->node, &rbd_client_list);
+ spin_unlock(&node_lock);
+
+ dout("rbd_client_create created %p\n", rbdc);
+ return rbdc;
+
+out_err:
+ ceph_destroy_client(rbdc->client);
+out_rbdc:
+ kfree(rbdc);
+out_opt:
+ if (opt)
+ ceph_destroy_options(opt);
+ return ERR_PTR(ret);
+}
+
+/*
+ * Find a ceph client with specific addr and configuration.
+ */
+static struct rbd_client *__rbd_client_find(struct ceph_options *opt)
+{
+ struct rbd_client *client_node;
+
+ if (opt->flags & CEPH_OPT_NOSHARE)
+ return NULL;
+
+ list_for_each_entry(client_node, &rbd_client_list, node)
+ if (ceph_compare_options(opt, client_node->client) == 0)
+ return client_node;
+ return NULL;
+}
+
+/*
+ * Get a ceph client with specific addr and configuration, if one does
+ * not exist create it.
+ */
+static int rbd_get_client(struct rbd_device *rbd_dev, const char *mon_addr,
+ char *options)
+{
+ struct rbd_client *rbdc;
+ struct ceph_options *opt;
+ int ret;
+
+ ret = ceph_parse_options(&opt, options, mon_addr,
+ mon_addr + strlen(mon_addr), NULL, NULL);
+ if (ret < 0)
+ return ret;
+
+ spin_lock(&node_lock);
+ rbdc = __rbd_client_find(opt);
+ if (rbdc) {
+ ceph_destroy_options(opt);
+
+ /* using an existing client */
+ kref_get(&rbdc->kref);
+ rbd_dev->rbd_client = rbdc;
+ rbd_dev->client = rbdc->client;
+ spin_unlock(&node_lock);
+ return 0;
+ }
+ spin_unlock(&node_lock);
+
+ rbdc = rbd_client_create(opt);
+ if (IS_ERR(rbdc))
+ return PTR_ERR(rbdc);
+
+ rbd_dev->rbd_client = rbdc;
+ rbd_dev->client = rbdc->client;
+ return 0;
+}
+
+/*
+ * Destroy ceph client
+ */
+static void rbd_client_release(struct kref *kref)
+{
+ struct rbd_client *rbdc = container_of(kref, struct rbd_client, kref);
+
+ dout("rbd_release_client %p\n", rbdc);
+ spin_lock(&node_lock);
+ list_del(&rbdc->node);
+ spin_unlock(&node_lock);
+
+ ceph_destroy_client(rbdc->client);
+ kfree(rbdc);
+}
+
+/*
+ * Drop reference to ceph client node. If it's not referenced anymore, release
+ * it.
+ */
+static void rbd_put_client(struct rbd_device *rbd_dev)
+{
+ kref_put(&rbd_dev->rbd_client->kref, rbd_client_release);
+ rbd_dev->rbd_client = NULL;
+ rbd_dev->client = NULL;
+}
+
+
+/*
+ * Create a new header structure, translate header format from the on-disk
+ * header.
+ */
+static int rbd_header_from_disk(struct rbd_image_header *header,
+ struct rbd_image_header_ondisk *ondisk,
+ int allocated_snaps,
+ gfp_t gfp_flags)
+{
+ int i;
+ u32 snap_count = le32_to_cpu(ondisk->snap_count);
+ int ret = -ENOMEM;
+
+ init_rwsem(&header->snap_rwsem);
+
+ header->snap_names_len = le64_to_cpu(ondisk->snap_names_len);
+ header->snapc = kmalloc(sizeof(struct ceph_snap_context) +
+ snap_count *
+ sizeof(struct rbd_image_snap_ondisk),
+ gfp_flags);
+ if (!header->snapc)
+ return -ENOMEM;
+ if (snap_count) {
+ header->snap_names = kmalloc(header->snap_names_len,
+ GFP_KERNEL);
+ if (!header->snap_names)
+ goto err_snapc;
+ header->snap_sizes = kmalloc(snap_count * sizeof(u64),
+ GFP_KERNEL);
+ if (!header->snap_sizes)
+ goto err_names;
+ } else {
+ header->snap_names = NULL;
+ header->snap_sizes = NULL;
+ }
+ memcpy(header->block_name, ondisk->block_name,
+ sizeof(ondisk->block_name));
+
+ header->image_size = le64_to_cpu(ondisk->image_size);
+ header->obj_order = ondisk->options.order;
+ header->crypt_type = ondisk->options.crypt_type;
+ header->comp_type = ondisk->options.comp_type;
+
+ atomic_set(&header->snapc->nref, 1);
+ header->snap_seq = le64_to_cpu(ondisk->snap_seq);
+ header->snapc->num_snaps = snap_count;
+ header->total_snaps = snap_count;
+
+ if (snap_count &&
+ allocated_snaps == snap_count) {
+ for (i = 0; i < snap_count; i++) {
+ header->snapc->snaps[i] =
+ le64_to_cpu(ondisk->snaps[i].id);
+ header->snap_sizes[i] =
+ le64_to_cpu(ondisk->snaps[i].image_size);
+ }
+
+ /* copy snapshot names */
+ memcpy(header->snap_names, &ondisk->snaps[i],
+ header->snap_names_len);
+ }
+
+ return 0;
+
+err_names:
+ kfree(header->snap_names);
+err_snapc:
+ kfree(header->snapc);
+ return ret;
+}
+
+static int snap_index(struct rbd_image_header *header, int snap_num)
+{
+ return header->total_snaps - snap_num;
+}
+
+static u64 cur_snap_id(struct rbd_device *rbd_dev)
+{
+ struct rbd_image_header *header = &rbd_dev->header;
+
+ if (!rbd_dev->cur_snap)
+ return 0;
+
+ return header->snapc->snaps[snap_index(header, rbd_dev->cur_snap)];
+}
+
+static int snap_by_name(struct rbd_image_header *header, const char *snap_name,
+ u64 *seq, u64 *size)
+{
+ int i;
+ char *p = header->snap_names;
+
+ for (i = 0; i < header->total_snaps; i++, p += strlen(p) + 1) {
+ if (strcmp(snap_name, p) == 0)
+ break;
+ }
+ if (i == header->total_snaps)
+ return -ENOENT;
+ if (seq)
+ *seq = header->snapc->snaps[i];
+
+ if (size)
+ *size = header->snap_sizes[i];
+
+ return i;
+}
+
+static int rbd_header_set_snap(struct rbd_device *dev,
+ const char *snap_name,
+ u64 *size)
+{
+ struct rbd_image_header *header = &dev->header;
+ struct ceph_snap_context *snapc = header->snapc;
+ int ret = -ENOENT;
+
+ down_write(&header->snap_rwsem);
+
+ if (!snap_name ||
+ !*snap_name ||
+ strcmp(snap_name, "-") == 0 ||
+ strcmp(snap_name, RBD_SNAP_HEAD_NAME) == 0) {
+ if (header->total_snaps)
+ snapc->seq = header->snap_seq;
+ else
+ snapc->seq = 0;
+ dev->cur_snap = 0;
+ dev->read_only = 0;
+ if (size)
+ *size = header->image_size;
+ } else {
+ ret = snap_by_name(header, snap_name, &snapc->seq, size);
+ if (ret < 0)
+ goto done;
+
+ dev->cur_snap = header->total_snaps - ret;
+ dev->read_only = 1;
+ }
+
+ ret = 0;
+done:
+ up_write(&header->snap_rwsem);
+ return ret;
+}
+
+static void rbd_header_free(struct rbd_image_header *header)
+{
+ kfree(header->snapc);
+ kfree(header->snap_names);
+ kfree(header->snap_sizes);
+}
+
+/*
+ * get the actual striped segment name, offset and length
+ */
+static u64 rbd_get_segment(struct rbd_image_header *header,
+ const char *block_name,
+ u64 ofs, u64 len,
+ char *seg_name, u64 *segofs)
+{
+ u64 seg = ofs >> header->obj_order;
+
+ if (seg_name)
+ snprintf(seg_name, RBD_MAX_SEG_NAME_LEN,
+ "%s.%012llx", block_name, seg);
+
+ ofs = ofs & ((1 << header->obj_order) - 1);
+ len = min_t(u64, len, (1 << header->obj_order) - ofs);
+
+ if (segofs)
+ *segofs = ofs;
+
+ return len;
+}
+
+/*
+ * bio helpers
+ */
+
+static void bio_chain_put(struct bio *chain)
+{
+ struct bio *tmp;
+
+ while (chain) {
+ tmp = chain;
+ chain = chain->bi_next;
+ bio_put(tmp);
+ }
+}
+
+/*
+ * zeros a bio chain, starting at specific offset
+ */
+static void zero_bio_chain(struct bio *chain, int start_ofs)
+{
+ struct bio_vec *bv;
+ unsigned long flags;
+ void *buf;
+ int i;
+ int pos = 0;
+
+ while (chain) {
+ bio_for_each_segment(bv, chain, i) {
+ if (pos + bv->bv_len > start_ofs) {
+ int remainder = max(start_ofs - pos, 0);
+ buf = bvec_kmap_irq(bv, &flags);
+ memset(buf + remainder, 0,
+ bv->bv_len - remainder);
+ bvec_kunmap_irq(buf, &flags);
+ }
+ pos += bv->bv_len;
+ }
+
+ chain = chain->bi_next;
+ }
+}
+
+/*
+ * bio_chain_clone - clone a chain of bios up to a certain length.
+ * might return a bio_pair that will need to be released.
+ */
+static struct bio *bio_chain_clone(struct bio **old, struct bio **next,
+ struct bio_pair **bp,
+ int len, gfp_t gfpmask)
+{
+ struct bio *tmp, *old_chain = *old, *new_chain = NULL, *tail = NULL;
+ int total = 0;
+
+ if (*bp) {
+ bio_pair_release(*bp);
+ *bp = NULL;
+ }
+
+ while (old_chain && (total < len)) {
+ tmp = bio_kmalloc(gfpmask, old_chain->bi_max_vecs);
+ if (!tmp)
+ goto err_out;
+
+ if (total + old_chain->bi_size > len) {
+ struct bio_pair *bp;
+
+ /*
+ * this split can only happen with a single paged bio,
+ * split_bio will BUG_ON if this is not the case
+ */
+ dout("bio_chain_clone split! total=%d remaining=%d"
+ "bi_size=%d\n",
+ (int)total, (int)len-total,
+ (int)old_chain->bi_size);
+
+ /* split the bio. We'll release it either in the next
+ call, or it will have to be released outside */
+ bp = bio_split(old_chain, (len - total) / 512ULL);
+ if (!bp)
+ goto err_out;
+
+ __bio_clone(tmp, &bp->bio1);
+
+ *next = &bp->bio2;
+ } else {
+ __bio_clone(tmp, old_chain);
+ *next = old_chain->bi_next;
+ }
+
+ tmp->bi_bdev = NULL;
+ gfpmask &= ~__GFP_WAIT;
+ tmp->bi_next = NULL;
+
+ if (!new_chain) {
+ new_chain = tail = tmp;
+ } else {
+ tail->bi_next = tmp;
+ tail = tmp;
+ }
+ old_chain = old_chain->bi_next;
+
+ total += tmp->bi_size;
+ }
+
+ BUG_ON(total < len);
+
+ if (tail)
+ tail->bi_next = NULL;
+
+ *old = old_chain;
+
+ return new_chain;
+
+err_out:
+ dout("bio_chain_clone with err\n");
+ bio_chain_put(new_chain);
+ return NULL;
+}
+
+/*
+ * helpers for osd request op vectors.
+ */
+static int rbd_create_rw_ops(struct ceph_osd_req_op **ops,
+ int num_ops,
+ int opcode,
+ u32 payload_len)
+{
+ *ops = kzalloc(sizeof(struct ceph_osd_req_op) * (num_ops + 1),
+ GFP_NOIO);
+ if (!*ops)
+ return -ENOMEM;
+ (*ops)[0].op = opcode;
+ /*
+ * op extent offset and length will be set later on
+ * in calc_raw_layout()
+ */
+ (*ops)[0].payload_len = payload_len;
+ return 0;
+}
+
+static void rbd_destroy_ops(struct ceph_osd_req_op *ops)
+{
+ kfree(ops);
+}
+
+/*
+ * Send ceph osd request
+ */
+static int rbd_do_request(struct request *rq,
+ struct rbd_device *dev,
+ struct ceph_snap_context *snapc,
+ u64 snapid,
+ const char *obj, u64 ofs, u64 len,
+ struct bio *bio,
+ struct page **pages,
+ int num_pages,
+ int flags,
+ struct ceph_osd_req_op *ops,
+ int num_reply,
+ void (*rbd_cb)(struct ceph_osd_request *req,
+ struct ceph_msg *msg))
+{
+ struct ceph_osd_request *req;
+ struct ceph_file_layout *layout;
+ int ret;
+ u64 bno;
+ struct timespec mtime = CURRENT_TIME;
+ struct rbd_request *req_data;
+ struct ceph_osd_request_head *reqhead;
+ struct rbd_image_header *header = &dev->header;
+
+ ret = -ENOMEM;
+ req_data = kzalloc(sizeof(*req_data), GFP_NOIO);
+ if (!req_data)
+ goto done;
+
+ dout("rbd_do_request len=%lld ofs=%lld\n", len, ofs);
+
+ down_read(&header->snap_rwsem);
+
+ req = ceph_osdc_alloc_request(&dev->client->osdc, flags,
+ snapc,
+ ops,
+ false,
+ GFP_NOIO, pages, bio);
+ if (IS_ERR(req)) {
+ up_read(&header->snap_rwsem);
+ ret = PTR_ERR(req);
+ goto done_pages;
+ }
+
+ req->r_callback = rbd_cb;
+
+ req_data->rq = rq;
+ req_data->bio = bio;
+ req_data->pages = pages;
+ req_data->len = len;
+
+ req->r_priv = req_data;
+
+ reqhead = req->r_request->front.iov_base;
+ reqhead->snapid = cpu_to_le64(CEPH_NOSNAP);
+
+ strncpy(req->r_oid, obj, sizeof(req->r_oid));
+ req->r_oid_len = strlen(req->r_oid);
+
+ layout = &req->r_file_layout;
+ memset(layout, 0, sizeof(*layout));
+ layout->fl_stripe_unit = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER);
+ layout->fl_stripe_count = cpu_to_le32(1);
+ layout->fl_object_size = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER);
+ layout->fl_pg_preferred = cpu_to_le32(-1);
+ layout->fl_pg_pool = cpu_to_le32(dev->poolid);
+ ceph_calc_raw_layout(&dev->client->osdc, layout, snapid,
+ ofs, &len, &bno, req, ops);
+
+ ceph_osdc_build_request(req, ofs, &len,
+ ops,
+ snapc,
+ &mtime,
+ req->r_oid, req->r_oid_len);
+ up_read(&header->snap_rwsem);
+
+ ret = ceph_osdc_start_request(&dev->client->osdc, req, false);
+ if (ret < 0)
+ goto done_err;
+
+ if (!rbd_cb) {
+ ret = ceph_osdc_wait_request(&dev->client->osdc, req);
+ ceph_osdc_put_request(req);
+ }
+ return ret;
+
+done_err:
+ bio_chain_put(req_data->bio);
+ ceph_osdc_put_request(req);
+done_pages:
+ kfree(req_data);
+done:
+ if (rq)
+ blk_end_request(rq, ret, len);
+ return ret;
+}
+
+/*
+ * Ceph osd op callback
+ */
+static void rbd_req_cb(struct ceph_osd_request *req, struct ceph_msg *msg)
+{
+ struct rbd_request *req_data = req->r_priv;
+ struct ceph_osd_reply_head *replyhead;
+ struct ceph_osd_op *op;
+ __s32 rc;
+ u64 bytes;
+ int read_op;
+
+ /* parse reply */
+ replyhead = msg->front.iov_base;
+ WARN_ON(le32_to_cpu(replyhead->num_ops) == 0);
+ op = (void *)(replyhead + 1);
+ rc = le32_to_cpu(replyhead->result);
+ bytes = le64_to_cpu(op->extent.length);
+ read_op = (le32_to_cpu(op->op) == CEPH_OSD_OP_READ);
+
+ dout("rbd_req_cb bytes=%lld readop=%d rc=%d\n", bytes, read_op, rc);
+
+ if (rc == -ENOENT && read_op) {
+ zero_bio_chain(req_data->bio, 0);
+ rc = 0;
+ } else if (rc == 0 && read_op && bytes < req_data->len) {
+ zero_bio_chain(req_data->bio, bytes);
+ bytes = req_data->len;
+ }
+
+ blk_end_request(req_data->rq, rc, bytes);
+
+ if (req_data->bio)
+ bio_chain_put(req_data->bio);
+
+ ceph_osdc_put_request(req);
+ kfree(req_data);
+}
+
+/*
+ * Do a synchronous ceph osd operation
+ */
+static int rbd_req_sync_op(struct rbd_device *dev,
+ struct ceph_snap_context *snapc,
+ u64 snapid,
+ int opcode,
+ int flags,
+ struct ceph_osd_req_op *orig_ops,
+ int num_reply,
+ const char *obj,
+ u64 ofs, u64 len,
+ char *buf)
+{
+ int ret;
+ struct page **pages;
+ int num_pages;
+ struct ceph_osd_req_op *ops = orig_ops;
+ u32 payload_len;
+
+ num_pages = calc_pages_for(ofs , len);
+ pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL);
+ if (IS_ERR(pages))
+ return PTR_ERR(pages);
+
+ if (!orig_ops) {
+ payload_len = (flags & CEPH_OSD_FLAG_WRITE ? len : 0);
+ ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len);
+ if (ret < 0)
+ goto done;
+
+ if ((flags & CEPH_OSD_FLAG_WRITE) && buf) {
+ ret = ceph_copy_to_page_vector(pages, buf, ofs, len);
+ if (ret < 0)
+ goto done_ops;
+ }
+ }
+
+ ret = rbd_do_request(NULL, dev, snapc, snapid,
+ obj, ofs, len, NULL,
+ pages, num_pages,
+ flags,
+ ops,
+ 2,
+ NULL);
+ if (ret < 0)
+ goto done_ops;
+
+ if ((flags & CEPH_OSD_FLAG_READ) && buf)
+ ret = ceph_copy_from_page_vector(pages, buf, ofs, ret);
+
+done_ops:
+ if (!orig_ops)
+ rbd_destroy_ops(ops);
+done:
+ ceph_release_page_vector(pages, num_pages);
+ return ret;
+}
+
+/*
+ * Do an asynchronous ceph osd operation
+ */
+static int rbd_do_op(struct request *rq,
+ struct rbd_device *rbd_dev ,
+ struct ceph_snap_context *snapc,
+ u64 snapid,
+ int opcode, int flags, int num_reply,
+ u64 ofs, u64 len,
+ struct bio *bio)
+{
+ char *seg_name;
+ u64 seg_ofs;
+ u64 seg_len;
+ int ret;
+ struct ceph_osd_req_op *ops;
+ u32 payload_len;
+
+ seg_name = kmalloc(RBD_MAX_SEG_NAME_LEN + 1, GFP_NOIO);
+ if (!seg_name)
+ return -ENOMEM;
+
+ seg_len = rbd_get_segment(&rbd_dev->header,
+ rbd_dev->header.block_name,
+ ofs, len,
+ seg_name, &seg_ofs);
+
+ payload_len = (flags & CEPH_OSD_FLAG_WRITE ? seg_len : 0);
+
+ ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len);
+ if (ret < 0)
+ goto done;
+
+ /* we've taken care of segment sizes earlier when we
+ cloned the bios. We should never have a segment
+ truncated at this point */
+ BUG_ON(seg_len < len);
+
+ ret = rbd_do_request(rq, rbd_dev, snapc, snapid,
+ seg_name, seg_ofs, seg_len,
+ bio,
+ NULL, 0,
+ flags,
+ ops,
+ num_reply,
+ rbd_req_cb);
+done:
+ kfree(seg_name);
+ return ret;
+}
+
+/*
+ * Request async osd write
+ */
+static int rbd_req_write(struct request *rq,
+ struct rbd_device *rbd_dev,
+ struct ceph_snap_context *snapc,
+ u64 ofs, u64 len,
+ struct bio *bio)
+{
+ return rbd_do_op(rq, rbd_dev, snapc, CEPH_NOSNAP,
+ CEPH_OSD_OP_WRITE,
+ CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+ 2,
+ ofs, len, bio);
+}
+
+/*
+ * Request async osd read
+ */
+static int rbd_req_read(struct request *rq,
+ struct rbd_device *rbd_dev,
+ u64 snapid,
+ u64 ofs, u64 len,
+ struct bio *bio)
+{
+ return rbd_do_op(rq, rbd_dev, NULL,
+ (snapid ? snapid : CEPH_NOSNAP),
+ CEPH_OSD_OP_READ,
+ CEPH_OSD_FLAG_READ,
+ 2,
+ ofs, len, bio);
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_read(struct rbd_device *dev,
+ struct ceph_snap_context *snapc,
+ u64 snapid,
+ const char *obj,
+ u64 ofs, u64 len,
+ char *buf)
+{
+ return rbd_req_sync_op(dev, NULL,
+ (snapid ? snapid : CEPH_NOSNAP),
+ CEPH_OSD_OP_READ,
+ CEPH_OSD_FLAG_READ,
+ NULL,
+ 1, obj, ofs, len, buf);
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_rollback_obj(struct rbd_device *dev,
+ u64 snapid,
+ const char *obj)
+{
+ struct ceph_osd_req_op *ops;
+ int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_ROLLBACK, 0);
+ if (ret < 0)
+ return ret;
+
+ ops[0].snap.snapid = snapid;
+
+ ret = rbd_req_sync_op(dev, NULL,
+ CEPH_NOSNAP,
+ 0,
+ CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+ ops,
+ 1, obj, 0, 0, NULL);
+
+ rbd_destroy_ops(ops);
+
+ if (ret < 0)
+ return ret;
+
+ return ret;
+}
+
+/*
+ * Request sync osd read
+ */
+static int rbd_req_sync_exec(struct rbd_device *dev,
+ const char *obj,
+ const char *cls,
+ const char *method,
+ const char *data,
+ int len)
+{
+ struct ceph_osd_req_op *ops;
+ int cls_len = strlen(cls);
+ int method_len = strlen(method);
+ int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_CALL,
+ cls_len + method_len + len);
+ if (ret < 0)
+ return ret;
+
+ ops[0].cls.class_name = cls;
+ ops[0].cls.class_len = (__u8)cls_len;
+ ops[0].cls.method_name = method;
+ ops[0].cls.method_len = (__u8)method_len;
+ ops[0].cls.argc = 0;
+ ops[0].cls.indata = data;
+ ops[0].cls.indata_len = len;
+
+ ret = rbd_req_sync_op(dev, NULL,
+ CEPH_NOSNAP,
+ 0,
+ CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK,
+ ops,
+ 1, obj, 0, 0, NULL);
+
+ rbd_destroy_ops(ops);
+
+ dout("cls_exec returned %d\n", ret);
+ return ret;
+}
+
+/*
+ * block device queue callback
+ */
+static void rbd_rq_fn(struct request_queue *q)
+{
+ struct rbd_device *rbd_dev = q->queuedata;
+ struct request *rq;
+ struct bio_pair *bp = NULL;
+
+ rq = blk_fetch_request(q);
+
+ while (1) {
+ struct bio *bio;
+ struct bio *rq_bio, *next_bio = NULL;
+ bool do_write;
+ int size, op_size = 0;
+ u64 ofs;
+
+ /* peek at request from block layer */
+ if (!rq)
+ break;
+
+ dout("fetched request\n");
+
+ /* filter out block requests we don't understand */
+ if ((rq->cmd_type != REQ_TYPE_FS)) {
+ __blk_end_request_all(rq, 0);
+ goto next;
+ }
+
+ /* deduce our operation (read, write) */
+ do_write = (rq_data_dir(rq) == WRITE);
+
+ size = blk_rq_bytes(rq);
+ ofs = blk_rq_pos(rq) * 512ULL;
+ rq_bio = rq->bio;
+ if (do_write && rbd_dev->read_only) {
+ __blk_end_request_all(rq, -EROFS);
+ goto next;
+ }
+
+ spin_unlock_irq(q->queue_lock);
+
+ dout("%s 0x%x bytes at 0x%llx\n",
+ do_write ? "write" : "read",
+ size, blk_rq_pos(rq) * 512ULL);
+
+ do {
+ /* a bio clone to be passed down to OSD req */
+ dout("rq->bio->bi_vcnt=%d\n", rq->bio->bi_vcnt);
+ op_size = rbd_get_segment(&rbd_dev->header,
+ rbd_dev->header.block_name,
+ ofs, size,
+ NULL, NULL);
+ bio = bio_chain_clone(&rq_bio, &next_bio, &bp,
+ op_size, GFP_ATOMIC);
+ if (!bio) {
+ spin_lock_irq(q->queue_lock);
+ __blk_end_request_all(rq, -ENOMEM);
+ goto next;
+ }
+
+ /* init OSD command: write or read */
+ if (do_write)
+ rbd_req_write(rq, rbd_dev,
+ rbd_dev->header.snapc,
+ ofs,
+ op_size, bio);
+ else
+ rbd_req_read(rq, rbd_dev,
+ cur_snap_id(rbd_dev),
+ ofs,
+ op_size, bio);
+
+ size -= op_size;
+ ofs += op_size;
+
+ rq_bio = next_bio;
+ } while (size > 0);
+
+ if (bp)
+ bio_pair_release(bp);
+
+ spin_lock_irq(q->queue_lock);
+next:
+ rq = blk_fetch_request(q);
+ }
+}
+
+/*
+ * a queue callback. Makes sure that we don't create a bio that spans across
+ * multiple osd objects. One exception would be with a single page bios,
+ * which we handle later at bio_chain_clone
+ */
+static int rbd_merge_bvec(struct request_queue *q, struct bvec_merge_data *bmd,
+ struct bio_vec *bvec)
+{
+ struct rbd_device *rbd_dev = q->queuedata;
+ unsigned int chunk_sectors = 1 << (rbd_dev->header.obj_order - 9);
+ sector_t sector = bmd->bi_sector + get_start_sect(bmd->bi_bdev);
+ unsigned int bio_sectors = bmd->bi_size >> 9;
+ int max;
+
+ max = (chunk_sectors - ((sector & (chunk_sectors - 1))
+ + bio_sectors)) << 9;
+ if (max < 0)
+ max = 0; /* bio_add cannot handle a negative return */
+ if (max <= bvec->bv_len && bio_sectors == 0)
+ return bvec->bv_len;
+ return max;
+}
+
+static void rbd_free_disk(struct rbd_device *rbd_dev)
+{
+ struct gendisk *disk = rbd_dev->disk;
+
+ if (!disk)
+ return;
+
+ rbd_header_free(&rbd_dev->header);
+
+ if (disk->flags & GENHD_FL_UP)
+ del_gendisk(disk);
+ if (disk->queue)
+ blk_cleanup_queue(disk->queue);
+ put_disk(disk);
+}
+
+/*
+ * reload the ondisk the header
+ */
+static int rbd_read_header(struct rbd_device *rbd_dev,
+ struct rbd_image_header *header)
+{
+ ssize_t rc;
+ struct rbd_image_header_ondisk *dh;
+ int snap_count = 0;
+ u64 snap_names_len = 0;
+
+ while (1) {
+ int len = sizeof(*dh) +
+ snap_count * sizeof(struct rbd_image_snap_ondisk) +
+ snap_names_len;
+
+ rc = -ENOMEM;
+ dh = kmalloc(len, GFP_KERNEL);
+ if (!dh)
+ return -ENOMEM;
+
+ rc = rbd_req_sync_read(rbd_dev,
+ NULL, CEPH_NOSNAP,
+ rbd_dev->obj_md_name,
+ 0, len,
+ (char *)dh);
+ if (rc < 0)
+ goto out_dh;
+
+ rc = rbd_header_from_disk(header, dh, snap_count, GFP_KERNEL);
+ if (rc < 0)
+ goto out_dh;
+
+ if (snap_count != header->total_snaps) {
+ snap_count = header->total_snaps;
+ snap_names_len = header->snap_names_len;
+ rbd_header_free(header);
+ kfree(dh);
+ continue;
+ }
+ break;
+ }
+
+out_dh:
+ kfree(dh);
+ return rc;
+}
+
+/*
+ * create a snapshot
+ */
+static int rbd_header_add_snap(struct rbd_device *dev,
+ const char *snap_name,
+ gfp_t gfp_flags)
+{
+ int name_len = strlen(snap_name);
+ u64 new_snapid;
+ int ret;
+ void *data, *data_start, *data_end;
+
+ /* we should create a snapshot only if we're pointing at the head */
+ if (dev->cur_snap)
+ return -EINVAL;
+
+ ret = ceph_monc_create_snapid(&dev->client->monc, dev->poolid,
+ &new_snapid);
+ dout("created snapid=%lld\n", new_snapid);
+ if (ret < 0)
+ return ret;
+
+ data = kmalloc(name_len + 16, gfp_flags);
+ if (!data)
+ return -ENOMEM;
+
+ data_start = data;
+ data_end = data + name_len + 16;
+
+ ceph_encode_string_safe(&data, data_end, snap_name, name_len, bad);
+ ceph_encode_64_safe(&data, data_end, new_snapid, bad);
+
+ ret = rbd_req_sync_exec(dev, dev->obj_md_name, "rbd", "snap_add",
+ data_start, data - data_start);
+
+ kfree(data_start);
+
+ if (ret < 0)
+ return ret;
+
+ dev->header.snapc->seq = new_snapid;
+
+ return 0;
+bad:
+ return -ERANGE;
+}
+
+/*
+ * only read the first part of the ondisk header, without the snaps info
+ */
+static int rbd_update_snaps(struct rbd_device *rbd_dev)
+{
+ int ret;
+ struct rbd_image_header h;
+ u64 snap_seq;
+
+ ret = rbd_read_header(rbd_dev, &h);
+ if (ret < 0)
+ return ret;
+
+ down_write(&rbd_dev->header.snap_rwsem);
+
+ snap_seq = rbd_dev->header.snapc->seq;
+
+ kfree(rbd_dev->header.snapc);
+ kfree(rbd_dev->header.snap_names);
+ kfree(rbd_dev->header.snap_sizes);
+
+ rbd_dev->header.total_snaps = h.total_snaps;
+ rbd_dev->header.snapc = h.snapc;
+ rbd_dev->header.snap_names = h.snap_names;
+ rbd_dev->header.snap_sizes = h.snap_sizes;
+ rbd_dev->header.snapc->seq = snap_seq;
+
+ up_write(&rbd_dev->header.snap_rwsem);
+
+ return 0;
+}
+
+static int rbd_init_disk(struct rbd_device *rbd_dev)
+{
+ struct gendisk *disk;
+ struct request_queue *q;
+ int rc;
+ u64 total_size = 0;
+
+ /* contact OSD, request size info about the object being mapped */
+ rc = rbd_read_header(rbd_dev, &rbd_dev->header);
+ if (rc)
+ return rc;
+
+ rc = rbd_header_set_snap(rbd_dev, rbd_dev->snap_name, &total_size);
+ if (rc)
+ return rc;
+
+ /* create gendisk info */
+ rc = -ENOMEM;
+ disk = alloc_disk(RBD_MINORS_PER_MAJOR);
+ if (!disk)
+ goto out;
+
+ sprintf(disk->disk_name, DRV_NAME "%d", rbd_dev->id);
+ disk->major = rbd_dev->major;
+ disk->first_minor = 0;
+ disk->fops = &rbd_bd_ops;
+ disk->private_data = rbd_dev;
+
+ /* init rq */
+ rc = -ENOMEM;
+ q = blk_init_queue(rbd_rq_fn, &rbd_dev->lock);
+ if (!q)
+ goto out_disk;
+ blk_queue_merge_bvec(q, rbd_merge_bvec);
+ disk->queue = q;
+
+ q->queuedata = rbd_dev;
+
+ rbd_dev->disk = disk;
+ rbd_dev->q = q;
+
+ /* finally, announce the disk to the world */
+ set_capacity(disk, total_size / 512ULL);
+ add_disk(disk);
+
+ pr_info("%s: added with size 0x%llx\n",
+ disk->disk_name, (unsigned long long)total_size);
+ return 0;
+
+out_disk:
+ put_disk(disk);
+out:
+ return rc;
+}
+
+/********************************************************************
+ * /sys/class/rbd/
+ * add map rados objects to blkdev
+ * remove unmap rados objects
+ * list show mappings
+ *******************************************************************/
+
+static void class_rbd_release(struct class *cls)
+{
+ kfree(cls);
+}
+
+static ssize_t class_rbd_list(struct class *c,
+ struct class_attribute *attr,
+ char *data)
+{
+ int n = 0;
+ struct list_head *tmp;
+ int max = PAGE_SIZE;
+
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ n += snprintf(data, max,
+ "#id\tmajor\tclient_name\tpool\tname\tsnap\tKB\n");
+
+ list_for_each(tmp, &rbd_dev_list) {
+ struct rbd_device *rbd_dev;
+
+ rbd_dev = list_entry(tmp, struct rbd_device, node);
+ n += snprintf(data+n, max-n,
+ "%d\t%d\tclient%lld\t%s\t%s\t%s\t%lld\n",
+ rbd_dev->id,
+ rbd_dev->major,
+ ceph_client_id(rbd_dev->client),
+ rbd_dev->pool_name,
+ rbd_dev->obj, rbd_dev->snap_name,
+ rbd_dev->header.image_size >> 10);
+ if (n == max)
+ break;
+ }
+
+ mutex_unlock(&ctl_mutex);
+ return n;
+}
+
+static ssize_t class_rbd_add(struct class *c,
+ struct class_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct ceph_osd_client *osdc;
+ struct rbd_device *rbd_dev;
+ ssize_t rc = -ENOMEM;
+ int irc, new_id = 0;
+ struct list_head *tmp;
+ char *mon_dev_name;
+ char *options;
+
+ if (!try_module_get(THIS_MODULE))
+ return -ENODEV;
+
+ mon_dev_name = kmalloc(RBD_MAX_OPT_LEN, GFP_KERNEL);
+ if (!mon_dev_name)
+ goto err_out_mod;
+
+ options = kmalloc(RBD_MAX_OPT_LEN, GFP_KERNEL);
+ if (!options)
+ goto err_mon_dev;
+
+ /* new rbd_device object */
+ rbd_dev = kzalloc(sizeof(*rbd_dev), GFP_KERNEL);
+ if (!rbd_dev)
+ goto err_out_opt;
+
+ /* static rbd_device initialization */
+ spin_lock_init(&rbd_dev->lock);
+ INIT_LIST_HEAD(&rbd_dev->node);
+
+ /* generate unique id: find highest unique id, add one */
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ list_for_each(tmp, &rbd_dev_list) {
+ struct rbd_device *rbd_dev;
+
+ rbd_dev = list_entry(tmp, struct rbd_device, node);
+ if (rbd_dev->id >= new_id)
+ new_id = rbd_dev->id + 1;
+ }
+
+ rbd_dev->id = new_id;
+
+ /* add to global list */
+ list_add_tail(&rbd_dev->node, &rbd_dev_list);
+
+ /* parse add command */
+ if (sscanf(buf, "%" __stringify(RBD_MAX_OPT_LEN) "s "
+ "%" __stringify(RBD_MAX_OPT_LEN) "s "
+ "%" __stringify(RBD_MAX_POOL_NAME_LEN) "s "
+ "%" __stringify(RBD_MAX_OBJ_NAME_LEN) "s"
+ "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+ mon_dev_name, options, rbd_dev->pool_name,
+ rbd_dev->obj, rbd_dev->snap_name) < 4) {
+ rc = -EINVAL;
+ goto err_out_slot;
+ }
+
+ if (rbd_dev->snap_name[0] == 0)
+ rbd_dev->snap_name[0] = '-';
+
+ rbd_dev->obj_len = strlen(rbd_dev->obj);
+ snprintf(rbd_dev->obj_md_name, sizeof(rbd_dev->obj_md_name), "%s%s",
+ rbd_dev->obj, RBD_SUFFIX);
+
+ /* initialize rest of new object */
+ snprintf(rbd_dev->name, DEV_NAME_LEN, DRV_NAME "%d", rbd_dev->id);
+ rc = rbd_get_client(rbd_dev, mon_dev_name, options);
+ if (rc < 0)
+ goto err_out_slot;
+
+ mutex_unlock(&ctl_mutex);
+
+ /* pick the pool */
+ osdc = &rbd_dev->client->osdc;
+ rc = ceph_pg_poolid_by_name(osdc->osdmap, rbd_dev->pool_name);
+ if (rc < 0)
+ goto err_out_client;
+ rbd_dev->poolid = rc;
+
+ /* register our block device */
+ irc = register_blkdev(0, rbd_dev->name);
+ if (irc < 0) {
+ rc = irc;
+ goto err_out_client;
+ }
+ rbd_dev->major = irc;
+
+ /* set up and announce blkdev mapping */
+ rc = rbd_init_disk(rbd_dev);
+ if (rc)
+ goto err_out_blkdev;
+
+ return count;
+
+err_out_blkdev:
+ unregister_blkdev(rbd_dev->major, rbd_dev->name);
+err_out_client:
+ rbd_put_client(rbd_dev);
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+err_out_slot:
+ list_del_init(&rbd_dev->node);
+ mutex_unlock(&ctl_mutex);
+
+ kfree(rbd_dev);
+err_out_opt:
+ kfree(options);
+err_mon_dev:
+ kfree(mon_dev_name);
+err_out_mod:
+ dout("Error adding device %s\n", buf);
+ module_put(THIS_MODULE);
+ return rc;
+}
+
+static struct rbd_device *__rbd_get_dev(unsigned long id)
+{
+ struct list_head *tmp;
+ struct rbd_device *rbd_dev;
+
+ list_for_each(tmp, &rbd_dev_list) {
+ rbd_dev = list_entry(tmp, struct rbd_device, node);
+ if (rbd_dev->id == id)
+ return rbd_dev;
+ }
+ return NULL;
+}
+
+static ssize_t class_rbd_remove(struct class *c,
+ struct class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rbd_device *rbd_dev = NULL;
+ int target_id, rc;
+ unsigned long ul;
+
+ rc = strict_strtoul(buf, 10, &ul);
+ if (rc)
+ return rc;
+
+ /* convert to int; abort if we lost anything in the conversion */
+ target_id = (int) ul;
+ if (target_id != ul)
+ return -EINVAL;
+
+ /* remove object from list immediately */
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ rbd_dev = __rbd_get_dev(target_id);
+ if (rbd_dev)
+ list_del_init(&rbd_dev->node);
+
+ mutex_unlock(&ctl_mutex);
+
+ if (!rbd_dev)
+ return -ENOENT;
+
+ rbd_put_client(rbd_dev);
+
+ /* clean up and free blkdev */
+ rbd_free_disk(rbd_dev);
+ unregister_blkdev(rbd_dev->major, rbd_dev->name);
+ kfree(rbd_dev);
+
+ /* release module ref */
+ module_put(THIS_MODULE);
+
+ return count;
+}
+
+static ssize_t class_rbd_snaps_list(struct class *c,
+ struct class_attribute *attr,
+ char *data)
+{
+ struct rbd_device *rbd_dev = NULL;
+ struct list_head *tmp;
+ struct rbd_image_header *header;
+ int i, n = 0, max = PAGE_SIZE;
+ int ret;
+
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ n += snprintf(data, max, "#id\tsnap\tKB\n");
+
+ list_for_each(tmp, &rbd_dev_list) {
+ char *names, *p;
+ struct ceph_snap_context *snapc;
+
+ rbd_dev = list_entry(tmp, struct rbd_device, node);
+ header = &rbd_dev->header;
+
+ down_read(&header->snap_rwsem);
+
+ names = header->snap_names;
+ snapc = header->snapc;
+
+ n += snprintf(data + n, max - n, "%d\t%s\t%lld%s\n",
+ rbd_dev->id, RBD_SNAP_HEAD_NAME,
+ header->image_size >> 10,
+ (!rbd_dev->cur_snap ? " (*)" : ""));
+ if (n == max)
+ break;
+
+ p = names;
+ for (i = 0; i < header->total_snaps; i++, p += strlen(p) + 1) {
+ n += snprintf(data + n, max - n, "%d\t%s\t%lld%s\n",
+ rbd_dev->id, p, header->snap_sizes[i] >> 10,
+ (rbd_dev->cur_snap &&
+ (snap_index(header, i) == rbd_dev->cur_snap) ?
+ " (*)" : ""));
+ if (n == max)
+ break;
+ }
+
+ up_read(&header->snap_rwsem);
+ }
+
+
+ ret = n;
+ mutex_unlock(&ctl_mutex);
+ return ret;
+}
+
+static ssize_t class_rbd_snaps_refresh(struct class *c,
+ struct class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rbd_device *rbd_dev = NULL;
+ int target_id, rc;
+ unsigned long ul;
+ int ret = count;
+
+ rc = strict_strtoul(buf, 10, &ul);
+ if (rc)
+ return rc;
+
+ /* convert to int; abort if we lost anything in the conversion */
+ target_id = (int) ul;
+ if (target_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;
+ }
+
+ rc = rbd_update_snaps(rbd_dev);
+ if (rc < 0)
+ ret = rc;
+
+done:
+ mutex_unlock(&ctl_mutex);
+ return ret;
+}
+
+static ssize_t class_rbd_snap_create(struct class *c,
+ struct class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rbd_device *rbd_dev = NULL;
+ int target_id, ret;
+ char *name;
+
+ name = kmalloc(RBD_MAX_SNAP_NAME_LEN + 1, GFP_KERNEL);
+ if (!name)
+ return -ENOMEM;
+
+ /* parse snaps add command */
+ if (sscanf(buf, "%d "
+ "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+ &target_id,
+ name) != 2) {
+ ret = -EINVAL;
+ goto done;
+ }
+
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ rbd_dev = __rbd_get_dev(target_id);
+ if (!rbd_dev) {
+ ret = -ENOENT;
+ goto done_unlock;
+ }
+
+ ret = rbd_header_add_snap(rbd_dev,
+ name, GFP_KERNEL);
+ if (ret < 0)
+ goto done_unlock;
+
+ ret = rbd_update_snaps(rbd_dev);
+ if (ret < 0)
+ goto done_unlock;
+
+ ret = count;
+done_unlock:
+ mutex_unlock(&ctl_mutex);
+done:
+ kfree(name);
+ return ret;
+}
+
+static ssize_t class_rbd_rollback(struct class *c,
+ struct class_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct rbd_device *rbd_dev = NULL;
+ int target_id, ret;
+ u64 snapid;
+ char snap_name[RBD_MAX_SNAP_NAME_LEN];
+ u64 cur_ofs;
+ char *seg_name;
+
+ /* parse snaps add command */
+ if (sscanf(buf, "%d "
+ "%" __stringify(RBD_MAX_SNAP_NAME_LEN) "s",
+ &target_id,
+ snap_name) != 2) {
+ return -EINVAL;
+ }
+
+ ret = -ENOMEM;
+ seg_name = kmalloc(RBD_MAX_SEG_NAME_LEN + 1, GFP_NOIO);
+ if (!seg_name)
+ return ret;
+
+ mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+ rbd_dev = __rbd_get_dev(target_id);
+ if (!rbd_dev) {
+ ret = -ENOENT;
+ goto done_unlock;
+ }
+
+ ret = snap_by_name(&rbd_dev->header, snap_name, &snapid, NULL);
+ if (ret < 0)
+ goto done_unlock;
+
+ dout("snapid=%lld\n", snapid);
+
+ cur_ofs = 0;
+ while (cur_ofs < rbd_dev->header.image_size) {
+ cur_ofs += rbd_get_segment(&rbd_dev->header,
+ rbd_dev->obj,
+ cur_ofs, (u64)-1,
+ seg_name, NULL);
+ dout("seg_name=%s\n", seg_name);
+
+ ret = rbd_req_sync_rollback_obj(rbd_dev, snapid, seg_name);
+ if (ret < 0)
+ pr_warning("could not roll back obj %s err=%d\n",
+ seg_name, ret);
+ }
+
+ ret = rbd_update_snaps(rbd_dev);
+ if (ret < 0)
+ goto done_unlock;
+
+ ret = count;
+
+done_unlock:
+ mutex_unlock(&ctl_mutex);
+ kfree(seg_name);
+
+ return ret;
+}
+
+static struct class_attribute class_rbd_attrs[] = {
+ __ATTR(add, 0200, NULL, class_rbd_add),
+ __ATTR(remove, 0200, NULL, class_rbd_remove),
+ __ATTR(list, 0444, class_rbd_list, NULL),
+ __ATTR(snaps_refresh, 0200, NULL, class_rbd_snaps_refresh),
+ __ATTR(snap_create, 0200, NULL, class_rbd_snap_create),
+ __ATTR(snaps_list, 0444, class_rbd_snaps_list, NULL),
+ __ATTR(snap_rollback, 0200, NULL, class_rbd_rollback),
+ __ATTR_NULL
+};
+
+/*
+ * create control files in sysfs
+ * /sys/class/rbd/...
+ */
+static int rbd_sysfs_init(void)
+{
+ int ret = -ENOMEM;
+
+ class_rbd = kzalloc(sizeof(*class_rbd), GFP_KERNEL);
+ if (!class_rbd)
+ goto out;
+
+ class_rbd->name = DRV_NAME;
+ class_rbd->owner = THIS_MODULE;
+ class_rbd->class_release = class_rbd_release;
+ class_rbd->class_attrs = class_rbd_attrs;
+
+ ret = class_register(class_rbd);
+ if (ret)
+ goto out_class;
+ return 0;
+
+out_class:
+ kfree(class_rbd);
+ class_rbd = NULL;
+ pr_err(DRV_NAME ": failed to create class rbd\n");
+out:
+ return ret;
+}
+
+static void rbd_sysfs_cleanup(void)
+{
+ if (class_rbd)
+ class_destroy(class_rbd);
+ class_rbd = NULL;
+}
+
+int __init rbd_init(void)
+{
+ int rc;
+
+ rc = rbd_sysfs_init();
+ if (rc)
+ return rc;
+ spin_lock_init(&node_lock);
+ pr_info("loaded " DRV_NAME_LONG "\n");
+ return 0;
+}
+
+void __exit rbd_exit(void)
+{
+ rbd_sysfs_cleanup();
+}
+
+module_init(rbd_init);
+module_exit(rbd_exit);
+
+MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
+MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
+MODULE_DESCRIPTION("rados block device");
+
+/* following authorship retained from original osdblk.c */
+MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>");
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/block/rbd_types.h b/drivers/block/rbd_types.h
new file mode 100644
index 000000000000..fc6c678aa2cb
--- /dev/null
+++ b/drivers/block/rbd_types.h
@@ -0,0 +1,73 @@
+/*
+ * Ceph - scalable distributed file system
+ *
+ * Copyright (C) 2004-2010 Sage Weil <sage@newdream.net>
+ *
+ * This is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License version 2.1, as published by the Free Software
+ * Foundation. See file COPYING.
+ *
+ */
+
+#ifndef CEPH_RBD_TYPES_H
+#define CEPH_RBD_TYPES_H
+
+#include <linux/types.h>
+
+/*
+ * rbd image 'foo' consists of objects
+ * foo.rbd - image metadata
+ * foo.00000000
+ * foo.00000001
+ * ... - data
+ */
+
+#define RBD_SUFFIX ".rbd"
+#define RBD_DIRECTORY "rbd_directory"
+#define RBD_INFO "rbd_info"
+
+#define RBD_DEFAULT_OBJ_ORDER 22 /* 4MB */
+#define RBD_MIN_OBJ_ORDER 16
+#define RBD_MAX_OBJ_ORDER 30
+
+#define RBD_MAX_OBJ_NAME_LEN 96
+#define RBD_MAX_SEG_NAME_LEN 128
+
+#define RBD_COMP_NONE 0
+#define RBD_CRYPT_NONE 0
+
+#define RBD_HEADER_TEXT "<<< Rados Block Device Image >>>\n"
+#define RBD_HEADER_SIGNATURE "RBD"
+#define RBD_HEADER_VERSION "001.005"
+
+struct rbd_info {
+ __le64 max_id;
+} __attribute__ ((packed));
+
+struct rbd_image_snap_ondisk {
+ __le64 id;
+ __le64 image_size;
+} __attribute__((packed));
+
+struct rbd_image_header_ondisk {
+ char text[40];
+ char block_name[24];
+ char signature[4];
+ char version[8];
+ struct {
+ __u8 order;
+ __u8 crypt_type;
+ __u8 comp_type;
+ __u8 unused;
+ } __attribute__((packed)) options;
+ __le64 image_size;
+ __le64 snap_seq;
+ __le32 snap_count;
+ __le32 reserved;
+ __le64 snap_names_len;
+ struct rbd_image_snap_ondisk snaps[0];
+} __attribute__((packed));
+
+
+#endif
diff --git a/drivers/block/swim.c b/drivers/block/swim.c
index 2e46815876df..75333d0a3327 100644
--- a/drivers/block/swim.c
+++ b/drivers/block/swim.c
@@ -20,7 +20,7 @@
#include <linux/fd.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/hdreg.h>
#include <linux/kernel.h>
#include <linux/delay.h>
@@ -222,6 +222,7 @@ extern int swim_read_sector_header(struct swim __iomem *base,
extern int swim_read_sector_data(struct swim __iomem *base,
unsigned char *data);
+static DEFINE_MUTEX(swim_mutex);
static inline void set_swim_mode(struct swim __iomem *base, int enable)
{
struct iwm __iomem *iwm_base;
@@ -666,9 +667,9 @@ static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&swim_mutex);
ret = floppy_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&swim_mutex);
return ret;
}
@@ -678,7 +679,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
struct floppy_state *fs = disk->private_data;
struct swim __iomem *base = fs->swd->base;
- lock_kernel();
+ mutex_lock(&swim_mutex);
if (fs->ref_count < 0)
fs->ref_count = 0;
else if (fs->ref_count > 0)
@@ -686,7 +687,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
if (fs->ref_count == 0)
swim_motor(base, OFF);
- unlock_kernel();
+ mutex_unlock(&swim_mutex);
return 0;
}
@@ -704,9 +705,9 @@ static int floppy_ioctl(struct block_device *bdev, fmode_t mode,
case FDEJECT:
if (fs->ref_count != 1)
return -EBUSY;
- lock_kernel();
+ mutex_lock(&swim_mutex);
err = floppy_eject(fs);
- unlock_kernel();
+ mutex_unlock(&swim_mutex);
return err;
case FDGETPRM:
diff --git a/drivers/block/swim3.c b/drivers/block/swim3.c
index cc6a3864822c..bf3a5b859299 100644
--- a/drivers/block/swim3.c
+++ b/drivers/block/swim3.c
@@ -25,7 +25,7 @@
#include <linux/ioctl.h>
#include <linux/blkdev.h>
#include <linux/interrupt.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/module.h>
#include <linux/spinlock.h>
#include <asm/io.h>
@@ -36,6 +36,7 @@
#include <asm/machdep.h>
#include <asm/pmac_feature.h>
+static DEFINE_MUTEX(swim3_mutex);
static struct request_queue *swim3_queue;
static struct gendisk *disks[2];
static struct request *fd_req;
@@ -873,9 +874,9 @@ static int floppy_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&swim3_mutex);
ret = floppy_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
+ mutex_unlock(&swim3_mutex);
return ret;
}
@@ -953,9 +954,9 @@ static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&swim3_mutex);
ret = floppy_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&swim3_mutex);
return ret;
}
@@ -964,13 +965,13 @@ static int floppy_release(struct gendisk *disk, fmode_t mode)
{
struct floppy_state *fs = disk->private_data;
struct swim3 __iomem *sw = fs->swim3;
- lock_kernel();
+ mutex_lock(&swim3_mutex);
if (fs->ref_count > 0 && --fs->ref_count == 0) {
swim3_action(fs, MOTOR_OFF);
out_8(&sw->control_bic, 0xff);
swim3_select(fs, RELAX);
}
- unlock_kernel();
+ mutex_unlock(&swim3_mutex);
return 0;
}
diff --git a/drivers/block/ub.c b/drivers/block/ub.c
index c48e14878582..b5690a045a01 100644
--- a/drivers/block/ub.c
+++ b/drivers/block/ub.c
@@ -28,7 +28,7 @@
#include <linux/timer.h>
#include <linux/scatterlist.h>
#include <linux/slab.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <scsi/scsi.h>
#define DRV_NAME "ub"
@@ -248,6 +248,7 @@ struct ub_completion {
spinlock_t lock;
};
+static DEFINE_MUTEX(ub_mutex);
static inline void ub_init_completion(struct ub_completion *x)
{
x->done = 0;
@@ -1715,9 +1716,9 @@ static int ub_bd_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&ub_mutex);
ret = ub_bd_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&ub_mutex);
return ret;
}
@@ -1730,9 +1731,9 @@ static int ub_bd_release(struct gendisk *disk, fmode_t mode)
struct ub_lun *lun = disk->private_data;
struct ub_dev *sc = lun->udev;
- lock_kernel();
+ mutex_lock(&ub_mutex);
ub_put(sc);
- unlock_kernel();
+ mutex_unlock(&ub_mutex);
return 0;
}
@@ -1747,9 +1748,9 @@ static int ub_bd_ioctl(struct block_device *bdev, fmode_t mode,
void __user *usermem = (void __user *) arg;
int ret;
- lock_kernel();
+ mutex_lock(&ub_mutex);
ret = scsi_cmd_ioctl(disk->queue, disk, mode, cmd, usermem);
- unlock_kernel();
+ mutex_unlock(&ub_mutex);
return ret;
}
diff --git a/drivers/block/viodasd.c b/drivers/block/viodasd.c
index f651e51a3319..e2ff697697c2 100644
--- a/drivers/block/viodasd.c
+++ b/drivers/block/viodasd.c
@@ -41,7 +41,7 @@
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/string.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/dma-mapping.h>
#include <linux/completion.h>
#include <linux/device.h>
@@ -73,6 +73,7 @@ enum {
MAX_DISK_NAME = FIELD_SIZEOF(struct gendisk, disk_name)
};
+static DEFINE_MUTEX(viodasd_mutex);
static DEFINE_SPINLOCK(viodasd_spinlock);
#define VIOMAXREQ 16
@@ -180,9 +181,9 @@ static int viodasd_unlocked_open(struct block_device *bdev, fmode_t mode)
{
int ret;
- lock_kernel();
+ mutex_lock(&viodasd_mutex);
ret = viodasd_open(bdev, mode);
- unlock_kernel();
+ mutex_unlock(&viodasd_mutex);
return ret;
}
@@ -196,7 +197,7 @@ static int viodasd_release(struct gendisk *disk, fmode_t mode)
struct viodasd_device *d = disk->private_data;
HvLpEvent_Rc hvrc;
- lock_kernel();
+ mutex_lock(&viodasd_mutex);
/* Send the event to OS/400. We DON'T expect a response */
hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
HvLpEvent_Type_VirtualIo,
@@ -210,7 +211,7 @@ static int viodasd_release(struct gendisk *disk, fmode_t mode)
if (hvrc != 0)
pr_warning("HV close call failed %d\n", (int)hvrc);
- unlock_kernel();
+ mutex_unlock(&viodasd_mutex);
return 0;
}
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 2aafafca2b13..8320490226b7 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -2,7 +2,6 @@
#include <linux/spinlock.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
#include <linux/hdreg.h>
#include <linux/virtio.h>
#include <linux/virtio_blk.h>
@@ -202,6 +201,7 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str)
struct virtio_blk *vblk = disk->private_data;
struct request *req;
struct bio *bio;
+ int err;
bio = bio_map_kern(vblk->disk->queue, id_str, VIRTIO_BLK_ID_BYTES,
GFP_KERNEL);
@@ -215,11 +215,14 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str)
}
req->cmd_type = REQ_TYPE_SPECIAL;
- return blk_execute_rq(vblk->disk->queue, vblk->disk, req, false);
+ err = blk_execute_rq(vblk->disk->queue, vblk->disk, req, false);
+ blk_put_request(req);
+
+ return err;
}
-static int virtblk_locked_ioctl(struct block_device *bdev, fmode_t mode,
- unsigned cmd, unsigned long data)
+static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
+ unsigned int cmd, unsigned long data)
{
struct gendisk *disk = bdev->bd_disk;
struct virtio_blk *vblk = disk->private_data;
@@ -234,18 +237,6 @@ static int virtblk_locked_ioctl(struct block_device *bdev, fmode_t mode,
(void __user *)data);
}
-static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
- unsigned int cmd, unsigned long param)
-{
- int ret;
-
- lock_kernel();
- ret = virtblk_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
-
- return ret;
-}
-
/* We provide getgeo only to please some old bootloader/partitioning tools */
static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo)
{
diff --git a/drivers/block/xd.c b/drivers/block/xd.c
index d5a3cd750561..4abd2bcd20fb 100644
--- a/drivers/block/xd.c
+++ b/drivers/block/xd.c
@@ -46,7 +46,7 @@
#include <linux/init.h>
#include <linux/wait.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/blkpg.h>
#include <linux/delay.h>
#include <linux/io.h>
@@ -58,6 +58,7 @@
#include "xd.h"
+static DEFINE_MUTEX(xd_mutex);
static void __init do_xd_setup (int *integers);
#ifdef MODULE
static int xd[5] = { -1,-1,-1,-1, };
@@ -381,9 +382,9 @@ static int xd_ioctl(struct block_device *bdev, fmode_t mode,
{
int ret;
- lock_kernel();
+ mutex_lock(&xd_mutex);
ret = xd_locked_ioctl(bdev, mode, cmd, param);
- unlock_kernel();
+ mutex_unlock(&xd_mutex);
return ret;
}
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
index ac1b682edecb..3ff06f475eef 100644
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -41,7 +41,7 @@
#include <linux/cdrom.h>
#include <linux/module.h>
#include <linux/slab.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/scatterlist.h>
#include <xen/xen.h>
@@ -69,6 +69,7 @@ struct blk_shadow {
unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST];
};
+static DEFINE_MUTEX(blkfront_mutex);
static const struct block_device_operations xlvbd_block_fops;
#define BLK_RING_SIZE __RING_SIZE((struct blkif_sring *)0, PAGE_SIZE)
@@ -834,7 +835,7 @@ static int blkfront_probe(struct xenbus_device *dev,
char *type;
int len;
/* no unplug has been done: do not hook devices != xen vbds */
- if (xen_platform_pci_unplug & XEN_UNPLUG_IGNORE) {
+ if (xen_platform_pci_unplug & XEN_UNPLUG_UNNECESSARY) {
int major;
if (!VDEV_IS_EXTENDED(vdevice))
@@ -1201,7 +1202,7 @@ static int blkif_open(struct block_device *bdev, fmode_t mode)
struct blkfront_info *info;
int err = 0;
- lock_kernel();
+ mutex_lock(&blkfront_mutex);
info = disk->private_data;
if (!info) {
@@ -1219,7 +1220,7 @@ static int blkif_open(struct block_device *bdev, fmode_t mode)
mutex_unlock(&info->mutex);
out:
- unlock_kernel();
+ mutex_unlock(&blkfront_mutex);
return err;
}
@@ -1229,7 +1230,7 @@ static int blkif_release(struct gendisk *disk, fmode_t mode)
struct block_device *bdev;
struct xenbus_device *xbdev;
- lock_kernel();
+ mutex_lock(&blkfront_mutex);
bdev = bdget_disk(disk, 0);
bdput(bdev);
@@ -1263,7 +1264,7 @@ static int blkif_release(struct gendisk *disk, fmode_t mode)
}
out:
- unlock_kernel();
+ mutex_unlock(&blkfront_mutex);
return 0;
}
diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c
index 057413bb16e2..6e968cd4893c 100644
--- a/drivers/block/xsysace.c
+++ b/drivers/block/xsysace.c
@@ -89,7 +89,7 @@
#include <linux/delay.h>
#include <linux/slab.h>
#include <linux/blkdev.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/ata.h>
#include <linux/hdreg.h>
#include <linux/platform_device.h>
@@ -214,6 +214,7 @@ struct ace_device {
u16 cf_id[ATA_ID_WORDS];
};
+static DEFINE_MUTEX(xsysace_mutex);
static int ace_major;
/* ---------------------------------------------------------------------
@@ -903,13 +904,13 @@ static int ace_open(struct block_device *bdev, fmode_t mode)
dev_dbg(ace->dev, "ace_open() users=%i\n", ace->users + 1);
- lock_kernel();
+ mutex_lock(&xsysace_mutex);
spin_lock_irqsave(&ace->lock, flags);
ace->users++;
spin_unlock_irqrestore(&ace->lock, flags);
check_disk_change(bdev);
- unlock_kernel();
+ mutex_unlock(&xsysace_mutex);
return 0;
}
@@ -922,7 +923,7 @@ static int ace_release(struct gendisk *disk, fmode_t mode)
dev_dbg(ace->dev, "ace_release() users=%i\n", ace->users - 1);
- lock_kernel();
+ mutex_lock(&xsysace_mutex);
spin_lock_irqsave(&ace->lock, flags);
ace->users--;
if (ace->users == 0) {
@@ -930,7 +931,7 @@ static int ace_release(struct gendisk *disk, fmode_t mode)
ace_out(ace, ACE_CTRL, val & ~ACE_CTRL_LOCKREQ);
}
spin_unlock_irqrestore(&ace->lock, flags);
- unlock_kernel();
+ mutex_unlock(&xsysace_mutex);
return 0;
}
diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c
index d75b2bb601ad..dcd4cfcf4126 100644
--- a/drivers/block/z2ram.c
+++ b/drivers/block/z2ram.c
@@ -33,7 +33,7 @@
#include <linux/module.h>
#include <linux/blkdev.h>
#include <linux/bitops.h>
-#include <linux/smp_lock.h>
+#include <linux/mutex.h>
#include <linux/slab.h>
#include <asm/setup.h>
@@ -57,6 +57,7 @@ extern struct mem_info m68k_memory[NUM_MEMINFO];
#define Z2RAM_CHUNK1024 ( Z2RAM_CHUNKSIZE >> 10 )
+static DEFINE_MUTEX(z2ram_mutex);
static u_long *z2ram_map = NULL;
static u_long z2ram_size = 0;
static int z2_count = 0;
@@ -154,7 +155,7 @@ static int z2_open(struct block_device *bdev, fmode_t mode)
device = MINOR(bdev->bd_dev);
- lock_kernel();
+ mutex_lock(&z2ram_mutex);
if ( current_device != -1 && current_device != device )
{
rc = -EBUSY;
@@ -296,25 +297,25 @@ static int z2_open(struct block_device *bdev, fmode_t mode)
set_capacity(z2ram_gendisk, z2ram_size >> 9);
}
- unlock_kernel();
+ mutex_unlock(&z2ram_mutex);
return 0;
err_out_kfree:
kfree(z2ram_map);
err_out:
- unlock_kernel();
+ mutex_unlock(&z2ram_mutex);
return rc;
}
static int
z2_release(struct gendisk *disk, fmode_t mode)
{
- lock_kernel();
+ mutex_lock(&z2ram_mutex);
if ( current_device == -1 ) {
- unlock_kernel();
+ mutex_unlock(&z2ram_mutex);
return 0;
}
- unlock_kernel();
+ mutex_unlock(&z2ram_mutex);
/*
* FIXME: unmap memory
*/