exofs: Multi-device mirror support
This patch changes on-disk format, it is accompanied with a parallel
patch to mkfs.exofs that enables multi-device capabilities.
After this patch, old exofs will refuse to mount a new formatted FS and
new exofs will refuse an old format. This is done by moving the magic
field offset inside the FSCB. A new FSCB *version* field was added. In
the future, exofs will refuse to mount unmatched FSCB version. To
up-grade or down-grade an exofs one must use mkfs.exofs --upgrade option
before mounting.
Introduced, a new object that contains a *device-table*. This object
contains the default *data-map* and a linear array of devices
information, which identifies the devices used in the filesystem. This
object is only written to offline by mkfs.exofs. This is why it is kept
separate from the FSCB, since the later is written to while mounted.
Same partition number, same object number is used on all devices only
the device varies.
* define the new format, then load the device table on mount time make
sure every thing is supported.
* Change I/O engine to now support Mirror IO, .i.e write same data
to multiple devices, read from a random device to spread the
read-load from multiple clients (TODO: stripe read)
Implementation notes:
A few points introduced in previous patch should be mentioned here:
* Special care was made so absolutlly all operation that have any chance
of failing are done before any osd-request is executed. This is to
minimize the need for a data consistency recovery, to only real IO
errors.
* Each IO state has a kref. It starts at 1, any osd-request executed
will increment the kref, finally when all are executed the first ref
is dropped. At IO-done, each request completion decrements the kref,
the last one to return executes the internal _last_io() routine.
_last_io() will call the registered io_state_done. On sync mode a
caller does not supply a done method, indicating a synchronous
request, the caller is put to sleep and a special io_state_done is
registered that will awaken the caller. Though also in sync mode all
operations are executed in parallel.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
diff --git a/fs/exofs/common.h b/fs/exofs/common.h
index ce1c716..b1b178e 100644
--- a/fs/exofs/common.h
+++ b/fs/exofs/common.h
@@ -49,6 +49,7 @@
#define EXOFS_MIN_PID 0x10000 /* Smallest partition ID */
#define EXOFS_OBJ_OFF 0x10000 /* offset for objects */
#define EXOFS_SUPER_ID 0x10000 /* object ID for on-disk superblock */
+#define EXOFS_DEVTABLE_ID 0x10001 /* object ID for on-disk device table */
#define EXOFS_ROOT_ID 0x10002 /* object ID for root directory */
/* exofs Application specific page/attribute */
@@ -78,17 +79,67 @@
#define EXOFS_SUPER_MAGIC 0x5DF5
/*
- * The file system control block - stored in an object's data (mainly, the one
- * with ID EXOFS_SUPER_ID). This is where the in-memory superblock is stored
- * on disk. Right now it just has a magic value, which is basically a sanity
- * check on our ability to communicate with the object store.
+ * The file system control block - stored in object EXOFS_SUPER_ID's data.
+ * This is where the in-memory superblock is stored on disk.
*/
+enum {EXOFS_FSCB_VER = 1, EXOFS_DT_VER = 1};
struct exofs_fscb {
__le64 s_nextid; /* Highest object ID used */
- __le32 s_numfiles; /* Number of files on fs */
+ __le64 s_numfiles; /* Number of files on fs */
+ __le32 s_version; /* == EXOFS_FSCB_VER */
__le16 s_magic; /* Magic signature */
__le16 s_newfs; /* Non-zero if this is a new fs */
-};
+
+ /* From here on it's a static part, only written by mkexofs */
+ __le64 s_dev_table_oid; /* Resurved, not used */
+ __le64 s_dev_table_count; /* == 0 means no dev_table */
+} __packed;
+
+/*
+ * Describes the raid used in the FS. It is part of the device table.
+ * This here is taken from the pNFS-objects definition. In exofs we
+ * use one raid policy through-out the filesystem. (NOTE: the funny
+ * alignment at begining. We take care of it at exofs_device_table.
+ */
+struct exofs_dt_data_map {
+ __le32 cb_num_comps;
+ __le64 cb_stripe_unit;
+ __le32 cb_group_width;
+ __le32 cb_group_depth;
+ __le32 cb_mirror_cnt;
+ __le32 cb_raid_algorithm;
+} __packed;
+
+/*
+ * This is an osd device information descriptor. It is a single entry in
+ * the exofs device table. It describes an osd target lun which
+ * contains data belonging to this FS. (Same partition_id on all devices)
+ */
+struct exofs_dt_device_info {
+ __le32 systemid_len;
+ u8 systemid[OSD_SYSTEMID_LEN];
+ __le64 long_name_offset; /* If !0 then offset-in-file */
+ __le32 osdname_len; /* */
+ u8 osdname[44]; /* Embbeded, Ususally an asci uuid */
+} __packed;
+
+/*
+ * The EXOFS device table - stored in object EXOFS_DEVTABLE_ID's data.
+ * It contains the raid used for this multy-device FS and an array of
+ * participating devices.
+ */
+struct exofs_device_table {
+ __le32 dt_version; /* == EXOFS_DT_VER */
+ struct exofs_dt_data_map dt_data_map; /* Raid policy to use */
+
+ /* Resurved space For future use. Total includeing this:
+ * (8 * sizeof(le64))
+ */
+ __le64 __Resurved[4];
+
+ __le64 dt_num_devices; /* Array size */
+ struct exofs_dt_device_info dt_dev_table[]; /* Array of devices */
+} __packed;
/****************************************************************************
* inode-related things
diff --git a/fs/exofs/exofs.h b/fs/exofs/exofs.h
index 2e08859..c35fd46 100644
--- a/fs/exofs/exofs.h
+++ b/fs/exofs/exofs.h
@@ -37,6 +37,11 @@
#include <linux/time.h>
#include "common.h"
+/* FIXME: Remove once pnfs hits mainline
+ * #include <linux/exportfs/pnfs_osd_xdr.h>
+ */
+#include "pnfs.h"
+
#define EXOFS_ERR(fmt, a...) printk(KERN_ERR "exofs: " fmt, ##a)
#ifdef CONFIG_EXOFS_DEBUG
@@ -54,7 +59,6 @@
* our extension to the in-memory superblock
*/
struct exofs_sb_info {
- struct osd_dev *s_dev; /* returned by get_osd_dev */
struct exofs_fscb s_fscb; /* Written often, pre-allocate*/
osd_id s_pid; /* partition ID of file system*/
int s_timeout; /* timeout for OSD operations */
@@ -63,7 +67,11 @@
spinlock_t s_next_gen_lock; /* spinlock for gen # update */
u32 s_next_generation; /* next gen # to use */
atomic_t s_curr_pending; /* number of pending commands */
- uint8_t s_cred[OSD_CAP_LEN]; /* all-powerful credential */
+ uint8_t s_cred[OSD_CAP_LEN]; /* credential for the fscb */
+
+ struct pnfs_osd_data_map data_map; /* Default raid to use */
+ unsigned s_numdevs; /* Num of devices in array */
+ struct osd_dev *s_ods[1]; /* Variable length, minimum 1 */
};
/*
diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c
index 7578950..698a863 100644
--- a/fs/exofs/inode.c
+++ b/fs/exofs/inode.c
@@ -62,7 +62,10 @@
struct exofs_sb_info *sbi = inode->i_sb->s_fs_info;
pcol->sbi = sbi;
- pcol->req_q = osd_request_queue(sbi->s_dev);
+ /* Create master bios on first Q, later on cloning, each clone will be
+ * allocated on it's destination Q
+ */
+ pcol->req_q = osd_request_queue(sbi->s_ods[0]);
pcol->inode = inode;
pcol->expected_pages = expected_pages;
diff --git a/fs/exofs/ios.c b/fs/exofs/ios.c
index bb2f9d3..5bad01f 100644
--- a/fs/exofs/ios.c
+++ b/fs/exofs/ios.c
@@ -71,7 +71,7 @@
/*TODO: Maybe use kmem_cach per sbi of size
* exofs_io_state_size(sbi->s_numdevs)
*/
- ios = kzalloc(exofs_io_state_size(1), GFP_KERNEL);
+ ios = kzalloc(exofs_io_state_size(sbi->s_numdevs), GFP_KERNEL);
if (unlikely(!ios)) {
*pios = NULL;
return -ENOMEM;
@@ -209,10 +209,10 @@
{
int i, ret;
- for (i = 0; i < 1; i++) {
+ for (i = 0; i < ios->sbi->s_numdevs; i++) {
struct osd_request *or;
- or = osd_start_request(ios->sbi->s_dev, GFP_KERNEL);
+ or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
if (unlikely(!or)) {
EXOFS_ERR("%s: osd_start_request failed\n", __func__);
ret = -ENOMEM;
@@ -233,10 +233,10 @@
{
int i, ret;
- for (i = 0; i < 1; i++) {
+ for (i = 0; i < ios->sbi->s_numdevs; i++) {
struct osd_request *or;
- or = osd_start_request(ios->sbi->s_dev, GFP_KERNEL);
+ or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
if (unlikely(!or)) {
EXOFS_ERR("%s: osd_start_request failed\n", __func__);
ret = -ENOMEM;
@@ -257,10 +257,10 @@
{
int i, ret;
- for (i = 0; i < 1; i++) {
+ for (i = 0; i < ios->sbi->s_numdevs; i++) {
struct osd_request *or;
- or = osd_start_request(ios->sbi->s_dev, GFP_KERNEL);
+ or = osd_start_request(ios->sbi->s_ods[i], GFP_KERNEL);
if (unlikely(!or)) {
EXOFS_ERR("%s: osd_start_request failed\n", __func__);
ret = -ENOMEM;
@@ -272,7 +272,21 @@
if (ios->bio) {
struct bio *bio;
- bio = ios->bio;
+ if (i != 0) {
+ bio = bio_kmalloc(GFP_KERNEL,
+ ios->bio->bi_max_vecs);
+ if (unlikely(!bio)) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ __bio_clone(bio, ios->bio);
+ bio->bi_bdev = NULL;
+ bio->bi_next = NULL;
+ ios->per_dev[i].bio = bio;
+ } else {
+ bio = ios->bio;
+ }
osd_req_write(or, &ios->obj, ios->offset, bio,
ios->length);
@@ -306,8 +320,10 @@
for (i = 0; i < 1; i++) {
struct osd_request *or;
+ unsigned first_dev = (unsigned)ios->obj.id;
- or = osd_start_request(ios->sbi->s_dev, GFP_KERNEL);
+ first_dev %= ios->sbi->s_numdevs;
+ or = osd_start_request(ios->sbi->s_ods[first_dev], GFP_KERNEL);
if (unlikely(!or)) {
EXOFS_ERR("%s: osd_start_request failed\n", __func__);
ret = -ENOMEM;
@@ -382,10 +398,10 @@
attr = g_attr_logical_length;
attr.val_ptr = &newsize;
- for (i = 0; i < 1; i++) {
+ for (i = 0; i < sbi->s_numdevs; i++) {
struct osd_request *or;
- or = osd_start_request(sbi->s_dev, GFP_KERNEL);
+ or = osd_start_request(sbi->s_ods[i], GFP_KERNEL);
if (unlikely(!or)) {
EXOFS_ERR("%s: osd_start_request failed\n", __func__);
ret = -ENOMEM;
diff --git a/fs/exofs/pnfs.h b/fs/exofs/pnfs.h
new file mode 100644
index 0000000..423033a
--- /dev/null
+++ b/fs/exofs/pnfs.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright (C) 2008, 2009
+ * Boaz Harrosh <bharrosh@panasas.com>
+ *
+ * This file is part of exofs.
+ *
+ * exofs is free software; you can redistribute it and/or modify it under the
+ * terms of the GNU General Public License version 2 as published by the Free
+ * Software Foundation.
+ *
+ */
+
+/* FIXME: Remove this file once pnfs hits mainline */
+
+#ifndef __EXOFS_PNFS_H__
+#define __EXOFS_PNFS_H__
+
+#if defined(CONFIG_PNFS)
+
+
+/* FIXME: move this file to: linux/exportfs/pnfs_osd_xdr.h */
+#include "../nfs/objlayout/pnfs_osd_xdr.h"
+
+#else /* defined(CONFIG_PNFS) */
+
+enum pnfs_iomode {
+ IOMODE_READ = 1,
+ IOMODE_RW = 2,
+ IOMODE_ANY = 3,
+};
+
+/* Layout Structure */
+enum pnfs_osd_raid_algorithm4 {
+ PNFS_OSD_RAID_0 = 1,
+ PNFS_OSD_RAID_4 = 2,
+ PNFS_OSD_RAID_5 = 3,
+ PNFS_OSD_RAID_PQ = 4 /* Reed-Solomon P+Q */
+};
+
+struct pnfs_osd_data_map {
+ u32 odm_num_comps;
+ u64 odm_stripe_unit;
+ u32 odm_group_width;
+ u32 odm_group_depth;
+ u32 odm_mirror_cnt;
+ u32 odm_raid_algorithm;
+};
+
+#endif /* else defined(CONFIG_PNFS) */
+
+#endif /* __EXOFS_PNFS_H__ */
diff --git a/fs/exofs/super.c b/fs/exofs/super.c
index 4cd97f5..a1d1e77 100644
--- a/fs/exofs/super.c
+++ b/fs/exofs/super.c
@@ -214,12 +214,17 @@
if (ret)
goto out;
- ios->length = sizeof(*fscb);
+ /* Note: We only write the changing part of the fscb. .i.e upto the
+ * the fscb->s_dev_table_oid member. There is no read-modify-write
+ * here.
+ */
+ ios->length = offsetof(struct exofs_fscb, s_dev_table_oid);
memset(fscb, 0, ios->length);
fscb->s_nextid = cpu_to_le64(sbi->s_nextid);
fscb->s_numfiles = cpu_to_le32(sbi->s_numfiles);
fscb->s_magic = cpu_to_le16(sb->s_magic);
fscb->s_newfs = 0;
+ fscb->s_version = EXOFS_FSCB_VER;
ios->obj.id = EXOFS_SUPER_ID;
ios->offset = 0;
@@ -257,6 +262,20 @@
msg, dev_path ?: "", odi->osdname, _LLU(pid));
}
+void exofs_free_sbi(struct exofs_sb_info *sbi)
+{
+ while (sbi->s_numdevs) {
+ int i = --sbi->s_numdevs;
+ struct osd_dev *od = sbi->s_ods[i];
+
+ if (od) {
+ sbi->s_ods[i] = NULL;
+ osduld_put_device(od);
+ }
+ }
+ kfree(sbi);
+}
+
/*
* This function is called when the vfs is freeing the superblock. We just
* need to free our own part.
@@ -279,12 +298,182 @@
msecs_to_jiffies(100));
}
- _exofs_print_device("Unmounting", NULL, sbi->s_dev, sbi->s_pid);
- osduld_put_device(sbi->s_dev);
- kfree(sb->s_fs_info);
+ _exofs_print_device("Unmounting", NULL, sbi->s_ods[0], sbi->s_pid);
+
+ exofs_free_sbi(sbi);
sb->s_fs_info = NULL;
}
+static int _read_and_match_data_map(struct exofs_sb_info *sbi, unsigned numdevs,
+ struct exofs_device_table *dt)
+{
+ sbi->data_map.odm_num_comps =
+ le32_to_cpu(dt->dt_data_map.cb_num_comps);
+ sbi->data_map.odm_stripe_unit =
+ le64_to_cpu(dt->dt_data_map.cb_stripe_unit);
+ sbi->data_map.odm_group_width =
+ le32_to_cpu(dt->dt_data_map.cb_group_width);
+ sbi->data_map.odm_group_depth =
+ le32_to_cpu(dt->dt_data_map.cb_group_depth);
+ sbi->data_map.odm_mirror_cnt =
+ le32_to_cpu(dt->dt_data_map.cb_mirror_cnt);
+ sbi->data_map.odm_raid_algorithm =
+ le32_to_cpu(dt->dt_data_map.cb_raid_algorithm);
+
+/* FIXME: Hard coded mirror only for now. if not so do not mount */
+ if ((sbi->data_map.odm_num_comps != numdevs) ||
+ (sbi->data_map.odm_stripe_unit != EXOFS_BLKSIZE) ||
+ (sbi->data_map.odm_raid_algorithm != PNFS_OSD_RAID_0) ||
+ (sbi->data_map.odm_mirror_cnt != (numdevs - 1)))
+ return -EINVAL;
+ else
+ return 0;
+}
+
+/* @odi is valid only as long as @fscb_dev is valid */
+static int exofs_devs_2_odi(struct exofs_dt_device_info *dt_dev,
+ struct osd_dev_info *odi)
+{
+ odi->systemid_len = le32_to_cpu(dt_dev->systemid_len);
+ memcpy(odi->systemid, dt_dev->systemid, odi->systemid_len);
+
+ odi->osdname_len = le32_to_cpu(dt_dev->osdname_len);
+ odi->osdname = dt_dev->osdname;
+
+ /* FIXME support long names. Will need a _put function */
+ if (dt_dev->long_name_offset)
+ return -EINVAL;
+
+ /* Make sure osdname is printable!
+ * mkexofs should give us space for a null-terminator else the
+ * device-table is invalid.
+ */
+ if (unlikely(odi->osdname_len >= sizeof(dt_dev->osdname)))
+ odi->osdname_len = sizeof(dt_dev->osdname) - 1;
+ dt_dev->osdname[odi->osdname_len] = 0;
+
+ /* If it's all zeros something is bad we read past end-of-obj */
+ return !(odi->systemid_len || odi->osdname_len);
+}
+
+static int exofs_read_lookup_dev_table(struct exofs_sb_info **psbi,
+ unsigned table_count)
+{
+ struct exofs_sb_info *sbi = *psbi;
+ struct osd_dev *fscb_od;
+ struct osd_obj_id obj = {.partition = sbi->s_pid,
+ .id = EXOFS_DEVTABLE_ID};
+ struct exofs_device_table *dt;
+ unsigned table_bytes = table_count * sizeof(dt->dt_dev_table[0]) +
+ sizeof(*dt);
+ unsigned numdevs, i;
+ int ret;
+
+ dt = kmalloc(table_bytes, GFP_KERNEL);
+ if (unlikely(!dt)) {
+ EXOFS_ERR("ERROR: allocating %x bytes for device table\n",
+ table_bytes);
+ return -ENOMEM;
+ }
+
+ fscb_od = sbi->s_ods[0];
+ sbi->s_ods[0] = NULL;
+ sbi->s_numdevs = 0;
+ ret = exofs_read_kern(fscb_od, sbi->s_cred, &obj, 0, dt, table_bytes);
+ if (unlikely(ret)) {
+ EXOFS_ERR("ERROR: reading device table\n");
+ goto out;
+ }
+
+ numdevs = le64_to_cpu(dt->dt_num_devices);
+ if (unlikely(!numdevs)) {
+ ret = -EINVAL;
+ goto out;
+ }
+ WARN_ON(table_count != numdevs);
+
+ ret = _read_and_match_data_map(sbi, numdevs, dt);
+ if (unlikely(ret))
+ goto out;
+
+ if (likely(numdevs > 1)) {
+ unsigned size = numdevs * sizeof(sbi->s_ods[0]);
+
+ sbi = krealloc(sbi, sizeof(*sbi) + size, GFP_KERNEL);
+ if (unlikely(!sbi)) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ memset(&sbi->s_ods[1], 0, size - sizeof(sbi->s_ods[0]));
+ *psbi = sbi;
+ }
+
+ for (i = 0; i < numdevs; i++) {
+ struct exofs_fscb fscb;
+ struct osd_dev_info odi;
+ struct osd_dev *od;
+
+ if (exofs_devs_2_odi(&dt->dt_dev_table[i], &odi)) {
+ EXOFS_ERR("ERROR: Read all-zeros device entry\n");
+ ret = -EINVAL;
+ goto out;
+ }
+
+ printk(KERN_NOTICE "Add device[%d]: osd_name-%s\n",
+ i, odi.osdname);
+
+ /* On all devices the device table is identical. The user can
+ * specify any one of the participating devices on the command
+ * line. We always keep them in device-table order.
+ */
+ if (fscb_od && osduld_device_same(fscb_od, &odi)) {
+ sbi->s_ods[i] = fscb_od;
+ ++sbi->s_numdevs;
+ fscb_od = NULL;
+ continue;
+ }
+
+ od = osduld_info_lookup(&odi);
+ if (unlikely(IS_ERR(od))) {
+ ret = PTR_ERR(od);
+ EXOFS_ERR("ERROR: device requested is not found "
+ "osd_name-%s =>%d\n", odi.osdname, ret);
+ goto out;
+ }
+
+ sbi->s_ods[i] = od;
+ ++sbi->s_numdevs;
+
+ /* Read the fscb of the other devices to make sure the FS
+ * partition is there.
+ */
+ ret = exofs_read_kern(od, sbi->s_cred, &obj, 0, &fscb,
+ sizeof(fscb));
+ if (unlikely(ret)) {
+ EXOFS_ERR("ERROR: Malformed participating device "
+ "error reading fscb osd_name-%s\n",
+ odi.osdname);
+ goto out;
+ }
+
+ /* TODO: verify other information is correct and FS-uuid
+ * matches. Benny what did you say about device table
+ * generation and old devices?
+ */
+ }
+
+out:
+ kfree(dt);
+ if (unlikely(!ret && fscb_od)) {
+ EXOFS_ERR(
+ "ERROR: Bad device-table container device not present\n");
+ osduld_put_device(fscb_od);
+ ret = -EINVAL;
+ }
+
+ return ret;
+}
+
/*
* Read the superblock from the OSD and fill in the fields
*/
@@ -296,6 +485,7 @@
struct osd_dev *od; /* Master device */
struct exofs_fscb fscb; /*on-disk superblock info */
struct osd_obj_id obj;
+ unsigned table_count;
int ret;
sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
@@ -309,7 +499,8 @@
goto free_sbi;
}
- sbi->s_dev = od;
+ sbi->s_ods[0] = od;
+ sbi->s_numdevs = 1;
sbi->s_pid = opts->pid;
sbi->s_timeout = opts->timeout;
@@ -342,11 +533,24 @@
ret = -EINVAL;
goto free_sbi;
}
+ if (le32_to_cpu(fscb.s_version) != EXOFS_FSCB_VER) {
+ EXOFS_ERR("ERROR: Bad FSCB version expected-%d got-%d\n",
+ EXOFS_FSCB_VER, le32_to_cpu(fscb.s_version));
+ ret = -EINVAL;
+ goto free_sbi;
+ }
/* start generation numbers from a random point */
get_random_bytes(&sbi->s_next_generation, sizeof(u32));
spin_lock_init(&sbi->s_next_gen_lock);
+ table_count = le64_to_cpu(fscb.s_dev_table_count);
+ if (table_count) {
+ ret = exofs_read_lookup_dev_table(&sbi, table_count);
+ if (unlikely(ret))
+ goto free_sbi;
+ }
+
/* set up operation vectors */
sb->s_fs_info = sbi;
sb->s_op = &exofs_sops;
@@ -374,14 +578,14 @@
goto free_sbi;
}
- _exofs_print_device("Mounting", opts->dev_name, sbi->s_dev, sbi->s_pid);
+ _exofs_print_device("Mounting", opts->dev_name, sbi->s_ods[0],
+ sbi->s_pid);
return 0;
free_sbi:
EXOFS_ERR("Unable to mount exofs on %s pid=0x%llx err=%d\n",
opts->dev_name, sbi->s_pid, ret);
- osduld_put_device(sbi->s_dev); /* NULL safe */
- kfree(sbi);
+ exofs_free_sbi(sbi);
return ret;
}
