Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1622276 > unrolled thread

RFC: drop the T10 OSD code and its users

Back to article view | Back to linux.kernel

Contents

  RFC: drop the T10 OSD code and its users Christoph Hellwig <hch@lst.de> - 2017-04-12 18:10 +0200
    [PATCH 1/4] block: remove the osdblk driver Christoph Hellwig <hch@lst.de> - 2017-04-12 18:10 +0200
      Re: [PATCH 1/4] block: remove the osdblk driver Boaz Harrosh <ooo@electrozaur.com> - 2017-04-19 09:30 +0200
    Re: RFC: drop the T10 OSD code and its users "Martin K. Petersen" <martin.petersen@oracle.com> - 2017-04-14 04:30 +0200
    Re: RFC: drop the T10 OSD code and its users Boaz Harrosh <ooo@electrozaur.com> - 2017-04-18 18:30 +0200
      Re: RFC: drop the T10 OSD code and its users "Chandy, John" <john.chandy@uconn.edu> - 2017-04-18 19:10 +0200
    Re: RFC: drop the T10 OSD code and its users Jens Axboe <axboe@kernel.dk> - 2017-04-19 17:20 +0200
    Re: [PATCH 3/4] nfs: remove the objlayout driver Trond Myklebust <trondmy@primarydata.com> - 2017-04-20 22:20 +0200
      Re: [PATCH 3/4] nfs: remove the objlayout driver Christoph Hellwig <hch@infradead.org> - 2017-04-21 13:00 +0200
      Re: [PATCH 3/4] nfs: remove the objlayout driver Boaz Harrosh <ooo@electrozaur.com> - 2017-04-21 15:20 +0200
        Re: [PATCH 3/4] nfs: remove the objlayout driver Boaz Harrosh <ooo@electrozaur.com> - 2017-04-23 11:40 +0200
    Re: [PATCH 3/4] nfs: remove the objlayout driver Christoph Hellwig <hch@infradead.org> - 2017-04-21 13:00 +0200

#1622276 — RFC: drop the T10 OSD code and its users

The only real user of the T10 OSD protocol, the pNFS object layout
driver never went to the point of having shipping products, and the
other two users (osdblk and exofs) were simple example of it's usage.

The code has been mostly unmaintained for years and is getting in the
way of block / SCSI changes, so I think it's finally time to drop it.

These patches are against Jens' block for-next tree as that already
has various modifications of the SCSI code.

[toc] | [next] | [standalone]

#1622283 — [PATCH 1/4] block: remove the osdblk driver

This was just a proof of concept user for the SCSI OSD library, and
never had any real users.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/block/Kconfig  |  16 --
 drivers/block/Makefile |   1 -
 drivers/block/osdblk.c | 693 -------------------------------------------------
 3 files changed, 710 deletions(-)
 delete mode 100644 drivers/block/osdblk.c

diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index a1c2e816128f..58e24c354933 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -312,22 +312,6 @@ config BLK_DEV_SKD
 
 	Use device /dev/skd$N amd /dev/skd$Np$M.
 
-config BLK_DEV_OSD
-	tristate "OSD object-as-blkdev support"
-	depends on SCSI_OSD_ULD
-	---help---
-	  Saying Y or M here will allow the exporting of a single SCSI
-	  OSD (object-based storage) object as a Linux block device.
-
-	  For example, if you create a 2G object on an OSD device,
-	  you can then use this module to present that 2G object as
-	  a Linux block device.
-
-	  To compile this driver as a module, choose M here: the
-	  module will be called osdblk.
-
-	  If unsure, say N.
-
 config BLK_DEV_SX8
 	tristate "Promise SATA SX8 support"
 	depends on PCI
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index b12c772bbeb3..42e8cee1cbc7 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -22,7 +22,6 @@ obj-$(CONFIG_CDROM_PKTCDVD)	+= pktcdvd.o
 obj-$(CONFIG_MG_DISK)		+= mg_disk.o
 obj-$(CONFIG_SUNVDC)		+= sunvdc.o
 obj-$(CONFIG_BLK_DEV_SKD)	+= skd.o
-obj-$(CONFIG_BLK_DEV_OSD)	+= osdblk.o
 
 obj-$(CONFIG_BLK_DEV_UMEM)	+= umem.o
 obj-$(CONFIG_BLK_DEV_NBD)	+= nbd.o
diff --git a/drivers/block/osdblk.c b/drivers/block/osdblk.c
deleted file mode 100644
index 8127b8201a01..000000000000
--- a/drivers/block/osdblk.c
+++ /dev/null
@@ -1,693 +0,0 @@
-
-/*
-   osdblk.c -- Export a single SCSI OSD object as a Linux block device
-
-
-   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 OSD object.
-
-      In this example, we will use partition id 1234, object id 5678,
-      OSD device /dev/osd1.
-
-      $ echo "1234 5678 /dev/osd1" > /sys/class/osdblk/add
-
-
-   2) List all active blkdev<->object mappings.
-
-      In this example, we have performed step #1 twice, creating two blkdevs,
-      mapped to two separate OSD objects.
-
-      $ cat /sys/class/osdblk/list
-      0 174 1234 5678 /dev/osd1
-      1 179 1994 897123 /dev/osd0
-
-      The columns, in order, are:
-      - blkdev unique id
-      - blkdev assigned major
-      - OSD object partition id
-      - OSD object id
-      - OSD device
-
-
-   3) Remove an active blkdev<->object mapping.
-
-      In this example, we remove the mapping with blkdev unique id 1.
-
-      $ echo 1 > /sys/class/osdblk/remove
-
-
-   NOTE:  The actual creation and deletion of OSD objects is outside the scope
-   of this driver.
-
- */
-
-#include <linux/kernel.h>
-#include <linux/device.h>
-#include <linux/module.h>
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <scsi/osd_initiator.h>
-#include <scsi/osd_attributes.h>
-#include <scsi/osd_sec.h>
-#include <scsi/scsi_device.h>
-
-#define DRV_NAME "osdblk"
-#define PFX DRV_NAME ": "
-
-/* #define _OSDBLK_DEBUG */
-#ifdef _OSDBLK_DEBUG
-#define OSDBLK_DEBUG(fmt, a...) \
-	printk(KERN_NOTICE "osdblk @%s:%d: " fmt, __func__, __LINE__, ##a)
-#else
-#define OSDBLK_DEBUG(fmt, a...) \
-	do { if (0) printk(fmt, ##a); } while (0)
-#endif
-
-MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>");
-MODULE_DESCRIPTION("block device inside an OSD object osdblk.ko");
-MODULE_LICENSE("GPL");
-
-struct osdblk_device;
-
-enum {
-	OSDBLK_MINORS_PER_MAJOR	= 256,		/* max minors per blkdev */
-	OSDBLK_MAX_REQ		= 32,		/* max parallel requests */
-	OSDBLK_OP_TIMEOUT	= 4 * 60,	/* sync OSD req timeout */
-};
-
-struct osdblk_request {
-	struct request		*rq;		/* blk layer request */
-	struct bio		*bio;		/* cloned bio */
-	struct osdblk_device	*osdev;		/* associated blkdev */
-};
-
-struct osdblk_device {
-	int			id;		/* blkdev unique id */
-
-	int			major;		/* blkdev assigned major */
-	struct gendisk		*disk;		/* blkdev's gendisk and rq */
-	struct request_queue	*q;
-
-	struct osd_dev		*osd;		/* associated OSD */
-
-	char			name[32];	/* blkdev name, e.g. osdblk34 */
-
-	spinlock_t		lock;		/* queue lock */
-
-	struct osd_obj_id	obj;		/* OSD partition, obj id */
-	uint8_t			obj_cred[OSD_CAP_LEN]; /* OSD cred */
-
-	struct osdblk_request	req[OSDBLK_MAX_REQ]; /* request table */
-
-	struct list_head	node;
-
-	char			osd_path[0];	/* OSD device path */
-};
-
-static struct class *class_osdblk;		/* /sys/class/osdblk */
-static DEFINE_MUTEX(ctl_mutex);	/* Serialize open/close/setup/teardown */
-static LIST_HEAD(osdblkdev_list);
-
-static const struct block_device_operations osdblk_bd_ops = {
-	.owner		= THIS_MODULE,
-};
-
-static const struct osd_attr g_attr_logical_length = ATTR_DEF(
-	OSD_APAGE_OBJECT_INFORMATION, OSD_ATTR_OI_LOGICAL_LENGTH, 8);
-
-static void osdblk_make_credential(u8 cred_a[OSD_CAP_LEN],
-				   const struct osd_obj_id *obj)
-{
-	osd_sec_init_nosec_doall_caps(cred_a, obj, false, true);
-}
-
-/* copied from exofs; move to libosd? */
-/*
- * Perform a synchronous OSD operation.  copied from exofs; move to libosd?
- */
-static int osd_sync_op(struct osd_request *or, int timeout, uint8_t *credential)
-{
-	int ret;
-
-	or->timeout = timeout;
-	ret = osd_finalize_request(or, 0, credential, NULL);
-	if (ret)
-		return ret;
-
-	ret = osd_execute_request(or);
-
-	/* osd_req_decode_sense(or, ret); */
-	return ret;
-}
-
-/*
- * Perform an asynchronous OSD operation.  copied from exofs; move to libosd?
- */
-static int osd_async_op(struct osd_request *or, osd_req_done_fn *async_done,
-		   void *caller_context, u8 *cred)
-{
-	int ret;
-
-	ret = osd_finalize_request(or, 0, cred, NULL);
-	if (ret)
-		return ret;
-
-	ret = osd_execute_request_async(or, async_done, caller_context);
-
-	return ret;
-}
-
-/* copied from exofs; move to libosd? */
-static int extract_attr_from_req(struct osd_request *or, struct osd_attr *attr)
-{
-	struct osd_attr cur_attr = {.attr_page = 0}; /* start with zeros */
-	void *iter = NULL;
-	int nelem;
-
-	do {
-		nelem = 1;
-		osd_req_decode_get_attr_list(or, &cur_attr, &nelem, &iter);
-		if ((cur_attr.attr_page == attr->attr_page) &&
-		    (cur_attr.attr_id == attr->attr_id)) {
-			attr->len = cur_attr.len;
-			attr->val_ptr = cur_attr.val_ptr;
-			return 0;
-		}
-	} while (iter);
-
-	return -EIO;
-}
-
-static int osdblk_get_obj_size(struct osdblk_device *osdev, u64 *size_out)
-{
-	struct osd_request *or;
-	struct osd_attr attr;
-	int ret;
-
-	/* start request */
-	or = osd_start_request(osdev->osd, GFP_KERNEL);
-	if (!or)
-		return -ENOMEM;
-
-	/* create a get-attributes(length) request */
-	osd_req_get_attributes(or, &osdev->obj);
-
-	osd_req_add_get_attr_list(or, &g_attr_logical_length, 1);
-
-	/* execute op synchronously */
-	ret = osd_sync_op(or, OSDBLK_OP_TIMEOUT, osdev->obj_cred);
-	if (ret)
-		goto out;
-
-	/* extract length from returned attribute info */
-	attr = g_attr_logical_length;
-	ret = extract_attr_from_req(or, &attr);
-	if (ret)
-		goto out;
-
-	*size_out = get_unaligned_be64(attr.val_ptr);
-
-out:
-	osd_end_request(or);
-	return ret;
-
-}
-
-static void osdblk_osd_complete(struct osd_request *or, void *private)
-{
-	struct osdblk_request *orq = private;
-	struct osd_sense_info osi;
-	int ret = osd_req_decode_sense(or, &osi);
-
-	if (ret) {
-		ret = -EIO;
-		OSDBLK_DEBUG("osdblk_osd_complete with err=%d\n", ret);
-	}
-
-	/* complete OSD request */
-	osd_end_request(or);
-
-	/* complete request passed to osdblk by block layer */
-	__blk_end_request_all(orq->rq, ret);
-}
-
-static void bio_chain_put(struct bio *chain)
-{
-	struct bio *tmp;
-
-	while (chain) {
-		tmp = chain;
-		chain = chain->bi_next;
-
-		bio_put(tmp);
-	}
-}
-
-static struct bio *bio_chain_clone(struct bio *old_chain, gfp_t gfpmask)
-{
-	struct bio *tmp, *new_chain = NULL, *tail = NULL;
-
-	while (old_chain) {
-		tmp = bio_clone_kmalloc(old_chain, gfpmask);
-		if (!tmp)
-			goto err_out;
-
-		tmp->bi_bdev = NULL;
-		gfpmask &= ~__GFP_DIRECT_RECLAIM;
-		tmp->bi_next = NULL;
-
-		if (!new_chain)
-			new_chain = tail = tmp;
-		else {
-			tail->bi_next = tmp;
-			tail = tmp;
-		}
-
-		old_chain = old_chain->bi_next;
-	}
-
-	return new_chain;
-
-err_out:
-	OSDBLK_DEBUG("bio_chain_clone with err\n");
-	bio_chain_put(new_chain);
-	return NULL;
-}
-
-static void osdblk_rq_fn(struct request_queue *q)
-{
-	struct osdblk_device *osdev = q->queuedata;
-
-	while (1) {
-		struct request *rq;
-		struct osdblk_request *orq;
-		struct osd_request *or;
-		struct bio *bio;
-		bool do_write, do_flush;
-
-		/* peek at request from block layer */
-		rq = blk_fetch_request(q);
-		if (!rq)
-			break;
-
-		/* deduce our operation (read, write, flush) */
-		/* I wish the block layer simplified cmd_type/cmd_flags/cmd[]
-		 * into a clearly defined set of RPC commands:
-		 * read, write, flush, scsi command, power mgmt req,
-		 * driver-specific, etc.
-		 */
-
-		do_flush = (req_op(rq) == REQ_OP_FLUSH);
-		do_write = (rq_data_dir(rq) == WRITE);
-
-		if (!do_flush) { /* osd_flush does not use a bio */
-			/* a bio clone to be passed down to OSD request */
-			bio = bio_chain_clone(rq->bio, GFP_ATOMIC);
-			if (!bio)
-				break;
-		} else
-			bio = NULL;
-
-		/* alloc internal OSD request, for OSD command execution */
-		or = osd_start_request(osdev->osd, GFP_ATOMIC);
-		if (!or) {
-			bio_chain_put(bio);
-			OSDBLK_DEBUG("osd_start_request with err\n");
-			break;
-		}
-
-		orq = &osdev->req[rq->tag];
-		orq->rq = rq;
-		orq->bio = bio;
-		orq->osdev = osdev;
-
-		/* init OSD command: flush, write or read */
-		if (do_flush)
-			osd_req_flush_object(or, &osdev->obj,
-					     OSD_CDB_FLUSH_ALL, 0, 0);
-		else if (do_write)
-			osd_req_write(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
-				      bio, blk_rq_bytes(rq));
-		else
-			osd_req_read(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
-				     bio, blk_rq_bytes(rq));
-
-		OSDBLK_DEBUG("%s 0x%x bytes at 0x%llx\n",
-			do_flush ? "flush" : do_write ?
-				"write" : "read", blk_rq_bytes(rq),
-			blk_rq_pos(rq) * 512ULL);
-
-		/* begin OSD command execution */
-		if (osd_async_op(or, osdblk_osd_complete, orq,
-				 osdev->obj_cred)) {
-			osd_end_request(or);
-			blk_requeue_request(q, rq);
-			bio_chain_put(bio);
-			OSDBLK_DEBUG("osd_execute_request_async with err\n");
-			break;
-		}
-
-		/* remove the special 'flush' marker, now that the command
-		 * is executing
-		 */
-		rq->special = NULL;
-	}
-}
-
-static void osdblk_free_disk(struct osdblk_device *osdev)
-{
-	struct gendisk *disk = osdev->disk;
-
-	if (!disk)
-		return;
-
-	if (disk->flags & GENHD_FL_UP)
-		del_gendisk(disk);
-	if (disk->queue)
-		blk_cleanup_queue(disk->queue);
-	put_disk(disk);
-}
-
-static int osdblk_init_disk(struct osdblk_device *osdev)
-{
-	struct gendisk *disk;
-	struct request_queue *q;
-	int rc;
-	u64 obj_size = 0;
-
-	/* contact OSD, request size info about the object being mapped */
-	rc = osdblk_get_obj_size(osdev, &obj_size);
-	if (rc)
-		return rc;
-
-	/* create gendisk info */
-	disk = alloc_disk(OSDBLK_MINORS_PER_MAJOR);
-	if (!disk)
-		return -ENOMEM;
-
-	sprintf(disk->disk_name, DRV_NAME "%d", osdev->id);
-	disk->major = osdev->major;
-	disk->first_minor = 0;
-	disk->fops = &osdblk_bd_ops;
-	disk->private_data = osdev;
-
-	/* init rq */
-	q = blk_init_queue(osdblk_rq_fn, &osdev->lock);
-	if (!q) {
-		put_disk(disk);
-		return -ENOMEM;
-	}
-
-	/* switch queue to TCQ mode; allocate tag map */
-	rc = blk_queue_init_tags(q, OSDBLK_MAX_REQ, NULL, BLK_TAG_ALLOC_FIFO);
-	if (rc) {
-		blk_cleanup_queue(q);
-		put_disk(disk);
-		return rc;
-	}
-
-	/* Set our limits to the lower device limits, because osdblk cannot
-	 * sleep when allocating a lower-request and therefore cannot be
-	 * bouncing.
-	 */
-	blk_queue_stack_limits(q, osd_request_queue(osdev->osd));
-
-	blk_queue_prep_rq(q, blk_queue_start_tag);
-	blk_queue_write_cache(q, true, false);
-
-	disk->queue = q;
-
-	q->queuedata = osdev;
-
-	osdev->disk = disk;
-	osdev->q = q;
-
-	/* finally, announce the disk to the world */
-	set_capacity(disk, obj_size / 512ULL);
-	add_disk(disk);
-
-	printk(KERN_INFO "%s: Added of size 0x%llx\n",
-		disk->disk_name, (unsigned long long)obj_size);
-
-	return 0;
-}
-
-/********************************************************************
- * /sys/class/osdblk/
- *                   add	map OSD object to blkdev
- *                   remove	unmap OSD object
- *                   list	show mappings
- *******************************************************************/
-
-static void class_osdblk_release(struct class *cls)
-{
-	kfree(cls);
-}
-
-static ssize_t class_osdblk_list(struct class *c,
-				struct class_attribute *attr,
-				char *data)
-{
-	int n = 0;
-	struct list_head *tmp;
-
-	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-
-	list_for_each(tmp, &osdblkdev_list) {
-		struct osdblk_device *osdev;
-
-		osdev = list_entry(tmp, struct osdblk_device, node);
-
-		n += sprintf(data+n, "%d %d %llu %llu %s\n",
-			osdev->id,
-			osdev->major,
-			osdev->obj.partition,
-			osdev->obj.id,
-			osdev->osd_path);
-	}
-
-	mutex_unlock(&ctl_mutex);
-	return n;
-}
-
-static ssize_t class_osdblk_add(struct class *c,
-				struct class_attribute *attr,
-				const char *buf, size_t count)
-{
-	struct osdblk_device *osdev;
-	ssize_t rc;
-	int irc, new_id = 0;
-	struct list_head *tmp;
-
-	if (!try_module_get(THIS_MODULE))
-		return -ENODEV;
-
-	/* new osdblk_device object */
-	osdev = kzalloc(sizeof(*osdev) + strlen(buf) + 1, GFP_KERNEL);
-	if (!osdev) {
-		rc = -ENOMEM;
-		goto err_out_mod;
-	}
-
-	/* static osdblk_device initialization */
-	spin_lock_init(&osdev->lock);
-	INIT_LIST_HEAD(&osdev->node);
-
-	/* generate unique id: find highest unique id, add one */
-
-	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-
-	list_for_each(tmp, &osdblkdev_list) {
-		struct osdblk_device *osdev;
-
-		osdev = list_entry(tmp, struct osdblk_device, node);
-		if (osdev->id > new_id)
-			new_id = osdev->id + 1;
-	}
-
-	osdev->id = new_id;
-
-	/* add to global list */
-	list_add_tail(&osdev->node, &osdblkdev_list);
-
-	mutex_unlock(&ctl_mutex);
-
-	/* parse add command */
-	if (sscanf(buf, "%llu %llu %s", &osdev->obj.partition, &osdev->obj.id,
-		   osdev->osd_path) != 3) {
-		rc = -EINVAL;
-		goto err_out_slot;
-	}
-
-	/* initialize rest of new object */
-	sprintf(osdev->name, DRV_NAME "%d", osdev->id);
-
-	/* contact requested OSD */
-	osdev->osd = osduld_path_lookup(osdev->osd_path);
-	if (IS_ERR(osdev->osd)) {
-		rc = PTR_ERR(osdev->osd);
-		goto err_out_slot;
-	}
-
-	/* build OSD credential */
-	osdblk_make_credential(osdev->obj_cred, &osdev->obj);
-
-	/* register our block device */
-	irc = register_blkdev(0, osdev->name);
-	if (irc < 0) {
-		rc = irc;
-		goto err_out_osd;
-	}
-
-	osdev->major = irc;
-
-	/* set up and announce blkdev mapping */
-	rc = osdblk_init_disk(osdev);
-	if (rc)
-		goto err_out_blkdev;
-
-	return count;
-
-err_out_blkdev:
-	unregister_blkdev(osdev->major, osdev->name);
-err_out_osd:
-	osduld_put_device(osdev->osd);
-err_out_slot:
-	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
-	list_del_init(&osdev->node);
-	mutex_unlock(&ctl_mutex);
-
-	kfree(osdev);
-err_out_mod:
-	OSDBLK_DEBUG("Error adding device %s\n", buf);
-	module_put(THIS_MODULE);
-	return rc;
-}
-
-static ssize_t class_osdblk_remove(struct class *c,
-					struct class_attribute *attr,
-					const char *buf,
-					size_t count)
-{
-	struct osdblk_device *osdev = NULL;
-	int target_id, rc;
-	unsigned long ul;
-	struct list_head *tmp;
-
-	rc = kstrtoul(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);
-
-	list_for_each(tmp, &osdblkdev_list) {
-		osdev = list_entry(tmp, struct osdblk_device, node);
-		if (osdev->id == target_id) {
-			list_del_init(&osdev->node);
-			break;
-		}
-		osdev = NULL;
-	}
-
-	mutex_unlock(&ctl_mutex);
-
-	if (!osdev)
-		return -ENOENT;
-
-	/* clean up and free blkdev and associated OSD connection */
-	osdblk_free_disk(osdev);
-	unregister_blkdev(osdev->major, osdev->name);
-	osduld_put_device(osdev->osd);
-	kfree(osdev);
-
-	/* release module ref */
-	module_put(THIS_MODULE);
-
-	return count;
-}
-
-static struct class_attribute class_osdblk_attrs[] = {
-	__ATTR(add,	0200, NULL, class_osdblk_add),
-	__ATTR(remove,	0200, NULL, class_osdblk_remove),
-	__ATTR(list,	0444, class_osdblk_list, NULL),
-	__ATTR_NULL
-};
-
-static int osdblk_sysfs_init(void)
-{
-	int ret = 0;
-
-	/*
-	 * create control files in sysfs
-	 * /sys/class/osdblk/...
-	 */
-	class_osdblk = kzalloc(sizeof(*class_osdblk), GFP_KERNEL);
-	if (!class_osdblk)
-		return -ENOMEM;
-
-	class_osdblk->name = DRV_NAME;
-	class_osdblk->owner = THIS_MODULE;
-	class_osdblk->class_release = class_osdblk_release;
-	class_osdblk->class_attrs = class_osdblk_attrs;
-
-	ret = class_register(class_osdblk);
-	if (ret) {
-		kfree(class_osdblk);
-		class_osdblk = NULL;
-		printk(PFX "failed to create class osdblk\n");
-		return ret;
-	}
-
-	return 0;
-}
-
-static void osdblk_sysfs_cleanup(void)
-{
-	if (class_osdblk)
-		class_destroy(class_osdblk);
-	class_osdblk = NULL;
-}
-
-static int __init osdblk_init(void)
-{
-	int rc;
-
-	rc = osdblk_sysfs_init();
-	if (rc)
-		return rc;
-
-	return 0;
-}
-
-static void __exit osdblk_exit(void)
-{
-	osdblk_sysfs_cleanup();
-}
-
-module_init(osdblk_init);
-module_exit(osdblk_exit);
-
-- 
2.11.0

[toc] | [prev] | [next] | [standalone]

#1625880 — Re: [PATCH 1/4] block: remove the osdblk driver

On 04/12/2017 07:01 PM, Christoph Hellwig wrote:
> This was just a proof of concept user for the SCSI OSD library, and
> never had any real users.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Yes please remove this driver

ACK-by Boaz Harrosh <ooo@electrozaur.com>

> ---
>  drivers/block/Kconfig  |  16 --
>  drivers/block/Makefile |   1 -
>  drivers/block/osdblk.c | 693 -------------------------------------------------
>  3 files changed, 710 deletions(-)
>  delete mode 100644 drivers/block/osdblk.c
> 
> diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
> index a1c2e816128f..58e24c354933 100644
> --- a/drivers/block/Kconfig
> +++ b/drivers/block/Kconfig
> @@ -312,22 +312,6 @@ config BLK_DEV_SKD
>  
>  	Use device /dev/skd$N amd /dev/skd$Np$M.
>  
> -config BLK_DEV_OSD
> -	tristate "OSD object-as-blkdev support"
> -	depends on SCSI_OSD_ULD
> -	---help---
> -	  Saying Y or M here will allow the exporting of a single SCSI
> -	  OSD (object-based storage) object as a Linux block device.
> -
> -	  For example, if you create a 2G object on an OSD device,
> -	  you can then use this module to present that 2G object as
> -	  a Linux block device.
> -
> -	  To compile this driver as a module, choose M here: the
> -	  module will be called osdblk.
> -
> -	  If unsure, say N.
> -
>  config BLK_DEV_SX8
>  	tristate "Promise SATA SX8 support"
>  	depends on PCI
> diff --git a/drivers/block/Makefile b/drivers/block/Makefile
> index b12c772bbeb3..42e8cee1cbc7 100644
> --- a/drivers/block/Makefile
> +++ b/drivers/block/Makefile
> @@ -22,7 +22,6 @@ obj-$(CONFIG_CDROM_PKTCDVD)	+= pktcdvd.o
>  obj-$(CONFIG_MG_DISK)		+= mg_disk.o
>  obj-$(CONFIG_SUNVDC)		+= sunvdc.o
>  obj-$(CONFIG_BLK_DEV_SKD)	+= skd.o
> -obj-$(CONFIG_BLK_DEV_OSD)	+= osdblk.o
>  
>  obj-$(CONFIG_BLK_DEV_UMEM)	+= umem.o
>  obj-$(CONFIG_BLK_DEV_NBD)	+= nbd.o
> diff --git a/drivers/block/osdblk.c b/drivers/block/osdblk.c
> deleted file mode 100644
> index 8127b8201a01..000000000000
> --- a/drivers/block/osdblk.c
> +++ /dev/null
> @@ -1,693 +0,0 @@
> -
> -/*
> -   osdblk.c -- Export a single SCSI OSD object as a Linux block device
> -
> -
> -   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 OSD object.
> -
> -      In this example, we will use partition id 1234, object id 5678,
> -      OSD device /dev/osd1.
> -
> -      $ echo "1234 5678 /dev/osd1" > /sys/class/osdblk/add
> -
> -
> -   2) List all active blkdev<->object mappings.
> -
> -      In this example, we have performed step #1 twice, creating two blkdevs,
> -      mapped to two separate OSD objects.
> -
> -      $ cat /sys/class/osdblk/list
> -      0 174 1234 5678 /dev/osd1
> -      1 179 1994 897123 /dev/osd0
> -
> -      The columns, in order, are:
> -      - blkdev unique id
> -      - blkdev assigned major
> -      - OSD object partition id
> -      - OSD object id
> -      - OSD device
> -
> -
> -   3) Remove an active blkdev<->object mapping.
> -
> -      In this example, we remove the mapping with blkdev unique id 1.
> -
> -      $ echo 1 > /sys/class/osdblk/remove
> -
> -
> -   NOTE:  The actual creation and deletion of OSD objects is outside the scope
> -   of this driver.
> -
> - */
> -
> -#include <linux/kernel.h>
> -#include <linux/device.h>
> -#include <linux/module.h>
> -#include <linux/fs.h>
> -#include <linux/slab.h>
> -#include <scsi/osd_initiator.h>
> -#include <scsi/osd_attributes.h>
> -#include <scsi/osd_sec.h>
> -#include <scsi/scsi_device.h>
> -
> -#define DRV_NAME "osdblk"
> -#define PFX DRV_NAME ": "
> -
> -/* #define _OSDBLK_DEBUG */
> -#ifdef _OSDBLK_DEBUG
> -#define OSDBLK_DEBUG(fmt, a...) \
> -	printk(KERN_NOTICE "osdblk @%s:%d: " fmt, __func__, __LINE__, ##a)
> -#else
> -#define OSDBLK_DEBUG(fmt, a...) \
> -	do { if (0) printk(fmt, ##a); } while (0)
> -#endif
> -
> -MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>");
> -MODULE_DESCRIPTION("block device inside an OSD object osdblk.ko");
> -MODULE_LICENSE("GPL");
> -
> -struct osdblk_device;
> -
> -enum {
> -	OSDBLK_MINORS_PER_MAJOR	= 256,		/* max minors per blkdev */
> -	OSDBLK_MAX_REQ		= 32,		/* max parallel requests */
> -	OSDBLK_OP_TIMEOUT	= 4 * 60,	/* sync OSD req timeout */
> -};
> -
> -struct osdblk_request {
> -	struct request		*rq;		/* blk layer request */
> -	struct bio		*bio;		/* cloned bio */
> -	struct osdblk_device	*osdev;		/* associated blkdev */
> -};
> -
> -struct osdblk_device {
> -	int			id;		/* blkdev unique id */
> -
> -	int			major;		/* blkdev assigned major */
> -	struct gendisk		*disk;		/* blkdev's gendisk and rq */
> -	struct request_queue	*q;
> -
> -	struct osd_dev		*osd;		/* associated OSD */
> -
> -	char			name[32];	/* blkdev name, e.g. osdblk34 */
> -
> -	spinlock_t		lock;		/* queue lock */
> -
> -	struct osd_obj_id	obj;		/* OSD partition, obj id */
> -	uint8_t			obj_cred[OSD_CAP_LEN]; /* OSD cred */
> -
> -	struct osdblk_request	req[OSDBLK_MAX_REQ]; /* request table */
> -
> -	struct list_head	node;
> -
> -	char			osd_path[0];	/* OSD device path */
> -};
> -
> -static struct class *class_osdblk;		/* /sys/class/osdblk */
> -static DEFINE_MUTEX(ctl_mutex);	/* Serialize open/close/setup/teardown */
> -static LIST_HEAD(osdblkdev_list);
> -
> -static const struct block_device_operations osdblk_bd_ops = {
> -	.owner		= THIS_MODULE,
> -};
> -
> -static const struct osd_attr g_attr_logical_length = ATTR_DEF(
> -	OSD_APAGE_OBJECT_INFORMATION, OSD_ATTR_OI_LOGICAL_LENGTH, 8);
> -
> -static void osdblk_make_credential(u8 cred_a[OSD_CAP_LEN],
> -				   const struct osd_obj_id *obj)
> -{
> -	osd_sec_init_nosec_doall_caps(cred_a, obj, false, true);
> -}
> -
> -/* copied from exofs; move to libosd? */
> -/*
> - * Perform a synchronous OSD operation.  copied from exofs; move to libosd?
> - */
> -static int osd_sync_op(struct osd_request *or, int timeout, uint8_t *credential)
> -{
> -	int ret;
> -
> -	or->timeout = timeout;
> -	ret = osd_finalize_request(or, 0, credential, NULL);
> -	if (ret)
> -		return ret;
> -
> -	ret = osd_execute_request(or);
> -
> -	/* osd_req_decode_sense(or, ret); */
> -	return ret;
> -}
> -
> -/*
> - * Perform an asynchronous OSD operation.  copied from exofs; move to libosd?
> - */
> -static int osd_async_op(struct osd_request *or, osd_req_done_fn *async_done,
> -		   void *caller_context, u8 *cred)
> -{
> -	int ret;
> -
> -	ret = osd_finalize_request(or, 0, cred, NULL);
> -	if (ret)
> -		return ret;
> -
> -	ret = osd_execute_request_async(or, async_done, caller_context);
> -
> -	return ret;
> -}
> -
> -/* copied from exofs; move to libosd? */
> -static int extract_attr_from_req(struct osd_request *or, struct osd_attr *attr)
> -{
> -	struct osd_attr cur_attr = {.attr_page = 0}; /* start with zeros */
> -	void *iter = NULL;
> -	int nelem;
> -
> -	do {
> -		nelem = 1;
> -		osd_req_decode_get_attr_list(or, &cur_attr, &nelem, &iter);
> -		if ((cur_attr.attr_page == attr->attr_page) &&
> -		    (cur_attr.attr_id == attr->attr_id)) {
> -			attr->len = cur_attr.len;
> -			attr->val_ptr = cur_attr.val_ptr;
> -			return 0;
> -		}
> -	} while (iter);
> -
> -	return -EIO;
> -}
> -
> -static int osdblk_get_obj_size(struct osdblk_device *osdev, u64 *size_out)
> -{
> -	struct osd_request *or;
> -	struct osd_attr attr;
> -	int ret;
> -
> -	/* start request */
> -	or = osd_start_request(osdev->osd, GFP_KERNEL);
> -	if (!or)
> -		return -ENOMEM;
> -
> -	/* create a get-attributes(length) request */
> -	osd_req_get_attributes(or, &osdev->obj);
> -
> -	osd_req_add_get_attr_list(or, &g_attr_logical_length, 1);
> -
> -	/* execute op synchronously */
> -	ret = osd_sync_op(or, OSDBLK_OP_TIMEOUT, osdev->obj_cred);
> -	if (ret)
> -		goto out;
> -
> -	/* extract length from returned attribute info */
> -	attr = g_attr_logical_length;
> -	ret = extract_attr_from_req(or, &attr);
> -	if (ret)
> -		goto out;
> -
> -	*size_out = get_unaligned_be64(attr.val_ptr);
> -
> -out:
> -	osd_end_request(or);
> -	return ret;
> -
> -}
> -
> -static void osdblk_osd_complete(struct osd_request *or, void *private)
> -{
> -	struct osdblk_request *orq = private;
> -	struct osd_sense_info osi;
> -	int ret = osd_req_decode_sense(or, &osi);
> -
> -	if (ret) {
> -		ret = -EIO;
> -		OSDBLK_DEBUG("osdblk_osd_complete with err=%d\n", ret);
> -	}
> -
> -	/* complete OSD request */
> -	osd_end_request(or);
> -
> -	/* complete request passed to osdblk by block layer */
> -	__blk_end_request_all(orq->rq, ret);
> -}
> -
> -static void bio_chain_put(struct bio *chain)
> -{
> -	struct bio *tmp;
> -
> -	while (chain) {
> -		tmp = chain;
> -		chain = chain->bi_next;
> -
> -		bio_put(tmp);
> -	}
> -}
> -
> -static struct bio *bio_chain_clone(struct bio *old_chain, gfp_t gfpmask)
> -{
> -	struct bio *tmp, *new_chain = NULL, *tail = NULL;
> -
> -	while (old_chain) {
> -		tmp = bio_clone_kmalloc(old_chain, gfpmask);
> -		if (!tmp)
> -			goto err_out;
> -
> -		tmp->bi_bdev = NULL;
> -		gfpmask &= ~__GFP_DIRECT_RECLAIM;
> -		tmp->bi_next = NULL;
> -
> -		if (!new_chain)
> -			new_chain = tail = tmp;
> -		else {
> -			tail->bi_next = tmp;
> -			tail = tmp;
> -		}
> -
> -		old_chain = old_chain->bi_next;
> -	}
> -
> -	return new_chain;
> -
> -err_out:
> -	OSDBLK_DEBUG("bio_chain_clone with err\n");
> -	bio_chain_put(new_chain);
> -	return NULL;
> -}
> -
> -static void osdblk_rq_fn(struct request_queue *q)
> -{
> -	struct osdblk_device *osdev = q->queuedata;
> -
> -	while (1) {
> -		struct request *rq;
> -		struct osdblk_request *orq;
> -		struct osd_request *or;
> -		struct bio *bio;
> -		bool do_write, do_flush;
> -
> -		/* peek at request from block layer */
> -		rq = blk_fetch_request(q);
> -		if (!rq)
> -			break;
> -
> -		/* deduce our operation (read, write, flush) */
> -		/* I wish the block layer simplified cmd_type/cmd_flags/cmd[]
> -		 * into a clearly defined set of RPC commands:
> -		 * read, write, flush, scsi command, power mgmt req,
> -		 * driver-specific, etc.
> -		 */
> -
> -		do_flush = (req_op(rq) == REQ_OP_FLUSH);
> -		do_write = (rq_data_dir(rq) == WRITE);
> -
> -		if (!do_flush) { /* osd_flush does not use a bio */
> -			/* a bio clone to be passed down to OSD request */
> -			bio = bio_chain_clone(rq->bio, GFP_ATOMIC);
> -			if (!bio)
> -				break;
> -		} else
> -			bio = NULL;
> -
> -		/* alloc internal OSD request, for OSD command execution */
> -		or = osd_start_request(osdev->osd, GFP_ATOMIC);
> -		if (!or) {
> -			bio_chain_put(bio);
> -			OSDBLK_DEBUG("osd_start_request with err\n");
> -			break;
> -		}
> -
> -		orq = &osdev->req[rq->tag];
> -		orq->rq = rq;
> -		orq->bio = bio;
> -		orq->osdev = osdev;
> -
> -		/* init OSD command: flush, write or read */
> -		if (do_flush)
> -			osd_req_flush_object(or, &osdev->obj,
> -					     OSD_CDB_FLUSH_ALL, 0, 0);
> -		else if (do_write)
> -			osd_req_write(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
> -				      bio, blk_rq_bytes(rq));
> -		else
> -			osd_req_read(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
> -				     bio, blk_rq_bytes(rq));
> -
> -		OSDBLK_DEBUG("%s 0x%x bytes at 0x%llx\n",
> -			do_flush ? "flush" : do_write ?
> -				"write" : "read", blk_rq_bytes(rq),
> -			blk_rq_pos(rq) * 512ULL);
> -
> -		/* begin OSD command execution */
> -		if (osd_async_op(or, osdblk_osd_complete, orq,
> -				 osdev->obj_cred)) {
> -			osd_end_request(or);
> -			blk_requeue_request(q, rq);
> -			bio_chain_put(bio);
> -			OSDBLK_DEBUG("osd_execute_request_async with err\n");
> -			break;
> -		}
> -
> -		/* remove the special 'flush' marker, now that the command
> -		 * is executing
> -		 */
> -		rq->special = NULL;
> -	}
> -}
> -
> -static void osdblk_free_disk(struct osdblk_device *osdev)
> -{
> -	struct gendisk *disk = osdev->disk;
> -
> -	if (!disk)
> -		return;
> -
> -	if (disk->flags & GENHD_FL_UP)
> -		del_gendisk(disk);
> -	if (disk->queue)
> -		blk_cleanup_queue(disk->queue);
> -	put_disk(disk);
> -}
> -
> -static int osdblk_init_disk(struct osdblk_device *osdev)
> -{
> -	struct gendisk *disk;
> -	struct request_queue *q;
> -	int rc;
> -	u64 obj_size = 0;
> -
> -	/* contact OSD, request size info about the object being mapped */
> -	rc = osdblk_get_obj_size(osdev, &obj_size);
> -	if (rc)
> -		return rc;
> -
> -	/* create gendisk info */
> -	disk = alloc_disk(OSDBLK_MINORS_PER_MAJOR);
> -	if (!disk)
> -		return -ENOMEM;
> -
> -	sprintf(disk->disk_name, DRV_NAME "%d", osdev->id);
> -	disk->major = osdev->major;
> -	disk->first_minor = 0;
> -	disk->fops = &osdblk_bd_ops;
> -	disk->private_data = osdev;
> -
> -	/* init rq */
> -	q = blk_init_queue(osdblk_rq_fn, &osdev->lock);
> -	if (!q) {
> -		put_disk(disk);
> -		return -ENOMEM;
> -	}
> -
> -	/* switch queue to TCQ mode; allocate tag map */
> -	rc = blk_queue_init_tags(q, OSDBLK_MAX_REQ, NULL, BLK_TAG_ALLOC_FIFO);
> -	if (rc) {
> -		blk_cleanup_queue(q);
> -		put_disk(disk);
> -		return rc;
> -	}
> -
> -	/* Set our limits to the lower device limits, because osdblk cannot
> -	 * sleep when allocating a lower-request and therefore cannot be
> -	 * bouncing.
> -	 */
> -	blk_queue_stack_limits(q, osd_request_queue(osdev->osd));
> -
> -	blk_queue_prep_rq(q, blk_queue_start_tag);
> -	blk_queue_write_cache(q, true, false);
> -
> -	disk->queue = q;
> -
> -	q->queuedata = osdev;
> -
> -	osdev->disk = disk;
> -	osdev->q = q;
> -
> -	/* finally, announce the disk to the world */
> -	set_capacity(disk, obj_size / 512ULL);
> -	add_disk(disk);
> -
> -	printk(KERN_INFO "%s: Added of size 0x%llx\n",
> -		disk->disk_name, (unsigned long long)obj_size);
> -
> -	return 0;
> -}
> -
> -/********************************************************************
> - * /sys/class/osdblk/
> - *                   add	map OSD object to blkdev
> - *                   remove	unmap OSD object
> - *                   list	show mappings
> - *******************************************************************/
> -
> -static void class_osdblk_release(struct class *cls)
> -{
> -	kfree(cls);
> -}
> -
> -static ssize_t class_osdblk_list(struct class *c,
> -				struct class_attribute *attr,
> -				char *data)
> -{
> -	int n = 0;
> -	struct list_head *tmp;
> -
> -	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
> -
> -	list_for_each(tmp, &osdblkdev_list) {
> -		struct osdblk_device *osdev;
> -
> -		osdev = list_entry(tmp, struct osdblk_device, node);
> -
> -		n += sprintf(data+n, "%d %d %llu %llu %s\n",
> -			osdev->id,
> -			osdev->major,
> -			osdev->obj.partition,
> -			osdev->obj.id,
> -			osdev->osd_path);
> -	}
> -
> -	mutex_unlock(&ctl_mutex);
> -	return n;
> -}
> -
> -static ssize_t class_osdblk_add(struct class *c,
> -				struct class_attribute *attr,
> -				const char *buf, size_t count)
> -{
> -	struct osdblk_device *osdev;
> -	ssize_t rc;
> -	int irc, new_id = 0;
> -	struct list_head *tmp;
> -
> -	if (!try_module_get(THIS_MODULE))
> -		return -ENODEV;
> -
> -	/* new osdblk_device object */
> -	osdev = kzalloc(sizeof(*osdev) + strlen(buf) + 1, GFP_KERNEL);
> -	if (!osdev) {
> -		rc = -ENOMEM;
> -		goto err_out_mod;
> -	}
> -
> -	/* static osdblk_device initialization */
> -	spin_lock_init(&osdev->lock);
> -	INIT_LIST_HEAD(&osdev->node);
> -
> -	/* generate unique id: find highest unique id, add one */
> -
> -	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
> -
> -	list_for_each(tmp, &osdblkdev_list) {
> -		struct osdblk_device *osdev;
> -
> -		osdev = list_entry(tmp, struct osdblk_device, node);
> -		if (osdev->id > new_id)
> -			new_id = osdev->id + 1;
> -	}
> -
> -	osdev->id = new_id;
> -
> -	/* add to global list */
> -	list_add_tail(&osdev->node, &osdblkdev_list);
> -
> -	mutex_unlock(&ctl_mutex);
> -
> -	/* parse add command */
> -	if (sscanf(buf, "%llu %llu %s", &osdev->obj.partition, &osdev->obj.id,
> -		   osdev->osd_path) != 3) {
> -		rc = -EINVAL;
> -		goto err_out_slot;
> -	}
> -
> -	/* initialize rest of new object */
> -	sprintf(osdev->name, DRV_NAME "%d", osdev->id);
> -
> -	/* contact requested OSD */
> -	osdev->osd = osduld_path_lookup(osdev->osd_path);
> -	if (IS_ERR(osdev->osd)) {
> -		rc = PTR_ERR(osdev->osd);
> -		goto err_out_slot;
> -	}
> -
> -	/* build OSD credential */
> -	osdblk_make_credential(osdev->obj_cred, &osdev->obj);
> -
> -	/* register our block device */
> -	irc = register_blkdev(0, osdev->name);
> -	if (irc < 0) {
> -		rc = irc;
> -		goto err_out_osd;
> -	}
> -
> -	osdev->major = irc;
> -
> -	/* set up and announce blkdev mapping */
> -	rc = osdblk_init_disk(osdev);
> -	if (rc)
> -		goto err_out_blkdev;
> -
> -	return count;
> -
> -err_out_blkdev:
> -	unregister_blkdev(osdev->major, osdev->name);
> -err_out_osd:
> -	osduld_put_device(osdev->osd);
> -err_out_slot:
> -	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
> -	list_del_init(&osdev->node);
> -	mutex_unlock(&ctl_mutex);
> -
> -	kfree(osdev);
> -err_out_mod:
> -	OSDBLK_DEBUG("Error adding device %s\n", buf);
> -	module_put(THIS_MODULE);
> -	return rc;
> -}
> -
> -static ssize_t class_osdblk_remove(struct class *c,
> -					struct class_attribute *attr,
> -					const char *buf,
> -					size_t count)
> -{
> -	struct osdblk_device *osdev = NULL;
> -	int target_id, rc;
> -	unsigned long ul;
> -	struct list_head *tmp;
> -
> -	rc = kstrtoul(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);
> -
> -	list_for_each(tmp, &osdblkdev_list) {
> -		osdev = list_entry(tmp, struct osdblk_device, node);
> -		if (osdev->id == target_id) {
> -			list_del_init(&osdev->node);
> -			break;
> -		}
> -		osdev = NULL;
> -	}
> -
> -	mutex_unlock(&ctl_mutex);
> -
> -	if (!osdev)
> -		return -ENOENT;
> -
> -	/* clean up and free blkdev and associated OSD connection */
> -	osdblk_free_disk(osdev);
> -	unregister_blkdev(osdev->major, osdev->name);
> -	osduld_put_device(osdev->osd);
> -	kfree(osdev);
> -
> -	/* release module ref */
> -	module_put(THIS_MODULE);
> -
> -	return count;
> -}
> -
> -static struct class_attribute class_osdblk_attrs[] = {
> -	__ATTR(add,	0200, NULL, class_osdblk_add),
> -	__ATTR(remove,	0200, NULL, class_osdblk_remove),
> -	__ATTR(list,	0444, class_osdblk_list, NULL),
> -	__ATTR_NULL
> -};
> -
> -static int osdblk_sysfs_init(void)
> -{
> -	int ret = 0;
> -
> -	/*
> -	 * create control files in sysfs
> -	 * /sys/class/osdblk/...
> -	 */
> -	class_osdblk = kzalloc(sizeof(*class_osdblk), GFP_KERNEL);
> -	if (!class_osdblk)
> -		return -ENOMEM;
> -
> -	class_osdblk->name = DRV_NAME;
> -	class_osdblk->owner = THIS_MODULE;
> -	class_osdblk->class_release = class_osdblk_release;
> -	class_osdblk->class_attrs = class_osdblk_attrs;
> -
> -	ret = class_register(class_osdblk);
> -	if (ret) {
> -		kfree(class_osdblk);
> -		class_osdblk = NULL;
> -		printk(PFX "failed to create class osdblk\n");
> -		return ret;
> -	}
> -
> -	return 0;
> -}
> -
> -static void osdblk_sysfs_cleanup(void)
> -{
> -	if (class_osdblk)
> -		class_destroy(class_osdblk);
> -	class_osdblk = NULL;
> -}
> -
> -static int __init osdblk_init(void)
> -{
> -	int rc;
> -
> -	rc = osdblk_sysfs_init();
> -	if (rc)
> -		return rc;
> -
> -	return 0;
> -}
> -
> -static void __exit osdblk_exit(void)
> -{
> -	osdblk_sysfs_cleanup();
> -}
> -
> -module_init(osdblk_init);
> -module_exit(osdblk_exit);
> -
> 

[toc] | [prev] | [next] | [standalone]

#1623482

Christoph,

> The only real user of the T10 OSD protocol, the pNFS object layout
> driver never went to the point of having shipping products, and the
> other two users (osdblk and exofs) were simple example of it's usage.
>
> The code has been mostly unmaintained for years and is getting in the
> way of block / SCSI changes, so I think it's finally time to drop it.

No particular objections from me.

-- 
Martin K. Petersen	Oracle Linux Engineering

[toc] | [prev] | [next] | [standalone]

#1625434

On 04/12/2017 07:01 PM, Christoph Hellwig wrote:

Hi Sir Christoph

> The only real user of the T10 OSD protocol, the pNFS object layout
> driver never went to the point of having shipping products, and the
> other two users (osdblk and exofs) were simple example of it's usage.
> 

I understand why osdblk might be a pain, and was broken from day one, and
should by all means go away ASAP.

But exofs should not be bothering anyone, and as far as I know does
not use any special API's except the osd_uld code of course.

> The code has been mostly unmaintained for years and is getting in the
> way of block / SCSI changes, so I think it's finally time to drop it.
> 

Please tell me what are those changes you are talking about? I might be
able to help in conversion. I guess you mean osd_uld and the Upper SCSI API.
Just point me at a tree where osd_uld is broken, and perhaps with a little
guidance from you I can do a satisfactory conversion.

Is true that no new code went in for a long while, but I still from time
to time run a setup and test that the all stack, like iscsi-bidi and so on still
works.

That said, yes only a stand alone exofs was tested for a long time, a full
pnfs setup is missing any supporting server. So Yes I admit that pnfs-obj is
getting very rotten. And is most probably broken, on the pnfs side of things.
[Which I admit makes my little plea kind of mute ;-) ]

Every once in a while I get emails from Students basing all kind of interesting
experiments on top of the exofs and object base storage. So for the sake of academics
and for the sake of a true bidi-stack testing, might we want to evaluate what is the
up coming cost, and what is a minimum set we are willing to keep?

Please advise?

thanks
Boaz

> These patches are against Jens' block for-next tree as that already
> has various modifications of the SCSI code.
> 

[toc] | [prev] | [next] | [standalone]

#1625466

As one of those academics that Boaz was talking about, we do use the OSD driver for various research projects including support for OSD in parallel file systems like OrangeFS and Lustre.

John.

> On Apr 18, 2017, at 12:24 PM, Boaz Harrosh <ooo@electrozaur.com> wrote:
> 
> On 04/12/2017 07:01 PM, Christoph Hellwig wrote:
> 
> Hi Sir Christoph
> 
>> The only real user of the T10 OSD protocol, the pNFS object layout
>> driver never went to the point of having shipping products, and the
>> other two users (osdblk and exofs) were simple example of it's usage.
>> 
> 
> I understand why osdblk might be a pain, and was broken from day one, and
> should by all means go away ASAP.
> 
> But exofs should not be bothering anyone, and as far as I know does
> not use any special API's except the osd_uld code of course.
> 
>> The code has been mostly unmaintained for years and is getting in the
>> way of block / SCSI changes, so I think it's finally time to drop it.
>> 
> 
> Please tell me what are those changes you are talking about? I might be
> able to help in conversion. I guess you mean osd_uld and the Upper SCSI API.
> Just point me at a tree where osd_uld is broken, and perhaps with a little
> guidance from you I can do a satisfactory conversion.
> 
> Is true that no new code went in for a long while, but I still from time
> to time run a setup and test that the all stack, like iscsi-bidi and so on still
> works.
> 
> That said, yes only a stand alone exofs was tested for a long time, a full
> pnfs setup is missing any supporting server. So Yes I admit that pnfs-obj is
> getting very rotten. And is most probably broken, on the pnfs side of things.
> [Which I admit makes my little plea kind of mute ;-) ]
> 
> Every once in a while I get emails from Students basing all kind of interesting
> experiments on top of the exofs and object base storage. So for the sake of academics
> and for the sake of a true bidi-stack testing, might we want to evaluate what is the
> up coming cost, and what is a minimum set we are willing to keep?
> 
> Please advise?
> 
> thanks
> Boaz
> 
>> These patches are against Jens' block for-next tree as that already
>> has various modifications of the SCSI code.
>> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[toc] | [prev] | [next] | [standalone]

#1626346

On Wed, Apr 12 2017, Christoph Hellwig wrote:
> The only real user of the T10 OSD protocol, the pNFS object layout
> driver never went to the point of having shipping products, and the
> other two users (osdblk and exofs) were simple example of it's usage.
> 
> The code has been mostly unmaintained for years and is getting in the
> way of block / SCSI changes, so I think it's finally time to drop it.
> 
> These patches are against Jens' block for-next tree as that already

I've added 1/4 for now, as that one should obviously go. If we have
consensus to kill the rest, I'm all for it.

-- 
Jens Axboe

[toc] | [prev] | [next] | [standalone]

#1627785 — Re: [PATCH 3/4] nfs: remove the objlayout driver

On Wed, 2017-04-12 at 18:01 +0200, Christoph Hellwig wrote:
> The objlayout code has been in the tree, but it's been unmaintained
> and
> no server product for it actually ever shipped.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  Documentation/admin-guide/kernel-parameters.txt |   6 -
>  Documentation/filesystems/nfs/pnfs.txt          |  37 --
>  fs/nfs/Kconfig                                  |   5 -
>  fs/nfs/Makefile                                 |   1 -
>  fs/nfs/objlayout/Kbuild                         |   5 -
>  fs/nfs/objlayout/objio_osd.c                    | 675 --------------
> --------
>  fs/nfs/objlayout/objlayout.c                    | 706 --------------
> ----------
>  fs/nfs/objlayout/objlayout.h                    | 183 ------
>  fs/nfs/objlayout/pnfs_osd_xdr_cli.c             | 415 --------------
>  9 files changed, 2033 deletions(-)
>  delete mode 100644 fs/nfs/objlayout/Kbuild
>  delete mode 100644 fs/nfs/objlayout/objio_osd.c
>  delete mode 100644 fs/nfs/objlayout/objlayout.c
>  delete mode 100644 fs/nfs/objlayout/objlayout.h
>  delete mode 100644 fs/nfs/objlayout/pnfs_osd_xdr_cli.c
> 

OK. I'm applying this patch for the 4.12 merge window. If, as Boaz
suggests, there is still an interest in exofs, then I suggest we put
that to the test by moving it into the STAGING area, to see if someone
will step up to maintain it.

Cheers
  Trond

-- 
Trond Myklebust
Linux NFS client maintainer, PrimaryData
trond.myklebust@primarydata.com

[toc] | [prev] | [next] | [standalone]

#1628129 — Re: [PATCH 3/4] nfs: remove the objlayout driver

On Thu, Apr 20, 2017 at 08:18:28PM +0000, Trond Myklebust wrote:
> OK. I'm applying this patch for the 4.12 merge window. If, as Boaz
> suggests, there is still an interest in exofs, then I suggest we put
> that to the test by moving it into the STAGING area, to see if someone
> will step up to maintain it.

I think that would be a good idea, I'll can prepare a patch for the
net merge window.

[toc] | [prev] | [next] | [standalone]

#1628243 — Re: [PATCH 3/4] nfs: remove the objlayout driver

On 04/20/2017 11:18 PM, Trond Myklebust wrote:
<>
> 
> OK. I'm applying this patch for the 4.12 merge window. 

That is understandable this code was not tested for a long while

> If, as Boaz
> suggests, there is still an interest in exofs, then I suggest we put
> that to the test by moving it into the STAGING area, to see if someone
> will step up to maintain it.
> 

I do not understand what you mean by "someone will maintain it"
What is it that is missing in exofs that you see needs development?

As I said I regularly run and test this FS and update as things advance
there where no new fixtures for a while, but no regressions either.

Please explain what it means "will maintain it"

> Cheers
>   Trond
> 

Thanks
Boaz

[toc] | [prev] | [next] | [standalone]

#1628962 — Re: [PATCH 3/4] nfs: remove the objlayout driver

On 04/21/2017 05:00 PM, Trond Myklebust wrote:
> Maintenance is not development. It’s about doing all the followup
> _after_ the feature is declared to be developed. That’s been missing
> for quite some time in the case of the OSD pNFS code, which is why
> I’m not even bothering to consider staging. If you are saying you are
> still maintaining exofs, and want to continue doing so, then great,
> but note that there is a file called BUGS in that directory that has
> been untouched since 2008, and that’s why I thing staging is a good
> idea.
> 

No, the BUGS file is just stale. As you said was not ever updated. All
the bugs (1) in there do no longer exist for a long long time.
I will send a patch to remove the file.

Yes I maintain this fs. It has the complete fixture list, of what was
first intended. I keep running it every major kernel release to make
sure it actually runs, and there are no regressions.

If this FS stands in the way of any new development please anyone
contact me and I will help in the conversion. Of exofs and the
osd-uld.

Thanks
Boaz

[toc] | [prev] | [next] | [standalone]

#1628128 — Re: [PATCH 3/4] nfs: remove the objlayout driver

On Fri, Apr 21, 2017 at 09:27:11AM +0200, Valentin Rothberg wrote:
> Hi Christoph,
> 
> I just came across this patch in linux-next commit 6d22323b2e9f.
> scripts/checkkconfigsymbols.py reports that there is a leftover
> reference on PNFS_OBJLAYOUT in fs/exofs/Kconfig.ore line 10 (depends
> on EXOFS_FS || PNFS_OBJLAYOUT). I assume that this reference can
> safely be removed now? If you want to I can send a trivial patch
> removing it.

Yes, please do.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web