lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
/*
|
|
|
|
|
* Copyright (C) 2015 IT University of Copenhagen (rrpc.c)
|
|
|
|
|
* Copyright (C) 2016 CNEX Labs
|
|
|
|
|
* Initial release: Javier Gonzalez <javier@cnexlabs.com>
|
|
|
|
|
* Matias Bjorling <matias@cnexlabs.com>
|
|
|
|
|
*
|
|
|
|
|
* This program 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.
|
|
|
|
|
*
|
|
|
|
|
* 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.
|
|
|
|
|
*
|
|
|
|
|
* Implementation of a physical block-device target for Open-channel SSDs.
|
|
|
|
|
*
|
|
|
|
|
* pblk-init.c - pblk's initialization.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include "pblk.h"
|
|
|
|
|
|
2018-06-05 17:15:12 +08:00
|
|
|
static unsigned int write_buffer_size;
|
2018-06-01 22:41:10 +08:00
|
|
|
|
|
|
|
|
module_param(write_buffer_size, uint, 0644);
|
|
|
|
|
MODULE_PARM_DESC(write_buffer_size, "number of entries in a write buffer");
|
|
|
|
|
|
2017-10-13 20:46:07 +08:00
|
|
|
static struct kmem_cache *pblk_ws_cache, *pblk_rec_cache, *pblk_g_rq_cache,
|
2017-10-13 20:46:09 +08:00
|
|
|
*pblk_w_rq_cache;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
static DECLARE_RWSEM(pblk_lock);
|
2018-05-21 06:25:50 +08:00
|
|
|
struct bio_set pblk_bio_set;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
static int pblk_rw_io(struct request_queue *q, struct pblk *pblk,
|
|
|
|
|
struct bio *bio)
|
|
|
|
|
{
|
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
|
|
/* Read requests must be <= 256kb due to NVMe's 64 bit completion bitmap
|
|
|
|
|
* constraint. Writes can be of arbitrary size.
|
|
|
|
|
*/
|
|
|
|
|
if (bio_data_dir(bio) == READ) {
|
2017-06-18 12:38:57 +08:00
|
|
|
blk_queue_split(q, &bio);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
ret = pblk_submit_read(pblk, bio);
|
|
|
|
|
if (ret == NVM_IO_DONE && bio_flagged(bio, BIO_CLONED))
|
|
|
|
|
bio_put(bio);
|
|
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Prevent deadlock in the case of a modest LUN configuration and large
|
|
|
|
|
* user I/Os. Unless stalled, the rate limiter leaves at least 256KB
|
|
|
|
|
* available for user I/O.
|
|
|
|
|
*/
|
2017-10-13 20:46:05 +08:00
|
|
|
if (pblk_get_secs(bio) > pblk_rl_max_io(&pblk->rl))
|
2017-06-18 12:38:57 +08:00
|
|
|
blk_queue_split(q, &bio);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
return pblk_write_to_cache(pblk, bio, PBLK_IOTYPE_USER);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static blk_qc_t pblk_make_rq(struct request_queue *q, struct bio *bio)
|
|
|
|
|
{
|
|
|
|
|
struct pblk *pblk = q->queuedata;
|
|
|
|
|
|
|
|
|
|
if (bio_op(bio) == REQ_OP_DISCARD) {
|
|
|
|
|
pblk_discard(pblk, bio);
|
|
|
|
|
if (!(bio->bi_opf & REQ_PREFLUSH)) {
|
|
|
|
|
bio_endio(bio);
|
|
|
|
|
return BLK_QC_T_NONE;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
switch (pblk_rw_io(q, pblk, bio)) {
|
|
|
|
|
case NVM_IO_ERR:
|
|
|
|
|
bio_io_error(bio);
|
|
|
|
|
break;
|
|
|
|
|
case NVM_IO_DONE:
|
|
|
|
|
bio_endio(bio);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return BLK_QC_T_NONE;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-13 20:46:40 +08:00
|
|
|
static size_t pblk_trans_map_size(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
int entry_size = 8;
|
|
|
|
|
|
2018-03-30 06:05:19 +08:00
|
|
|
if (pblk->addrf_len < 32)
|
2017-10-13 20:46:40 +08:00
|
|
|
entry_size = 4;
|
|
|
|
|
|
|
|
|
|
return entry_size * pblk->rl.nr_secs;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_NVM_DEBUG
|
|
|
|
|
static u32 pblk_l2p_crc(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
size_t map_size;
|
|
|
|
|
u32 crc = ~(u32)0;
|
|
|
|
|
|
|
|
|
|
map_size = pblk_trans_map_size(pblk);
|
|
|
|
|
crc = crc32_le(crc, pblk->trans_map, map_size);
|
|
|
|
|
return crc;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
static void pblk_l2p_free(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
vfree(pblk->trans_map);
|
|
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
static int pblk_l2p_recover(struct pblk *pblk, bool factory_init)
|
|
|
|
|
{
|
|
|
|
|
struct pblk_line *line = NULL;
|
|
|
|
|
|
|
|
|
|
if (factory_init) {
|
|
|
|
|
pblk_setup_uuid(pblk);
|
|
|
|
|
} else {
|
|
|
|
|
line = pblk_recov_l2p(pblk);
|
|
|
|
|
if (IS_ERR(line)) {
|
|
|
|
|
pr_err("pblk: could not recover l2p table\n");
|
|
|
|
|
return -EFAULT;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_NVM_DEBUG
|
|
|
|
|
pr_info("pblk init: L2P CRC: %x\n", pblk_l2p_crc(pblk));
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
/* Free full lines directly as GC has not been started yet */
|
|
|
|
|
pblk_gc_free_full_lines(pblk);
|
|
|
|
|
|
|
|
|
|
if (!line) {
|
|
|
|
|
/* Configure next line for user data */
|
|
|
|
|
line = pblk_line_get_first_data(pblk);
|
2018-06-01 21:04:16 +08:00
|
|
|
if (!line)
|
2018-03-30 06:05:09 +08:00
|
|
|
return -EFAULT;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_l2p_init(struct pblk *pblk, bool factory_init)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
|
|
|
|
sector_t i;
|
|
|
|
|
struct ppa_addr ppa;
|
2017-10-13 20:46:40 +08:00
|
|
|
size_t map_size;
|
2018-06-01 21:04:16 +08:00
|
|
|
int ret = 0;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-10-13 20:46:40 +08:00
|
|
|
map_size = pblk_trans_map_size(pblk);
|
|
|
|
|
pblk->trans_map = vmalloc(map_size);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (!pblk->trans_map)
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
|
|
pblk_ppa_set_empty(&ppa);
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < pblk->rl.nr_secs; i++)
|
|
|
|
|
pblk_trans_map_set(pblk, i, ppa);
|
|
|
|
|
|
2018-06-01 21:04:16 +08:00
|
|
|
ret = pblk_l2p_recover(pblk, factory_init);
|
|
|
|
|
if (ret)
|
|
|
|
|
vfree(pblk->trans_map);
|
|
|
|
|
|
|
|
|
|
return ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_rwb_free(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
if (pblk_rb_tear_down_check(&pblk->rwb))
|
|
|
|
|
pr_err("pblk: write buffer error on tear down\n");
|
|
|
|
|
|
|
|
|
|
pblk_rb_data_free(&pblk->rwb);
|
|
|
|
|
vfree(pblk_rb_entries_ref(&pblk->rwb));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_rwb_init(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_rb_entry *entries;
|
2018-06-01 22:41:10 +08:00
|
|
|
unsigned long nr_entries, buffer_size;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
unsigned int power_size, power_seg_sz;
|
|
|
|
|
|
2018-06-01 22:41:10 +08:00
|
|
|
if (write_buffer_size && (write_buffer_size > pblk->pgs_in_buffer))
|
|
|
|
|
buffer_size = write_buffer_size;
|
|
|
|
|
else
|
|
|
|
|
buffer_size = pblk->pgs_in_buffer;
|
|
|
|
|
|
|
|
|
|
nr_entries = pblk_rb_calculate_size(buffer_size);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
entries = vzalloc(nr_entries * sizeof(struct pblk_rb_entry));
|
|
|
|
|
if (!entries)
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
|
|
power_size = get_count_order(nr_entries);
|
2018-03-30 06:05:10 +08:00
|
|
|
power_seg_sz = get_count_order(geo->csecs);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
return pblk_rb_init(&pblk->rwb, entries, power_size, power_seg_sz);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Minimum pages needed within a lun */
|
|
|
|
|
#define ADDR_POOL_SIZE 64
|
|
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
static int pblk_set_addrf_12(struct nvm_geo *geo, struct nvm_addrf_12 *dst)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
2018-03-30 06:05:10 +08:00
|
|
|
struct nvm_addrf_12 *src = (struct nvm_addrf_12 *)&geo->addrf;
|
|
|
|
|
int power_len;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
/* Re-calculate channel and lun format to adapt to configuration */
|
2018-03-30 06:05:14 +08:00
|
|
|
power_len = get_count_order(geo->num_ch);
|
|
|
|
|
if (1 << power_len != geo->num_ch) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
pr_err("pblk: supports only power-of-two channel config.\n");
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
2018-03-30 06:05:10 +08:00
|
|
|
dst->ch_len = power_len;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:14 +08:00
|
|
|
power_len = get_count_order(geo->num_lun);
|
|
|
|
|
if (1 << power_len != geo->num_lun) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
pr_err("pblk: supports only power-of-two LUN config.\n");
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
2018-03-30 06:05:10 +08:00
|
|
|
dst->lun_len = power_len;
|
|
|
|
|
|
|
|
|
|
dst->blk_len = src->blk_len;
|
|
|
|
|
dst->pg_len = src->pg_len;
|
|
|
|
|
dst->pln_len = src->pln_len;
|
2018-03-30 06:05:14 +08:00
|
|
|
dst->sec_len = src->sec_len;
|
2018-03-30 06:05:10 +08:00
|
|
|
|
2018-03-30 06:05:14 +08:00
|
|
|
dst->sec_offset = 0;
|
|
|
|
|
dst->pln_offset = dst->sec_len;
|
2018-03-30 06:05:10 +08:00
|
|
|
dst->ch_offset = dst->pln_offset + dst->pln_len;
|
|
|
|
|
dst->lun_offset = dst->ch_offset + dst->ch_len;
|
|
|
|
|
dst->pg_offset = dst->lun_offset + dst->lun_len;
|
|
|
|
|
dst->blk_offset = dst->pg_offset + dst->pg_len;
|
|
|
|
|
|
2018-03-30 06:05:14 +08:00
|
|
|
dst->sec_mask = ((1ULL << dst->sec_len) - 1) << dst->sec_offset;
|
2018-03-30 06:05:10 +08:00
|
|
|
dst->pln_mask = ((1ULL << dst->pln_len) - 1) << dst->pln_offset;
|
|
|
|
|
dst->ch_mask = ((1ULL << dst->ch_len) - 1) << dst->ch_offset;
|
|
|
|
|
dst->lun_mask = ((1ULL << dst->lun_len) - 1) << dst->lun_offset;
|
|
|
|
|
dst->pg_mask = ((1ULL << dst->pg_len) - 1) << dst->pg_offset;
|
|
|
|
|
dst->blk_mask = ((1ULL << dst->blk_len) - 1) << dst->blk_offset;
|
|
|
|
|
|
|
|
|
|
return dst->blk_offset + src->blk_len;
|
|
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:21 +08:00
|
|
|
static int pblk_set_addrf_20(struct nvm_geo *geo, struct nvm_addrf *adst,
|
|
|
|
|
struct pblk_addrf *udst)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_addrf *src = &geo->addrf;
|
|
|
|
|
|
|
|
|
|
adst->ch_len = get_count_order(geo->num_ch);
|
|
|
|
|
adst->lun_len = get_count_order(geo->num_lun);
|
|
|
|
|
adst->chk_len = src->chk_len;
|
|
|
|
|
adst->sec_len = src->sec_len;
|
|
|
|
|
|
|
|
|
|
adst->sec_offset = 0;
|
|
|
|
|
adst->ch_offset = adst->sec_len;
|
|
|
|
|
adst->lun_offset = adst->ch_offset + adst->ch_len;
|
|
|
|
|
adst->chk_offset = adst->lun_offset + adst->lun_len;
|
|
|
|
|
|
|
|
|
|
adst->sec_mask = ((1ULL << adst->sec_len) - 1) << adst->sec_offset;
|
|
|
|
|
adst->chk_mask = ((1ULL << adst->chk_len) - 1) << adst->chk_offset;
|
|
|
|
|
adst->lun_mask = ((1ULL << adst->lun_len) - 1) << adst->lun_offset;
|
|
|
|
|
adst->ch_mask = ((1ULL << adst->ch_len) - 1) << adst->ch_offset;
|
|
|
|
|
|
|
|
|
|
udst->sec_stripe = geo->ws_opt;
|
|
|
|
|
udst->ch_stripe = geo->num_ch;
|
|
|
|
|
udst->lun_stripe = geo->num_lun;
|
|
|
|
|
|
|
|
|
|
udst->sec_lun_stripe = udst->sec_stripe * udst->ch_stripe;
|
|
|
|
|
udst->sec_ws_stripe = udst->sec_lun_stripe * udst->lun_stripe;
|
|
|
|
|
|
|
|
|
|
return adst->chk_offset + adst->chk_len;
|
|
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:19 +08:00
|
|
|
static int pblk_set_addrf(struct pblk *pblk)
|
2018-03-30 06:05:10 +08:00
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
int mod;
|
|
|
|
|
|
2018-03-30 06:05:21 +08:00
|
|
|
switch (geo->version) {
|
|
|
|
|
case NVM_OCSSD_SPEC_12:
|
|
|
|
|
div_u64_rem(geo->clba, pblk->min_write_pgs, &mod);
|
|
|
|
|
if (mod) {
|
|
|
|
|
pr_err("pblk: bad configuration of sectors/pages\n");
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
pblk->addrf_len = pblk_set_addrf_12(geo, (void *)&pblk->addrf);
|
|
|
|
|
break;
|
|
|
|
|
case NVM_OCSSD_SPEC_20:
|
|
|
|
|
pblk->addrf_len = pblk_set_addrf_20(geo, (void *)&pblk->addrf,
|
|
|
|
|
&pblk->uaddrf);
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
pr_err("pblk: OCSSD revision not supported (%d)\n",
|
|
|
|
|
geo->version);
|
2018-03-30 06:05:10 +08:00
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_init_global_caches(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
down_write(&pblk_lock);
|
2017-10-13 20:46:07 +08:00
|
|
|
pblk_ws_cache = kmem_cache_create("pblk_blk_ws",
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
sizeof(struct pblk_line_ws), 0, 0, NULL);
|
2017-10-13 20:46:07 +08:00
|
|
|
if (!pblk_ws_cache) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
pblk_rec_cache = kmem_cache_create("pblk_rec",
|
|
|
|
|
sizeof(struct pblk_rec_ctx), 0, 0, NULL);
|
|
|
|
|
if (!pblk_rec_cache) {
|
2017-10-13 20:46:07 +08:00
|
|
|
kmem_cache_destroy(pblk_ws_cache);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
2017-06-27 06:27:13 +08:00
|
|
|
pblk_g_rq_cache = kmem_cache_create("pblk_g_rq", pblk_g_rq_size,
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
0, 0, NULL);
|
2017-06-27 06:27:13 +08:00
|
|
|
if (!pblk_g_rq_cache) {
|
2017-10-13 20:46:07 +08:00
|
|
|
kmem_cache_destroy(pblk_ws_cache);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
kmem_cache_destroy(pblk_rec_cache);
|
|
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
pblk_w_rq_cache = kmem_cache_create("pblk_w_rq", pblk_w_rq_size,
|
|
|
|
|
0, 0, NULL);
|
|
|
|
|
if (!pblk_w_rq_cache) {
|
2017-10-13 20:46:07 +08:00
|
|
|
kmem_cache_destroy(pblk_ws_cache);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
kmem_cache_destroy(pblk_rec_cache);
|
2017-06-27 06:27:13 +08:00
|
|
|
kmem_cache_destroy(pblk_g_rq_cache);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-13 20:46:33 +08:00
|
|
|
static void pblk_free_global_caches(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
kmem_cache_destroy(pblk_ws_cache);
|
|
|
|
|
kmem_cache_destroy(pblk_rec_cache);
|
|
|
|
|
kmem_cache_destroy(pblk_g_rq_cache);
|
|
|
|
|
kmem_cache_destroy(pblk_w_rq_cache);
|
|
|
|
|
}
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
static int pblk_core_init(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
2018-05-21 06:25:50 +08:00
|
|
|
int ret, max_write_ppas;
|
2018-03-30 06:05:09 +08:00
|
|
|
|
|
|
|
|
atomic64_set(&pblk->user_wa, 0);
|
|
|
|
|
atomic64_set(&pblk->pad_wa, 0);
|
|
|
|
|
atomic64_set(&pblk->gc_wa, 0);
|
|
|
|
|
pblk->user_rst_wa = 0;
|
|
|
|
|
pblk->pad_rst_wa = 0;
|
|
|
|
|
pblk->gc_rst_wa = 0;
|
|
|
|
|
|
|
|
|
|
atomic64_set(&pblk->nr_flush, 0);
|
|
|
|
|
pblk->nr_flush_rst = 0;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
pblk->pgs_in_buffer = geo->mw_cunits * geo->all_luns;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
pblk->min_write_pgs = geo->ws_opt * (geo->csecs / PAGE_SIZE);
|
2018-03-30 06:05:09 +08:00
|
|
|
max_write_ppas = pblk->min_write_pgs * geo->all_luns;
|
|
|
|
|
pblk->max_write_pgs = min_t(int, max_write_ppas, NVM_MAX_VLBA);
|
|
|
|
|
pblk_set_sec_per_write(pblk, pblk->min_write_pgs);
|
|
|
|
|
|
|
|
|
|
if (pblk->max_write_pgs > PBLK_MAX_REQ_ADDRS) {
|
|
|
|
|
pr_err("pblk: vector list too big(%u > %u)\n",
|
|
|
|
|
pblk->max_write_pgs, PBLK_MAX_REQ_ADDRS);
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
treewide: kzalloc() -> kcalloc()
The kzalloc() function has a 2-factor argument form, kcalloc(). This
patch replaces cases of:
kzalloc(a * b, gfp)
with:
kcalloc(a * b, gfp)
as well as handling cases of:
kzalloc(a * b * c, gfp)
with:
kzalloc(array3_size(a, b, c), gfp)
as it's slightly less ugly than:
kzalloc_array(array_size(a, b), c, gfp)
This does, however, attempt to ignore constant size factors like:
kzalloc(4 * 1024, gfp)
though any constants defined via macros get caught up in the conversion.
Any factors with a sizeof() of "unsigned char", "char", and "u8" were
dropped, since they're redundant.
The Coccinelle script used for this was:
// Fix redundant parens around sizeof().
@@
type TYPE;
expression THING, E;
@@
(
kzalloc(
- (sizeof(TYPE)) * E
+ sizeof(TYPE) * E
, ...)
|
kzalloc(
- (sizeof(THING)) * E
+ sizeof(THING) * E
, ...)
)
// Drop single-byte sizes and redundant parens.
@@
expression COUNT;
typedef u8;
typedef __u8;
@@
(
kzalloc(
- sizeof(u8) * (COUNT)
+ COUNT
, ...)
|
kzalloc(
- sizeof(__u8) * (COUNT)
+ COUNT
, ...)
|
kzalloc(
- sizeof(char) * (COUNT)
+ COUNT
, ...)
|
kzalloc(
- sizeof(unsigned char) * (COUNT)
+ COUNT
, ...)
|
kzalloc(
- sizeof(u8) * COUNT
+ COUNT
, ...)
|
kzalloc(
- sizeof(__u8) * COUNT
+ COUNT
, ...)
|
kzalloc(
- sizeof(char) * COUNT
+ COUNT
, ...)
|
kzalloc(
- sizeof(unsigned char) * COUNT
+ COUNT
, ...)
)
// 2-factor product with sizeof(type/expression) and identifier or constant.
@@
type TYPE;
expression THING;
identifier COUNT_ID;
constant COUNT_CONST;
@@
(
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * (COUNT_ID)
+ COUNT_ID, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * COUNT_ID
+ COUNT_ID, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * (COUNT_CONST)
+ COUNT_CONST, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * COUNT_CONST
+ COUNT_CONST, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * (COUNT_ID)
+ COUNT_ID, sizeof(THING)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * COUNT_ID
+ COUNT_ID, sizeof(THING)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * (COUNT_CONST)
+ COUNT_CONST, sizeof(THING)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * COUNT_CONST
+ COUNT_CONST, sizeof(THING)
, ...)
)
// 2-factor product, only identifiers.
@@
identifier SIZE, COUNT;
@@
- kzalloc
+ kcalloc
(
- SIZE * COUNT
+ COUNT, SIZE
, ...)
// 3-factor product with 1 sizeof(type) or sizeof(expression), with
// redundant parens removed.
@@
expression THING;
identifier STRIDE, COUNT;
type TYPE;
@@
(
kzalloc(
- sizeof(TYPE) * (COUNT) * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kzalloc(
- sizeof(TYPE) * (COUNT) * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kzalloc(
- sizeof(TYPE) * COUNT * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kzalloc(
- sizeof(TYPE) * COUNT * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kzalloc(
- sizeof(THING) * (COUNT) * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kzalloc(
- sizeof(THING) * (COUNT) * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kzalloc(
- sizeof(THING) * COUNT * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kzalloc(
- sizeof(THING) * COUNT * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
)
// 3-factor product with 2 sizeof(variable), with redundant parens removed.
@@
expression THING1, THING2;
identifier COUNT;
type TYPE1, TYPE2;
@@
(
kzalloc(
- sizeof(TYPE1) * sizeof(TYPE2) * COUNT
+ array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2))
, ...)
|
kzalloc(
- sizeof(TYPE1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2))
, ...)
|
kzalloc(
- sizeof(THING1) * sizeof(THING2) * COUNT
+ array3_size(COUNT, sizeof(THING1), sizeof(THING2))
, ...)
|
kzalloc(
- sizeof(THING1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(THING1), sizeof(THING2))
, ...)
|
kzalloc(
- sizeof(TYPE1) * sizeof(THING2) * COUNT
+ array3_size(COUNT, sizeof(TYPE1), sizeof(THING2))
, ...)
|
kzalloc(
- sizeof(TYPE1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(TYPE1), sizeof(THING2))
, ...)
)
// 3-factor product, only identifiers, with redundant parens removed.
@@
identifier STRIDE, SIZE, COUNT;
@@
(
kzalloc(
- (COUNT) * STRIDE * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- COUNT * (STRIDE) * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- COUNT * STRIDE * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- (COUNT) * (STRIDE) * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- COUNT * (STRIDE) * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- (COUNT) * STRIDE * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- (COUNT) * (STRIDE) * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kzalloc(
- COUNT * STRIDE * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
)
// Any remaining multi-factor products, first at least 3-factor products,
// when they're not all constants...
@@
expression E1, E2, E3;
constant C1, C2, C3;
@@
(
kzalloc(C1 * C2 * C3, ...)
|
kzalloc(
- (E1) * E2 * E3
+ array3_size(E1, E2, E3)
, ...)
|
kzalloc(
- (E1) * (E2) * E3
+ array3_size(E1, E2, E3)
, ...)
|
kzalloc(
- (E1) * (E2) * (E3)
+ array3_size(E1, E2, E3)
, ...)
|
kzalloc(
- E1 * E2 * E3
+ array3_size(E1, E2, E3)
, ...)
)
// And then all remaining 2 factors products when they're not all constants,
// keeping sizeof() as the second factor argument.
@@
expression THING, E1, E2;
type TYPE;
constant C1, C2, C3;
@@
(
kzalloc(sizeof(THING) * C2, ...)
|
kzalloc(sizeof(TYPE) * C2, ...)
|
kzalloc(C1 * C2 * C3, ...)
|
kzalloc(C1 * C2, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * (E2)
+ E2, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(TYPE) * E2
+ E2, sizeof(TYPE)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * (E2)
+ E2, sizeof(THING)
, ...)
|
- kzalloc
+ kcalloc
(
- sizeof(THING) * E2
+ E2, sizeof(THING)
, ...)
|
- kzalloc
+ kcalloc
(
- (E1) * E2
+ E1, E2
, ...)
|
- kzalloc
+ kcalloc
(
- (E1) * (E2)
+ E1, E2
, ...)
|
- kzalloc
+ kcalloc
(
- E1 * E2
+ E1, E2
, ...)
)
Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 05:03:40 +08:00
|
|
|
pblk->pad_dist = kcalloc(pblk->min_write_pgs - 1, sizeof(atomic64_t),
|
2018-03-30 06:05:09 +08:00
|
|
|
GFP_KERNEL);
|
|
|
|
|
if (!pblk->pad_dist)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
if (pblk_init_global_caches(pblk))
|
|
|
|
|
goto fail_free_pad_dist;
|
|
|
|
|
|
2017-10-13 20:46:07 +08:00
|
|
|
/* Internal bios can be at most the sectors signaled by the device. */
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_page_pool(&pblk->page_bio_pool, NVM_MAX_VLBA, 0);
|
|
|
|
|
if (ret)
|
2017-10-13 20:46:33 +08:00
|
|
|
goto free_global_caches;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_slab_pool(&pblk->gen_ws_pool, PBLK_GEN_WS_POOL_SIZE,
|
|
|
|
|
pblk_ws_cache);
|
|
|
|
|
if (ret)
|
2017-10-13 20:46:06 +08:00
|
|
|
goto free_page_bio_pool;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_slab_pool(&pblk->rec_pool, geo->all_luns,
|
|
|
|
|
pblk_rec_cache);
|
|
|
|
|
if (ret)
|
2017-10-13 20:46:07 +08:00
|
|
|
goto free_gen_ws_pool;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_slab_pool(&pblk->r_rq_pool, geo->all_luns,
|
|
|
|
|
pblk_g_rq_cache);
|
|
|
|
|
if (ret)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
goto free_rec_pool;
|
|
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_slab_pool(&pblk->e_rq_pool, geo->all_luns,
|
|
|
|
|
pblk_g_rq_cache);
|
|
|
|
|
if (ret)
|
2017-10-13 20:46:08 +08:00
|
|
|
goto free_r_rq_pool;
|
|
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = mempool_init_slab_pool(&pblk->w_rq_pool, geo->all_luns,
|
|
|
|
|
pblk_w_rq_cache);
|
|
|
|
|
if (ret)
|
2017-10-13 20:46:08 +08:00
|
|
|
goto free_e_rq_pool;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-06-26 17:57:28 +08:00
|
|
|
pblk->close_wq = alloc_workqueue("pblk-close-wq",
|
|
|
|
|
WQ_MEM_RECLAIM | WQ_UNBOUND, PBLK_NR_CLOSE_JOBS);
|
|
|
|
|
if (!pblk->close_wq)
|
2017-10-13 20:46:09 +08:00
|
|
|
goto free_w_rq_pool;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-06-26 17:57:28 +08:00
|
|
|
pblk->bb_wq = alloc_workqueue("pblk-bb-wq",
|
|
|
|
|
WQ_MEM_RECLAIM | WQ_UNBOUND, 0);
|
|
|
|
|
if (!pblk->bb_wq)
|
|
|
|
|
goto free_close_wq;
|
|
|
|
|
|
2017-10-13 20:46:23 +08:00
|
|
|
pblk->r_end_wq = alloc_workqueue("pblk-read-end-wq",
|
|
|
|
|
WQ_MEM_RECLAIM | WQ_UNBOUND, 0);
|
|
|
|
|
if (!pblk->r_end_wq)
|
2017-06-26 17:57:28 +08:00
|
|
|
goto free_bb_wq;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:19 +08:00
|
|
|
if (pblk_set_addrf(pblk))
|
2017-10-13 20:46:23 +08:00
|
|
|
goto free_r_end_wq;
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
INIT_LIST_HEAD(&pblk->compl_list);
|
2018-06-01 22:41:05 +08:00
|
|
|
INIT_LIST_HEAD(&pblk->resubmit_list);
|
2018-03-30 06:05:09 +08:00
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return 0;
|
|
|
|
|
|
2017-10-13 20:46:23 +08:00
|
|
|
free_r_end_wq:
|
|
|
|
|
destroy_workqueue(pblk->r_end_wq);
|
2017-06-26 17:57:28 +08:00
|
|
|
free_bb_wq:
|
|
|
|
|
destroy_workqueue(pblk->bb_wq);
|
|
|
|
|
free_close_wq:
|
|
|
|
|
destroy_workqueue(pblk->close_wq);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
free_w_rq_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->w_rq_pool);
|
2017-10-13 20:46:08 +08:00
|
|
|
free_e_rq_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->e_rq_pool);
|
2017-10-13 20:46:08 +08:00
|
|
|
free_r_rq_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->r_rq_pool);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
free_rec_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->rec_pool);
|
2017-10-13 20:46:07 +08:00
|
|
|
free_gen_ws_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->gen_ws_pool);
|
2017-10-13 20:46:06 +08:00
|
|
|
free_page_bio_pool:
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->page_bio_pool);
|
2017-10-13 20:46:33 +08:00
|
|
|
free_global_caches:
|
|
|
|
|
pblk_free_global_caches(pblk);
|
2018-03-30 06:05:09 +08:00
|
|
|
fail_free_pad_dist:
|
|
|
|
|
kfree(pblk->pad_dist);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_core_free(struct pblk *pblk)
|
|
|
|
|
{
|
2017-06-26 17:57:28 +08:00
|
|
|
if (pblk->close_wq)
|
|
|
|
|
destroy_workqueue(pblk->close_wq);
|
|
|
|
|
|
2017-10-13 20:46:23 +08:00
|
|
|
if (pblk->r_end_wq)
|
|
|
|
|
destroy_workqueue(pblk->r_end_wq);
|
|
|
|
|
|
2017-06-26 17:57:28 +08:00
|
|
|
if (pblk->bb_wq)
|
|
|
|
|
destroy_workqueue(pblk->bb_wq);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
mempool_exit(&pblk->page_bio_pool);
|
|
|
|
|
mempool_exit(&pblk->gen_ws_pool);
|
|
|
|
|
mempool_exit(&pblk->rec_pool);
|
|
|
|
|
mempool_exit(&pblk->r_rq_pool);
|
|
|
|
|
mempool_exit(&pblk->e_rq_pool);
|
|
|
|
|
mempool_exit(&pblk->w_rq_pool);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-10-13 20:46:33 +08:00
|
|
|
pblk_free_global_caches(pblk);
|
2018-03-30 06:05:09 +08:00
|
|
|
kfree(pblk->pad_dist);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:04:59 +08:00
|
|
|
static void pblk_line_mg_free(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
kfree(l_mg->bb_template);
|
|
|
|
|
kfree(l_mg->bb_aux);
|
|
|
|
|
kfree(l_mg->vsc_list);
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < PBLK_DATA_LINES; i++) {
|
|
|
|
|
kfree(l_mg->sline_meta[i]);
|
|
|
|
|
pblk_mfree(l_mg->eline_meta[i]->buf, l_mg->emeta_alloc_type);
|
|
|
|
|
kfree(l_mg->eline_meta[i]);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2018-06-01 22:41:06 +08:00
|
|
|
static void pblk_line_meta_free(struct pblk_line_mgmt *l_mg,
|
|
|
|
|
struct pblk_line *line)
|
2017-06-26 17:57:22 +08:00
|
|
|
{
|
2018-06-01 22:41:06 +08:00
|
|
|
struct pblk_w_err_gc *w_err_gc = line->w_err_gc;
|
|
|
|
|
|
2017-06-26 17:57:22 +08:00
|
|
|
kfree(line->blk_bitmap);
|
|
|
|
|
kfree(line->erase_bitmap);
|
2018-03-30 06:05:20 +08:00
|
|
|
kfree(line->chks);
|
2018-06-01 22:41:06 +08:00
|
|
|
|
|
|
|
|
pblk_mfree(w_err_gc->lba_list, l_mg->emeta_alloc_type);
|
|
|
|
|
kfree(w_err_gc);
|
2017-06-26 17:57:22 +08:00
|
|
|
}
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
static void pblk_lines_free(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct pblk_line *line;
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
spin_lock(&l_mg->free_lock);
|
|
|
|
|
for (i = 0; i < l_mg->nr_lines; i++) {
|
|
|
|
|
line = &pblk->lines[i];
|
|
|
|
|
|
2018-06-01 21:04:22 +08:00
|
|
|
pblk_line_free(line);
|
2018-06-01 22:41:06 +08:00
|
|
|
pblk_line_meta_free(l_mg, line);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
spin_unlock(&l_mg->free_lock);
|
2018-03-30 06:05:09 +08:00
|
|
|
|
|
|
|
|
pblk_line_mg_free(pblk);
|
|
|
|
|
|
|
|
|
|
kfree(pblk->luns);
|
|
|
|
|
kfree(pblk->lines);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:04:59 +08:00
|
|
|
static int pblk_bb_get_tbl(struct nvm_tgt_dev *dev, struct pblk_lun *rlun,
|
|
|
|
|
u8 *blks, int nr_blks)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
|
|
|
|
struct ppa_addr ppa;
|
2018-03-30 06:04:59 +08:00
|
|
|
int ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
ppa.ppa = 0;
|
|
|
|
|
ppa.g.ch = rlun->bppa.g.ch;
|
|
|
|
|
ppa.g.lun = rlun->bppa.g.lun;
|
|
|
|
|
|
|
|
|
|
ret = nvm_get_tgt_bb_tbl(dev, ppa, blks);
|
|
|
|
|
if (ret)
|
2018-03-30 06:04:59 +08:00
|
|
|
return ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
nr_blks = nvm_bb_tbl_fold(dev->parent, blks, nr_blks);
|
2018-03-30 06:04:59 +08:00
|
|
|
if (nr_blks < 0)
|
|
|
|
|
return -EIO;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-04-26 00:15:30 +08:00
|
|
|
return 0;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
static void *pblk_bb_get_meta(struct pblk *pblk)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
2017-06-26 17:57:22 +08:00
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
2018-03-30 06:05:20 +08:00
|
|
|
u8 *meta;
|
2018-03-30 06:04:59 +08:00
|
|
|
int i, nr_blks, blk_per_lun;
|
|
|
|
|
int ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:14 +08:00
|
|
|
blk_per_lun = geo->num_chk * geo->pln_mode;
|
2018-03-30 06:04:59 +08:00
|
|
|
nr_blks = blk_per_lun * geo->all_luns;
|
2017-06-26 17:57:22 +08:00
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
meta = kmalloc(nr_blks, GFP_KERNEL);
|
|
|
|
|
if (!meta)
|
2018-03-30 06:04:59 +08:00
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < geo->all_luns; i++) {
|
|
|
|
|
struct pblk_lun *rlun = &pblk->luns[i];
|
2018-03-30 06:05:20 +08:00
|
|
|
u8 *meta_pos = meta + i * blk_per_lun;
|
2018-03-30 06:04:59 +08:00
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
ret = pblk_bb_get_tbl(dev, rlun, meta_pos, blk_per_lun);
|
2018-03-30 06:04:59 +08:00
|
|
|
if (ret) {
|
2018-03-30 06:05:20 +08:00
|
|
|
kfree(meta);
|
2018-03-30 06:04:59 +08:00
|
|
|
return ERR_PTR(-EIO);
|
|
|
|
|
}
|
2017-06-26 17:57:22 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
return meta;
|
2017-06-26 17:57:22 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
static void *pblk_chunk_get_meta(struct pblk *pblk)
|
2017-06-26 17:57:22 +08:00
|
|
|
{
|
2018-03-30 06:04:59 +08:00
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
2017-06-26 17:57:22 +08:00
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
if (geo->version == NVM_OCSSD_SPEC_12)
|
|
|
|
|
return pblk_bb_get_meta(pblk);
|
|
|
|
|
else
|
|
|
|
|
return pblk_chunk_get_info(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
static int pblk_luns_init(struct pblk *pblk)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_lun *rlun;
|
2018-03-30 06:04:59 +08:00
|
|
|
int i;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
/* TODO: Implement unbalanced LUN support */
|
2018-03-30 06:05:14 +08:00
|
|
|
if (geo->num_lun < 0) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
pr_err("pblk: unbalanced LUN config.\n");
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
2018-01-05 21:16:03 +08:00
|
|
|
pblk->luns = kcalloc(geo->all_luns, sizeof(struct pblk_lun),
|
|
|
|
|
GFP_KERNEL);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (!pblk->luns)
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
2018-01-05 21:16:03 +08:00
|
|
|
for (i = 0; i < geo->all_luns; i++) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
/* Stripe across channels */
|
2018-03-30 06:05:14 +08:00
|
|
|
int ch = i % geo->num_ch;
|
|
|
|
|
int lun_raw = i / geo->num_ch;
|
|
|
|
|
int lunid = lun_raw + ch * geo->num_lun;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
rlun = &pblk->luns[i];
|
2018-03-30 06:05:09 +08:00
|
|
|
rlun->bppa = dev->luns[lunid];
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
sema_init(&rlun->wr_sem, 1);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* See comment over struct line_emeta definition */
|
2017-06-26 17:57:17 +08:00
|
|
|
static unsigned int calc_emeta_len(struct pblk *pblk)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
2017-06-26 17:57:17 +08:00
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
|
|
|
|
|
/* Round to sector size so that lba_list starts on its own sector */
|
|
|
|
|
lm->emeta_sec[1] = DIV_ROUND_UP(
|
2018-03-30 06:04:52 +08:00
|
|
|
sizeof(struct line_emeta) + lm->blk_bitmap_len +
|
2018-03-30 06:05:10 +08:00
|
|
|
sizeof(struct wa_counters), geo->csecs);
|
|
|
|
|
lm->emeta_len[1] = lm->emeta_sec[1] * geo->csecs;
|
2017-06-26 17:57:17 +08:00
|
|
|
|
|
|
|
|
/* Round to sector size so that vsc_list starts on its own sector */
|
|
|
|
|
lm->dsec_per_line = lm->sec_per_line - lm->emeta_sec[0];
|
|
|
|
|
lm->emeta_sec[2] = DIV_ROUND_UP(lm->dsec_per_line * sizeof(u64),
|
2018-03-30 06:05:10 +08:00
|
|
|
geo->csecs);
|
|
|
|
|
lm->emeta_len[2] = lm->emeta_sec[2] * geo->csecs;
|
2017-06-26 17:57:17 +08:00
|
|
|
|
|
|
|
|
lm->emeta_sec[3] = DIV_ROUND_UP(l_mg->nr_lines * sizeof(u32),
|
2018-03-30 06:05:10 +08:00
|
|
|
geo->csecs);
|
|
|
|
|
lm->emeta_len[3] = lm->emeta_sec[3] * geo->csecs;
|
2017-06-26 17:57:17 +08:00
|
|
|
|
|
|
|
|
lm->vsc_list_len = l_mg->nr_lines * sizeof(u32);
|
|
|
|
|
|
|
|
|
|
return (lm->emeta_len[1] + lm->emeta_len[2] + lm->emeta_len[3]);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_set_provision(struct pblk *pblk, long nr_free_blks)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
2018-01-05 21:16:13 +08:00
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
sector_t provisioned;
|
2018-01-05 21:16:13 +08:00
|
|
|
int sec_meta, blk_meta;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-01-05 21:16:14 +08:00
|
|
|
if (geo->op == NVM_TARGET_DEFAULT_OP)
|
|
|
|
|
pblk->op = PBLK_DEFAULT_OP;
|
|
|
|
|
else
|
|
|
|
|
pblk->op = geo->op;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
provisioned = nr_free_blks;
|
2018-01-05 21:16:13 +08:00
|
|
|
provisioned *= (100 - pblk->op);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
sector_div(provisioned, 100);
|
|
|
|
|
|
2018-01-05 21:16:13 +08:00
|
|
|
pblk->op_blks = nr_free_blks - provisioned;
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
/* Internally pblk manages all free blocks, but all calculations based
|
|
|
|
|
* on user capacity consider only provisioned blocks
|
|
|
|
|
*/
|
|
|
|
|
pblk->rl.total_blocks = nr_free_blks;
|
2018-03-30 06:05:10 +08:00
|
|
|
pblk->rl.nr_secs = nr_free_blks * geo->clba;
|
2018-01-05 21:16:13 +08:00
|
|
|
|
|
|
|
|
/* Consider sectors used for metadata */
|
|
|
|
|
sec_meta = (lm->smeta_sec + lm->emeta_sec[0]) * l_mg->nr_free_lines;
|
2018-03-30 06:05:10 +08:00
|
|
|
blk_meta = DIV_ROUND_UP(sec_meta, geo->clba);
|
2018-01-05 21:16:13 +08:00
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
pblk->capacity = (provisioned - blk_meta) * geo->clba;
|
2018-01-05 21:16:13 +08:00
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
atomic_set(&pblk->rl.free_blocks, nr_free_blks);
|
2018-01-05 21:16:13 +08:00
|
|
|
atomic_set(&pblk->rl.free_user_blocks, nr_free_blks);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
static int pblk_setup_line_meta_12(struct pblk *pblk, struct pblk_line *line,
|
|
|
|
|
void *chunk_meta)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
int i, chk_per_lun, nr_bad_chks = 0;
|
|
|
|
|
|
|
|
|
|
chk_per_lun = geo->num_chk * geo->pln_mode;
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < lm->blk_per_line; i++) {
|
|
|
|
|
struct pblk_lun *rlun = &pblk->luns[i];
|
|
|
|
|
struct nvm_chk_meta *chunk;
|
|
|
|
|
int pos = pblk_ppa_to_pos(geo, rlun->bppa);
|
|
|
|
|
u8 *lun_bb_meta = chunk_meta + pos * chk_per_lun;
|
|
|
|
|
|
|
|
|
|
chunk = &line->chks[pos];
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* In 1.2 spec. chunk state is not persisted by the device. Thus
|
|
|
|
|
* some of the values are reset each time pblk is instantiated.
|
|
|
|
|
*/
|
|
|
|
|
if (lun_bb_meta[line->id] == NVM_BLK_T_FREE)
|
|
|
|
|
chunk->state = NVM_CHK_ST_FREE;
|
|
|
|
|
else
|
|
|
|
|
chunk->state = NVM_CHK_ST_OFFLINE;
|
|
|
|
|
|
|
|
|
|
chunk->type = NVM_CHK_TP_W_SEQ;
|
|
|
|
|
chunk->wi = 0;
|
|
|
|
|
chunk->slba = -1;
|
|
|
|
|
chunk->cnlb = geo->clba;
|
|
|
|
|
chunk->wp = 0;
|
|
|
|
|
|
|
|
|
|
if (!(chunk->state & NVM_CHK_ST_OFFLINE))
|
|
|
|
|
continue;
|
|
|
|
|
|
|
|
|
|
set_bit(pos, line->blk_bitmap);
|
|
|
|
|
nr_bad_chks++;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return nr_bad_chks;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_setup_line_meta_20(struct pblk *pblk, struct pblk_line *line,
|
|
|
|
|
struct nvm_chk_meta *meta)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
int i, nr_bad_chks = 0;
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < lm->blk_per_line; i++) {
|
|
|
|
|
struct pblk_lun *rlun = &pblk->luns[i];
|
|
|
|
|
struct nvm_chk_meta *chunk;
|
|
|
|
|
struct nvm_chk_meta *chunk_meta;
|
|
|
|
|
struct ppa_addr ppa;
|
|
|
|
|
int pos;
|
|
|
|
|
|
|
|
|
|
ppa = rlun->bppa;
|
|
|
|
|
pos = pblk_ppa_to_pos(geo, ppa);
|
|
|
|
|
chunk = &line->chks[pos];
|
|
|
|
|
|
|
|
|
|
ppa.m.chk = line->id;
|
|
|
|
|
chunk_meta = pblk_chunk_get_off(pblk, meta, ppa);
|
|
|
|
|
|
|
|
|
|
chunk->state = chunk_meta->state;
|
|
|
|
|
chunk->type = chunk_meta->type;
|
|
|
|
|
chunk->wi = chunk_meta->wi;
|
|
|
|
|
chunk->slba = chunk_meta->slba;
|
|
|
|
|
chunk->cnlb = chunk_meta->cnlb;
|
|
|
|
|
chunk->wp = chunk_meta->wp;
|
|
|
|
|
|
|
|
|
|
if (chunk->type & NVM_CHK_TP_SZ_SPEC) {
|
|
|
|
|
WARN_ONCE(1, "pblk: custom-sized chunks unsupported\n");
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
2018-06-01 21:04:23 +08:00
|
|
|
if (!(chunk->state & NVM_CHK_ST_OFFLINE))
|
|
|
|
|
continue;
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
set_bit(pos, line->blk_bitmap);
|
|
|
|
|
nr_bad_chks++;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return nr_bad_chks;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static long pblk_setup_line_meta(struct pblk *pblk, struct pblk_line *line,
|
|
|
|
|
void *chunk_meta, int line_id)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
long nr_bad_chks, chk_in_line;
|
|
|
|
|
|
|
|
|
|
line->pblk = pblk;
|
|
|
|
|
line->id = line_id;
|
|
|
|
|
line->type = PBLK_LINETYPE_FREE;
|
|
|
|
|
line->state = PBLK_LINESTATE_NEW;
|
|
|
|
|
line->gc_group = PBLK_LINEGC_NONE;
|
|
|
|
|
line->vsc = &l_mg->vsc_list[line_id];
|
|
|
|
|
spin_lock_init(&line->lock);
|
|
|
|
|
|
|
|
|
|
if (geo->version == NVM_OCSSD_SPEC_12)
|
|
|
|
|
nr_bad_chks = pblk_setup_line_meta_12(pblk, line, chunk_meta);
|
|
|
|
|
else
|
|
|
|
|
nr_bad_chks = pblk_setup_line_meta_20(pblk, line, chunk_meta);
|
|
|
|
|
|
|
|
|
|
chk_in_line = lm->blk_per_line - nr_bad_chks;
|
|
|
|
|
if (nr_bad_chks < 0 || nr_bad_chks > lm->blk_per_line ||
|
|
|
|
|
chk_in_line < lm->min_blk_line) {
|
|
|
|
|
line->state = PBLK_LINESTATE_BAD;
|
|
|
|
|
list_add_tail(&line->list, &l_mg->bad_list);
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
atomic_set(&line->blk_in_line, chk_in_line);
|
|
|
|
|
list_add_tail(&line->list, &l_mg->free_list);
|
|
|
|
|
l_mg->nr_free_lines++;
|
|
|
|
|
|
|
|
|
|
return chk_in_line;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_alloc_line_meta(struct pblk *pblk, struct pblk_line *line)
|
2018-03-30 06:05:09 +08:00
|
|
|
{
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
|
|
|
|
|
line->blk_bitmap = kzalloc(lm->blk_bitmap_len, GFP_KERNEL);
|
|
|
|
|
if (!line->blk_bitmap)
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
|
|
line->erase_bitmap = kzalloc(lm->blk_bitmap_len, GFP_KERNEL);
|
2018-06-01 22:41:06 +08:00
|
|
|
if (!line->erase_bitmap)
|
|
|
|
|
goto free_blk_bitmap;
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
|
treewide: kmalloc() -> kmalloc_array()
The kmalloc() function has a 2-factor argument form, kmalloc_array(). This
patch replaces cases of:
kmalloc(a * b, gfp)
with:
kmalloc_array(a * b, gfp)
as well as handling cases of:
kmalloc(a * b * c, gfp)
with:
kmalloc(array3_size(a, b, c), gfp)
as it's slightly less ugly than:
kmalloc_array(array_size(a, b), c, gfp)
This does, however, attempt to ignore constant size factors like:
kmalloc(4 * 1024, gfp)
though any constants defined via macros get caught up in the conversion.
Any factors with a sizeof() of "unsigned char", "char", and "u8" were
dropped, since they're redundant.
The tools/ directory was manually excluded, since it has its own
implementation of kmalloc().
The Coccinelle script used for this was:
// Fix redundant parens around sizeof().
@@
type TYPE;
expression THING, E;
@@
(
kmalloc(
- (sizeof(TYPE)) * E
+ sizeof(TYPE) * E
, ...)
|
kmalloc(
- (sizeof(THING)) * E
+ sizeof(THING) * E
, ...)
)
// Drop single-byte sizes and redundant parens.
@@
expression COUNT;
typedef u8;
typedef __u8;
@@
(
kmalloc(
- sizeof(u8) * (COUNT)
+ COUNT
, ...)
|
kmalloc(
- sizeof(__u8) * (COUNT)
+ COUNT
, ...)
|
kmalloc(
- sizeof(char) * (COUNT)
+ COUNT
, ...)
|
kmalloc(
- sizeof(unsigned char) * (COUNT)
+ COUNT
, ...)
|
kmalloc(
- sizeof(u8) * COUNT
+ COUNT
, ...)
|
kmalloc(
- sizeof(__u8) * COUNT
+ COUNT
, ...)
|
kmalloc(
- sizeof(char) * COUNT
+ COUNT
, ...)
|
kmalloc(
- sizeof(unsigned char) * COUNT
+ COUNT
, ...)
)
// 2-factor product with sizeof(type/expression) and identifier or constant.
@@
type TYPE;
expression THING;
identifier COUNT_ID;
constant COUNT_CONST;
@@
(
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * (COUNT_ID)
+ COUNT_ID, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * COUNT_ID
+ COUNT_ID, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * (COUNT_CONST)
+ COUNT_CONST, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * COUNT_CONST
+ COUNT_CONST, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * (COUNT_ID)
+ COUNT_ID, sizeof(THING)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * COUNT_ID
+ COUNT_ID, sizeof(THING)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * (COUNT_CONST)
+ COUNT_CONST, sizeof(THING)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * COUNT_CONST
+ COUNT_CONST, sizeof(THING)
, ...)
)
// 2-factor product, only identifiers.
@@
identifier SIZE, COUNT;
@@
- kmalloc
+ kmalloc_array
(
- SIZE * COUNT
+ COUNT, SIZE
, ...)
// 3-factor product with 1 sizeof(type) or sizeof(expression), with
// redundant parens removed.
@@
expression THING;
identifier STRIDE, COUNT;
type TYPE;
@@
(
kmalloc(
- sizeof(TYPE) * (COUNT) * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kmalloc(
- sizeof(TYPE) * (COUNT) * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kmalloc(
- sizeof(TYPE) * COUNT * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kmalloc(
- sizeof(TYPE) * COUNT * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(TYPE))
, ...)
|
kmalloc(
- sizeof(THING) * (COUNT) * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kmalloc(
- sizeof(THING) * (COUNT) * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kmalloc(
- sizeof(THING) * COUNT * (STRIDE)
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
|
kmalloc(
- sizeof(THING) * COUNT * STRIDE
+ array3_size(COUNT, STRIDE, sizeof(THING))
, ...)
)
// 3-factor product with 2 sizeof(variable), with redundant parens removed.
@@
expression THING1, THING2;
identifier COUNT;
type TYPE1, TYPE2;
@@
(
kmalloc(
- sizeof(TYPE1) * sizeof(TYPE2) * COUNT
+ array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2))
, ...)
|
kmalloc(
- sizeof(TYPE1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2))
, ...)
|
kmalloc(
- sizeof(THING1) * sizeof(THING2) * COUNT
+ array3_size(COUNT, sizeof(THING1), sizeof(THING2))
, ...)
|
kmalloc(
- sizeof(THING1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(THING1), sizeof(THING2))
, ...)
|
kmalloc(
- sizeof(TYPE1) * sizeof(THING2) * COUNT
+ array3_size(COUNT, sizeof(TYPE1), sizeof(THING2))
, ...)
|
kmalloc(
- sizeof(TYPE1) * sizeof(THING2) * (COUNT)
+ array3_size(COUNT, sizeof(TYPE1), sizeof(THING2))
, ...)
)
// 3-factor product, only identifiers, with redundant parens removed.
@@
identifier STRIDE, SIZE, COUNT;
@@
(
kmalloc(
- (COUNT) * STRIDE * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- COUNT * (STRIDE) * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- COUNT * STRIDE * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- (COUNT) * (STRIDE) * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- COUNT * (STRIDE) * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- (COUNT) * STRIDE * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- (COUNT) * (STRIDE) * (SIZE)
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
|
kmalloc(
- COUNT * STRIDE * SIZE
+ array3_size(COUNT, STRIDE, SIZE)
, ...)
)
// Any remaining multi-factor products, first at least 3-factor products,
// when they're not all constants...
@@
expression E1, E2, E3;
constant C1, C2, C3;
@@
(
kmalloc(C1 * C2 * C3, ...)
|
kmalloc(
- (E1) * E2 * E3
+ array3_size(E1, E2, E3)
, ...)
|
kmalloc(
- (E1) * (E2) * E3
+ array3_size(E1, E2, E3)
, ...)
|
kmalloc(
- (E1) * (E2) * (E3)
+ array3_size(E1, E2, E3)
, ...)
|
kmalloc(
- E1 * E2 * E3
+ array3_size(E1, E2, E3)
, ...)
)
// And then all remaining 2 factors products when they're not all constants,
// keeping sizeof() as the second factor argument.
@@
expression THING, E1, E2;
type TYPE;
constant C1, C2, C3;
@@
(
kmalloc(sizeof(THING) * C2, ...)
|
kmalloc(sizeof(TYPE) * C2, ...)
|
kmalloc(C1 * C2 * C3, ...)
|
kmalloc(C1 * C2, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * (E2)
+ E2, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(TYPE) * E2
+ E2, sizeof(TYPE)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * (E2)
+ E2, sizeof(THING)
, ...)
|
- kmalloc
+ kmalloc_array
(
- sizeof(THING) * E2
+ E2, sizeof(THING)
, ...)
|
- kmalloc
+ kmalloc_array
(
- (E1) * E2
+ E1, E2
, ...)
|
- kmalloc
+ kmalloc_array
(
- (E1) * (E2)
+ E1, E2
, ...)
|
- kmalloc
+ kmalloc_array
(
- E1 * E2
+ E1, E2
, ...)
)
Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 04:55:00 +08:00
|
|
|
line->chks = kmalloc_array(lm->blk_per_line,
|
|
|
|
|
sizeof(struct nvm_chk_meta), GFP_KERNEL);
|
2018-06-01 22:41:06 +08:00
|
|
|
if (!line->chks)
|
|
|
|
|
goto free_erase_bitmap;
|
|
|
|
|
|
|
|
|
|
line->w_err_gc = kzalloc(sizeof(struct pblk_w_err_gc), GFP_KERNEL);
|
|
|
|
|
if (!line->w_err_gc)
|
|
|
|
|
goto free_chks;
|
2018-03-30 06:05:09 +08:00
|
|
|
|
|
|
|
|
return 0;
|
2018-06-01 22:41:06 +08:00
|
|
|
|
|
|
|
|
free_chks:
|
|
|
|
|
kfree(line->chks);
|
|
|
|
|
free_erase_bitmap:
|
|
|
|
|
kfree(line->erase_bitmap);
|
|
|
|
|
free_blk_bitmap:
|
|
|
|
|
kfree(line->blk_bitmap);
|
|
|
|
|
return -ENOMEM;
|
2018-03-30 06:05:09 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_line_mg_init(struct pblk *pblk)
|
2017-06-26 17:57:17 +08:00
|
|
|
{
|
2018-03-30 06:05:09 +08:00
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
2017-06-26 17:57:17 +08:00
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
2018-03-30 06:05:09 +08:00
|
|
|
int i, bb_distance;
|
|
|
|
|
|
2018-03-30 06:05:14 +08:00
|
|
|
l_mg->nr_lines = geo->num_chk;
|
2018-03-30 06:05:09 +08:00
|
|
|
l_mg->log_line = l_mg->data_line = NULL;
|
|
|
|
|
l_mg->l_seq_nr = l_mg->d_seq_nr = 0;
|
|
|
|
|
l_mg->nr_free_lines = 0;
|
|
|
|
|
bitmap_zero(&l_mg->meta_bitmap, PBLK_DATA_LINES);
|
|
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->free_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->corrupt_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->bad_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->gc_full_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->gc_high_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->gc_mid_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->gc_low_list);
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->gc_empty_list);
|
2018-06-01 22:41:06 +08:00
|
|
|
INIT_LIST_HEAD(&l_mg->gc_werr_list);
|
2018-03-30 06:05:09 +08:00
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&l_mg->emeta_list);
|
|
|
|
|
|
2018-06-01 22:41:06 +08:00
|
|
|
l_mg->gc_lists[0] = &l_mg->gc_werr_list;
|
|
|
|
|
l_mg->gc_lists[1] = &l_mg->gc_high_list;
|
|
|
|
|
l_mg->gc_lists[2] = &l_mg->gc_mid_list;
|
|
|
|
|
l_mg->gc_lists[3] = &l_mg->gc_low_list;
|
2018-03-30 06:05:09 +08:00
|
|
|
|
|
|
|
|
spin_lock_init(&l_mg->free_lock);
|
|
|
|
|
spin_lock_init(&l_mg->close_lock);
|
|
|
|
|
spin_lock_init(&l_mg->gc_lock);
|
|
|
|
|
|
|
|
|
|
l_mg->vsc_list = kcalloc(l_mg->nr_lines, sizeof(__le32), GFP_KERNEL);
|
|
|
|
|
if (!l_mg->vsc_list)
|
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
|
|
l_mg->bb_template = kzalloc(lm->sec_bitmap_len, GFP_KERNEL);
|
|
|
|
|
if (!l_mg->bb_template)
|
|
|
|
|
goto fail_free_vsc_list;
|
|
|
|
|
|
|
|
|
|
l_mg->bb_aux = kzalloc(lm->sec_bitmap_len, GFP_KERNEL);
|
|
|
|
|
if (!l_mg->bb_aux)
|
|
|
|
|
goto fail_free_bb_template;
|
2017-06-26 17:57:17 +08:00
|
|
|
|
|
|
|
|
/* smeta is always small enough to fit on a kmalloc memory allocation,
|
|
|
|
|
* emeta depends on the number of LUNs allocated to the pblk instance
|
|
|
|
|
*/
|
|
|
|
|
for (i = 0; i < PBLK_DATA_LINES; i++) {
|
|
|
|
|
l_mg->sline_meta[i] = kmalloc(lm->smeta_len, GFP_KERNEL);
|
|
|
|
|
if (!l_mg->sline_meta[i])
|
|
|
|
|
goto fail_free_smeta;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* emeta allocates three different buffers for managing metadata with
|
|
|
|
|
* in-memory and in-media layouts
|
|
|
|
|
*/
|
|
|
|
|
for (i = 0; i < PBLK_DATA_LINES; i++) {
|
|
|
|
|
struct pblk_emeta *emeta;
|
|
|
|
|
|
|
|
|
|
emeta = kmalloc(sizeof(struct pblk_emeta), GFP_KERNEL);
|
|
|
|
|
if (!emeta)
|
|
|
|
|
goto fail_free_emeta;
|
|
|
|
|
|
|
|
|
|
if (lm->emeta_len[0] > KMALLOC_MAX_CACHE_SIZE) {
|
|
|
|
|
l_mg->emeta_alloc_type = PBLK_VMALLOC_META;
|
|
|
|
|
|
|
|
|
|
emeta->buf = vmalloc(lm->emeta_len[0]);
|
|
|
|
|
if (!emeta->buf) {
|
|
|
|
|
kfree(emeta);
|
|
|
|
|
goto fail_free_emeta;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
emeta->nr_entries = lm->emeta_sec[0];
|
|
|
|
|
l_mg->eline_meta[i] = emeta;
|
|
|
|
|
} else {
|
|
|
|
|
l_mg->emeta_alloc_type = PBLK_KMALLOC_META;
|
|
|
|
|
|
|
|
|
|
emeta->buf = kmalloc(lm->emeta_len[0], GFP_KERNEL);
|
|
|
|
|
if (!emeta->buf) {
|
|
|
|
|
kfree(emeta);
|
|
|
|
|
goto fail_free_emeta;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
emeta->nr_entries = lm->emeta_sec[0];
|
|
|
|
|
l_mg->eline_meta[i] = emeta;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < l_mg->nr_lines; i++)
|
|
|
|
|
l_mg->vsc_list[i] = cpu_to_le32(EMPTY_ENTRY);
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
bb_distance = (geo->all_luns) * geo->ws_opt;
|
|
|
|
|
for (i = 0; i < lm->sec_per_line; i += bb_distance)
|
|
|
|
|
bitmap_set(l_mg->bb_template, i, geo->ws_opt);
|
|
|
|
|
|
2017-06-26 17:57:17 +08:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
fail_free_emeta:
|
|
|
|
|
while (--i >= 0) {
|
2017-10-13 20:45:54 +08:00
|
|
|
if (l_mg->emeta_alloc_type == PBLK_VMALLOC_META)
|
|
|
|
|
vfree(l_mg->eline_meta[i]->buf);
|
|
|
|
|
else
|
|
|
|
|
kfree(l_mg->eline_meta[i]->buf);
|
2017-06-26 17:57:21 +08:00
|
|
|
kfree(l_mg->eline_meta[i]);
|
2017-06-26 17:57:17 +08:00
|
|
|
}
|
|
|
|
|
fail_free_smeta:
|
|
|
|
|
for (i = 0; i < PBLK_DATA_LINES; i++)
|
2017-06-26 17:57:21 +08:00
|
|
|
kfree(l_mg->sline_meta[i]);
|
2018-03-30 06:05:09 +08:00
|
|
|
kfree(l_mg->bb_aux);
|
|
|
|
|
fail_free_bb_template:
|
|
|
|
|
kfree(l_mg->bb_template);
|
|
|
|
|
fail_free_vsc_list:
|
|
|
|
|
kfree(l_mg->vsc_list);
|
|
|
|
|
fail:
|
2017-06-26 17:57:17 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
static int pblk_line_meta_init(struct pblk *pblk)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
|
|
|
|
struct nvm_tgt_dev *dev = pblk->dev;
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct pblk_line_meta *lm = &pblk->lm;
|
|
|
|
|
unsigned int smeta_len, emeta_len;
|
2018-03-30 06:05:09 +08:00
|
|
|
int i;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
lm->sec_per_line = geo->clba * geo->all_luns;
|
2018-01-05 21:16:03 +08:00
|
|
|
lm->blk_per_line = geo->all_luns;
|
|
|
|
|
lm->blk_bitmap_len = BITS_TO_LONGS(geo->all_luns) * sizeof(long);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
lm->sec_bitmap_len = BITS_TO_LONGS(lm->sec_per_line) * sizeof(long);
|
2018-01-05 21:16:03 +08:00
|
|
|
lm->lun_bitmap_len = BITS_TO_LONGS(geo->all_luns) * sizeof(long);
|
2017-10-13 20:46:28 +08:00
|
|
|
lm->mid_thrs = lm->sec_per_line / 2;
|
|
|
|
|
lm->high_thrs = lm->sec_per_line / 4;
|
2018-01-05 21:16:03 +08:00
|
|
|
lm->meta_distance = (geo->all_luns / 2) * pblk->min_write_pgs;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
/* Calculate necessary pages for smeta. See comment over struct
|
|
|
|
|
* line_smeta definition
|
|
|
|
|
*/
|
|
|
|
|
i = 1;
|
|
|
|
|
add_smeta_page:
|
2018-03-30 06:05:10 +08:00
|
|
|
lm->smeta_sec = i * geo->ws_opt;
|
|
|
|
|
lm->smeta_len = lm->smeta_sec * geo->csecs;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-06-26 17:57:17 +08:00
|
|
|
smeta_len = sizeof(struct line_smeta) + lm->lun_bitmap_len;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (smeta_len > lm->smeta_len) {
|
|
|
|
|
i++;
|
|
|
|
|
goto add_smeta_page;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Calculate necessary pages for emeta. See comment over struct
|
|
|
|
|
* line_emeta definition
|
|
|
|
|
*/
|
|
|
|
|
i = 1;
|
|
|
|
|
add_emeta_page:
|
2018-03-30 06:05:10 +08:00
|
|
|
lm->emeta_sec[0] = i * geo->ws_opt;
|
|
|
|
|
lm->emeta_len[0] = lm->emeta_sec[0] * geo->csecs;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2017-06-26 17:57:17 +08:00
|
|
|
emeta_len = calc_emeta_len(pblk);
|
|
|
|
|
if (emeta_len > lm->emeta_len[0]) {
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
i++;
|
|
|
|
|
goto add_emeta_page;
|
|
|
|
|
}
|
|
|
|
|
|
2018-01-05 21:16:03 +08:00
|
|
|
lm->emeta_bb = geo->all_luns > i ? geo->all_luns - i : 0;
|
lightnvm: pblk: enable 1 LUN configuration
Metadata I/Os are scheduled to minimize their impact on user data I/Os.
When there are enough LUNs instantiated (i.e., enough bandwidth), it is
easy to interleave metadata and data one after the other so that
metadata I/Os are the ones being blocked and not vice-versa.
We do this by calculating the distance between the I/Os in terms of the
LUNs that are not in used, and selecting a free LUN that satisfies a
the simple heuristic that metadata is scheduled behind. The per-LUN
semaphores guarantee consistency. This works fine on >1 LUN
configuration. However, when a single LUN is instantiated, this design
leads to a deadlock, where metadata waits to be scheduled on a free LUN.
This patch implements the 1 LUN case by simply scheduling the metadada
I/O after the data I/O. In the process, we refactor the way a line is
replaced to ensure that metadata writes are submitted after data writes
in order to guarantee block sequentiality. Note that, since there is
only one LUN, both I/Os will block each other by design. However, such
configuration only pursues tight read latencies, not write bandwidth.
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-10-13 20:46:26 +08:00
|
|
|
|
|
|
|
|
lm->min_blk_line = 1;
|
2018-01-05 21:16:03 +08:00
|
|
|
if (geo->all_luns > 1)
|
lightnvm: pblk: enable 1 LUN configuration
Metadata I/Os are scheduled to minimize their impact on user data I/Os.
When there are enough LUNs instantiated (i.e., enough bandwidth), it is
easy to interleave metadata and data one after the other so that
metadata I/Os are the ones being blocked and not vice-versa.
We do this by calculating the distance between the I/Os in terms of the
LUNs that are not in used, and selecting a free LUN that satisfies a
the simple heuristic that metadata is scheduled behind. The per-LUN
semaphores guarantee consistency. This works fine on >1 LUN
configuration. However, when a single LUN is instantiated, this design
leads to a deadlock, where metadata waits to be scheduled on a free LUN.
This patch implements the 1 LUN case by simply scheduling the metadada
I/O after the data I/O. In the process, we refactor the way a line is
replaced to ensure that metadata writes are submitted after data writes
in order to guarantee block sequentiality. Note that, since there is
only one LUN, both I/Os will block each other by design. However, such
configuration only pursues tight read latencies, not write bandwidth.
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-10-13 20:46:26 +08:00
|
|
|
lm->min_blk_line += DIV_ROUND_UP(lm->smeta_sec +
|
2018-03-30 06:05:10 +08:00
|
|
|
lm->emeta_sec[0], geo->clba);
|
lightnvm: pblk: enable 1 LUN configuration
Metadata I/Os are scheduled to minimize their impact on user data I/Os.
When there are enough LUNs instantiated (i.e., enough bandwidth), it is
easy to interleave metadata and data one after the other so that
metadata I/Os are the ones being blocked and not vice-versa.
We do this by calculating the distance between the I/Os in terms of the
LUNs that are not in used, and selecting a free LUN that satisfies a
the simple heuristic that metadata is scheduled behind. The per-LUN
semaphores guarantee consistency. This works fine on >1 LUN
configuration. However, when a single LUN is instantiated, this design
leads to a deadlock, where metadata waits to be scheduled on a free LUN.
This patch implements the 1 LUN case by simply scheduling the metadada
I/O after the data I/O. In the process, we refactor the way a line is
replaced to ensure that metadata writes are submitted after data writes
in order to guarantee block sequentiality. Note that, since there is
only one LUN, both I/Os will block each other by design. However, such
configuration only pursues tight read latencies, not write bandwidth.
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-10-13 20:46:26 +08:00
|
|
|
|
2017-06-30 23:56:41 +08:00
|
|
|
if (lm->min_blk_line > lm->blk_per_line) {
|
|
|
|
|
pr_err("pblk: config. not supported. Min. LUN in line:%d\n",
|
|
|
|
|
lm->blk_per_line);
|
2018-03-30 06:04:59 +08:00
|
|
|
return -EINVAL;
|
2017-06-30 23:56:41 +08:00
|
|
|
}
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_lines_init(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
struct pblk_line_mgmt *l_mg = &pblk->l_mg;
|
|
|
|
|
struct pblk_line *line;
|
2018-03-30 06:05:20 +08:00
|
|
|
void *chunk_meta;
|
|
|
|
|
long nr_free_chks = 0;
|
2018-03-30 06:05:09 +08:00
|
|
|
int i, ret;
|
|
|
|
|
|
|
|
|
|
ret = pblk_line_meta_init(pblk);
|
2017-06-26 17:57:17 +08:00
|
|
|
if (ret)
|
2018-03-30 06:04:59 +08:00
|
|
|
return ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
ret = pblk_line_mg_init(pblk);
|
|
|
|
|
if (ret)
|
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
|
|
ret = pblk_luns_init(pblk);
|
|
|
|
|
if (ret)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
goto fail_free_meta;
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
chunk_meta = pblk_chunk_get_meta(pblk);
|
|
|
|
|
if (IS_ERR(chunk_meta)) {
|
|
|
|
|
ret = PTR_ERR(chunk_meta);
|
2018-03-30 06:05:09 +08:00
|
|
|
goto fail_free_luns;
|
2017-04-16 02:55:53 +08:00
|
|
|
}
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
pblk->lines = kcalloc(l_mg->nr_lines, sizeof(struct pblk_line),
|
|
|
|
|
GFP_KERNEL);
|
2017-04-16 02:55:53 +08:00
|
|
|
if (!pblk->lines) {
|
|
|
|
|
ret = -ENOMEM;
|
2018-03-30 06:05:20 +08:00
|
|
|
goto fail_free_chunk_meta;
|
2018-03-30 06:04:59 +08:00
|
|
|
}
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
for (i = 0; i < l_mg->nr_lines; i++) {
|
|
|
|
|
line = &pblk->lines[i];
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
ret = pblk_alloc_line_meta(pblk, line);
|
2017-06-26 17:57:22 +08:00
|
|
|
if (ret)
|
2018-03-30 06:05:09 +08:00
|
|
|
goto fail_free_lines;
|
2017-06-26 17:57:22 +08:00
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
nr_free_chks += pblk_setup_line_meta(pblk, line, chunk_meta, i);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-06-01 21:04:15 +08:00
|
|
|
if (!nr_free_chks) {
|
|
|
|
|
pr_err("pblk: too many bad blocks prevent for sane instance\n");
|
|
|
|
|
return -EINTR;
|
|
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
pblk_set_provision(pblk, nr_free_chks);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-03-30 06:05:20 +08:00
|
|
|
kfree(chunk_meta);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return 0;
|
2018-03-30 06:04:59 +08:00
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
fail_free_lines:
|
2017-06-26 17:57:22 +08:00
|
|
|
while (--i >= 0)
|
2018-06-01 22:41:06 +08:00
|
|
|
pblk_line_meta_free(l_mg, &pblk->lines[i]);
|
2018-03-30 06:05:09 +08:00
|
|
|
kfree(pblk->lines);
|
2018-03-30 06:05:20 +08:00
|
|
|
fail_free_chunk_meta:
|
|
|
|
|
kfree(chunk_meta);
|
2018-03-30 06:05:09 +08:00
|
|
|
fail_free_luns:
|
|
|
|
|
kfree(pblk->luns);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
fail_free_meta:
|
2018-03-30 06:04:59 +08:00
|
|
|
pblk_line_mg_free(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int pblk_writer_init(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
pblk->writer_ts = kthread_create(pblk_write_ts, pblk, "pblk-writer-t");
|
|
|
|
|
if (IS_ERR(pblk->writer_ts)) {
|
2018-01-05 21:16:17 +08:00
|
|
|
int err = PTR_ERR(pblk->writer_ts);
|
|
|
|
|
|
|
|
|
|
if (err != -EINTR)
|
|
|
|
|
pr_err("pblk: could not allocate writer kthread (%d)\n",
|
|
|
|
|
err);
|
|
|
|
|
return err;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-01-05 21:16:17 +08:00
|
|
|
timer_setup(&pblk->wtimer, pblk_write_timer_fn, 0);
|
|
|
|
|
mod_timer(&pblk->wtimer, jiffies + msecs_to_jiffies(100));
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_writer_stop(struct pblk *pblk)
|
|
|
|
|
{
|
2017-06-30 23:56:40 +08:00
|
|
|
/* The pipeline must be stopped and the write buffer emptied before the
|
|
|
|
|
* write thread is stopped
|
|
|
|
|
*/
|
|
|
|
|
WARN(pblk_rb_read_count(&pblk->rwb),
|
|
|
|
|
"Stopping not fully persisted write buffer\n");
|
|
|
|
|
|
|
|
|
|
WARN(pblk_rb_sync_count(&pblk->rwb),
|
|
|
|
|
"Stopping not fully synced write buffer\n");
|
|
|
|
|
|
2018-03-30 06:04:56 +08:00
|
|
|
del_timer_sync(&pblk->wtimer);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (pblk->writer_ts)
|
|
|
|
|
kthread_stop(pblk->writer_ts);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_free(struct pblk *pblk)
|
|
|
|
|
{
|
|
|
|
|
pblk_lines_free(pblk);
|
|
|
|
|
pblk_l2p_free(pblk);
|
2018-03-30 06:05:09 +08:00
|
|
|
pblk_rwb_free(pblk);
|
|
|
|
|
pblk_core_free(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
|
|
|
|
kfree(pblk);
|
|
|
|
|
}
|
|
|
|
|
|
2018-06-01 21:04:24 +08:00
|
|
|
static void pblk_tear_down(struct pblk *pblk, bool graceful)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
2018-06-01 21:04:24 +08:00
|
|
|
if (graceful)
|
|
|
|
|
__pblk_pipeline_flush(pblk);
|
|
|
|
|
__pblk_pipeline_stop(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
pblk_writer_stop(pblk);
|
|
|
|
|
pblk_rb_sync_l2p(&pblk->rwb);
|
|
|
|
|
pblk_rl_free(&pblk->rl);
|
|
|
|
|
|
2018-06-01 21:04:24 +08:00
|
|
|
pr_debug("pblk: consistent tear down (graceful:%d)\n", graceful);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-06-01 21:04:24 +08:00
|
|
|
static void pblk_exit(void *private, bool graceful)
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
{
|
|
|
|
|
struct pblk *pblk = private;
|
|
|
|
|
|
|
|
|
|
down_write(&pblk_lock);
|
2018-06-01 21:04:24 +08:00
|
|
|
pblk_gc_exit(pblk, graceful);
|
|
|
|
|
pblk_tear_down(pblk, graceful);
|
2017-10-13 20:46:40 +08:00
|
|
|
|
|
|
|
|
#ifdef CONFIG_NVM_DEBUG
|
|
|
|
|
pr_info("pblk exit: L2P CRC: %x\n", pblk_l2p_crc(pblk));
|
|
|
|
|
#endif
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
pblk_free(pblk);
|
|
|
|
|
up_write(&pblk_lock);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static sector_t pblk_capacity(void *private)
|
|
|
|
|
{
|
|
|
|
|
struct pblk *pblk = private;
|
|
|
|
|
|
|
|
|
|
return pblk->capacity * NR_PHY_IN_LOG;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void *pblk_init(struct nvm_tgt_dev *dev, struct gendisk *tdisk,
|
|
|
|
|
int flags)
|
|
|
|
|
{
|
|
|
|
|
struct nvm_geo *geo = &dev->geo;
|
|
|
|
|
struct request_queue *bqueue = dev->q;
|
|
|
|
|
struct request_queue *tqueue = tdisk->queue;
|
|
|
|
|
struct pblk *pblk;
|
|
|
|
|
int ret;
|
|
|
|
|
|
2018-03-30 06:05:21 +08:00
|
|
|
/* pblk supports 1.2 and 2.0 versions */
|
|
|
|
|
if (!(geo->version == NVM_OCSSD_SPEC_12 ||
|
|
|
|
|
geo->version == NVM_OCSSD_SPEC_20)) {
|
2018-03-30 06:05:18 +08:00
|
|
|
pr_err("pblk: OCSSD version not supported (%u)\n",
|
|
|
|
|
geo->version);
|
|
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (geo->version == NVM_OCSSD_SPEC_12 && geo->dom & NVM_RSP_L2P) {
|
2017-10-13 20:45:57 +08:00
|
|
|
pr_err("pblk: host-side L2P table not supported. (%x)\n",
|
2018-03-30 06:05:18 +08:00
|
|
|
geo->dom);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return ERR_PTR(-EINVAL);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
pblk = kzalloc(sizeof(struct pblk), GFP_KERNEL);
|
|
|
|
|
if (!pblk)
|
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
|
|
pblk->dev = dev;
|
|
|
|
|
pblk->disk = tdisk;
|
2017-06-26 17:57:29 +08:00
|
|
|
pblk->state = PBLK_STATE_RUNNING;
|
2017-10-13 20:46:34 +08:00
|
|
|
pblk->gc.gc_enabled = 0;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-06-01 22:41:05 +08:00
|
|
|
spin_lock_init(&pblk->resubmit_lock);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
spin_lock_init(&pblk->trans_lock);
|
|
|
|
|
spin_lock_init(&pblk->lock);
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_NVM_DEBUG
|
|
|
|
|
atomic_long_set(&pblk->inflight_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->padded_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->padded_wb, 0);
|
|
|
|
|
atomic_long_set(&pblk->req_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->sub_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->sync_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->inflight_reads, 0);
|
2017-06-26 17:57:13 +08:00
|
|
|
atomic_long_set(&pblk->cache_reads, 0);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
atomic_long_set(&pblk->sync_reads, 0);
|
|
|
|
|
atomic_long_set(&pblk->recov_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->recov_writes, 0);
|
|
|
|
|
atomic_long_set(&pblk->recov_gc_writes, 0);
|
2017-10-13 20:46:01 +08:00
|
|
|
atomic_long_set(&pblk->recov_gc_reads, 0);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
atomic_long_set(&pblk->read_failed, 0);
|
|
|
|
|
atomic_long_set(&pblk->read_empty, 0);
|
|
|
|
|
atomic_long_set(&pblk->read_high_ecc, 0);
|
|
|
|
|
atomic_long_set(&pblk->read_failed_gc, 0);
|
|
|
|
|
atomic_long_set(&pblk->write_failed, 0);
|
|
|
|
|
atomic_long_set(&pblk->erase_failed, 0);
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
ret = pblk_core_init(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (ret) {
|
2018-03-30 06:05:09 +08:00
|
|
|
pr_err("pblk: could not initialize core\n");
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
goto fail;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ret = pblk_lines_init(pblk);
|
|
|
|
|
if (ret) {
|
|
|
|
|
pr_err("pblk: could not initialize lines\n");
|
2018-03-30 06:05:09 +08:00
|
|
|
goto fail_free_core;
|
2018-03-30 06:04:55 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
ret = pblk_rwb_init(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (ret) {
|
2018-03-30 06:05:09 +08:00
|
|
|
pr_err("pblk: could not initialize write buffer\n");
|
|
|
|
|
goto fail_free_lines;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
2018-03-30 06:05:09 +08:00
|
|
|
ret = pblk_l2p_init(pblk, flags & NVM_TARGET_FACTORY);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
if (ret) {
|
|
|
|
|
pr_err("pblk: could not initialize maps\n");
|
2018-03-30 06:05:09 +08:00
|
|
|
goto fail_free_rwb;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ret = pblk_writer_init(pblk);
|
|
|
|
|
if (ret) {
|
2018-01-05 21:16:17 +08:00
|
|
|
if (ret != -EINTR)
|
|
|
|
|
pr_err("pblk: could not initialize write thread\n");
|
2018-03-30 06:05:09 +08:00
|
|
|
goto fail_free_l2p;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ret = pblk_gc_init(pblk);
|
|
|
|
|
if (ret) {
|
|
|
|
|
pr_err("pblk: could not initialize gc\n");
|
|
|
|
|
goto fail_stop_writer;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* inherit the size from the underlying device */
|
|
|
|
|
blk_queue_logical_block_size(tqueue, queue_physical_block_size(bqueue));
|
|
|
|
|
blk_queue_max_hw_sectors(tqueue, queue_max_hw_sectors(bqueue));
|
|
|
|
|
|
|
|
|
|
blk_queue_write_cache(tqueue, true, false);
|
|
|
|
|
|
2018-03-30 06:05:10 +08:00
|
|
|
tqueue->limits.discard_granularity = geo->clba * geo->csecs;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
tqueue->limits.discard_alignment = 0;
|
|
|
|
|
blk_queue_max_discard_sectors(tqueue, UINT_MAX >> 9);
|
2018-03-08 09:10:10 +08:00
|
|
|
blk_queue_flag_set(QUEUE_FLAG_DISCARD, tqueue);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
|
2018-01-05 21:16:19 +08:00
|
|
|
pr_info("pblk(%s): luns:%u, lines:%d, secs:%llu, buf entries:%u\n",
|
|
|
|
|
tdisk->disk_name,
|
2018-01-05 21:16:03 +08:00
|
|
|
geo->all_luns, pblk->l_mg.nr_lines,
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
(unsigned long long)pblk->rl.nr_secs,
|
|
|
|
|
pblk->rwb.nr_entries);
|
|
|
|
|
|
|
|
|
|
wake_up_process(pblk->writer_ts);
|
2017-10-13 20:46:37 +08:00
|
|
|
|
|
|
|
|
/* Check if we need to start GC */
|
|
|
|
|
pblk_gc_should_kick(pblk);
|
|
|
|
|
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
return pblk;
|
|
|
|
|
|
|
|
|
|
fail_stop_writer:
|
|
|
|
|
pblk_writer_stop(pblk);
|
|
|
|
|
fail_free_l2p:
|
|
|
|
|
pblk_l2p_free(pblk);
|
2018-03-30 06:05:09 +08:00
|
|
|
fail_free_rwb:
|
|
|
|
|
pblk_rwb_free(pblk);
|
|
|
|
|
fail_free_lines:
|
|
|
|
|
pblk_lines_free(pblk);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
fail_free_core:
|
|
|
|
|
pblk_core_free(pblk);
|
|
|
|
|
fail:
|
|
|
|
|
kfree(pblk);
|
|
|
|
|
return ERR_PTR(ret);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* physical block device target */
|
|
|
|
|
static struct nvm_tgt_type tt_pblk = {
|
|
|
|
|
.name = "pblk",
|
|
|
|
|
.version = {1, 0, 0},
|
|
|
|
|
|
|
|
|
|
.make_rq = pblk_make_rq,
|
|
|
|
|
.capacity = pblk_capacity,
|
|
|
|
|
|
|
|
|
|
.init = pblk_init,
|
|
|
|
|
.exit = pblk_exit,
|
|
|
|
|
|
|
|
|
|
.sysfs_init = pblk_sysfs_init,
|
|
|
|
|
.sysfs_exit = pblk_sysfs_exit,
|
2017-10-13 20:45:50 +08:00
|
|
|
.owner = THIS_MODULE,
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
static int __init pblk_module_init(void)
|
|
|
|
|
{
|
2017-06-18 12:38:58 +08:00
|
|
|
int ret;
|
|
|
|
|
|
2018-05-21 06:25:50 +08:00
|
|
|
ret = bioset_init(&pblk_bio_set, BIO_POOL_SIZE, 0, 0);
|
|
|
|
|
if (ret)
|
|
|
|
|
return ret;
|
2017-06-18 12:38:58 +08:00
|
|
|
ret = nvm_register_tgt_type(&tt_pblk);
|
|
|
|
|
if (ret)
|
2018-05-21 06:25:50 +08:00
|
|
|
bioset_exit(&pblk_bio_set);
|
2017-06-18 12:38:58 +08:00
|
|
|
return ret;
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void pblk_module_exit(void)
|
|
|
|
|
{
|
2018-05-21 06:25:50 +08:00
|
|
|
bioset_exit(&pblk_bio_set);
|
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-16 02:55:50 +08:00
|
|
|
nvm_unregister_tgt_type(&tt_pblk);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
module_init(pblk_module_init);
|
|
|
|
|
module_exit(pblk_module_exit);
|
|
|
|
|
MODULE_AUTHOR("Javier Gonzalez <javier@cnexlabs.com>");
|
|
|
|
|
MODULE_AUTHOR("Matias Bjorling <matias@cnexlabs.com>");
|
|
|
|
|
MODULE_LICENSE("GPL v2");
|
|
|
|
|
MODULE_DESCRIPTION("Physical Block-Device for Open-Channel SSDs");
|
