openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity
linux/drivers/lightnvm/core.c

1133 lines
25 KiB
C
Raw Normal View History

lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
/*
* Copyright (C) 2015 IT University of Copenhagen. All rights reserved.
* Initial release: Matias Bjorling <m@bjorling.me>
*
* 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.
*
* You should have received a copy of the GNU General Public License
* along with this program; see the file COPYING. If not, write to
* the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139,
* USA.
*
*/
#include <linux/list.h>
#include <linux/types.h>
#include <linux/sem.h>
#include <linux/bitmap.h>
lightnvm: make core.c explicitly non-modular The Kconfig currently controlling compilation of this code is: drivers/lightnvm/Kconfig:menuconfig NVM drivers/lightnvm/Kconfig: bool "Open-Channel SSD target support" ...meaning that it currently is not being built as a module by anyone. Lets remove the modular code that is essentially orphaned, so that when reading the driver there is no doubt it is builtin-only. Since module_misc_driver translates to device_initcall in the non-modular case, the init ordering remains unchanged with this commit. We also delete the MODULE_LICENSE tag etc. since all that information is already contained at the top of the file in the comments. We replace module.h with moduleparam.h because this file still uses module params to control behaviour. Also note that MODULE_ALIAS is a no-op for non-modular code. Cc: Matias Bjorling <mb@lightnvm.io> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-10-30 04:38:41 +08:00
#include <linux/moduleparam.h>
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
#include <linux/miscdevice.h>
#include <linux/lightnvm.h>
lightnvm: refactor end_io functions for sync To implement sync I/O support within the LightNVM core, the end_io functions are refactored to take an end_io function pointer instead of testing for initialized media manager, followed by calling its end_io function. Sync I/O can then be implemented using a callback that signal I/O completion. This is similar to the logic found in blk_to_execute_io(). By implementing it this way, the underlying device I/Os submission logic is abstracted away from core, targets, and media managers. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:21 +08:00
#include <linux/sched/sysctl.h>
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
static LIST_HEAD(nvm_tgt_types);
lightnvm: remove nested lock conflict with mm If a media manager tries to initialize it targets upon media manager initialization, the media manager will need to know which target types are available in LightNVM. The lists of which managers and target types are available shares the same lock. Therefore, on initialization, the nvm_lock is taken by LightNVM core, which later leads to a deadlock when target types are enumerated by the media manager. Add an exclusive lock for target types to resolve this conflict. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:17 +08:00
static DECLARE_RWSEM(nvm_tgtt_lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
static LIST_HEAD(nvm_mgrs);
static LIST_HEAD(nvm_devices);
static DECLARE_RWSEM(nvm_lock);
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
struct nvm_tgt_type *nvm_find_target_type(const char *name, int lock)
lightnvm: refactor dev->online_target to global nvm_targets A target name must be unique. However, a per-device registration of targets is maintained on a dev->online_targets list, with a per-device search for targets upon registration. This results in a name collision when two targets, with the same name, are created on two different targets, where the per-device list is not shared. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:03 +08:00
{
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
struct nvm_tgt_type *tmp, *tt = NULL;
lightnvm: refactor dev->online_target to global nvm_targets A target name must be unique. However, a per-device registration of targets is maintained on a dev->online_targets list, with a per-device search for targets upon registration. This results in a name collision when two targets, with the same name, are created on two different targets, where the per-device list is not shared. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:03 +08:00
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
if (lock)
lightnvm: remove nested lock conflict with mm If a media manager tries to initialize it targets upon media manager initialization, the media manager will need to know which target types are available in LightNVM. The lists of which managers and target types are available shares the same lock. Therefore, on initialization, the nvm_lock is taken by LightNVM core, which later leads to a deadlock when target types are enumerated by the media manager. Add an exclusive lock for target types to resolve this conflict. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:17 +08:00
down_write(&nvm_tgtt_lock);
lightnvm: refactor dev->online_target to global nvm_targets A target name must be unique. However, a per-device registration of targets is maintained on a dev->online_targets list, with a per-device search for targets upon registration. This results in a name collision when two targets, with the same name, are created on two different targets, where the per-device list is not shared. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:03 +08:00
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
list_for_each_entry(tmp, &nvm_tgt_types, list)
if (!strcmp(name, tmp->name)) {
tt = tmp;
break;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
if (lock)
lightnvm: remove nested lock conflict with mm If a media manager tries to initialize it targets upon media manager initialization, the media manager will need to know which target types are available in LightNVM. The lists of which managers and target types are available shares the same lock. Therefore, on initialization, the nvm_lock is taken by LightNVM core, which later leads to a deadlock when target types are enumerated by the media manager. Add an exclusive lock for target types to resolve this conflict. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:17 +08:00
up_write(&nvm_tgtt_lock);
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
return tt;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
EXPORT_SYMBOL(nvm_find_target_type);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
int nvm_register_tgt_type(struct nvm_tgt_type *tt)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
int ret = 0;
lightnvm: remove nested lock conflict with mm If a media manager tries to initialize it targets upon media manager initialization, the media manager will need to know which target types are available in LightNVM. The lists of which managers and target types are available shares the same lock. Therefore, on initialization, the nvm_lock is taken by LightNVM core, which later leads to a deadlock when target types are enumerated by the media manager. Add an exclusive lock for target types to resolve this conflict. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:17 +08:00
down_write(&nvm_tgtt_lock);
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
if (nvm_find_target_type(tt->name, 0))
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
ret = -EEXIST;
else
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
list_add(&tt->list, &nvm_tgt_types);
lightnvm: remove nested lock conflict with mm If a media manager tries to initialize it targets upon media manager initialization, the media manager will need to know which target types are available in LightNVM. The lists of which managers and target types are available shares the same lock. Therefore, on initialization, the nvm_lock is taken by LightNVM core, which later leads to a deadlock when target types are enumerated by the media manager. Add an exclusive lock for target types to resolve this conflict. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:17 +08:00
up_write(&nvm_tgtt_lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return ret;
}
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
EXPORT_SYMBOL(nvm_register_tgt_type);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
void nvm_unregister_tgt_type(struct nvm_tgt_type *tt)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
if (!tt)
return;
down_write(&nvm_lock);
list_del(&tt->list);
up_write(&nvm_lock);
}
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
EXPORT_SYMBOL(nvm_unregister_tgt_type);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
void *nvm_dev_dma_alloc(struct nvm_dev *dev, gfp_t mem_flags,
dma_addr_t *dma_handler)
{
lightnvm: rename dma helper functions Until now, the dma pool have been exclusively used to allocate the ppa list being sent to the device. In pblk (upcoming), we use these pools to allocate metadata too. Thus, we generalize the names of some variables on the dma helper functions to make the code more readable. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:13 +08:00
return dev->ops->dev_dma_alloc(dev, dev->dma_pool, mem_flags,
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
dma_handler);
}
EXPORT_SYMBOL(nvm_dev_dma_alloc);
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
void nvm_dev_dma_free(struct nvm_dev *dev, void *addr, dma_addr_t dma_handler)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
lightnvm: rename dma helper functions Until now, the dma pool have been exclusively used to allocate the ppa list being sent to the device. In pblk (upcoming), we use these pools to allocate metadata too. Thus, we generalize the names of some variables on the dma helper functions to make the code more readable. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:13 +08:00
dev->ops->dev_dma_free(dev->dma_pool, addr, dma_handler);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
EXPORT_SYMBOL(nvm_dev_dma_free);
static struct nvmm_type *nvm_find_mgr_type(const char *name)
{
struct nvmm_type *mt;
list_for_each_entry(mt, &nvm_mgrs, list)
if (!strcmp(name, mt->name))
return mt;
return NULL;
}
lightnvm: Make functions not used by ouside static Mark functions not used by ouside of thier implementing file as static. Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:11 +08:00
static struct nvmm_type *nvm_init_mgr(struct nvm_dev *dev)
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
{
struct nvmm_type *mt;
int ret;
lockdep_assert_held(&nvm_lock);
list_for_each_entry(mt, &nvm_mgrs, list) {
lightnvm: use system block for mm initialization Use system block information to register the appropriate media manager. This enables the LightNVM subsystem to instantiate a media manager selected by the user, instead of relying on automatic detection by each media manager loaded in the kernel. A device must now be initialized before it can proceed to initialize its media manager. Upon initialization, the configured media manager is automatically initialized as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:38 +08:00
if (strncmp(dev->sb.mmtype, mt->name, NVM_MMTYPE_LEN))
continue;
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
ret = mt->register_mgr(dev);
if (ret < 0) {
pr_err("nvm: media mgr failed to init (%d) on dev %s\n",
ret, dev->name);
return NULL; /* initialization failed */
} else if (ret > 0)
return mt;
}
return NULL;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
int nvm_register_mgr(struct nvmm_type *mt)
{
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
struct nvm_dev *dev;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
int ret = 0;
down_write(&nvm_lock);
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
if (nvm_find_mgr_type(mt->name)) {
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
ret = -EEXIST;
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
goto finish;
} else {
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
list_add(&mt->list, &nvm_mgrs);
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
}
/* try to register media mgr if any device have none configured */
list_for_each_entry(dev, &nvm_devices, devices) {
if (dev->mt)
continue;
dev->mt = nvm_init_mgr(dev);
}
finish:
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
up_write(&nvm_lock);
return ret;
}
EXPORT_SYMBOL(nvm_register_mgr);
void nvm_unregister_mgr(struct nvmm_type *mt)
{
if (!mt)
return;
down_write(&nvm_lock);
list_del(&mt->list);
up_write(&nvm_lock);
}
EXPORT_SYMBOL(nvm_unregister_mgr);
static struct nvm_dev *nvm_find_nvm_dev(const char *name)
{
struct nvm_dev *dev;
list_for_each_entry(dev, &nvm_devices, devices)
if (!strcmp(name, dev->name))
return dev;
return NULL;
}
lightnvm: transform target get/set bad block Since targets are given a virtual target device, it is necessary to translate all communication between targets and the backend device. Implement the translation layer for get/set bad block table. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:14 +08:00
static void nvm_tgt_generic_to_addr_mode(struct nvm_tgt_dev *tgt_dev,
struct nvm_rq *rqd)
{
struct nvm_dev *dev = tgt_dev->parent;
int i;
if (rqd->nr_ppas > 1) {
for (i = 0; i < rqd->nr_ppas; i++) {
rqd->ppa_list[i] = dev->mt->trans_ppa(tgt_dev,
rqd->ppa_list[i], TRANS_TGT_TO_DEV);
rqd->ppa_list[i] = generic_to_dev_addr(dev,
rqd->ppa_list[i]);
}
} else {
rqd->ppa_addr = dev->mt->trans_ppa(tgt_dev, rqd->ppa_addr,
TRANS_TGT_TO_DEV);
rqd->ppa_addr = generic_to_dev_addr(dev, rqd->ppa_addr);
}
}
lightnvm: export set bad block table Bad blocks should be managed by block owners. This would be either targets for data blocks or sysblk for system blocks. In order to support this, export two functions: One to mark a block as an specific type (e.g., bad block) and another to update the bad block table on the device. Move bad block management to rrpc. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:38:56 +08:00
int nvm_set_bb_tbl(struct nvm_dev *dev, struct ppa_addr *ppas, int nr_ppas,
int type)
{
struct nvm_rq rqd;
int ret;
if (nr_ppas > dev->ops->max_phys_sect) {
pr_err("nvm: unable to update all sysblocks atomically\n");
return -EINVAL;
}
memset(&rqd, 0, sizeof(struct nvm_rq));
nvm_set_rqd_ppalist(dev, &rqd, ppas, nr_ppas, 1);
nvm_generic_to_addr_mode(dev, &rqd);
ret = dev->ops->set_bb_tbl(dev, &rqd.ppa_addr, rqd.nr_ppas, type);
nvm_free_rqd_ppalist(dev, &rqd);
if (ret) {
pr_err("nvm: sysblk failed bb mark\n");
return -EINVAL;
}
return 0;
}
EXPORT_SYMBOL(nvm_set_bb_tbl);
lightnvm: transform target get/set bad block Since targets are given a virtual target device, it is necessary to translate all communication between targets and the backend device. Implement the translation layer for get/set bad block table. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:14 +08:00
int nvm_set_tgt_bb_tbl(struct nvm_tgt_dev *tgt_dev, struct ppa_addr *ppas,
int nr_ppas, int type)
{
struct nvm_dev *dev = tgt_dev->parent;
struct nvm_rq rqd;
int ret;
if (nr_ppas > dev->ops->max_phys_sect) {
pr_err("nvm: unable to update all blocks atomically\n");
return -EINVAL;
}
memset(&rqd, 0, sizeof(struct nvm_rq));
nvm_set_rqd_ppalist(dev, &rqd, ppas, nr_ppas, 1);
nvm_tgt_generic_to_addr_mode(tgt_dev, &rqd);
ret = dev->ops->set_bb_tbl(dev, &rqd.ppa_addr, rqd.nr_ppas, type);
nvm_free_rqd_ppalist(dev, &rqd);
if (ret) {
pr_err("nvm: sysblk failed bb mark\n");
return -EINVAL;
}
return 0;
}
EXPORT_SYMBOL(nvm_set_tgt_bb_tbl);
lightnvm: introduce max_phys_sects helper function Target devices do not have access to the device driver operations. Introduce a helper function that exposes the max. number of physical sectors supported by the underlying device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:12 +08:00
int nvm_max_phys_sects(struct nvm_tgt_dev *tgt_dev)
{
struct nvm_dev *dev = tgt_dev->parent;
return dev->ops->max_phys_sect;
}
EXPORT_SYMBOL(nvm_max_phys_sects);
lightnvm: eliminate nvm_lun abstraction in mm In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. Since targets own the LUNs the are instantiated on top of and manage the free block list internally, there is no need for a LUN abstraction in the media manager. LUNs are intrinsically managed as in the physical layout (ch:0,lun:0, ..., ch:0,lun:n, ch:1,lun:0, ch:1,lun:n, ..., ch:m,lun:0, ch:m,lun:n) and given to the targets based on the target creation ioctl. This simplifies LUN management and clears the path for a partition manager to sit directly underneath LightNVM targets. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:10 +08:00
int nvm_submit_io(struct nvm_tgt_dev *tgt_dev, struct nvm_rq *rqd)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
lightnvm: eliminate nvm_lun abstraction in mm In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. Since targets own the LUNs the are instantiated on top of and manage the free block list internally, there is no need for a LUN abstraction in the media manager. LUNs are intrinsically managed as in the physical layout (ch:0,lun:0, ..., ch:0,lun:n, ch:1,lun:0, ch:1,lun:n, ..., ch:m,lun:0, ch:m,lun:n) and given to the targets based on the target creation ioctl. This simplifies LUN management and clears the path for a partition manager to sit directly underneath LightNVM targets. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:10 +08:00
struct nvm_dev *dev = tgt_dev->parent;
return dev->mt->submit_io(tgt_dev, rqd);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
EXPORT_SYMBOL(nvm_submit_io);
lightnvm: eliminate nvm_lun abstraction in mm In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. Since targets own the LUNs the are instantiated on top of and manage the free block list internally, there is no need for a LUN abstraction in the media manager. LUNs are intrinsically managed as in the physical layout (ch:0,lun:0, ..., ch:0,lun:n, ch:1,lun:0, ch:1,lun:n, ..., ch:m,lun:0, ch:m,lun:n) and given to the targets based on the target creation ioctl. This simplifies LUN management and clears the path for a partition manager to sit directly underneath LightNVM targets. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:10 +08:00
int nvm_erase_blk(struct nvm_tgt_dev *tgt_dev, struct ppa_addr *p, int flags)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
lightnvm: eliminate nvm_lun abstraction in mm In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. Since targets own the LUNs the are instantiated on top of and manage the free block list internally, there is no need for a LUN abstraction in the media manager. LUNs are intrinsically managed as in the physical layout (ch:0,lun:0, ..., ch:0,lun:n, ch:1,lun:0, ch:1,lun:n, ..., ch:m,lun:0, ch:m,lun:n) and given to the targets based on the target creation ioctl. This simplifies LUN management and clears the path for a partition manager to sit directly underneath LightNVM targets. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:10 +08:00
struct nvm_dev *dev = tgt_dev->parent;
return dev->mt->erase_blk(tgt_dev, p, flags);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
EXPORT_SYMBOL(nvm_erase_blk);
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
int nvm_get_l2p_tbl(struct nvm_tgt_dev *tgt_dev, u64 slba, u32 nlb,
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
nvm_l2p_update_fn *update_l2p, void *priv)
{
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
struct nvm_dev *dev = tgt_dev->parent;
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
if (!dev->ops->get_l2p_tbl)
return 0;
return dev->ops->get_l2p_tbl(dev, slba, nlb, update_l2p, priv);
}
EXPORT_SYMBOL(nvm_get_l2p_tbl);
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
int nvm_get_area(struct nvm_tgt_dev *tgt_dev, sector_t *lba, sector_t len)
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
{
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
struct nvm_dev *dev = tgt_dev->parent;
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
return dev->mt->get_area(dev, lba, len);
}
EXPORT_SYMBOL(nvm_get_area);
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
void nvm_put_area(struct nvm_tgt_dev *tgt_dev, sector_t lba)
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
{
lightnvm: use target nvm on target-specific ops. On target-specific operations pass on nvm_tgt_dev instead of the generic nvm device. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:13 +08:00
struct nvm_dev *dev = tgt_dev->parent;
lightnvm: introduce helpers for generic ops in rrpc Avoid calling media manager and device-specific operations directly from rrpc. Create helper functions on lightnvm's core instead. Signed-off-by: Javier González <javier@cnexlabs.com> Made it work with null_blk as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:11 +08:00
dev->mt->put_area(dev, lba);
}
EXPORT_SYMBOL(nvm_put_area);
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
void nvm_addr_to_generic_mode(struct nvm_dev *dev, struct nvm_rq *rqd)
{
int i;
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
if (rqd->nr_ppas > 1) {
for (i = 0; i < rqd->nr_ppas; i++)
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
rqd->ppa_list[i] = dev_to_generic_addr(dev,
rqd->ppa_list[i]);
} else {
rqd->ppa_addr = dev_to_generic_addr(dev, rqd->ppa_addr);
}
}
EXPORT_SYMBOL(nvm_addr_to_generic_mode);
void nvm_generic_to_addr_mode(struct nvm_dev *dev, struct nvm_rq *rqd)
{
int i;
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
if (rqd->nr_ppas > 1) {
for (i = 0; i < rqd->nr_ppas; i++)
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
rqd->ppa_list[i] = generic_to_dev_addr(dev,
rqd->ppa_list[i]);
} else {
rqd->ppa_addr = generic_to_dev_addr(dev, rqd->ppa_addr);
}
}
EXPORT_SYMBOL(nvm_generic_to_addr_mode);
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
int nvm_set_rqd_ppalist(struct nvm_dev *dev, struct nvm_rq *rqd,
lightnvm: make ppa_list const in nvm_set_rqd_list The passed by reference ppa list in nvm_set_rqd_list() is updated when multiple planes are available. In that case, each PPA plane is incremented when the device side PPA list is created. This prevents the caller to rely on the PPA list to be unmodified after a call. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:22 +08:00
const struct ppa_addr *ppas, int nr_ppas, int vblk)
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
{
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
struct nvm_geo *geo = &dev->geo;
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
int i, plane_cnt, pl_idx;
lightnvm: make ppa_list const in nvm_set_rqd_list The passed by reference ppa list in nvm_set_rqd_list() is updated when multiple planes are available. In that case, each PPA plane is incremented when the device side PPA list is created. This prevents the caller to rely on the PPA list to be unmodified after a call. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:22 +08:00
struct ppa_addr ppa;
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
if ((!vblk || geo->plane_mode == NVM_PLANE_SINGLE) && nr_ppas == 1) {
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
rqd->nr_ppas = nr_ppas;
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
rqd->ppa_addr = ppas[0];
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
return 0;
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
}
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
rqd->nr_ppas = nr_ppas;
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
rqd->ppa_list = nvm_dev_dma_alloc(dev, GFP_KERNEL, &rqd->dma_ppa_list);
if (!rqd->ppa_list) {
pr_err("nvm: failed to allocate dma memory\n");
return -ENOMEM;
}
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
if (!vblk) {
for (i = 0; i < nr_ppas; i++)
rqd->ppa_list[i] = ppas[i];
} else {
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
plane_cnt = geo->plane_mode;
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
rqd->nr_ppas *= plane_cnt;
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
lightnvm: sectors first in ppa list The Westlake controller requires that the PPA list has sectors defined sequentially. Currently, the PPA list is created with planes first, then sectors. Change this to sectors first, then planes. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:26 +08:00
for (i = 0; i < nr_ppas; i++) {
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
for (pl_idx = 0; pl_idx < plane_cnt; pl_idx++) {
lightnvm: make ppa_list const in nvm_set_rqd_list The passed by reference ppa list in nvm_set_rqd_list() is updated when multiple planes are available. In that case, each PPA plane is incremented when the device side PPA list is created. This prevents the caller to rely on the PPA list to be unmodified after a call. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:22 +08:00
ppa = ppas[i];
ppa.g.pl = pl_idx;
rqd->ppa_list[(pl_idx * nr_ppas) + i] = ppa;
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
}
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
}
}
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
return 0;
}
EXPORT_SYMBOL(nvm_set_rqd_ppalist);
void nvm_free_rqd_ppalist(struct nvm_dev *dev, struct nvm_rq *rqd)
{
if (!rqd->ppa_list)
return;
nvm_dev_dma_free(dev, rqd->ppa_list, rqd->dma_ppa_list);
}
EXPORT_SYMBOL(nvm_free_rqd_ppalist);
lightnvm: enable to send hint to erase command Erases might be subject to host hints. An example is multi-plane programming to erase blocks in parallel. Enable targets to specify this hint. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:38:54 +08:00
int nvm_erase_ppa(struct nvm_dev *dev, struct ppa_addr *ppas, int nr_ppas,
int flags)
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
{
struct nvm_rq rqd;
int ret;
if (!dev->ops->erase_block)
return 0;
memset(&rqd, 0, sizeof(struct nvm_rq));
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
ret = nvm_set_rqd_ppalist(dev, &rqd, ppas, nr_ppas, 1);
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
if (ret)
return ret;
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
nvm_generic_to_addr_mode(dev, &rqd);
lightnvm: enable to send hint to erase command Erases might be subject to host hints. An example is multi-plane programming to erase blocks in parallel. Enable targets to specify this hint. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:38:54 +08:00
rqd.flags = flags;
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
ret = dev->ops->erase_block(dev, &rqd);
lightnvm: refactor rqd ppa list into set/free A device may be driven in single, double or quad plane mode. In that case, the rqd must have either one, two, or four PPAs set for a single PPA sent to the device. Refactor this logic into their own functions to be shared by program/erase/read in the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:20 +08:00
nvm_free_rqd_ppalist(dev, &rqd);
lightnvm: move ppa erase logic to core A device may function in single, dual or quad plane mode. The gennvm media manager manages this with explicit helpers. They convert a single ppa to 1, 2 or 4 separate ppas in a ppa list. To aid implementation of recovery and system blocks, this functionality can be moved directly into the core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:19 +08:00
return ret;
}
EXPORT_SYMBOL(nvm_erase_ppa);
lightnvm: refactor end_io functions for sync To implement sync I/O support within the LightNVM core, the end_io functions are refactored to take an end_io function pointer instead of testing for initialized media manager, followed by calling its end_io function. Sync I/O can then be implemented using a callback that signal I/O completion. This is similar to the logic found in blk_to_execute_io(). By implementing it this way, the underlying device I/Os submission logic is abstracted away from core, targets, and media managers. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:21 +08:00
void nvm_end_io(struct nvm_rq *rqd, int error)
{
lightnvm: move rq->error to nvm_rq->error Instead of passing request error into the LightNVM modules, incorporate it into the nvm_rq. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:29 +08:00
rqd->error = error;
rqd->end_io(rqd);
lightnvm: refactor end_io functions for sync To implement sync I/O support within the LightNVM core, the end_io functions are refactored to take an end_io function pointer instead of testing for initialized media manager, followed by calling its end_io function. Sync I/O can then be implemented using a callback that signal I/O completion. This is similar to the logic found in blk_to_execute_io(). By implementing it this way, the underlying device I/Os submission logic is abstracted away from core, targets, and media managers. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:21 +08:00
}
EXPORT_SYMBOL(nvm_end_io);
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
static void nvm_end_io_sync(struct nvm_rq *rqd)
lightnvm: refactor end_io functions for sync To implement sync I/O support within the LightNVM core, the end_io functions are refactored to take an end_io function pointer instead of testing for initialized media manager, followed by calling its end_io function. Sync I/O can then be implemented using a callback that signal I/O completion. This is similar to the logic found in blk_to_execute_io(). By implementing it this way, the underlying device I/Os submission logic is abstracted away from core, targets, and media managers. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:21 +08:00
{
struct completion *waiting = rqd->wait;
rqd->wait = NULL;
complete(waiting);
}
lightnvm: make __nvm_submit_ppa static The __nvm_submit_ppa() function is not used outside lightnvm core. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:23 +08:00
static int __nvm_submit_ppa(struct nvm_dev *dev, struct nvm_rq *rqd, int opcode,
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
int flags, void *buf, int len)
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
{
DECLARE_COMPLETION_ONSTACK(wait);
struct bio *bio;
int ret;
unsigned long hang_check;
bio = bio_map_kern(dev->q, buf, len, GFP_KERNEL);
if (IS_ERR_OR_NULL(bio))
return -ENOMEM;
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
nvm_generic_to_addr_mode(dev, rqd);
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
lightnvm: eliminate nvm_lun abstraction in mm In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. Since targets own the LUNs the are instantiated on top of and manage the free block list internally, there is no need for a LUN abstraction in the media manager. LUNs are intrinsically managed as in the physical layout (ch:0,lun:0, ..., ch:0,lun:n, ch:1,lun:0, ch:1,lun:n, ..., ch:m,lun:0, ch:m,lun:n) and given to the targets based on the target creation ioctl. This simplifies LUN management and clears the path for a partition manager to sit directly underneath LightNVM targets. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:10 +08:00
rqd->dev = NULL;
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
rqd->opcode = opcode;
rqd->flags = flags;
rqd->bio = bio;
rqd->wait = &wait;
rqd->end_io = nvm_end_io_sync;
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
ret = dev->ops->submit_io(dev, rqd);
lightnvm: handle submit_io failure The device ->submit_io() callback might fail to submit I/O to device. In that case, the nvm_submit_ppa function should not wait for completion. Instead return the ->submit_io() error. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:55 +08:00
if (ret) {
bio_put(bio);
return ret;
}
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
/* Prevent hang_check timer from firing at us during very long I/O */
hang_check = sysctl_hung_task_timeout_secs;
if (hang_check)
lightnvm: fix checkpatch terse errors A couple of small checkpatch fixups to stop it from complaining. ./drivers/lightnvm/core.c:360: WARNING: line over 80 characters ./drivers/lightnvm/core.c:360: ERROR: trailing statements should be on next line ./drivers/lightnvm/core.c:503: WARNING: Block comments use a trailing */ on a separate line Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:13 +08:00
while (!wait_for_completion_io_timeout(&wait,
hang_check * (HZ/2)))
;
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
else
wait_for_completion_io(&wait);
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
return rqd->error;
}
/**
* nvm_submit_ppa_list - submit user-defined ppa list to device. The user must
* take to free ppa list if necessary.
* @dev: device
* @ppa_list: user created ppa_list
* @nr_ppas: length of ppa_list
* @opcode: device opcode
* @flags: device flags
* @buf: data buffer
* @len: data buffer length
*/
int nvm_submit_ppa_list(struct nvm_dev *dev, struct ppa_addr *ppa_list,
int nr_ppas, int opcode, int flags, void *buf, int len)
{
struct nvm_rq rqd;
if (dev->ops->max_phys_sect < nr_ppas)
return -EINVAL;
memset(&rqd, 0, sizeof(struct nvm_rq));
lightnvm: rename nr_pages to nr_ppas on nvm_rq The number of ppas contained on a request is not necessarily the number of pages that it maps to neither on the target nor on the device side. In order to avoid confusion, rename nr_pages to nr_ppas since it is what the variable actually contains. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:20 +08:00
rqd.nr_ppas = nr_ppas;
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
if (nr_ppas > 1)
rqd.ppa_list = ppa_list;
else
rqd.ppa_addr = ppa_list[0];
return __nvm_submit_ppa(dev, &rqd, opcode, flags, buf, len);
}
EXPORT_SYMBOL(nvm_submit_ppa_list);
/**
* nvm_submit_ppa - submit PPAs to device. PPAs will automatically be unfolded
* as single, dual, quad plane PPAs depending on device type.
* @dev: device
* @ppa: user created ppa_list
* @nr_ppas: length of ppa_list
* @opcode: device opcode
* @flags: device flags
* @buf: data buffer
* @len: data buffer length
*/
int nvm_submit_ppa(struct nvm_dev *dev, struct ppa_addr *ppa, int nr_ppas,
int opcode, int flags, void *buf, int len)
{
struct nvm_rq rqd;
int ret;
memset(&rqd, 0, sizeof(struct nvm_rq));
lightnvm: make nvm_set_rqd_ppalist() aware of vblks A virtual block enables a block to identify multiple physical blocks. This is useful for metadata where a device media supports multiple planes. In that case, a block, with multiple planes can be managed as a single vblk. Reducing the metadata required by one forth. nvm_set_rqd_ppalist() takes care of expanding a ppa_list with vblks automatically. However, for some use-cases, where only a single physical block is required, the ppa_list should not be expanded. Therefore, add a vblk parameter to nvm_set_rqd_ppalist(), and only expand the ppa_list if vblk is set. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:07 +08:00
ret = nvm_set_rqd_ppalist(dev, &rqd, ppa, nr_ppas, 1);
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
if (ret)
return ret;
ret = __nvm_submit_ppa(dev, &rqd, opcode, flags, buf, len);
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
nvm_free_rqd_ppalist(dev, &rqd);
lightnvm: implement nvm_submit_ppa_list The nvm_submit_ppa function assumes that users manage all plane blocks as a single block. Extend the API with nvm_submit_ppa_list to allow the user to send its own ppa list. If the user submits more than a single PPA, the user must take care to allocate and free the corresponding ppa list. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:56 +08:00
return ret;
lightnvm: introduce nvm_submit_ppa Internal logic for both core and media managers, does not have a backing bio for issuing I/Os. Introduce nvm_submit_ppa to allow raw I/Os to be submitted to the underlying device driver. The function request the device, ppa, data buffer and its length and will submit the I/O synchronously to the device. The return value may therefore be used to detect any errors regarding the issued I/O. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:30 +08:00
}
EXPORT_SYMBOL(nvm_submit_ppa);
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
/*
* folds a bad block list from its plane representation to its virtual
* block representation. The fold is done in place and reduced size is
* returned.
*
* If any of the planes status are bad or grown bad block, the virtual block
* is marked bad. If not bad, the first plane state acts as the block state.
*/
int nvm_bb_tbl_fold(struct nvm_dev *dev, u8 *blks, int nr_blks)
{
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
struct nvm_geo *geo = &dev->geo;
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
int blk, offset, pl, blktype;
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
if (nr_blks != geo->blks_per_lun * geo->plane_mode)
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
return -EINVAL;
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
for (blk = 0; blk < geo->blks_per_lun; blk++) {
offset = blk * geo->plane_mode;
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
blktype = blks[offset];
/* Bad blocks on any planes take precedence over other types */
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
for (pl = 0; pl < geo->plane_mode; pl++) {
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
if (blks[offset + pl] &
(NVM_BLK_T_BAD|NVM_BLK_T_GRWN_BAD)) {
blktype = blks[offset + pl];
break;
}
}
blks[blk] = blktype;
}
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
return geo->blks_per_lun;
lightnvm: move block fold outside of get_bb_tbl() The get block table command returns a list of blocks and planes with their associated state. Users, such as gennvm and sysblk, manages all planes as a single virtual block. It was therefore natural to fold the bad block list before it is returned. However, to allow users, which manages on a per-plane block level, to also use the interface, the get_bb_tbl interface is changed to not fold by default and instead let the caller fold if necessary. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:58 +08:00
}
EXPORT_SYMBOL(nvm_bb_tbl_fold);
lightnvm: refactor device ops->get_bb_tbl() The device ops->get_bb_tbl() takes a callback, that allows the caller to use its own callback function to update its data structures in the returning function. This makes it difficult to send parameters to the callback, and usually is circumvented by small private structures, that both carry the callers state and any flags needed to fulfill the update. Refactor ops->get_bb_tbl() to fill a data buffer with the status of the blocks returned, and let the user call the callback function manually. That will provide the necessary flags and data structures and simplify the logic around ops->get_bb_tbl(). Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:05 +08:00
int nvm_get_bb_tbl(struct nvm_dev *dev, struct ppa_addr ppa, u8 *blks)
{
ppa = generic_to_dev_addr(dev, ppa);
return dev->ops->get_bb_tbl(dev, ppa, blks);
}
EXPORT_SYMBOL(nvm_get_bb_tbl);
lightnvm: transform target get/set bad block Since targets are given a virtual target device, it is necessary to translate all communication between targets and the backend device. Implement the translation layer for get/set bad block table. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:14 +08:00
int nvm_get_tgt_bb_tbl(struct nvm_tgt_dev *tgt_dev, struct ppa_addr ppa,
u8 *blks)
{
struct nvm_dev *dev = tgt_dev->parent;
ppa = dev->mt->trans_ppa(tgt_dev, ppa, TRANS_TGT_TO_DEV);
return nvm_get_bb_tbl(dev, ppa, blks);
}
EXPORT_SYMBOL(nvm_get_tgt_bb_tbl);
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
static int nvm_init_slc_tbl(struct nvm_dev *dev, struct nvm_id_group *grp)
{
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
struct nvm_geo *geo = &dev->geo;
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
int i;
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
dev->lps_per_blk = geo->pgs_per_blk;
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
dev->lptbl = kcalloc(dev->lps_per_blk, sizeof(int), GFP_KERNEL);
if (!dev->lptbl)
return -ENOMEM;
/* Just a linear array */
for (i = 0; i < dev->lps_per_blk; i++)
dev->lptbl[i] = i;
return 0;
}
static int nvm_init_mlc_tbl(struct nvm_dev *dev, struct nvm_id_group *grp)
{
int i, p;
struct nvm_id_lp_mlc *mlc = &grp->lptbl.mlc;
if (!mlc->num_pairs)
return 0;
dev->lps_per_blk = mlc->num_pairs;
dev->lptbl = kcalloc(dev->lps_per_blk, sizeof(int), GFP_KERNEL);
if (!dev->lptbl)
return -ENOMEM;
/* The lower page table encoding consists of a list of bytes, where each
* has a lower and an upper half. The first half byte maintains the
* increment value and every value after is an offset added to the
lightnvm: fix checkpatch terse errors A couple of small checkpatch fixups to stop it from complaining. ./drivers/lightnvm/core.c:360: WARNING: line over 80 characters ./drivers/lightnvm/core.c:360: ERROR: trailing statements should be on next line ./drivers/lightnvm/core.c:503: WARNING: Block comments use a trailing */ on a separate line Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:13 +08:00
* previous incrementation value
*/
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
dev->lptbl[0] = mlc->pairs[0] & 0xF;
for (i = 1; i < dev->lps_per_blk; i++) {
p = mlc->pairs[i >> 1];
if (i & 0x1) /* upper */
dev->lptbl[i] = dev->lptbl[i - 1] + ((p & 0xF0) >> 4);
else /* lower */
dev->lptbl[i] = dev->lptbl[i - 1] + (p & 0xF);
}
return 0;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
static int nvm_core_init(struct nvm_dev *dev)
{
struct nvm_id *id = &dev->identity;
struct nvm_id_group *grp = &id->groups[0];
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
struct nvm_geo *geo = &dev->geo;
lightnvm: avoid memory leak when lun_map kcalloc fails A memory leak occurs if the lower page table is initialized and the following dev->lun_map fails on allocation. Rearrange the initialization of lower page table to allow dev->lun_map to fail gracefully without memory leak. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Move kfree of dev->lun_map to nvm_free() Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:59 +08:00
int ret;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
/* Whole device values */
geo->nr_chnls = grp->num_ch;
geo->luns_per_chnl = grp->num_lun;
/* Generic device values */
geo->pgs_per_blk = grp->num_pg;
geo->blks_per_lun = grp->num_blk;
geo->nr_planes = grp->num_pln;
geo->fpg_size = grp->fpg_sz;
geo->pfpg_size = grp->fpg_sz * grp->num_pln;
geo->sec_size = grp->csecs;
geo->oob_size = grp->sos;
geo->sec_per_pg = grp->fpg_sz / grp->csecs;
geo->mccap = grp->mccap;
memcpy(&geo->ppaf, &id->ppaf, sizeof(struct nvm_addr_format));
geo->plane_mode = NVM_PLANE_SINGLE;
geo->max_rq_size = dev->ops->max_phys_sect * geo->sec_size;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (grp->mpos & 0x020202)
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
geo->plane_mode = NVM_PLANE_DOUBLE;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (grp->mpos & 0x040404)
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
geo->plane_mode = NVM_PLANE_QUAD;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: avoid memory leak when lun_map kcalloc fails A memory leak occurs if the lower page table is initialized and the following dev->lun_map fails on allocation. Rearrange the initialization of lower page table to allow dev->lun_map to fail gracefully without memory leak. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Move kfree of dev->lun_map to nvm_free() Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:59 +08:00
if (grp->mtype != 0) {
pr_err("nvm: memory type not supported\n");
return -EINVAL;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
/* calculated values */
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
geo->sec_per_pl = geo->sec_per_pg * geo->nr_planes;
geo->sec_per_blk = geo->sec_per_pl * geo->pgs_per_blk;
geo->sec_per_lun = geo->sec_per_blk * geo->blks_per_lun;
geo->nr_luns = geo->luns_per_chnl * geo->nr_chnls;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
dev->total_secs = geo->nr_luns * geo->sec_per_lun;
dev->lun_map = kcalloc(BITS_TO_LONGS(geo->nr_luns),
lightnvm: add a bitmap of luns Add a bitmap of luns to indicate the status of luns: inuse/available. When create targets do the necessary check to avoid allocating luns that are already allocated. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Freed dev->lun_map if nvm_core_init later failed in the init process. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-03-03 22:06:38 +08:00
sizeof(unsigned long), GFP_KERNEL);
if (!dev->lun_map)
return -ENOMEM;
lightnvm: avoid memory leak when lun_map kcalloc fails A memory leak occurs if the lower page table is initialized and the following dev->lun_map fails on allocation. Rearrange the initialization of lower page table to allow dev->lun_map to fail gracefully without memory leak. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Move kfree of dev->lun_map to nvm_free() Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:59 +08:00
switch (grp->fmtype) {
case NVM_ID_FMTYPE_SLC:
if (nvm_init_slc_tbl(dev, grp)) {
ret = -ENOMEM;
goto err_fmtype;
}
break;
case NVM_ID_FMTYPE_MLC:
if (nvm_init_mlc_tbl(dev, grp)) {
ret = -ENOMEM;
goto err_fmtype;
}
break;
default:
pr_err("nvm: flash type not supported\n");
ret = -EINVAL;
goto err_fmtype;
}
lightnvm: core on-disk initialization An Open-Channel SSD shall be initialized before use. To initialize, we define an on-disk format, that keeps a small set of metadata to bring up the media manager on top of the device. The initial step is introduced to allow a user to format the disks for a given media manager. During format, a system block is stored on one to three separate luns on the device. Each lun has the system block duplicated. During initialization, the system block can be retrieved and the appropriate media manager can initialized. The on-disk format currently covers (struct nvm_system_block): - Magic value "NVMS". - Monotonic increasing sequence number. - The physical block erase count. - Version of the system block format. - Media manager type. - Media manager superblock physical address. The interface provides three functions to manage the system block: int nvm_init_sysblock(struct nvm_dev *, struct nvm_sb_info *) int nvm_get_sysblock(struct nvm *dev, struct nvm_sb_info *) int nvm_update_sysblock(struct nvm *dev, struct nvm_sb_info *) Each implement a part of the logic to manage the system block. The initialization creates the first system blocks and mark them on the device. Get retrieves the latest system block by scanning all pages in the associated system blocks. The update sysblock writes new metadata and allocates new block if necessary. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:36 +08:00
mutex_init(&dev->mlock);
lightnvm: specify target's logical address area We can create more than one target on a lightnvm device by specifying its begin lun and end lun. But only specify the physical address area is not enough, we need to get the corresponding non- intersection logical address area division from the backend device's logcial address space. Otherwise the targets on the device might use the same logical addresses cause incorrect information in the device's l2p table. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-03-03 22:06:37 +08:00
spin_lock_init(&dev->lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
blk_queue_logical_block_size(dev->q, geo->sec_size);
nvme: refactor namespaces to support non-gendisk devices With LightNVM enabled namespaces, the gendisk structure is not exposed to the user. This prevents LightNVM users from accessing the NVMe device driver specific sysfs entries, and LightNVM namespace geometry. Refactor the revalidation process, so that a namespace, instead of a gendisk, is revalidated. This later allows patches to wire up the sysfs entries up to a non-gendisk namespace. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:04 +08:00
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return 0;
lightnvm: avoid memory leak when lun_map kcalloc fails A memory leak occurs if the lower page table is initialized and the following dev->lun_map fails on allocation. Rearrange the initialization of lower page table to allow dev->lun_map to fail gracefully without memory leak. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Move kfree of dev->lun_map to nvm_free() Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:59 +08:00
err_fmtype:
kfree(dev->lun_map);
return ret;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
lightnvm: remove mgt targets on mgt removal Targets associated with a device manager are not freed on device removal. They have to be manually removed before shutdown. Make sure any outstanding targets are freed upon shutdown. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:17 +08:00
static void nvm_free_mgr(struct nvm_dev *dev)
{
if (!dev->mt)
return;
dev->mt->unregister_mgr(dev);
dev->mt = NULL;
}
lightnvm: expose device geometry through sysfs For a host to access an Open-Channel SSD, it has to know its geometry, so that it writes and reads at the appropriate device bounds. Currently, the geometry information is kept within the kernel, and not exported to user-space for consumption. This patch exposes the configuration through sysfs and enables user-space libraries, such as liblightnvm, to use the sysfs implementation to get the geometry of an Open-Channel SSD. The sysfs entries are stored within the device hierarchy, and can be found using the "lightnvm" device type. An example configuration looks like this: /sys/class/nvme/ └── nvme0n1 ├── capabilities: 3 ├── device_mode: 1 ├── erase_max: 1000000 ├── erase_typ: 1000000 ├── flash_media_type: 0 ├── media_capabilities: 0x00000001 ├── media_type: 0 ├── multiplane: 0x00010101 ├── num_blocks: 1022 ├── num_channels: 1 ├── num_luns: 4 ├── num_pages: 64 ├── num_planes: 1 ├── page_size: 4096 ├── prog_max: 100000 ├── prog_typ: 100000 ├── read_max: 10000 ├── read_typ: 10000 ├── sector_oob_size: 0 ├── sector_size: 4096 ├── media_manager: gennvm ├── ppa_format: 0x380830082808001010102008 ├── vendor_opcode: 0 ├── max_phys_secs: 64 └── version: 1 Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:08 +08:00
void nvm_free(struct nvm_dev *dev)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
if (!dev)
return;
lightnvm: remove mgt targets on mgt removal Targets associated with a device manager are not freed on device removal. They have to be manually removed before shutdown. Make sure any outstanding targets are freed upon shutdown. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:17 +08:00
nvm_free_mgr(dev);
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
lightnvm: expose device geometry through sysfs For a host to access an Open-Channel SSD, it has to know its geometry, so that it writes and reads at the appropriate device bounds. Currently, the geometry information is kept within the kernel, and not exported to user-space for consumption. This patch exposes the configuration through sysfs and enables user-space libraries, such as liblightnvm, to use the sysfs implementation to get the geometry of an Open-Channel SSD. The sysfs entries are stored within the device hierarchy, and can be found using the "lightnvm" device type. An example configuration looks like this: /sys/class/nvme/ └── nvme0n1 ├── capabilities: 3 ├── device_mode: 1 ├── erase_max: 1000000 ├── erase_typ: 1000000 ├── flash_media_type: 0 ├── media_capabilities: 0x00000001 ├── media_type: 0 ├── multiplane: 0x00010101 ├── num_blocks: 1022 ├── num_channels: 1 ├── num_luns: 4 ├── num_pages: 64 ├── num_planes: 1 ├── page_size: 4096 ├── prog_max: 100000 ├── prog_typ: 100000 ├── read_max: 10000 ├── read_typ: 10000 ├── sector_oob_size: 0 ├── sector_size: 4096 ├── media_manager: gennvm ├── ppa_format: 0x380830082808001010102008 ├── vendor_opcode: 0 ├── max_phys_secs: 64 └── version: 1 Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:08 +08:00
if (dev->dma_pool)
dev->ops->destroy_dma_pool(dev->dma_pool);
lightnvm: introduce mlc lower page table mappings NAND MLC memories have both lower and upper pages. When programming, both of these must be written, before data can be read. However, these lower and upper pages might not placed at even and odd flash pages, but can be skipped. Therefore each flash memory has its lower pages defined, which can then be used when programming and to know when padding are necessary. This patch implements the lower page definition in the specification, and exposes it through a simple lookup table at dev->lptbl. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:35 +08:00
kfree(dev->lptbl);
lightnvm: avoid memory leak when lun_map kcalloc fails A memory leak occurs if the lower page table is initialized and the following dev->lun_map fails on allocation. Rearrange the initialization of lower page table to allow dev->lun_map to fail gracefully without memory leak. Reviewed by: Johannes Thumshirn <jthumshirn@suse.de> Move kfree of dev->lun_map to nvm_free() Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:02:59 +08:00
kfree(dev->lun_map);
lightnvm: expose device geometry through sysfs For a host to access an Open-Channel SSD, it has to know its geometry, so that it writes and reads at the appropriate device bounds. Currently, the geometry information is kept within the kernel, and not exported to user-space for consumption. This patch exposes the configuration through sysfs and enables user-space libraries, such as liblightnvm, to use the sysfs implementation to get the geometry of an Open-Channel SSD. The sysfs entries are stored within the device hierarchy, and can be found using the "lightnvm" device type. An example configuration looks like this: /sys/class/nvme/ └── nvme0n1 ├── capabilities: 3 ├── device_mode: 1 ├── erase_max: 1000000 ├── erase_typ: 1000000 ├── flash_media_type: 0 ├── media_capabilities: 0x00000001 ├── media_type: 0 ├── multiplane: 0x00010101 ├── num_blocks: 1022 ├── num_channels: 1 ├── num_luns: 4 ├── num_pages: 64 ├── num_planes: 1 ├── page_size: 4096 ├── prog_max: 100000 ├── prog_typ: 100000 ├── read_max: 10000 ├── read_typ: 10000 ├── sector_oob_size: 0 ├── sector_size: 4096 ├── media_manager: gennvm ├── ppa_format: 0x380830082808001010102008 ├── vendor_opcode: 0 ├── max_phys_secs: 64 └── version: 1 Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:08 +08:00
kfree(dev);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
static int nvm_init(struct nvm_dev *dev)
{
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
struct nvm_geo *geo = &dev->geo;
lightnvm: wrong return value and redundant free The return value should be non-zero under error conditions. Remove nvme_free(dev) to avoid free dev more than once. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-20 20:47:53 +08:00
int ret = -EINVAL;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (!dev->q || !dev->ops)
lightnvm: wrong return value and redundant free The return value should be non-zero under error conditions. Remove nvme_free(dev) to avoid free dev more than once. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-20 20:47:53 +08:00
return ret;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: replace req queue with nvmdev for lld In the case where a request queue is passed to the low lever lightnvm device drive integration, the device driver might pass its admin commands through another queue. Instead pass nvm_dev, and let the low level drive the appropriate queue. Reported-by: Christoph Hellwig <hch@infradead.org> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:48 +08:00
if (dev->ops->identity(dev, &dev->identity)) {
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
pr_err("nvm: device could not be identified\n");
goto err;
}
pr_debug("nvm: ver:%x nvm_vendor:%x groups:%u\n",
dev->identity.ver_id, dev->identity.vmnt,
dev->identity.cgrps);
if (dev->identity.ver_id != 1) {
pr_err("nvm: device not supported by kernel.");
goto err;
}
if (dev->identity.cgrps != 1) {
pr_err("nvm: only one group configuration supported.");
goto err;
}
ret = nvm_core_init(dev);
if (ret) {
pr_err("nvm: could not initialize core structures.\n");
goto err;
}
pr_info("nvm: registered %s [%u/%u/%u/%u/%u/%u]\n",
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
dev->name, geo->sec_per_pg, geo->nr_planes,
geo->pgs_per_blk, geo->blks_per_lun,
geo->nr_luns, geo->nr_chnls);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return 0;
err:
pr_err("nvm: failed to initialize nvm\n");
return ret;
}
lightnvm: control life of nvm_dev in driver LightNVM compatible device drivers does not have a method to expose LightNVM specific sysfs entries. To enable LightNVM sysfs entries to be exposed, lightnvm device drivers require a struct device to attach it to. To allow both the actual device driver and lightnvm sysfs entries to coexist, the device driver tracks the lifetime of the nvm_dev structure. This patch refactors NVMe and null_blk to handle the lifetime of struct nvm_dev, which eliminates the need for struct gendisk when a lightnvm compatible device is provided. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:07 +08:00
struct nvm_dev *nvm_alloc_dev(int node)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
lightnvm: control life of nvm_dev in driver LightNVM compatible device drivers does not have a method to expose LightNVM specific sysfs entries. To enable LightNVM sysfs entries to be exposed, lightnvm device drivers require a struct device to attach it to. To allow both the actual device driver and lightnvm sysfs entries to coexist, the device driver tracks the lifetime of the nvm_dev structure. This patch refactors NVMe and null_blk to handle the lifetime of struct nvm_dev, which eliminates the need for struct gendisk when a lightnvm compatible device is provided. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:07 +08:00
return kzalloc_node(sizeof(struct nvm_dev), GFP_KERNEL, node);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
lightnvm: control life of nvm_dev in driver LightNVM compatible device drivers does not have a method to expose LightNVM specific sysfs entries. To enable LightNVM sysfs entries to be exposed, lightnvm device drivers require a struct device to attach it to. To allow both the actual device driver and lightnvm sysfs entries to coexist, the device driver tracks the lifetime of the nvm_dev structure. This patch refactors NVMe and null_blk to handle the lifetime of struct nvm_dev, which eliminates the need for struct gendisk when a lightnvm compatible device is provided. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:07 +08:00
EXPORT_SYMBOL(nvm_alloc_dev);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
lightnvm: control life of nvm_dev in driver LightNVM compatible device drivers does not have a method to expose LightNVM specific sysfs entries. To enable LightNVM sysfs entries to be exposed, lightnvm device drivers require a struct device to attach it to. To allow both the actual device driver and lightnvm sysfs entries to coexist, the device driver tracks the lifetime of the nvm_dev structure. This patch refactors NVMe and null_blk to handle the lifetime of struct nvm_dev, which eliminates the need for struct gendisk when a lightnvm compatible device is provided. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:07 +08:00
int nvm_register(struct nvm_dev *dev)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
int ret;
ret = nvm_init(dev);
if (ret)
goto err_init;
lightnvm: do device max sectors boundary check first do device max_phys_sect boundary check first, otherwise we will allocate dma_pools for devices whose max sectors are beyond lightnvm support and register them. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:25 +08:00
if (dev->ops->max_phys_sect > 256) {
pr_info("nvm: max sectors supported is 256.\n");
ret = -EINVAL;
goto err_init;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (dev->ops->max_phys_sect > 1) {
lightnvm: rename dma helper functions Until now, the dma pool have been exclusively used to allocate the ppa list being sent to the device. In pblk (upcoming), we use these pools to allocate metadata too. Thus, we generalize the names of some variables on the dma helper functions to make the code more readable. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:13 +08:00
dev->dma_pool = dev->ops->create_dma_pool(dev, "ppalist");
if (!dev->dma_pool) {
pr_err("nvm: could not create dma pool\n");
lightnvm: missing free on init error If either max_phys_sect is out of bound, the nvm_dev structure is not freed. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-20 20:47:54 +08:00
ret = -ENOMEM;
goto err_init;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
}
lightnvm: allow to force mm initialization System block allows the device to initialize with its configured media manager. The system blocks is written to disk, and read again when media manager is determined. For this to work, the backend must store the data. Device drivers, such as null_blk, does not have any backend storage. This patch allows the media manager to be initialized without a storage backend. It also fix incorrect configuration of capabilities in null_blk, as it does not support get/set bad block interface. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-02-04 22:13:27 +08:00
if (dev->identity.cap & NVM_ID_DCAP_BBLKMGMT) {
ret = nvm_get_sysblock(dev, &dev->sb);
if (!ret)
pr_err("nvm: device not initialized.\n");
else if (ret < 0)
pr_err("nvm: err (%d) on device initialization\n", ret);
}
lightnvm: use system block for mm initialization Use system block information to register the appropriate media manager. This enables the LightNVM subsystem to instantiate a media manager selected by the user, instead of relying on automatic detection by each media manager loaded in the kernel. A device must now be initialized before it can proceed to initialize its media manager. Upon initialization, the configured media manager is automatically initialized as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:38 +08:00
lightnvm: fix media mgr registration This patch fixes two issues during media manager registration. 1. The ppa pool can be used at media manager registration. Allocate the ppa pool before that. 2. If a media manager can't be found, this should not lead to the device being unallocated. A media manager can be registered later, that can manage the device. Only warn if a media manager fails initialization. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-12-06 18:25:49 +08:00
/* register device with a supported media manager */
lightnvm: prematurely activate nvm_dev We register with nvm_devices when there registration can still fail. Move the final registration at the end of the nvm_register function to make sure we are fully registered when added to the nvm_devices list. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-16 22:34:42 +08:00
down_write(&nvm_lock);
lightnvm: use system block for mm initialization Use system block information to register the appropriate media manager. This enables the LightNVM subsystem to instantiate a media manager selected by the user, instead of relying on automatic detection by each media manager loaded in the kernel. A device must now be initialized before it can proceed to initialize its media manager. Upon initialization, the configured media manager is automatically initialized as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:38 +08:00
if (ret > 0)
dev->mt = nvm_init_mgr(dev);
lightnvm: prematurely activate nvm_dev We register with nvm_devices when there registration can still fail. Move the final registration at the end of the nvm_register function to make sure we are fully registered when added to the nvm_devices list. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-16 22:34:42 +08:00
list_add(&dev->devices, &nvm_devices);
up_write(&nvm_lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return 0;
err_init:
lightnvm: add a bitmap of luns Add a bitmap of luns to indicate the status of luns: inuse/available. When create targets do the necessary check to avoid allocating luns that are already allocated. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Freed dev->lun_map if nvm_core_init later failed in the init process. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-03-03 22:06:38 +08:00
kfree(dev->lun_map);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return ret;
}
EXPORT_SYMBOL(nvm_register);
lightnvm: control life of nvm_dev in driver LightNVM compatible device drivers does not have a method to expose LightNVM specific sysfs entries. To enable LightNVM sysfs entries to be exposed, lightnvm device drivers require a struct device to attach it to. To allow both the actual device driver and lightnvm sysfs entries to coexist, the device driver tracks the lifetime of the nvm_dev structure. This patch refactors NVMe and null_blk to handle the lifetime of struct nvm_dev, which eliminates the need for struct gendisk when a lightnvm compatible device is provided. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-09-16 20:25:07 +08:00
void nvm_unregister(struct nvm_dev *dev)
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
{
lightnvm: missing nvm_lock acquire To avoid race conditions, traverse dev, media manager, and target lists and also register, unregister entries to/from them, should be always under the nvm_lock control. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:28 +08:00
down_write(&nvm_lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
list_del(&dev->devices);
up_write(&nvm_lock);
lightnvm: prevent double free on init error Both the nvm_register and nvm_init does a kfree(dev) on error. Make sure to only free it once. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-16 22:34:43 +08:00
nvme: lightnvm: attach lightnvm sysfs to nvme block device Previously, LBA read and write were not supported in the lightnvm specification. Now that it supports it, lets use the traditional NVMe gendisk, and attach the lightnvm sysfs geometry export. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:38:53 +08:00
nvm_free(dev);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
EXPORT_SYMBOL(nvm_unregister);
static int __nvm_configure_create(struct nvm_ioctl_create *create)
{
struct nvm_dev *dev;
struct nvm_ioctl_create_simple *s;
lightnvm: missing nvm_lock acquire To avoid race conditions, traverse dev, media manager, and target lists and also register, unregister entries to/from them, should be always under the nvm_lock control. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:28 +08:00
down_write(&nvm_lock);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
dev = nvm_find_nvm_dev(create->dev);
lightnvm: missing nvm_lock acquire To avoid race conditions, traverse dev, media manager, and target lists and also register, unregister entries to/from them, should be always under the nvm_lock control. Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:28 +08:00
up_write(&nvm_lock);
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (!dev) {
pr_err("nvm: device not found\n");
return -EINVAL;
}
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
if (!dev->mt) {
pr_info("nvm: device has no media manager registered.\n");
return -ENODEV;
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (create->conf.type != NVM_CONFIG_TYPE_SIMPLE) {
pr_err("nvm: config type not valid\n");
return -EINVAL;
}
s = &create->conf.s;
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
if (s->lun_begin > s->lun_end || s->lun_end > dev->geo.nr_luns) {
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
pr_err("nvm: lun out of bound (%u:%u > %u)\n",
lightnvm: move block provisioning to targets In order to naturally support multi-target instances on an Open-Channel SSD, targets should own the LUNs they get blocks from and manage provisioning internally. This is done in several steps. This patch moves the block provisioning inside of the target and removes the get/put block interface from the media manager. Signed-off-by: Javier González <javier@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-11-29 05:39:06 +08:00
s->lun_begin, s->lun_end, dev->geo.nr_luns);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return -EINVAL;
}
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
return dev->mt->create_tgt(dev, create);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
static long nvm_ioctl_info(struct file *file, void __user *arg)
{
struct nvm_ioctl_info *info;
struct nvm_tgt_type *tt;
int tgt_iter = 0;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
info = memdup_user(arg, sizeof(struct nvm_ioctl_info));
if (IS_ERR(info))
return -EFAULT;
info->version[0] = NVM_VERSION_MAJOR;
info->version[1] = NVM_VERSION_MINOR;
info->version[2] = NVM_VERSION_PATCH;
down_write(&nvm_lock);
lightnvm: rename nvm_targets to nvm_tgt_type The functions nvm_register_target(), nvm_unregister_target() and associated list refers to a target type that is being registered by a target type module. Rename nvm_*_targets() to nvm_*_tgt_type(), so that the intension is clear. This enables target instances to use the _nvm_*_targets() naming. Signed-off-by: Simon A. F. Lund <slund@cnexlabs.com> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:02 +08:00
list_for_each_entry(tt, &nvm_tgt_types, list) {
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
struct nvm_ioctl_info_tgt *tgt = &info->tgts[tgt_iter];
tgt->version[0] = tt->version[0];
tgt->version[1] = tt->version[1];
tgt->version[2] = tt->version[2];
strncpy(tgt->tgtname, tt->name, NVM_TTYPE_NAME_MAX);
tgt_iter++;
}
info->tgtsize = tgt_iter;
up_write(&nvm_lock);
lightnvm: fix ioctl memory leaks If copy_to_user() fails we returned error but we missed releasing devices. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:24 +08:00
if (copy_to_user(arg, info, sizeof(struct nvm_ioctl_info))) {
kfree(info);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return -EFAULT;
lightnvm: fix ioctl memory leaks If copy_to_user() fails we returned error but we missed releasing devices. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:24 +08:00
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
kfree(info);
return 0;
}
static long nvm_ioctl_get_devices(struct file *file, void __user *arg)
{
struct nvm_ioctl_get_devices *devices;
struct nvm_dev *dev;
int i = 0;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
devices = kzalloc(sizeof(struct nvm_ioctl_get_devices), GFP_KERNEL);
if (!devices)
return -ENOMEM;
down_write(&nvm_lock);
list_for_each_entry(dev, &nvm_devices, devices) {
struct nvm_ioctl_device_info *info = &devices->info[i];
sprintf(info->devname, "%s", dev->name);
if (dev->mt) {
info->bmversion[0] = dev->mt->version[0];
info->bmversion[1] = dev->mt->version[1];
info->bmversion[2] = dev->mt->version[2];
sprintf(info->bmname, "%s", dev->mt->name);
} else {
sprintf(info->bmname, "none");
}
i++;
if (i > 31) {
pr_err("nvm: max 31 devices can be reported.\n");
break;
}
}
up_write(&nvm_lock);
devices->nr_devices = i;
lightnvm: fix ioctl memory leaks If copy_to_user() fails we returned error but we missed releasing devices. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:24 +08:00
if (copy_to_user(arg, devices,
sizeof(struct nvm_ioctl_get_devices))) {
kfree(devices);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
return -EFAULT;
lightnvm: fix ioctl memory leaks If copy_to_user() fails we returned error but we missed releasing devices. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-11-28 23:49:24 +08:00
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
kfree(devices);
return 0;
}
static long nvm_ioctl_dev_create(struct file *file, void __user *arg)
{
struct nvm_ioctl_create create;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&create, arg, sizeof(struct nvm_ioctl_create)))
return -EFAULT;
create.dev[DISK_NAME_LEN - 1] = '\0';
create.tgttype[NVM_TTYPE_NAME_MAX - 1] = '\0';
create.tgtname[DISK_NAME_LEN - 1] = '\0';
if (create.flags != 0) {
pr_err("nvm: no flags supported\n");
return -EINVAL;
}
return __nvm_configure_create(&create);
}
static long nvm_ioctl_dev_remove(struct file *file, void __user *arg)
{
struct nvm_ioctl_remove remove;
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
struct nvm_dev *dev;
int ret = 0;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&remove, arg, sizeof(struct nvm_ioctl_remove)))
return -EFAULT;
remove.tgtname[DISK_NAME_LEN - 1] = '\0';
if (remove.flags != 0) {
pr_err("nvm: no flags supported\n");
return -EINVAL;
}
lightnvm: move target mgmt into media mgr To enable persistent block management to easily control creation and removal of targets, we move target management into the media manager. The LightNVM core continues to maintain which target types are registered, while the media manager now keeps track of its initialized targets. Two new callbacks for the media manager are introduced. create_tgt and remove_tgt. Note that remove_tgt returns 0 on successfully removing a target, and returns 1 if the target was not found. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-07-07 15:54:16 +08:00
list_for_each_entry(dev, &nvm_devices, devices) {
ret = dev->mt->remove_tgt(dev, &remove);
if (!ret)
break;
}
return ret;
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
lightnvm: introduce ioctl to initialize device Based on the previous patch, we now introduce an ioctl to initialize the device using nvm_init_sysblock and create the necessary system blocks. The user may specify the media manager that they wish to instantiate on top. Default from user-space will be "gennvm". Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:37 +08:00
static void nvm_setup_nvm_sb_info(struct nvm_sb_info *info)
{
info->seqnr = 1;
info->erase_cnt = 0;
info->version = 1;
}
static long __nvm_ioctl_dev_init(struct nvm_ioctl_dev_init *init)
{
struct nvm_dev *dev;
struct nvm_sb_info info;
lightnvm: use system block for mm initialization Use system block information to register the appropriate media manager. This enables the LightNVM subsystem to instantiate a media manager selected by the user, instead of relying on automatic detection by each media manager loaded in the kernel. A device must now be initialized before it can proceed to initialize its media manager. Upon initialization, the configured media manager is automatically initialized as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:38 +08:00
int ret;
lightnvm: introduce ioctl to initialize device Based on the previous patch, we now introduce an ioctl to initialize the device using nvm_init_sysblock and create the necessary system blocks. The user may specify the media manager that they wish to instantiate on top. Default from user-space will be "gennvm". Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:37 +08:00
down_write(&nvm_lock);
dev = nvm_find_nvm_dev(init->dev);
up_write(&nvm_lock);
if (!dev) {
pr_err("nvm: device not found\n");
return -EINVAL;
}
nvm_setup_nvm_sb_info(&info);
strncpy(info.mmtype, init->mmtype, NVM_MMTYPE_LEN);
info.fs_ppa.ppa = -1;
lightnvm: allow to force mm initialization System block allows the device to initialize with its configured media manager. The system blocks is written to disk, and read again when media manager is determined. For this to work, the backend must store the data. Device drivers, such as null_blk, does not have any backend storage. This patch allows the media manager to be initialized without a storage backend. It also fix incorrect configuration of capabilities in null_blk, as it does not support get/set bad block interface. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-02-04 22:13:27 +08:00
if (dev->identity.cap & NVM_ID_DCAP_BBLKMGMT) {
ret = nvm_init_sysblock(dev, &info);
if (ret)
return ret;
}
lightnvm: use system block for mm initialization Use system block information to register the appropriate media manager. This enables the LightNVM subsystem to instantiate a media manager selected by the user, instead of relying on automatic detection by each media manager loaded in the kernel. A device must now be initialized before it can proceed to initialize its media manager. Upon initialization, the configured media manager is automatically initialized as well. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:38 +08:00
memcpy(&dev->sb, &info, sizeof(struct nvm_sb_info));
down_write(&nvm_lock);
dev->mt = nvm_init_mgr(dev);
up_write(&nvm_lock);
return 0;
lightnvm: introduce ioctl to initialize device Based on the previous patch, we now introduce an ioctl to initialize the device using nvm_init_sysblock and create the necessary system blocks. The user may specify the media manager that they wish to instantiate on top. Default from user-space will be "gennvm". Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:37 +08:00
}
static long nvm_ioctl_dev_init(struct file *file, void __user *arg)
{
struct nvm_ioctl_dev_init init;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&init, arg, sizeof(struct nvm_ioctl_dev_init)))
return -EFAULT;
if (init.flags != 0) {
pr_err("nvm: no flags supported\n");
return -EINVAL;
}
init.dev[DISK_NAME_LEN - 1] = '\0';
return __nvm_ioctl_dev_init(&init);
}
lightnvm: introduce factory reset Now that a device can be managed using the system blocks, a method to reset the device is necessary as well. This patch introduces logic to reset the device easily to factory state and exposes it through an ioctl. The ioctl takes the following flags: NVM_FACTORY_ERASE_ONLY_USER By default all blocks, except host-reserved blocks are erased upon factory reset. Instead of this, only erase host-reserved blocks. NVM_FACTORY_RESET_HOST_BLKS Mark host-reserved blocks to be erased and set their type to free. NVM_FACTORY_RESET_GRWN_BBLKS Mark "grown bad blocks" to be erased and set their type to free. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:39 +08:00
static long nvm_ioctl_dev_factory(struct file *file, void __user *arg)
{
struct nvm_ioctl_dev_factory fact;
struct nvm_dev *dev;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&fact, arg, sizeof(struct nvm_ioctl_dev_factory)))
return -EFAULT;
fact.dev[DISK_NAME_LEN - 1] = '\0';
if (fact.flags & ~(NVM_FACTORY_NR_BITS - 1))
return -EINVAL;
down_write(&nvm_lock);
dev = nvm_find_nvm_dev(fact.dev);
up_write(&nvm_lock);
if (!dev) {
pr_err("nvm: device not found\n");
return -EINVAL;
}
lightnvm: remove mgt targets on mgt removal Targets associated with a device manager are not freed on device removal. They have to be manually removed before shutdown. Make sure any outstanding targets are freed upon shutdown. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-05-07 02:03:17 +08:00
nvm_free_mgr(dev);
lightnvm: introduce factory reset Now that a device can be managed using the system blocks, a method to reset the device is necessary as well. This patch introduces logic to reset the device easily to factory state and exposes it through an ioctl. The ioctl takes the following flags: NVM_FACTORY_ERASE_ONLY_USER By default all blocks, except host-reserved blocks are erased upon factory reset. Instead of this, only erase host-reserved blocks. NVM_FACTORY_RESET_HOST_BLKS Mark host-reserved blocks to be erased and set their type to free. NVM_FACTORY_RESET_GRWN_BBLKS Mark "grown bad blocks" to be erased and set their type to free. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:39 +08:00
lightnvm: allow to force mm initialization System block allows the device to initialize with its configured media manager. The system blocks is written to disk, and read again when media manager is determined. For this to work, the backend must store the data. Device drivers, such as null_blk, does not have any backend storage. This patch allows the media manager to be initialized without a storage backend. It also fix incorrect configuration of capabilities in null_blk, as it does not support get/set bad block interface. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-02-04 22:13:27 +08:00
if (dev->identity.cap & NVM_ID_DCAP_BBLKMGMT)
return nvm_dev_factory(dev, fact.flags);
return 0;
lightnvm: introduce factory reset Now that a device can be managed using the system blocks, a method to reset the device is necessary as well. This patch introduces logic to reset the device easily to factory state and exposes it through an ioctl. The ioctl takes the following flags: NVM_FACTORY_ERASE_ONLY_USER By default all blocks, except host-reserved blocks are erased upon factory reset. Instead of this, only erase host-reserved blocks. NVM_FACTORY_RESET_HOST_BLKS Mark host-reserved blocks to be erased and set their type to free. NVM_FACTORY_RESET_GRWN_BBLKS Mark "grown bad blocks" to be erased and set their type to free. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:39 +08:00
}
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
static long nvm_ctl_ioctl(struct file *file, uint cmd, unsigned long arg)
{
void __user *argp = (void __user *)arg;
switch (cmd) {
case NVM_INFO:
return nvm_ioctl_info(file, argp);
case NVM_GET_DEVICES:
return nvm_ioctl_get_devices(file, argp);
case NVM_DEV_CREATE:
return nvm_ioctl_dev_create(file, argp);
case NVM_DEV_REMOVE:
return nvm_ioctl_dev_remove(file, argp);
lightnvm: introduce ioctl to initialize device Based on the previous patch, we now introduce an ioctl to initialize the device using nvm_init_sysblock and create the necessary system blocks. The user may specify the media manager that they wish to instantiate on top. Default from user-space will be "gennvm". Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:37 +08:00
case NVM_DEV_INIT:
return nvm_ioctl_dev_init(file, argp);
lightnvm: introduce factory reset Now that a device can be managed using the system blocks, a method to reset the device is necessary as well. This patch introduces logic to reset the device easily to factory state and exposes it through an ioctl. The ioctl takes the following flags: NVM_FACTORY_ERASE_ONLY_USER By default all blocks, except host-reserved blocks are erased upon factory reset. Instead of this, only erase host-reserved blocks. NVM_FACTORY_RESET_HOST_BLKS Mark host-reserved blocks to be erased and set their type to free. NVM_FACTORY_RESET_GRWN_BBLKS Mark "grown bad blocks" to be erased and set their type to free. Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2016-01-12 14:49:39 +08:00
case NVM_DEV_FACTORY:
return nvm_ioctl_dev_factory(file, argp);
lightnvm: Support for Open-Channel SSDs Open-channel SSDs are devices that share responsibilities with the host in order to implement and maintain features that typical SSDs keep strictly in firmware. These include (i) the Flash Translation Layer (FTL), (ii) bad block management, and (iii) hardware units such as the flash controller, the interface controller, and large amounts of flash chips. In this way, Open-channels SSDs exposes direct access to their physical flash storage, while keeping a subset of the internal features of SSDs. LightNVM is a specification that gives support to Open-channel SSDs LightNVM allows the host to manage data placement, garbage collection, and parallelism. Device specific responsibilities such as bad block management, FTL extensions to support atomic IOs, or metadata persistence are still handled by the device. The implementation of LightNVM consists of two parts: core and (multiple) targets. The core implements functionality shared across targets. This is initialization, teardown and statistics. The targets implement the interface that exposes physical flash to user-space applications. Examples of such targets include key-value store, object-store, as well as traditional block devices, which can be application-specific. Contributions in this patch from: Javier Gonzalez <jg@lightnvm.io> Dongsheng Yang <yangds.fnst@cn.fujitsu.com> Jesper Madsen <jmad@itu.dk> Signed-off-by: Matias Bjørling <m@bjorling.me> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-10-29 02:54:55 +08:00
}
return 0;
}
static const struct file_operations _ctl_fops = {
.open = nonseekable_open,
.unlocked_ioctl = nvm_ctl_ioctl,
.owner = THIS_MODULE,
.llseek = noop_llseek,
};
static struct miscdevice _nvm_misc = {
.minor = MISC_DYNAMIC_MINOR,
.name = "lightnvm",
.nodename = "lightnvm/control",
.fops = &_ctl_fops,
};
lightnvm: make core.c explicitly non-modular The Kconfig currently controlling compilation of this code is: drivers/lightnvm/Kconfig:menuconfig NVM drivers/lightnvm/Kconfig: bool "Open-Channel SSD target support" ...meaning that it currently is not being built as a module by anyone. Lets remove the modular code that is essentially orphaned, so that when reading the driver there is no doubt it is builtin-only. Since module_misc_driver translates to device_initcall in the non-modular case, the init ordering remains unchanged with this commit. We also delete the MODULE_LICENSE tag etc. since all that information is already contained at the top of the file in the comments. We replace module.h with moduleparam.h because this file still uses module params to control behaviour. Also note that MODULE_ALIAS is a no-op for non-modular code. Cc: Matias Bjorling <mb@lightnvm.io> Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-10-30 04:38:41 +08:00
builtin_misc_device(_nvm_misc);