[GIT PULL 00/17] LightNVM pull request for 4.8

[GIT PULL 00/17] LightNVM pull request for 4.8

[Matias Bjørling]

Hi Jens,

Please pick up the following patches.

Nothing major. A couple of bigger pieces are being prepared. Notably
pblk target, sysfs support and persistent block information.

The patches are also available from here:

  https://github.com/OpenChannelSSD/linux.git for-jens

Thank you,
Matias

Javier González (2):
  lightnvm: add media manager mark_blk helper
  lightnvm: initialize ppa_addr in dev_to_generic_addr()

Johannes Thumshirn (2):
  nvme: lightnvm: make MLC num_pairs little endian
  lightnvm: Make functions not used by ouside static

Matias Bjørling (12):
  lightnvm: remove checkpatch warning for unsigned ints
  lightnvm: fix checkpatch terse errors
  lightnvm: remove open/close statistics for gennvm
  lightnvm: rename gennvm and update description
  lightnvm: move target mgmt into media mgr
  lightnvm: remove nested lock conflict with mm
  lightnvm: remove unused lists from struct rrpc_block
  lightnvm: remove _unlocked variant of [get/put]_blk
  lightnvm: make rrpc_map_page call nvm_get_blk outside locks
  lightnvm: fix lun offset calculation for mark blk
  lightnvm: make ppa_list const in nvm_set_rqd_list
  lightnvm: make __nvm_submit_ppa static

Wenwei Tao (1):
  lightnvm: break the loop when rqd is not null

 drivers/lightnvm/Kconfig     |  10 +-
 drivers/lightnvm/core.c      | 242 +++++++--------------------
 drivers/lightnvm/gennvm.c    | 385 ++++++++++++++++++++++++++++---------------
 drivers/lightnvm/gennvm.h    |  10 +-
 drivers/lightnvm/rrpc.c      | 145 ++++++++--------
 drivers/lightnvm/rrpc.h      |  13 +-
 drivers/lightnvm/sysblk.c    |   8 +-
 drivers/nvme/host/lightnvm.c |   2 +-
 include/linux/lightnvm.h     |  36 ++--
 9 files changed, 420 insertions(+), 431 deletions(-)

-- 
2.1.4

[PATCH 01/17] lightnvm: break the loop when rqd is not null

[Matias Bjørling]

From: Wenwei Tao <ww.tao0320@gmail.com>

Break the loop when rqd is not null to reduce
an unnecessary schedule.

Signed-off-by: Wenwei Tao <ww.tao0320@gmail.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/rrpc.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index de86d72..736e669 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -96,10 +96,13 @@ static void rrpc_discard(struct rrpc *rrpc, struct bio *bio)
 	sector_t len = bio->bi_iter.bi_size / RRPC_EXPOSED_PAGE_SIZE;
 	struct nvm_rq *rqd;
 
-	do {
+	while (1) {
 		rqd = rrpc_inflight_laddr_acquire(rrpc, slba, len);
+		if (rqd)
+			break;
+
 		schedule();
-	} while (!rqd);
+	}
 
 	if (IS_ERR(rqd)) {
 		pr_err("rrpc: unable to acquire inflight IO\n");
-- 
2.1.4

[PATCH 02/17] lightnvm: add media manager mark_blk helper

[Matias Bjørling]

From: Javier González <jg@lightnvm.io>

Expose media manager mark_blk() to targets, as done for the rest of the
media manager callback functions.

Signed-off-by: Javier González <javier@cnexlabs.com>
Updated description
Signed-off-by: Matias Bjørling <m@bjorling.me>

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/core.c  | 6 ++++++
 include/linux/lightnvm.h | 2 ++
 2 files changed, 8 insertions(+)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 160c1a6..13993c9 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -210,6 +210,12 @@ void nvm_put_blk(struct nvm_dev *dev, struct nvm_block *blk)
 }
 EXPORT_SYMBOL(nvm_put_blk);
 
+void nvm_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type)
+{
+	return dev->mt->mark_blk(dev, ppa, type);
+}
+EXPORT_SYMBOL(nvm_mark_blk);
+
 int nvm_submit_io(struct nvm_dev *dev, struct nvm_rq *rqd)
 {
 	return dev->mt->submit_io(dev, rqd);
diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index ef2c7d2..9c56148 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -532,6 +532,8 @@ extern int nvm_register(struct request_queue *, char *,
 						struct nvm_dev_ops *);
 extern void nvm_unregister(char *);
 
+void nvm_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type);
+
 extern int nvm_submit_io(struct nvm_dev *, struct nvm_rq *);
 extern void nvm_generic_to_addr_mode(struct nvm_dev *, struct nvm_rq *);
 extern void nvm_addr_to_generic_mode(struct nvm_dev *, struct nvm_rq *);
-- 
2.1.4

[PATCH 03/17] lightnvm: initialize ppa_addr in dev_to_generic_addr()

[Matias Bjørling]

From: Javier González <javier@cnexlabs.com>

The ->reserved bit is not initialized when allocated on stack.
This may lead targets to misinterpret the PPA as cached.

Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 include/linux/lightnvm.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index 9c56148..cee4c8d 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -385,6 +385,7 @@ static inline struct ppa_addr dev_to_generic_addr(struct nvm_dev *dev,
 {
 	struct ppa_addr l;
 
+	l.ppa = 0;
 	/*
 	 * (r.ppa << X offset) & X len bitmask. X eq. blk, pg, etc.
 	 */
-- 
2.1.4

[PATCH 04/17] nvme: lightnvm: make MLC num_pairs little endian

[Matias Bjørling]

From: Johannes Thumshirn <jthumshirn@suse.de>

According to the OpenChannel SSD interface specification the NAND flash
MLC page pairing information's number of page page pairings field is the
first two bytes in the MLC Page Pairing data structure. The hardware's
data structure itself is little endian so annotate it as such, like the
rest of lighnvm's data structures.

Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/nvme/host/lightnvm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/host/lightnvm.c b/drivers/nvme/host/lightnvm.c
index a0af055..97fe610 100644
--- a/drivers/nvme/host/lightnvm.c
+++ b/drivers/nvme/host/lightnvm.c
@@ -156,7 +156,7 @@ struct nvme_nvm_completion {
 
 #define NVME_NVM_LP_MLC_PAIRS 886
 struct nvme_nvm_lp_mlc {
-	__u16			num_pairs;
+	__le16			num_pairs;
 	__u8			pairs[NVME_NVM_LP_MLC_PAIRS];
 };
 
-- 
2.1.4

[PATCH 05/17] lightnvm: Make functions not used by ouside static

[Matias Bjørling]

From: Johannes Thumshirn <jthumshirn@suse.de>

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>
---
 drivers/lightnvm/core.c   | 2 +-
 drivers/lightnvm/sysblk.c | 8 +++++---
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 13993c9..8beb9c0 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -110,7 +110,7 @@ static struct nvmm_type *nvm_find_mgr_type(const char *name)
 	return NULL;
 }
 
-struct nvmm_type *nvm_init_mgr(struct nvm_dev *dev)
+static struct nvmm_type *nvm_init_mgr(struct nvm_dev *dev)
 {
 	struct nvmm_type *mt;
 	int ret;
diff --git a/drivers/lightnvm/sysblk.c b/drivers/lightnvm/sysblk.c
index 994697a..a75bd28 100644
--- a/drivers/lightnvm/sysblk.c
+++ b/drivers/lightnvm/sysblk.c
@@ -39,7 +39,8 @@ static inline int scan_ppa_idx(int row, int blkid)
 	return (row * MAX_BLKS_PR_SYSBLK) + blkid;
 }
 
-void nvm_sysblk_to_cpu(struct nvm_sb_info *info, struct nvm_system_block *sb)
+static void nvm_sysblk_to_cpu(struct nvm_sb_info *info,
+			      struct nvm_system_block *sb)
 {
 	info->seqnr = be32_to_cpu(sb->seqnr);
 	info->erase_cnt = be32_to_cpu(sb->erase_cnt);
@@ -48,7 +49,8 @@ void nvm_sysblk_to_cpu(struct nvm_sb_info *info, struct nvm_system_block *sb)
 	info->fs_ppa.ppa = be64_to_cpu(sb->fs_ppa);
 }
 
-void nvm_cpu_to_sysblk(struct nvm_system_block *sb, struct nvm_sb_info *info)
+static void nvm_cpu_to_sysblk(struct nvm_system_block *sb,
+			      struct nvm_sb_info *info)
 {
 	sb->magic = cpu_to_be32(NVM_SYSBLK_MAGIC);
 	sb->seqnr = cpu_to_be32(info->seqnr);
@@ -86,7 +88,7 @@ static int nvm_setup_sysblks(struct nvm_dev *dev, struct ppa_addr *sysblk_ppas)
 	return nr_rows;
 }
 
-void nvm_setup_sysblk_scan(struct nvm_dev *dev, struct sysblk_scan *s,
+static void nvm_setup_sysblk_scan(struct nvm_dev *dev, struct sysblk_scan *s,
 						struct ppa_addr *sysblk_ppas)
 {
 	memset(s, 0, sizeof(struct sysblk_scan));
-- 
2.1.4

[PATCH 06/17] lightnvm: remove checkpatch warning for unsigned ints

[Matias Bjørling]

Checkpatch found two incidents where the type was preferred to be
written out in full.

./drivers/lightnvm/rrpc.h:184: WARNING: Prefer 'unsigned int' to bare
use of 'unsigned'
./drivers/lightnvm/rrpc.h:209: WARNING: Prefer 'unsigned int' to bare
use of 'unsigned'
./drivers/lightnvm/rrpc.c:51: WARNING: Prefer 'unsigned int' to bare use
of 'unsigned'

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/rrpc.c | 2 +-
 drivers/lightnvm/rrpc.h | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index 736e669..90c7cb4 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -48,7 +48,7 @@ static void rrpc_page_invalidate(struct rrpc *rrpc, struct rrpc_addr *a)
 }
 
 static void rrpc_invalidate_range(struct rrpc *rrpc, sector_t slba,
-								unsigned len)
+							unsigned int len)
 {
 	sector_t i;
 
diff --git a/drivers/lightnvm/rrpc.h b/drivers/lightnvm/rrpc.h
index 87e84b5..5797343 100644
--- a/drivers/lightnvm/rrpc.h
+++ b/drivers/lightnvm/rrpc.h
@@ -188,7 +188,7 @@ static inline int request_intersects(struct rrpc_inflight_rq *r,
 }
 
 static int __rrpc_lock_laddr(struct rrpc *rrpc, sector_t laddr,
-			     unsigned pages, struct rrpc_inflight_rq *r)
+			     unsigned int pages, struct rrpc_inflight_rq *r)
 {
 	sector_t laddr_end = laddr + pages - 1;
 	struct rrpc_inflight_rq *rtmp;
@@ -213,7 +213,7 @@ static int __rrpc_lock_laddr(struct rrpc *rrpc, sector_t laddr,
 }
 
 static inline int rrpc_lock_laddr(struct rrpc *rrpc, sector_t laddr,
-				 unsigned pages,
+				 unsigned int pages,
 				 struct rrpc_inflight_rq *r)
 {
 	BUG_ON((laddr + pages) > rrpc->nr_sects);
-- 
2.1.4

[PATCH 07/17] lightnvm: fix checkpatch terse errors

[Matias Bjørling]

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>
---
 drivers/lightnvm/core.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 8beb9c0..0da196f 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -373,7 +373,9 @@ int __nvm_submit_ppa(struct nvm_dev *dev, struct nvm_rq *rqd, int opcode,
 	/* Prevent hang_check timer from firing at us during very long I/O */
 	hang_check = sysctl_hung_task_timeout_secs;
 	if (hang_check)
-		while (!wait_for_completion_io_timeout(&wait, hang_check * (HZ/2)));
+		while (!wait_for_completion_io_timeout(&wait,
+							hang_check * (HZ/2)))
+			;
 	else
 		wait_for_completion_io(&wait);
 
@@ -516,7 +518,8 @@ static int nvm_init_mlc_tbl(struct nvm_dev *dev, struct nvm_id_group *grp)
 	/* 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
-	 * previous incrementation value */
+	 * previous incrementation value
+	 */
 	dev->lptbl[0] = mlc->pairs[0] & 0xF;
 	for (i = 1; i < dev->lps_per_blk; i++) {
 		p = mlc->pairs[i >> 1];
-- 
2.1.4

[PATCH 08/17] lightnvm: remove open/close statistics for gennvm

[Matias Bjørling]

The responsibility of the media manager is not to keep track of
open/closed blocks. This is better maintained within a target,
that already manages this information on writes.

Remove the statistics and merge the states NVM_BLK_ST_OPEN and
NVM_BLK_ST_CLOSED.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/gennvm.c | 31 ++++++-------------------------
 drivers/lightnvm/rrpc.c   |  5 -----
 include/linux/lightnvm.h  | 15 +++------------
 3 files changed, 9 insertions(+), 42 deletions(-)

diff --git a/drivers/lightnvm/gennvm.c b/drivers/lightnvm/gennvm.c
index ec9fb68..1a3e164 100644
--- a/drivers/lightnvm/gennvm.c
+++ b/drivers/lightnvm/gennvm.c
@@ -122,9 +122,6 @@ static int gennvm_luns_init(struct nvm_dev *dev, struct gen_nvm *gn)
 		lun->vlun.lun_id = i % dev->luns_per_chnl;
 		lun->vlun.chnl_id = i / dev->luns_per_chnl;
 		lun->vlun.nr_free_blocks = dev->blks_per_lun;
-		lun->vlun.nr_open_blocks = 0;
-		lun->vlun.nr_closed_blocks = 0;
-		lun->vlun.nr_bad_blocks = 0;
 	}
 	return 0;
 }
@@ -149,7 +146,6 @@ static int gennvm_block_bb(struct gen_nvm *gn, struct ppa_addr ppa,
 
 		blk = &lun->vlun.blocks[i];
 		list_move_tail(&blk->list, &lun->bb_list);
-		lun->vlun.nr_bad_blocks++;
 		lun->vlun.nr_free_blocks--;
 	}
 
@@ -200,9 +196,8 @@ static int gennvm_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
 			 * block state. The block is assumed to be open.
 			 */
 			list_move_tail(&blk->list, &lun->used_list);
-			blk->state = NVM_BLK_ST_OPEN;
+			blk->state = NVM_BLK_ST_TGT;
 			lun->vlun.nr_free_blocks--;
-			lun->vlun.nr_open_blocks++;
 		}
 	}
 
@@ -346,11 +341,10 @@ static struct nvm_block *gennvm_get_blk_unlocked(struct nvm_dev *dev,
 		goto out;
 
 	blk = list_first_entry(&lun->free_list, struct nvm_block, list);
+
 	list_move_tail(&blk->list, &lun->used_list);
-	blk->state = NVM_BLK_ST_OPEN;
-
+	blk->state = NVM_BLK_ST_TGT;
 	lun->vlun.nr_free_blocks--;
-	lun->vlun.nr_open_blocks++;
 
 out:
 	return blk;
@@ -374,27 +368,18 @@ static void gennvm_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
 
 	assert_spin_locked(&vlun->lock);
 
-	if (blk->state & NVM_BLK_ST_OPEN) {
+	if (blk->state & NVM_BLK_ST_TGT) {
 		list_move_tail(&blk->list, &lun->free_list);
-		lun->vlun.nr_open_blocks--;
-		lun->vlun.nr_free_blocks++;
-		blk->state = NVM_BLK_ST_FREE;
-	} else if (blk->state & NVM_BLK_ST_CLOSED) {
-		list_move_tail(&blk->list, &lun->free_list);
-		lun->vlun.nr_closed_blocks--;
 		lun->vlun.nr_free_blocks++;
 		blk->state = NVM_BLK_ST_FREE;
 	} else if (blk->state & NVM_BLK_ST_BAD) {
 		list_move_tail(&blk->list, &lun->bb_list);
-		lun->vlun.nr_bad_blocks++;
 		blk->state = NVM_BLK_ST_BAD;
 	} else {
 		WARN_ON_ONCE(1);
 		pr_err("gennvm: erroneous block type (%lu -> %u)\n",
 							blk->id, blk->state);
 		list_move_tail(&blk->list, &lun->bb_list);
-		lun->vlun.nr_bad_blocks++;
-		blk->state = NVM_BLK_ST_BAD;
 	}
 }
 
@@ -516,12 +501,8 @@ static void gennvm_lun_info_print(struct nvm_dev *dev)
 	gennvm_for_each_lun(gn, lun, i) {
 		spin_lock(&lun->vlun.lock);
 
-		pr_info("%s: lun%8u\t%u\t%u\t%u\t%u\n",
-				dev->name, i,
-				lun->vlun.nr_free_blocks,
-				lun->vlun.nr_open_blocks,
-				lun->vlun.nr_closed_blocks,
-				lun->vlun.nr_bad_blocks);
+		pr_info("%s: lun%8u\t%u\n", dev->name, i,
+						lun->vlun.nr_free_blocks);
 
 		spin_unlock(&lun->vlun.lock);
 	}
diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index 90c7cb4..c3a6f34 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -512,17 +512,12 @@ static void rrpc_gc_queue(struct work_struct *work)
 	struct rrpc_block *rblk = gcb->rblk;
 	struct rrpc_lun *rlun = rblk->rlun;
 	struct nvm_lun *lun = rblk->parent->lun;
-	struct nvm_block *blk = rblk->parent;
 
 	spin_lock(&rlun->lock);
 	list_add_tail(&rblk->prio, &rlun->prio_list);
 	spin_unlock(&rlun->lock);
 
 	spin_lock(&lun->lock);
-	lun->nr_open_blocks--;
-	lun->nr_closed_blocks++;
-	blk->state &= ~NVM_BLK_ST_OPEN;
-	blk->state |= NVM_BLK_ST_CLOSED;
 	list_move_tail(&rblk->list, &rlun->closed_list);
 	spin_unlock(&lun->lock);
 
diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index cee4c8d..8b51d57 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -269,24 +269,15 @@ struct nvm_lun {
 	int lun_id;
 	int chnl_id;
 
-	/* It is up to the target to mark blocks as closed. If the target does
-	 * not do it, all blocks are marked as open, and nr_open_blocks
-	 * represents the number of blocks in use
-	 */
-	unsigned int nr_open_blocks;	/* Number of used, writable blocks */
-	unsigned int nr_closed_blocks;	/* Number of used, read-only blocks */
+	spinlock_t lock;
+
 	unsigned int nr_free_blocks;	/* Number of unused blocks */
-	unsigned int nr_bad_blocks;	/* Number of bad blocks */
-
-	spinlock_t lock;
-
 	struct nvm_block *blocks;
 };
 
 enum {
 	NVM_BLK_ST_FREE =	0x1,	/* Free block */
-	NVM_BLK_ST_OPEN =	0x2,	/* Open block - read-write */
-	NVM_BLK_ST_CLOSED =	0x4,	/* Closed block - read-only */
+	NVM_BLK_ST_TGT =	0x2,	/* Block in use by target */
 	NVM_BLK_ST_BAD =	0x8,	/* Bad block */
 };
 
-- 
2.1.4

[PATCH 09/17] lightnvm: rename gennvm and update description

[Matias Bjørling]

The generic manager should be called the general media manager, and
instead of using the rather long name of "gennvm" in front of each data
structures, use "gen" instead to shorten it. Update the description of
the media manager as well to make the media manager purpose clearer.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/Kconfig  |  10 ++-
 drivers/lightnvm/gennvm.c | 184 +++++++++++++++++++++++-----------------------
 drivers/lightnvm/gennvm.h |   7 +-
 3 files changed, 102 insertions(+), 99 deletions(-)

diff --git a/drivers/lightnvm/Kconfig b/drivers/lightnvm/Kconfig
index 85a3390..61c68a1 100644
--- a/drivers/lightnvm/Kconfig
+++ b/drivers/lightnvm/Kconfig
@@ -27,11 +27,13 @@ config NVM_DEBUG
 	It is required to create/remove targets without IOCTLs.
 
 config NVM_GENNVM
-	tristate "Generic NVM manager for Open-Channel SSDs"
+	tristate "General Non-Volatile Memory Manager for Open-Channel SSDs"
 	---help---
-	NVM media manager for Open-Channel SSDs that offload management
-	functionality to device, while keeping data placement and garbage
-	collection decisions on the host.
+	Non-volatile memory media manager for Open-Channel SSDs that implements
+	physical media metadata management and block provisioning API.
+
+	This is the standard media manager for using Open-Channel SSDs, and
+	required for targets to be instantiated.
 
 config NVM_RRPC
 	tristate "Round-robin Hybrid Open-Channel SSD target"
diff --git a/drivers/lightnvm/gennvm.c b/drivers/lightnvm/gennvm.c
index 1a3e164..3d2762f 100644
--- a/drivers/lightnvm/gennvm.c
+++ b/drivers/lightnvm/gennvm.c
@@ -15,22 +15,22 @@
  * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139,
  * USA.
  *
- * Implementation of a generic nvm manager for Open-Channel SSDs.
+ * Implementation of a general nvm manager for Open-Channel SSDs.
  */
 
 #include "gennvm.h"
 
-static int gennvm_get_area(struct nvm_dev *dev, sector_t *lba, sector_t len)
+static int gen_get_area(struct nvm_dev *dev, sector_t *lba, sector_t len)
 {
-	struct gen_nvm *gn = dev->mp;
-	struct gennvm_area *area, *prev, *next;
+	struct gen_dev *gn = dev->mp;
+	struct gen_area *area, *prev, *next;
 	sector_t begin = 0;
 	sector_t max_sectors = (dev->sec_size * dev->total_secs) >> 9;
 
 	if (len > max_sectors)
 		return -EINVAL;
 
-	area = kmalloc(sizeof(struct gennvm_area), GFP_KERNEL);
+	area = kmalloc(sizeof(struct gen_area), GFP_KERNEL);
 	if (!area)
 		return -ENOMEM;
 
@@ -64,10 +64,10 @@ static int gennvm_get_area(struct nvm_dev *dev, sector_t *lba, sector_t len)
 	return 0;
 }
 
-static void gennvm_put_area(struct nvm_dev *dev, sector_t begin)
+static void gen_put_area(struct nvm_dev *dev, sector_t begin)
 {
-	struct gen_nvm *gn = dev->mp;
-	struct gennvm_area *area;
+	struct gen_dev *gn = dev->mp;
+	struct gen_area *area;
 
 	spin_lock(&dev->lock);
 	list_for_each_entry(area, &gn->area_list, list) {
@@ -82,27 +82,27 @@ static void gennvm_put_area(struct nvm_dev *dev, sector_t begin)
 	spin_unlock(&dev->lock);
 }
 
-static void gennvm_blocks_free(struct nvm_dev *dev)
+static void gen_blocks_free(struct nvm_dev *dev)
 {
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 	struct gen_lun *lun;
 	int i;
 
-	gennvm_for_each_lun(gn, lun, i) {
+	gen_for_each_lun(gn, lun, i) {
 		if (!lun->vlun.blocks)
 			break;
 		vfree(lun->vlun.blocks);
 	}
 }
 
-static void gennvm_luns_free(struct nvm_dev *dev)
+static void gen_luns_free(struct nvm_dev *dev)
 {
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 
 	kfree(gn->luns);
 }
 
-static int gennvm_luns_init(struct nvm_dev *dev, struct gen_nvm *gn)
+static int gen_luns_init(struct nvm_dev *dev, struct gen_dev *gn)
 {
 	struct gen_lun *lun;
 	int i;
@@ -111,7 +111,7 @@ static int gennvm_luns_init(struct nvm_dev *dev, struct gen_nvm *gn)
 	if (!gn->luns)
 		return -ENOMEM;
 
-	gennvm_for_each_lun(gn, lun, i) {
+	gen_for_each_lun(gn, lun, i) {
 		spin_lock_init(&lun->vlun.lock);
 		INIT_LIST_HEAD(&lun->free_list);
 		INIT_LIST_HEAD(&lun->used_list);
@@ -126,7 +126,7 @@ static int gennvm_luns_init(struct nvm_dev *dev, struct gen_nvm *gn)
 	return 0;
 }
 
-static int gennvm_block_bb(struct gen_nvm *gn, struct ppa_addr ppa,
+static int gen_block_bb(struct gen_dev *gn, struct ppa_addr ppa,
 							u8 *blks, int nr_blks)
 {
 	struct nvm_dev *dev = gn->dev;
@@ -152,10 +152,10 @@ static int gennvm_block_bb(struct gen_nvm *gn, struct ppa_addr ppa,
 	return 0;
 }
 
-static int gennvm_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
+static int gen_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
 {
 	struct nvm_dev *dev = private;
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 	u64 elba = slba + nlb;
 	struct gen_lun *lun;
 	struct nvm_block *blk;
@@ -163,7 +163,7 @@ static int gennvm_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
 	int lun_id;
 
 	if (unlikely(elba > dev->total_secs)) {
-		pr_err("gennvm: L2P data from device is out of bounds!\n");
+		pr_err("gen: L2P data from device is out of bounds!\n");
 		return -EINVAL;
 	}
 
@@ -171,7 +171,7 @@ static int gennvm_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
 		u64 pba = le64_to_cpu(entries[i]);
 
 		if (unlikely(pba >= dev->total_secs && pba != U64_MAX)) {
-			pr_err("gennvm: L2P data entry is out of bounds!\n");
+			pr_err("gen: L2P data entry is out of bounds!\n");
 			return -EINVAL;
 		}
 
@@ -204,7 +204,7 @@ static int gennvm_block_map(u64 slba, u32 nlb, __le64 *entries, void *private)
 	return 0;
 }
 
-static int gennvm_blocks_init(struct nvm_dev *dev, struct gen_nvm *gn)
+static int gen_blocks_init(struct nvm_dev *dev, struct gen_dev *gn)
 {
 	struct gen_lun *lun;
 	struct nvm_block *block;
@@ -217,7 +217,7 @@ static int gennvm_blocks_init(struct nvm_dev *dev, struct gen_nvm *gn)
 	if (!blks)
 		return -ENOMEM;
 
-	gennvm_for_each_lun(gn, lun, lun_iter) {
+	gen_for_each_lun(gn, lun, lun_iter) {
 		lun->vlun.blocks = vzalloc(sizeof(struct nvm_block) *
 							dev->blks_per_lun);
 		if (!lun->vlun.blocks) {
@@ -251,20 +251,20 @@ static int gennvm_blocks_init(struct nvm_dev *dev, struct gen_nvm *gn)
 
 			ret = nvm_get_bb_tbl(dev, ppa, blks);
 			if (ret)
-				pr_err("gennvm: could not get BB table\n");
+				pr_err("gen: could not get BB table\n");
 
-			ret = gennvm_block_bb(gn, ppa, blks, nr_blks);
+			ret = gen_block_bb(gn, ppa, blks, nr_blks);
 			if (ret)
-				pr_err("gennvm: BB table map failed\n");
+				pr_err("gen: BB table map failed\n");
 		}
 	}
 
 	if ((dev->identity.dom & NVM_RSP_L2P) && dev->ops->get_l2p_tbl) {
 		ret = dev->ops->get_l2p_tbl(dev, 0, dev->total_secs,
-							gennvm_block_map, dev);
+							gen_block_map, dev);
 		if (ret) {
-			pr_err("gennvm: could not read L2P table.\n");
-			pr_warn("gennvm: default block initialization");
+			pr_err("gen: could not read L2P table.\n");
+			pr_warn("gen: default block initialization");
 		}
 	}
 
@@ -272,23 +272,23 @@ static int gennvm_blocks_init(struct nvm_dev *dev, struct gen_nvm *gn)
 	return 0;
 }
 
-static void gennvm_free(struct nvm_dev *dev)
+static void gen_free(struct nvm_dev *dev)
 {
-	gennvm_blocks_free(dev);
-	gennvm_luns_free(dev);
+	gen_blocks_free(dev);
+	gen_luns_free(dev);
 	kfree(dev->mp);
 	dev->mp = NULL;
 }
 
-static int gennvm_register(struct nvm_dev *dev)
+static int gen_register(struct nvm_dev *dev)
 {
-	struct gen_nvm *gn;
+	struct gen_dev *gn;
 	int ret;
 
 	if (!try_module_get(THIS_MODULE))
 		return -ENODEV;
 
-	gn = kzalloc(sizeof(struct gen_nvm), GFP_KERNEL);
+	gn = kzalloc(sizeof(struct gen_dev), GFP_KERNEL);
 	if (!gn)
 		return -ENOMEM;
 
@@ -297,32 +297,32 @@ static int gennvm_register(struct nvm_dev *dev)
 	INIT_LIST_HEAD(&gn->area_list);
 	dev->mp = gn;
 
-	ret = gennvm_luns_init(dev, gn);
+	ret = gen_luns_init(dev, gn);
 	if (ret) {
-		pr_err("gennvm: could not initialize luns\n");
+		pr_err("gen: could not initialize luns\n");
 		goto err;
 	}
 
-	ret = gennvm_blocks_init(dev, gn);
+	ret = gen_blocks_init(dev, gn);
 	if (ret) {
-		pr_err("gennvm: could not initialize blocks\n");
+		pr_err("gen: could not initialize blocks\n");
 		goto err;
 	}
 
 	return 1;
 err:
-	gennvm_free(dev);
+	gen_free(dev);
 	module_put(THIS_MODULE);
 	return ret;
 }
 
-static void gennvm_unregister(struct nvm_dev *dev)
+static void gen_unregister(struct nvm_dev *dev)
 {
-	gennvm_free(dev);
+	gen_free(dev);
 	module_put(THIS_MODULE);
 }
 
-static struct nvm_block *gennvm_get_blk_unlocked(struct nvm_dev *dev,
+static struct nvm_block *gen_get_blk_unlocked(struct nvm_dev *dev,
 				struct nvm_lun *vlun, unsigned long flags)
 {
 	struct gen_lun *lun = container_of(vlun, struct gen_lun, vlun);
@@ -332,7 +332,7 @@ static struct nvm_block *gennvm_get_blk_unlocked(struct nvm_dev *dev,
 	assert_spin_locked(&vlun->lock);
 
 	if (list_empty(&lun->free_list)) {
-		pr_err_ratelimited("gennvm: lun %u have no free pages available",
+		pr_err_ratelimited("gen: lun %u have no free pages available",
 								lun->vlun.id);
 		goto out;
 	}
@@ -350,18 +350,18 @@ out:
 	return blk;
 }
 
-static struct nvm_block *gennvm_get_blk(struct nvm_dev *dev,
+static struct nvm_block *gen_get_blk(struct nvm_dev *dev,
 				struct nvm_lun *vlun, unsigned long flags)
 {
 	struct nvm_block *blk;
 
 	spin_lock(&vlun->lock);
-	blk = gennvm_get_blk_unlocked(dev, vlun, flags);
+	blk = gen_get_blk_unlocked(dev, vlun, flags);
 	spin_unlock(&vlun->lock);
 	return blk;
 }
 
-static void gennvm_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
+static void gen_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
 {
 	struct nvm_lun *vlun = blk->lun;
 	struct gen_lun *lun = container_of(vlun, struct gen_lun, vlun);
@@ -377,35 +377,35 @@ static void gennvm_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
 		blk->state = NVM_BLK_ST_BAD;
 	} else {
 		WARN_ON_ONCE(1);
-		pr_err("gennvm: erroneous block type (%lu -> %u)\n",
+		pr_err("gen: erroneous block type (%lu -> %u)\n",
 							blk->id, blk->state);
 		list_move_tail(&blk->list, &lun->bb_list);
 	}
 }
 
-static void gennvm_put_blk(struct nvm_dev *dev, struct nvm_block *blk)
+static void gen_put_blk(struct nvm_dev *dev, struct nvm_block *blk)
 {
 	struct nvm_lun *vlun = blk->lun;
 
 	spin_lock(&vlun->lock);
-	gennvm_put_blk_unlocked(dev, blk);
+	gen_put_blk_unlocked(dev, blk);
 	spin_unlock(&vlun->lock);
 }
 
-static void gennvm_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type)
+static void gen_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type)
 {
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 	struct gen_lun *lun;
 	struct nvm_block *blk;
 
-	pr_debug("gennvm: ppa  (ch: %u lun: %u blk: %u pg: %u) -> %u\n",
+	pr_debug("gen: ppa  (ch: %u lun: %u blk: %u pg: %u) -> %u\n",
 			ppa.g.ch, ppa.g.lun, ppa.g.blk, ppa.g.pg, type);
 
 	if (unlikely(ppa.g.ch > dev->nr_chnls ||
 					ppa.g.lun > dev->luns_per_chnl ||
 					ppa.g.blk > dev->blks_per_lun)) {
 		WARN_ON_ONCE(1);
-		pr_err("gennvm: ppa broken (ch: %u > %u lun: %u > %u blk: %u > %u",
+		pr_err("gen: ppa broken (ch: %u > %u lun: %u > %u blk: %u > %u",
 				ppa.g.ch, dev->nr_chnls,
 				ppa.g.lun, dev->luns_per_chnl,
 				ppa.g.blk, dev->blks_per_lun);
@@ -420,9 +420,9 @@ static void gennvm_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type)
 }
 
 /*
- * mark block bad in gennvm. It is expected that the target recovers separately
+ * mark block bad in gen. It is expected that the target recovers separately
  */
-static void gennvm_mark_blk_bad(struct nvm_dev *dev, struct nvm_rq *rqd)
+static void gen_mark_blk_bad(struct nvm_dev *dev, struct nvm_rq *rqd)
 {
 	int bit = -1;
 	int max_secs = dev->ops->max_phys_sect;
@@ -432,25 +432,25 @@ static void gennvm_mark_blk_bad(struct nvm_dev *dev, struct nvm_rq *rqd)
 
 	/* look up blocks and mark them as bad */
 	if (rqd->nr_ppas == 1) {
-		gennvm_mark_blk(dev, rqd->ppa_addr, NVM_BLK_ST_BAD);
+		gen_mark_blk(dev, rqd->ppa_addr, NVM_BLK_ST_BAD);
 		return;
 	}
 
 	while ((bit = find_next_bit(comp_bits, max_secs, bit + 1)) < max_secs)
-		gennvm_mark_blk(dev, rqd->ppa_list[bit], NVM_BLK_ST_BAD);
+		gen_mark_blk(dev, rqd->ppa_list[bit], NVM_BLK_ST_BAD);
 }
 
-static void gennvm_end_io(struct nvm_rq *rqd)
+static void gen_end_io(struct nvm_rq *rqd)
 {
 	struct nvm_tgt_instance *ins = rqd->ins;
 
 	if (rqd->error == NVM_RSP_ERR_FAILWRITE)
-		gennvm_mark_blk_bad(rqd->dev, rqd);
+		gen_mark_blk_bad(rqd->dev, rqd);
 
 	ins->tt->end_io(rqd);
 }
 
-static int gennvm_submit_io(struct nvm_dev *dev, struct nvm_rq *rqd)
+static int gen_submit_io(struct nvm_dev *dev, struct nvm_rq *rqd)
 {
 	if (!dev->ops->submit_io)
 		return -ENODEV;
@@ -459,11 +459,11 @@ static int gennvm_submit_io(struct nvm_dev *dev, struct nvm_rq *rqd)
 	nvm_generic_to_addr_mode(dev, rqd);
 
 	rqd->dev = dev;
-	rqd->end_io = gennvm_end_io;
+	rqd->end_io = gen_end_io;
 	return dev->ops->submit_io(dev, rqd);
 }
 
-static int gennvm_erase_blk(struct nvm_dev *dev, struct nvm_block *blk,
+static int gen_erase_blk(struct nvm_dev *dev, struct nvm_block *blk,
 							unsigned long flags)
 {
 	struct ppa_addr addr = block_to_ppa(dev, blk);
@@ -471,19 +471,19 @@ static int gennvm_erase_blk(struct nvm_dev *dev, struct nvm_block *blk,
 	return nvm_erase_ppa(dev, &addr, 1);
 }
 
-static int gennvm_reserve_lun(struct nvm_dev *dev, int lunid)
+static int gen_reserve_lun(struct nvm_dev *dev, int lunid)
 {
 	return test_and_set_bit(lunid, dev->lun_map);
 }
 
-static void gennvm_release_lun(struct nvm_dev *dev, int lunid)
+static void gen_release_lun(struct nvm_dev *dev, int lunid)
 {
 	WARN_ON(!test_and_clear_bit(lunid, dev->lun_map));
 }
 
-static struct nvm_lun *gennvm_get_lun(struct nvm_dev *dev, int lunid)
+static struct nvm_lun *gen_get_lun(struct nvm_dev *dev, int lunid)
 {
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 
 	if (unlikely(lunid >= dev->nr_luns))
 		return NULL;
@@ -491,14 +491,14 @@ static struct nvm_lun *gennvm_get_lun(struct nvm_dev *dev, int lunid)
 	return &gn->luns[lunid].vlun;
 }
 
-static void gennvm_lun_info_print(struct nvm_dev *dev)
+static void gen_lun_info_print(struct nvm_dev *dev)
 {
-	struct gen_nvm *gn = dev->mp;
+	struct gen_dev *gn = dev->mp;
 	struct gen_lun *lun;
 	unsigned int i;
 
 
-	gennvm_for_each_lun(gn, lun, i) {
+	gen_for_each_lun(gn, lun, i) {
 		spin_lock(&lun->vlun.lock);
 
 		pr_info("%s: lun%8u\t%u\n", dev->name, i,
@@ -508,45 +508,45 @@ static void gennvm_lun_info_print(struct nvm_dev *dev)
 	}
 }
 
-static struct nvmm_type gennvm = {
+static struct nvmm_type gen = {
 	.name			= "gennvm",
 	.version		= {0, 1, 0},
 
-	.register_mgr		= gennvm_register,
-	.unregister_mgr		= gennvm_unregister,
+	.register_mgr		= gen_register,
+	.unregister_mgr		= gen_unregister,
 
-	.get_blk_unlocked	= gennvm_get_blk_unlocked,
-	.put_blk_unlocked	= gennvm_put_blk_unlocked,
+	.get_blk_unlocked	= gen_get_blk_unlocked,
+	.put_blk_unlocked	= gen_put_blk_unlocked,
 
-	.get_blk		= gennvm_get_blk,
-	.put_blk		= gennvm_put_blk,
+	.get_blk		= gen_get_blk,
+	.put_blk		= gen_put_blk,
 
-	.submit_io		= gennvm_submit_io,
-	.erase_blk		= gennvm_erase_blk,
+	.submit_io		= gen_submit_io,
+	.erase_blk		= gen_erase_blk,
 
-	.mark_blk		= gennvm_mark_blk,
+	.mark_blk		= gen_mark_blk,
 
-	.get_lun		= gennvm_get_lun,
-	.reserve_lun		= gennvm_reserve_lun,
-	.release_lun		= gennvm_release_lun,
-	.lun_info_print		= gennvm_lun_info_print,
+	.get_lun		= gen_get_lun,
+	.reserve_lun		= gen_reserve_lun,
+	.release_lun		= gen_release_lun,
+	.lun_info_print		= gen_lun_info_print,
 
-	.get_area		= gennvm_get_area,
-	.put_area		= gennvm_put_area,
+	.get_area		= gen_get_area,
+	.put_area		= gen_put_area,
 
 };
 
-static int __init gennvm_module_init(void)
+static int __init gen_module_init(void)
 {
-	return nvm_register_mgr(&gennvm);
+	return nvm_register_mgr(&gen);
 }
 
-static void gennvm_module_exit(void)
+static void gen_module_exit(void)
 {
-	nvm_unregister_mgr(&gennvm);
+	nvm_unregister_mgr(&gen);
 }
 
-module_init(gennvm_module_init);
-module_exit(gennvm_module_exit);
+module_init(gen_module_init);
+module_exit(gen_module_exit);
 MODULE_LICENSE("GPL v2");
-MODULE_DESCRIPTION("Generic media manager for Open-Channel SSDs");
+MODULE_DESCRIPTION("General media manager for Open-Channel SSDs");
diff --git a/drivers/lightnvm/gennvm.h b/drivers/lightnvm/gennvm.h
index 04d7c23..bf06219 100644
--- a/drivers/lightnvm/gennvm.h
+++ b/drivers/lightnvm/gennvm.h
@@ -34,7 +34,7 @@ struct gen_lun {
 					 */
 };
 
-struct gen_nvm {
+struct gen_dev {
 	struct nvm_dev *dev;
 
 	int nr_luns;
@@ -42,12 +42,13 @@ struct gen_nvm {
 	struct list_head area_list;
 };
 
-struct gennvm_area {
+struct gen_area {
 	struct list_head list;
 	sector_t begin;
 	sector_t end;	/* end is excluded */
 };
-#define gennvm_for_each_lun(bm, lun, i) \
+
+#define gen_for_each_lun(bm, lun, i) \
 		for ((i) = 0, lun = &(bm)->luns[0]; \
 			(i) < (bm)->nr_luns; (i)++, lun = &(bm)->luns[(i)])
 
-- 
2.1.4

[PATCH 10/17] lightnvm: move target mgmt into media mgr

[Matias Bjørling]

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>
---
 drivers/lightnvm/core.c   | 198 +++++++++-------------------------------------
 drivers/lightnvm/gennvm.c | 154 ++++++++++++++++++++++++++++++++++++
 drivers/lightnvm/gennvm.h |   3 +
 include/linux/lightnvm.h  |  10 +++
 4 files changed, 206 insertions(+), 159 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 0da196f..8afb04c 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -18,8 +18,6 @@
  *
  */
 
-#include <linux/blkdev.h>
-#include <linux/blk-mq.h>
 #include <linux/list.h>
 #include <linux/types.h>
 #include <linux/sem.h>
@@ -28,42 +26,37 @@
 #include <linux/miscdevice.h>
 #include <linux/lightnvm.h>
 #include <linux/sched/sysctl.h>
-#include <uapi/linux/lightnvm.h>
 
 static LIST_HEAD(nvm_tgt_types);
 static LIST_HEAD(nvm_mgrs);
 static LIST_HEAD(nvm_devices);
-static LIST_HEAD(nvm_targets);
 static DECLARE_RWSEM(nvm_lock);
 
-static struct nvm_target *nvm_find_target(const char *name)
+struct nvm_tgt_type *nvm_find_target_type(const char *name, int lock)
 {
-	struct nvm_target *tgt;
+	struct nvm_tgt_type *tmp, *tt = NULL;
 
-	list_for_each_entry(tgt, &nvm_targets, list)
-		if (!strcmp(name, tgt->disk->disk_name))
-			return tgt;
+	if (lock)
+		down_write(&nvm_lock);
 
-	return NULL;
-}
-
-static struct nvm_tgt_type *nvm_find_target_type(const char *name)
-{
-	struct nvm_tgt_type *tt;
-
-	list_for_each_entry(tt, &nvm_tgt_types, list)
-		if (!strcmp(name, tt->name))
-			return tt;
+	list_for_each_entry(tmp, &nvm_tgt_types, list)
+		if (!strcmp(name, tmp->name)) {
+			tt = tmp;
+			break;
+		}
 
-	return NULL;
+	if (lock)
+		up_write(&nvm_lock);
+	return tt;
 }
+EXPORT_SYMBOL(nvm_find_target_type);
 
 int nvm_register_tgt_type(struct nvm_tgt_type *tt)
 {
 	int ret = 0;
 
 	down_write(&nvm_lock);
-	if (nvm_find_target_type(tt->name))
+	if (nvm_find_target_type(tt->name, 0))
 		ret = -EEXIST;
 	else
 		list_add(&tt->list, &nvm_tgt_types);
@@ -605,42 +598,11 @@ err_fmtype:
 	return ret;
 }
 
-static void nvm_remove_target(struct nvm_target *t)
-{
-	struct nvm_tgt_type *tt = t->type;
-	struct gendisk *tdisk = t->disk;
-	struct request_queue *q = tdisk->queue;
-
-	lockdep_assert_held(&nvm_lock);
-
-	del_gendisk(tdisk);
-	blk_cleanup_queue(q);
-
-	if (tt->exit)
-		tt->exit(tdisk->private_data);
-
-	put_disk(tdisk);
-
-	list_del(&t->list);
-	kfree(t);
-}
-
 static void nvm_free_mgr(struct nvm_dev *dev)
 {
-	struct nvm_target *tgt, *tmp;
-
 	if (!dev->mt)
 		return;
 
-	down_write(&nvm_lock);
-	list_for_each_entry_safe(tgt, tmp, &nvm_targets, list) {
-		if (tgt->dev != dev)
-			continue;
-
-		nvm_remove_target(tgt);
-	}
-	up_write(&nvm_lock);
-
 	dev->mt->unregister_mgr(dev);
 	dev->mt = NULL;
 }
@@ -787,91 +749,6 @@ void nvm_unregister(char *disk_name)
 }
 EXPORT_SYMBOL(nvm_unregister);
 
-static const struct block_device_operations nvm_fops = {
-	.owner		= THIS_MODULE,
-};
-
-static int nvm_create_target(struct nvm_dev *dev,
-						struct nvm_ioctl_create *create)
-{
-	struct nvm_ioctl_create_simple *s = &create->conf.s;
-	struct request_queue *tqueue;
-	struct gendisk *tdisk;
-	struct nvm_tgt_type *tt;
-	struct nvm_target *t;
-	void *targetdata;
-
-	if (!dev->mt) {
-		pr_info("nvm: device has no media manager registered.\n");
-		return -ENODEV;
-	}
-
-	down_write(&nvm_lock);
-	tt = nvm_find_target_type(create->tgttype);
-	if (!tt) {
-		pr_err("nvm: target type %s not found\n", create->tgttype);
-		up_write(&nvm_lock);
-		return -EINVAL;
-	}
-
-	t = nvm_find_target(create->tgtname);
-	if (t) {
-		pr_err("nvm: target name already exists.\n");
-		up_write(&nvm_lock);
-		return -EINVAL;
-	}
-	up_write(&nvm_lock);
-
-	t = kmalloc(sizeof(struct nvm_target), GFP_KERNEL);
-	if (!t)
-		return -ENOMEM;
-
-	tqueue = blk_alloc_queue_node(GFP_KERNEL, dev->q->node);
-	if (!tqueue)
-		goto err_t;
-	blk_queue_make_request(tqueue, tt->make_rq);
-
-	tdisk = alloc_disk(0);
-	if (!tdisk)
-		goto err_queue;
-
-	sprintf(tdisk->disk_name, "%s", create->tgtname);
-	tdisk->flags = GENHD_FL_EXT_DEVT;
-	tdisk->major = 0;
-	tdisk->first_minor = 0;
-	tdisk->fops = &nvm_fops;
-	tdisk->queue = tqueue;
-
-	targetdata = tt->init(dev, tdisk, s->lun_begin, s->lun_end);
-	if (IS_ERR(targetdata))
-		goto err_init;
-
-	tdisk->private_data = targetdata;
-	tqueue->queuedata = targetdata;
-
-	blk_queue_max_hw_sectors(tqueue, 8 * dev->ops->max_phys_sect);
-
-	set_capacity(tdisk, tt->capacity(targetdata));
-	add_disk(tdisk);
-
-	t->type = tt;
-	t->disk = tdisk;
-	t->dev = dev;
-
-	down_write(&nvm_lock);
-	list_add_tail(&t->list, &nvm_targets);
-	up_write(&nvm_lock);
-
-	return 0;
-err_init:
-	put_disk(tdisk);
-err_queue:
-	blk_cleanup_queue(tqueue);
-err_t:
-	kfree(t);
-	return -ENOMEM;
-}
-
 static int __nvm_configure_create(struct nvm_ioctl_create *create)
 {
 	struct nvm_dev *dev;
@@ -880,11 +757,17 @@ static int __nvm_configure_create(struct nvm_ioctl_create *create)
 	down_write(&nvm_lock);
 	dev = nvm_find_nvm_dev(create->dev);
 	up_write(&nvm_lock);
+
 	if (!dev) {
 		pr_err("nvm: device not found\n");
 		return -EINVAL;
 	}
 
+	if (!dev->mt) {
+		pr_info("nvm: device has no media manager registered.\n");
+		return -ENODEV;
+	}
+
 	if (create->conf.type != NVM_CONFIG_TYPE_SIMPLE) {
 		pr_err("nvm: config type not valid\n");
 		return -EINVAL;
@@ -897,25 +780,7 @@ static int __nvm_configure_create(struct nvm_ioctl_create *create)
 		return -EINVAL;
 	}
 
-	return nvm_create_target(dev, create);
-}
-
-static int __nvm_configure_remove(struct nvm_ioctl_remove *remove)
-{
-	struct nvm_target *t;
-
-	down_write(&nvm_lock);
-	t = nvm_find_target(remove->tgtname);
-	if (!t) {
-		pr_err("nvm: target \"%s\" doesn't exist.\n", remove->tgtname);
-		up_write(&nvm_lock);
-		return -EINVAL;
-	}
-
-	nvm_remove_target(t);
-	up_write(&nvm_lock);
-
-	return 0;
+	return dev->mt->create_tgt(dev, create);
 }
 
 #ifdef CONFIG_NVM_DEBUG
@@ -950,8 +815,9 @@ static int nvm_configure_show(const char *val)
 static int nvm_configure_remove(const char *val)
 {
 	struct nvm_ioctl_remove remove;
+	struct nvm_dev *dev;
 	char opcode;
-	int ret;
+	int ret = 0;
 
 	ret = sscanf(val, "%c %256s", &opcode, remove.tgtname);
 	if (ret != 2) {
@@ -961,7 +827,13 @@ static int nvm_configure_remove(const char *val)
 
 	remove.flags = 0;
 
-	return __nvm_configure_remove(&remove);
+	list_for_each_entry(dev, &nvm_devices, devices) {
+		ret = dev->mt->remove_tgt(dev, &remove);
+		if (!ret)
+			break;
+	}
+
+	return ret;
 }
 
 static int nvm_configure_create(const char *val)
@@ -1158,6 +1030,8 @@ static long nvm_ioctl_dev_create(struct file *file, void __user *arg)
 static long nvm_ioctl_dev_remove(struct file *file, void __user *arg)
 {
 	struct nvm_ioctl_remove remove;
+	struct nvm_dev *dev;
+	int ret = 0;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
@@ -1172,7 +1046,13 @@ static long nvm_ioctl_dev_remove(struct file *file, void __user *arg)
 		return -EINVAL;
 	}
 
-	return __nvm_configure_remove(&remove);
+	list_for_each_entry(dev, &nvm_devices, devices) {
+		ret = dev->mt->remove_tgt(dev, &remove);
+		if (!ret)
+			break;
+	}
+
+	return ret;
 }
 
 static void nvm_setup_nvm_sb_info(struct nvm_sb_info *info)
diff --git a/drivers/lightnvm/gennvm.c b/drivers/lightnvm/gennvm.c
index 3d2762f..41760b2 100644
--- a/drivers/lightnvm/gennvm.c
+++ b/drivers/lightnvm/gennvm.c
@@ -20,6 +20,144 @@
 
 #include "gennvm.h"
 
+static struct nvm_target *gen_find_target(struct gen_dev *gn, const char *name)
+{
+	struct nvm_target *tgt;
+
+	list_for_each_entry(tgt, &gn->targets, list)
+		if (!strcmp(name, tgt->disk->disk_name))
+			return tgt;
+
+	return NULL;
+}
+
+static const struct block_device_operations gen_fops = {
+	.owner		= THIS_MODULE,
+};
+
+static int gen_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
+{
+	struct gen_dev *gn = dev->mp;
+	struct nvm_ioctl_create_simple *s = &create->conf.s;
+	struct request_queue *tqueue;
+	struct gendisk *tdisk;
+	struct nvm_tgt_type *tt;
+	struct nvm_target *t;
+	void *targetdata;
+
+	tt = nvm_find_target_type(create->tgttype, 1);
+	if (!tt) {
+		pr_err("nvm: target type %s not found\n", create->tgttype);
+		return -EINVAL;
+	}
+
+	mutex_lock(&gn->lock);
+	t = gen_find_target(gn, create->tgtname);
+	if (t) {
+		pr_err("nvm: target name already exists.\n");
+		mutex_unlock(&gn->lock);
+		return -EINVAL;
+	}
+	mutex_unlock(&gn->lock);
+
+	t = kmalloc(sizeof(struct nvm_target), GFP_KERNEL);
+	if (!t)
+		return -ENOMEM;
+
+	tqueue = blk_alloc_queue_node(GFP_KERNEL, dev->q->node);
+	if (!tqueue)
+		goto err_t;
+	blk_queue_make_request(tqueue, tt->make_rq);
+
+	tdisk = alloc_disk(0);
+	if (!tdisk)
+		goto err_queue;
+
+	sprintf(tdisk->disk_name, "%s", create->tgtname);
+	tdisk->flags = GENHD_FL_EXT_DEVT;
+	tdisk->major = 0;
+	tdisk->first_minor = 0;
+	tdisk->fops = &gen_fops;
+	tdisk->queue = tqueue;
+
+	targetdata = tt->init(dev, tdisk, s->lun_begin, s->lun_end);
+	if (IS_ERR(targetdata))
+		goto err_init;
+
+	tdisk->private_data = targetdata;
+	tqueue->queuedata = targetdata;
+
+	blk_queue_max_hw_sectors(tqueue, 8 * dev->ops->max_phys_sect);
+
+	set_capacity(tdisk, tt->capacity(targetdata));
+	add_disk(tdisk);
+
+	t->type = tt;
+	t->disk = tdisk;
+	t->dev = dev;
+
+	mutex_lock(&gn->lock);
+	list_add_tail(&t->list, &gn->targets);
+	mutex_unlock(&gn->lock);
+
+	return 0;
+err_init:
+	put_disk(tdisk);
+err_queue:
+	blk_cleanup_queue(tqueue);
+err_t:
+	kfree(t);
+	return -ENOMEM;
+}
+
+static void __gen_remove_target(struct nvm_target *t)
+{
+	struct nvm_tgt_type *tt = t->type;
+	struct gendisk *tdisk = t->disk;
+	struct request_queue *q = tdisk->queue;
+
+	del_gendisk(tdisk);
+	blk_cleanup_queue(q);
+
+	if (tt->exit)
+		tt->exit(tdisk->private_data);
+
+	put_disk(tdisk);
+
+	list_del(&t->list);
+	kfree(t);
+}
+
+/**
+ * gen_remove_tgt - Removes a target from the media manager
+ * @dev:	device
+ * @remove:	ioctl structure with target name to remove.
+ *
+ * Returns:
+ * 0: on success
+ * 1: on not found
+ * <0: on error
+ */
+static int gen_remove_tgt(struct nvm_dev *dev, struct nvm_ioctl_remove *remove)
+{
+	struct gen_dev *gn = dev->mp;
+	struct nvm_target *t;
+
+	if (!gn)
+		return 1;
+
+	mutex_lock(&gn->lock);
+	t = gen_find_target(gn, remove->tgtname);
+	if (!t) {
+		mutex_unlock(&gn->lock);
+		return 1;
+	}
+	__gen_remove_target(t);
+	mutex_unlock(&gn->lock);
+
+	return 0;
+}
+
 static int gen_get_area(struct nvm_dev *dev, sector_t *lba, sector_t len)
 {
 	struct gen_dev *gn = dev->mp;
@@ -295,6 +433,8 @@ static int gen_register(struct nvm_dev *dev)
 	gn->dev = dev;
 	gn->nr_luns = dev->nr_luns;
 	INIT_LIST_HEAD(&gn->area_list);
+	mutex_init(&gn->lock);
+	INIT_LIST_HEAD(&gn->targets);
 	dev->mp = gn;
 
 	ret = gen_luns_init(dev, gn);
@@ -318,6 +458,17 @@ err:
 
 static void gen_unregister(struct nvm_dev *dev)
 {
+	struct gen_dev *gn = dev->mp;
+	struct nvm_target *t, *tmp;
+
+	mutex_lock(&gn->lock);
+	list_for_each_entry_safe(t, tmp, &gn->targets, list) {
+		if (t->dev != dev)
+			continue;
+		__gen_remove_target(t);
+	}
+	mutex_unlock(&gn->lock);
+
 	gen_free(dev);
 	module_put(THIS_MODULE);
 }
@@ -515,6 +666,9 @@ static struct nvmm_type gen = {
 	.register_mgr		= gen_register,
 	.unregister_mgr		= gen_unregister,
 
+	.create_tgt		= gen_create_tgt,
+	.remove_tgt		= gen_remove_tgt,
+
 	.get_blk_unlocked	= gen_get_blk_unlocked,
 	.put_blk_unlocked	= gen_put_blk_unlocked,
 
diff --git a/drivers/lightnvm/gennvm.h b/drivers/lightnvm/gennvm.h
index bf06219..8ecfa81 100644
--- a/drivers/lightnvm/gennvm.h
+++ b/drivers/lightnvm/gennvm.h
@@ -40,6 +40,9 @@ struct gen_dev {
 	int nr_luns;
 	struct gen_lun *luns;
 	struct list_head area_list;
+
+	struct mutex lock;
+	struct list_head targets;
 };
 
 struct gen_area {
diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index 8b51d57..d619f6d 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -1,7 +1,9 @@
 #ifndef NVM_H
 #define NVM_H
 
+#include <linux/blkdev.h>
 #include <linux/types.h>
+#include <uapi/linux/lightnvm.h>
 
 enum {
 	NVM_IO_OK = 0,
@@ -447,6 +449,8 @@ struct nvm_tgt_type {
 	struct list_head list;
 };
 
+extern struct nvm_tgt_type *nvm_find_target_type(const char *, int);
+
 extern int nvm_register_tgt_type(struct nvm_tgt_type *);
 extern void nvm_unregister_tgt_type(struct nvm_tgt_type *);
 
@@ -455,6 +459,9 @@ extern void nvm_dev_dma_free(struct nvm_dev *, void *, dma_addr_t);
 
 typedef int (nvmm_register_fn)(struct nvm_dev *);
 typedef void (nvmm_unregister_fn)(struct nvm_dev *);
+
+typedef int (nvmm_create_tgt_fn)(struct nvm_dev *, struct nvm_ioctl_create *);
+typedef int (nvmm_remove_tgt_fn)(struct nvm_dev *, struct nvm_ioctl_remove *);
 typedef struct nvm_block *(nvmm_get_blk_fn)(struct nvm_dev *,
 					      struct nvm_lun *, unsigned long);
 typedef void (nvmm_put_blk_fn)(struct nvm_dev *, struct nvm_block *);
@@ -480,6 +487,9 @@ struct nvmm_type {
 	nvmm_register_fn *register_mgr;
 	nvmm_unregister_fn *unregister_mgr;
 
+	nvmm_create_tgt_fn *create_tgt;
+	nvmm_remove_tgt_fn *remove_tgt;
+
 	/* Block administration callbacks */
 	nvmm_get_blk_fn *get_blk_unlocked;
 	nvmm_put_blk_fn *put_blk_unlocked;
-- 
2.1.4

[PATCH 11/17] lightnvm: remove nested lock conflict with mm

[Matias Bjørling]

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>
---
 drivers/lightnvm/core.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 8afb04c..04469e0 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -28,6 +28,7 @@
 #include <linux/sched/sysctl.h>
 
 static LIST_HEAD(nvm_tgt_types);
+static DECLARE_RWSEM(nvm_tgtt_lock);
 static LIST_HEAD(nvm_mgrs);
 static LIST_HEAD(nvm_devices);
 static DECLARE_RWSEM(nvm_lock);
@@ -37,7 +38,7 @@ struct nvm_tgt_type *nvm_find_target_type(const char *name, int lock)
 	struct nvm_tgt_type *tmp, *tt = NULL;
 
 	if (lock)
-		down_write(&nvm_lock);
+		down_write(&nvm_tgtt_lock);
 
 	list_for_each_entry(tmp, &nvm_tgt_types, list)
 		if (!strcmp(name, tmp->name)) {
@@ -46,7 +47,7 @@ struct nvm_tgt_type *nvm_find_target_type(const char *name, int lock)
 		}
 
 	if (lock)
-		up_write(&nvm_lock);
+		up_write(&nvm_tgtt_lock);
 	return tt;
 }
 EXPORT_SYMBOL(nvm_find_target_type);
@@ -55,12 +56,12 @@ int nvm_register_tgt_type(struct nvm_tgt_type *tt)
 {
 	int ret = 0;
 
-	down_write(&nvm_lock);
+	down_write(&nvm_tgtt_lock);
 	if (nvm_find_target_type(tt->name, 0))
 		ret = -EEXIST;
 	else
 		list_add(&tt->list, &nvm_tgt_types);
-	up_write(&nvm_lock);
+	up_write(&nvm_tgtt_lock);
 
 	return ret;
 }
-- 
2.1.4

[PATCH 12/17] lightnvm: remove unused lists from struct rrpc_block

[Matias Bjørling]

The ->list, ->open_list, and ->closed_list lists were previously used
for statistics. However, their usage have been removed, and thus these
can safely be removed.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/rrpc.c | 9 ---------
 drivers/lightnvm/rrpc.h | 8 --------
 2 files changed, 17 deletions(-)

diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index c3a6f34..10ed22b 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -205,7 +205,6 @@ static struct rrpc_block *rrpc_get_blk(struct rrpc *rrpc, struct rrpc_lun *rlun,
 	}
 
 	rblk = rrpc_get_rblk(rlun, blk->id);
-	list_add_tail(&rblk->list, &rlun->open_list);
 	spin_unlock(&lun->lock);
 
 	blk->priv = rblk;
@@ -224,7 +223,6 @@ static void rrpc_put_blk(struct rrpc *rrpc, struct rrpc_block *rblk)
 
 	spin_lock(&lun->lock);
 	nvm_put_blk_unlocked(rrpc->dev, rblk->parent);
-	list_del(&rblk->list);
 	spin_unlock(&lun->lock);
 }
 
@@ -511,16 +509,11 @@ static void rrpc_gc_queue(struct work_struct *work)
 	struct rrpc *rrpc = gcb->rrpc;
 	struct rrpc_block *rblk = gcb->rblk;
 	struct rrpc_lun *rlun = rblk->rlun;
-	struct nvm_lun *lun = rblk->parent->lun;
 
 	spin_lock(&rlun->lock);
 	list_add_tail(&rblk->prio, &rlun->prio_list);
 	spin_unlock(&rlun->lock);
 
-	spin_lock(&lun->lock);
-	list_move_tail(&rblk->list, &rlun->closed_list);
-	spin_unlock(&lun->lock);
-
 	mempool_free(gcb, rrpc->gcb_pool);
 	pr_debug("nvm: block '%lu' is full, allow GC (sched)\n",
 							rblk->parent->id);
@@ -1194,8 +1187,6 @@ static int rrpc_luns_init(struct rrpc *rrpc, int lun_begin, int lun_end)
 
 		rlun->rrpc = rrpc;
 		INIT_LIST_HEAD(&rlun->prio_list);
-		INIT_LIST_HEAD(&rlun->open_list);
-		INIT_LIST_HEAD(&rlun->closed_list);
 
 		INIT_WORK(&rlun->ws_gc, rrpc_lun_gc);
 		spin_lock_init(&rlun->lock);
diff --git a/drivers/lightnvm/rrpc.h b/drivers/lightnvm/rrpc.h
index 5797343..448e39a 100644
--- a/drivers/lightnvm/rrpc.h
+++ b/drivers/lightnvm/rrpc.h
@@ -56,7 +56,6 @@ struct rrpc_block {
 	struct nvm_block *parent;
 	struct rrpc_lun *rlun;
 	struct list_head prio;
-	struct list_head list;
 
 #define MAX_INVALID_PAGES_STORAGE 8
 	/* Bitmap for invalid page intries */
@@ -77,13 +76,6 @@ struct rrpc_lun {
 	struct rrpc_block *blocks;	/* Reference to block allocation */
 
 	struct list_head prio_list;	/* Blocks that may be GC'ed */
-	struct list_head open_list;	/* In-use open blocks. These are blocks
-					 * that can be both written to and read
-					 * from
-					 */
-	struct list_head closed_list;	/* In-use closed blocks. These are
-					 * blocks that can _only_ be read from
-					 */
 
 	struct work_struct ws_gc;
 
-- 
2.1.4

[PATCH 13/17] lightnvm: remove _unlocked variant of [get/put]_blk

[Matias Bjørling]

The [get/put]_blk API enables targets to get ownership of blocks at
runtime. This information is currently not recorded on disk, and the
information is therefore lost on power failure. To restore the
metadata, the [get/put]_blk must persist its metadata. In that case,
we need to control the outer lock, so that we can disable them while
updating the on-disk metadata. Fortunately, the _unlocked versions can
be removed, which allows us to move the lock into the [get/put]_blk
functions.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/core.c   | 14 --------------
 drivers/lightnvm/gennvm.c | 32 ++++----------------------------
 drivers/lightnvm/rrpc.c   | 14 ++------------
 include/linux/lightnvm.h  |  6 ------
 4 files changed, 6 insertions(+), 60 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 04469e0..ddc8098 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -176,20 +176,6 @@ static struct nvm_dev *nvm_find_nvm_dev(const char *name)
 	return NULL;
 }
 
-struct nvm_block *nvm_get_blk_unlocked(struct nvm_dev *dev, struct nvm_lun *lun,
-							unsigned long flags)
-{
-	return dev->mt->get_blk_unlocked(dev, lun, flags);
-}
-EXPORT_SYMBOL(nvm_get_blk_unlocked);
-
-/* Assumes that all valid pages have already been moved on release to bm */
-void nvm_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
-{
-	return dev->mt->put_blk_unlocked(dev, blk);
-}
-EXPORT_SYMBOL(nvm_put_blk_unlocked);
-
 struct nvm_block *nvm_get_blk(struct nvm_dev *dev, struct nvm_lun *lun,
 							unsigned long flags)
 {
diff --git a/drivers/lightnvm/gennvm.c b/drivers/lightnvm/gennvm.c
index 41760b2..c65fb67 100644
--- a/drivers/lightnvm/gennvm.c
+++ b/drivers/lightnvm/gennvm.c
@@ -473,15 +473,14 @@ static void gen_unregister(struct nvm_dev *dev)
 	module_put(THIS_MODULE);
 }
 
-static struct nvm_block *gen_get_blk_unlocked(struct nvm_dev *dev,
+static struct nvm_block *gen_get_blk(struct nvm_dev *dev,
 				struct nvm_lun *vlun, unsigned long flags)
 {
 	struct gen_lun *lun = container_of(vlun, struct gen_lun, vlun);
 	struct nvm_block *blk = NULL;
 	int is_gc = flags & NVM_IOTYPE_GC;
 
-	assert_spin_locked(&vlun->lock);
-
+	spin_lock(&vlun->lock);
 	if (list_empty(&lun->free_list)) {
 		pr_err_ratelimited("gen: lun %u have no free pages available",
 								lun->vlun.id);
@@ -496,29 +495,17 @@ static struct nvm_block *gen_get_blk_unlocked(struct nvm_dev *dev,
 	list_move_tail(&blk->list, &lun->used_list);
 	blk->state = NVM_BLK_ST_TGT;
 	lun->vlun.nr_free_blocks--;
-
 out:
-	return blk;
-}
-
-static struct nvm_block *gen_get_blk(struct nvm_dev *dev,
-				struct nvm_lun *vlun, unsigned long flags)
-{
-	struct nvm_block *blk;
-
-	spin_lock(&vlun->lock);
-	blk = gen_get_blk_unlocked(dev, vlun, flags);
 	spin_unlock(&vlun->lock);
 	return blk;
 }
 
-static void gen_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
+static void gen_put_blk(struct nvm_dev *dev, struct nvm_block *blk)
 {
 	struct nvm_lun *vlun = blk->lun;
 	struct gen_lun *lun = container_of(vlun, struct gen_lun, vlun);
 
-	assert_spin_locked(&vlun->lock);
-
+	spin_lock(&vlun->lock);
 	if (blk->state & NVM_BLK_ST_TGT) {
 		list_move_tail(&blk->list, &lun->free_list);
 		lun->vlun.nr_free_blocks++;
@@ -532,14 +519,6 @@ static void gen_put_blk_unlocked(struct nvm_dev *dev, struct nvm_block *blk)
 							blk->id, blk->state);
 		list_move_tail(&blk->list, &lun->bb_list);
 	}
-}
-
-static void gen_put_blk(struct nvm_dev *dev, struct nvm_block *blk)
-{
-	struct nvm_lun *vlun = blk->lun;
-
-	spin_lock(&vlun->lock);
-	gen_put_blk_unlocked(dev, blk);
 	spin_unlock(&vlun->lock);
 }
 
@@ -669,9 +648,6 @@ static struct nvmm_type gen = {
 	.create_tgt		= gen_create_tgt,
 	.remove_tgt		= gen_remove_tgt,
 
-	.get_blk_unlocked	= gen_get_blk_unlocked,
-	.put_blk_unlocked	= gen_put_blk_unlocked,
-
 	.get_blk		= gen_get_blk,
 	.put_blk		= gen_put_blk,
 
diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index 10ed22b..fa8d5be 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -192,21 +192,16 @@ static void rrpc_set_lun_cur(struct rrpc_lun *rlun, struct rrpc_block *rblk)
 static struct rrpc_block *rrpc_get_blk(struct rrpc *rrpc, struct rrpc_lun *rlun,
 							unsigned long flags)
 {
-	struct nvm_lun *lun = rlun->parent;
 	struct nvm_block *blk;
 	struct rrpc_block *rblk;
 
-	spin_lock(&lun->lock);
-	blk = nvm_get_blk_unlocked(rrpc->dev, rlun->parent, flags);
+	blk = nvm_get_blk(rrpc->dev, rlun->parent, flags);
 	if (!blk) {
 		pr_err("nvm: rrpc: cannot get new block from media manager\n");
-		spin_unlock(&lun->lock);
 		return NULL;
 	}
 
 	rblk = rrpc_get_rblk(rlun, blk->id);
-	spin_unlock(&lun->lock);
-
 	blk->priv = rblk;
 	bitmap_zero(rblk->invalid_pages, rrpc->dev->sec_per_blk);
 	rblk->next_page = 0;
@@ -218,12 +213,7 @@ static struct rrpc_block *rrpc_get_blk(struct rrpc *rrpc, struct rrpc_lun *rlun,
 
 static void rrpc_put_blk(struct rrpc *rrpc, struct rrpc_block *rblk)
 {
-	struct rrpc_lun *rlun = rblk->rlun;
-	struct nvm_lun *lun = rlun->parent;
-
-	spin_lock(&lun->lock);
-	nvm_put_blk_unlocked(rrpc->dev, rblk->parent);
-	spin_unlock(&lun->lock);
+	nvm_put_blk(rrpc->dev, rblk->parent);
 }
 
 static void rrpc_put_blks(struct rrpc *rrpc)
diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index d619f6d..e9836cf 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -491,8 +491,6 @@ struct nvmm_type {
 	nvmm_remove_tgt_fn *remove_tgt;
 
 	/* Block administration callbacks */
-	nvmm_get_blk_fn *get_blk_unlocked;
-	nvmm_put_blk_fn *put_blk_unlocked;
 	nvmm_get_blk_fn *get_blk;
 	nvmm_put_blk_fn *put_blk;
 	nvmm_open_blk_fn *open_blk;
@@ -522,10 +520,6 @@ struct nvmm_type {
 extern int nvm_register_mgr(struct nvmm_type *);
 extern void nvm_unregister_mgr(struct nvmm_type *);
 
-extern struct nvm_block *nvm_get_blk_unlocked(struct nvm_dev *,
-					struct nvm_lun *, unsigned long);
-extern void nvm_put_blk_unlocked(struct nvm_dev *, struct nvm_block *);
-
 extern struct nvm_block *nvm_get_blk(struct nvm_dev *, struct nvm_lun *,
 								unsigned long);
 extern void nvm_put_blk(struct nvm_dev *, struct nvm_block *);
-- 
2.1.4

[PATCH 14/17] lightnvm: make rrpc_map_page call nvm_get_blk outside locks

[Matias Bjørling]

The nvm_get_blk() function is called with rlun->lock held. This is ok
when the media manager implementation doesn't go out of its atomic
context. However, if a media manager persists its metadata, and
guarantees that the block is given to the target, this is no longer
a viable approach. Therefore, clean up the flow of rrpc_map_page,
and make sure that nvm_get_blk() is called without any locks acquired.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/rrpc.c | 108 +++++++++++++++++++++++++++++-------------------
 drivers/lightnvm/rrpc.h |   1 +
 2 files changed, 66 insertions(+), 43 deletions(-)

diff --git a/drivers/lightnvm/rrpc.c b/drivers/lightnvm/rrpc.c
index fa8d5be..fa1ab04 100644
--- a/drivers/lightnvm/rrpc.c
+++ b/drivers/lightnvm/rrpc.c
@@ -175,18 +175,17 @@ static struct ppa_addr rrpc_ppa_to_gaddr(struct nvm_dev *dev, u64 addr)
 }
 
 /* requires lun->lock taken */
-static void rrpc_set_lun_cur(struct rrpc_lun *rlun, struct rrpc_block *rblk)
+static void rrpc_set_lun_cur(struct rrpc_lun *rlun, struct rrpc_block *new_rblk,
+						struct rrpc_block **cur_rblk)
 {
 	struct rrpc *rrpc = rlun->rrpc;
 
-	BUG_ON(!rblk);
-
-	if (rlun->cur) {
-		spin_lock(&rlun->cur->lock);
-		WARN_ON(!block_is_full(rrpc, rlun->cur));
-		spin_unlock(&rlun->cur->lock);
+	if (*cur_rblk) {
+		spin_lock(&(*cur_rblk)->lock);
+		WARN_ON(!block_is_full(rrpc, *cur_rblk));
+		spin_unlock(&(*cur_rblk)->lock);
 	}
-	rlun->cur = rblk;
+	*cur_rblk = new_rblk;
 }
 
 static struct rrpc_block *rrpc_get_blk(struct rrpc *rrpc, struct rrpc_lun *rlun,
@@ -577,21 +576,20 @@ out:
 	return addr;
 }
 
-/* Simple round-robin Logical to physical address translation.
+/* Map logical address to a physical page. The mapping implements a round robin
+ * approach and allocates a page from the next lun available.
  *
- * Retrieve the mapping using the active append point. Then update the ap for
- * the next write to the disk.
- *
- * Returns rrpc_addr with the physical address and block. Remember to return to
- * rrpc->addr_cache when request is finished.
+ * Returns rrpc_addr with the physical address and block. Returns NULL if no
+ * blocks in the next rlun are available.
  */
 static struct rrpc_addr *rrpc_map_page(struct rrpc *rrpc, sector_t laddr,
 								int is_gc)
 {
 	struct rrpc_lun *rlun;
-	struct rrpc_block *rblk;
+	struct rrpc_block *rblk, **cur_rblk;
 	struct nvm_lun *lun;
 	u64 paddr;
+	int gc_force = 0;
 
 	rlun = rrpc_get_lun_rr(rrpc, is_gc);
 	lun = rlun->parent;
@@ -599,41 +597,65 @@ static struct rrpc_addr *rrpc_map_page(struct rrpc *rrpc, sector_t laddr,
 	if (!is_gc && lun->nr_free_blocks < rrpc->nr_luns * 4)
 		return NULL;
 
-	spin_lock(&rlun->lock);
+	/*
+	 * page allocation steps:
+	 * 1. Try to allocate new page from current rblk
+	 * 2a. If succeed, proceed to map it in and return
+	 * 2b. If fail, first try to allocate a new block from media manger,
+	 *     and then retry step 1. Retry until the normal block pool is
+	 *     exhausted.
+	 * 3. If exhausted, and garbage collector is requesting the block,
+	 *    go to the reserved block and retry step 1.
+	 *    In the case that this fails as well, or it is not GC
+	 *    requesting, report not able to retrieve a block and let the
+	 *    caller handle further processing.
+	 */
 
+	spin_lock(&rlun->lock);
+	cur_rblk = &rlun->cur;
 	rblk = rlun->cur;
 retry:
 	paddr = rrpc_alloc_addr(rrpc, rblk);
 
-	if (paddr == ADDR_EMPTY) {
-		rblk = rrpc_get_blk(rrpc, rlun, 0);
-		if (rblk) {
-			rrpc_set_lun_cur(rlun, rblk);
-			goto retry;
-		}
+	if (paddr != ADDR_EMPTY)
+		goto done;
 
-		if (is_gc) {
-			/* retry from emergency gc block */
-			paddr = rrpc_alloc_addr(rrpc, rlun->gc_cur);
-			if (paddr == ADDR_EMPTY) {
-				rblk = rrpc_get_blk(rrpc, rlun, 1);
-				if (!rblk) {
-					pr_err("rrpc: no more blocks");
-					goto err;
-				}
+	if (!list_empty(&rlun->wblk_list)) {
+new_blk:
+		rblk = list_first_entry(&rlun->wblk_list, struct rrpc_block,
+									prio);
+		rrpc_set_lun_cur(rlun, rblk, cur_rblk);
+		list_del(&rblk->prio);
+		goto retry;
+	}
+	spin_unlock(&rlun->lock);
 
-				rlun->gc_cur = rblk;
-				paddr = rrpc_alloc_addr(rrpc, rlun->gc_cur);
-			}
-			rblk = rlun->gc_cur;
-		}
+	rblk = rrpc_get_blk(rrpc, rlun, gc_force);
+	if (rblk) {
+		spin_lock(&rlun->lock);
+		list_add_tail(&rblk->prio, &rlun->wblk_list);
+		/*
+		 * another thread might already have added a new block,
+		 * Therefore, make sure that one is used, instead of the
+		 * one just added.
+		 */
+		goto new_blk;
 	}
 
-	spin_unlock(&rlun->lock);
-	return rrpc_update_map(rrpc, laddr, rblk, paddr);
-err:
-	spin_unlock(&rlun->lock);
+	if (unlikely(is_gc) && !gc_force) {
+		/* retry from emergency gc block */
+		cur_rblk = &rlun->gc_cur;
+		rblk = rlun->gc_cur;
+		gc_force = 1;
+		spin_lock(&rlun->lock);
+		goto retry;
+	}
+
+	pr_err("rrpc: failed to allocate new block\n");
 	return NULL;
+done:
+	spin_unlock(&rlun->lock);
+	return rrpc_update_map(rrpc, laddr, rblk, paddr);
 }
 
 static void rrpc_run_gc(struct rrpc *rrpc, struct rrpc_block *rblk)
@@ -1177,6 +1199,7 @@ static int rrpc_luns_init(struct rrpc *rrpc, int lun_begin, int lun_end)
 
 		rlun->rrpc = rrpc;
 		INIT_LIST_HEAD(&rlun->prio_list);
+		INIT_LIST_HEAD(&rlun->wblk_list);
 
 		INIT_WORK(&rlun->ws_gc, rrpc_lun_gc);
 		spin_lock_init(&rlun->lock);
@@ -1317,14 +1340,13 @@ static int rrpc_luns_configure(struct rrpc *rrpc)
 		rblk = rrpc_get_blk(rrpc, rlun, 0);
 		if (!rblk)
 			goto err;
-
-		rrpc_set_lun_cur(rlun, rblk);
+		rrpc_set_lun_cur(rlun, rblk, &rlun->cur);
 
 		/* Emergency gc block */
 		rblk = rrpc_get_blk(rrpc, rlun, 1);
 		if (!rblk)
 			goto err;
-		rlun->gc_cur = rblk;
+		rrpc_set_lun_cur(rlun, rblk, &rlun->gc_cur);
 	}
 
 	return 0;
diff --git a/drivers/lightnvm/rrpc.h b/drivers/lightnvm/rrpc.h
index 448e39a..5e87d52 100644
--- a/drivers/lightnvm/rrpc.h
+++ b/drivers/lightnvm/rrpc.h
@@ -76,6 +76,7 @@ struct rrpc_lun {
 	struct rrpc_block *blocks;	/* Reference to block allocation */
 
 	struct list_head prio_list;	/* Blocks that may be GC'ed */
+	struct list_head wblk_list;	/* Queued blocks to be written to */
 
 	struct work_struct ws_gc;
 
-- 
2.1.4

[PATCH 15/17] lightnvm: fix lun offset calculation for mark blk

[Matias Bjørling]

The gen_mark_blk_bad function marks the wrong block when a block is on
a different channel. Fix the index calculation, so that it updates the
correct block.

Reported-by: Javier Gonzalez <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/gennvm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/lightnvm/gennvm.c b/drivers/lightnvm/gennvm.c
index c65fb67..b74174c 100644
--- a/drivers/lightnvm/gennvm.c
+++ b/drivers/lightnvm/gennvm.c
@@ -542,7 +542,7 @@ static void gen_mark_blk(struct nvm_dev *dev, struct ppa_addr ppa, int type)
 		return;
 	}
 
-	lun = &gn->luns[ppa.g.lun * ppa.g.ch];
+	lun = &gn->luns[(dev->luns_per_chnl * ppa.g.ch) + ppa.g.lun];
 	blk = &lun->vlun.blocks[ppa.g.blk];
 
 	/* will be moved to bb list on put_blk from target */
-- 
2.1.4

[PATCH 16/17] lightnvm: make ppa_list const in nvm_set_rqd_list

[Matias Bjørling]

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>
---
 drivers/lightnvm/core.c  | 8 +++++---
 include/linux/lightnvm.h | 2 +-
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index ddc8098..00b64f7 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -237,9 +237,10 @@ void nvm_generic_to_addr_mode(struct nvm_dev *dev, struct nvm_rq *rqd)
 EXPORT_SYMBOL(nvm_generic_to_addr_mode);
 
 int nvm_set_rqd_ppalist(struct nvm_dev *dev, struct nvm_rq *rqd,
-				struct ppa_addr *ppas, int nr_ppas, int vblk)
+			const struct ppa_addr *ppas, int nr_ppas, int vblk)
 {
 	int i, plane_cnt, pl_idx;
+	struct ppa_addr ppa;
 
 	if ((!vblk || dev->plane_mode == NVM_PLANE_SINGLE) && nr_ppas == 1) {
 		rqd->nr_ppas = nr_ppas;
@@ -264,8 +265,9 @@ int nvm_set_rqd_ppalist(struct nvm_dev *dev, struct nvm_rq *rqd,
 
 		for (i = 0; i < nr_ppas; i++) {
 			for (pl_idx = 0; pl_idx < plane_cnt; pl_idx++) {
-				ppas[i].g.pl = pl_idx;
-				rqd->ppa_list[(pl_idx * nr_ppas) + i] = ppas[i];
+				ppa = ppas[i];
+				ppa.g.pl = pl_idx;
+				rqd->ppa_list[(pl_idx * nr_ppas) + i] = ppa;
 			}
 		}
 	}
diff --git a/include/linux/lightnvm.h b/include/linux/lightnvm.h
index e9836cf..ba78b83 100644
--- a/include/linux/lightnvm.h
+++ b/include/linux/lightnvm.h
@@ -534,7 +534,7 @@ extern int nvm_submit_io(struct nvm_dev *, struct nvm_rq *);
 extern void nvm_generic_to_addr_mode(struct nvm_dev *, struct nvm_rq *);
 extern void nvm_addr_to_generic_mode(struct nvm_dev *, struct nvm_rq *);
 extern int nvm_set_rqd_ppalist(struct nvm_dev *, struct nvm_rq *,
-						struct ppa_addr *, int, int);
+					const struct ppa_addr *, int, int);
 extern void nvm_free_rqd_ppalist(struct nvm_dev *, struct nvm_rq *);
 extern int nvm_erase_ppa(struct nvm_dev *, struct ppa_addr *, int);
 extern int nvm_erase_blk(struct nvm_dev *, struct nvm_block *);
-- 
2.1.4

[PATCH 17/17] lightnvm: make __nvm_submit_ppa static

[Matias Bjørling]

The __nvm_submit_ppa() function is not used outside lightnvm core.

Signed-off-by: Matias Bjørling <m@bjorling.me>
---
 drivers/lightnvm/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index 00b64f7..9ebd2cf 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -325,7 +325,7 @@ static void nvm_end_io_sync(struct nvm_rq *rqd)
 	complete(waiting);
 }
 
-int __nvm_submit_ppa(struct nvm_dev *dev, struct nvm_rq *rqd, int opcode,
+static int __nvm_submit_ppa(struct nvm_dev *dev, struct nvm_rq *rqd, int opcode,
 						int flags, void *buf, int len)
 {
 	DECLARE_COMPLETION_ONSTACK(wait);
-- 
2.1.4

Re: [GIT PULL 00/17] LightNVM pull request for 4.8

[Jens Axboe]

On 07/07/2016 01:54 AM, Matias Bjørling wrote:
> Hi Jens,
>
> Please pick up the following patches.
>
> Nothing major. A couple of bigger pieces are being prepared. Notably
> pblk target, sysfs support and persistent block information.

Pulled for 4.8, thanks.

-- 
Jens Axboe