From 5ee0524ba137fe928a88b440d014e3c8451fb32c Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 28 Feb 2018 10:15:31 -0800 Subject: block: Add 'lock' as third argument to blk_alloc_queue_node() This patch does not change any functionality. Signed-off-by: Bart Van Assche Reviewed-by: Joseph Qi Cc: Christoph Hellwig Cc: Philipp Reisner Cc: Ulf Hansson Cc: Kees Cook Signed-off-by: Jens Axboe --- include/linux/blkdev.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/linux/blkdev.h') diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index ed63f3b69c12..667a9b0053d9 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -1321,7 +1321,8 @@ extern long nr_blockdev_pages(void); bool __must_check blk_get_queue(struct request_queue *); struct request_queue *blk_alloc_queue(gfp_t); -struct request_queue *blk_alloc_queue_node(gfp_t, int); +struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id, + spinlock_t *lock); extern void blk_put_queue(struct request_queue *); extern void blk_set_queue_dying(struct request_queue *); -- cgit v1.2.3 From 66f91322f39cd18a01524264464c2ff4c98c936e Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 7 Mar 2018 17:10:02 -0800 Subject: block: Reorder the queue flag manipulation function definitions Move the definition of queue_flag_clear_unlocked() up and move the definition of queue_in_flight() down such that all queue flag manipulation function definitions become contiguous. This patch does not change any functionality. Cc: Christoph Hellwig Cc: Hannes Reinecke Cc: Ming Lei Reviewed-by: Johannes Thumshirn Reviewed-by: Martin K. Petersen Signed-off-by: Bart Van Assche Signed-off-by: Jens Axboe --- include/linux/blkdev.h | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) (limited to 'include/linux/blkdev.h') diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 667a9b0053d9..c351aaec3ca7 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -726,6 +726,12 @@ static inline void queue_flag_set_unlocked(unsigned int flag, __set_bit(flag, &q->queue_flags); } +static inline void queue_flag_clear_unlocked(unsigned int flag, + struct request_queue *q) +{ + __clear_bit(flag, &q->queue_flags); +} + static inline int queue_flag_test_and_clear(unsigned int flag, struct request_queue *q) { @@ -758,17 +764,6 @@ static inline void queue_flag_set(unsigned int flag, struct request_queue *q) __set_bit(flag, &q->queue_flags); } -static inline void queue_flag_clear_unlocked(unsigned int flag, - struct request_queue *q) -{ - __clear_bit(flag, &q->queue_flags); -} - -static inline int queue_in_flight(struct request_queue *q) -{ - return q->in_flight[0] + q->in_flight[1]; -} - static inline void queue_flag_clear(unsigned int flag, struct request_queue *q) { queue_lockdep_assert_held(q); @@ -804,6 +799,11 @@ static inline void queue_flag_clear(unsigned int flag, struct request_queue *q) extern int blk_set_preempt_only(struct request_queue *q); extern void blk_clear_preempt_only(struct request_queue *q); +static inline int queue_in_flight(struct request_queue *q) +{ + return q->in_flight[0] + q->in_flight[1]; +} + static inline bool blk_account_rq(struct request *rq) { return (rq->rq_flags & RQF_STARTED) && !blk_rq_is_passthrough(rq); -- cgit v1.2.3 From 8814ce8a0f680599a837af18aefdec774e5c7b97 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 7 Mar 2018 17:10:04 -0800 Subject: block: Introduce blk_queue_flag_{set,clear,test_and_{set,clear}}() Introduce functions that modify the queue flags and that protect these modifications with the request queue lock. Except for moving one wake_up_all() call from inside to outside a critical section, this patch does not change any functionality. Cc: Christoph Hellwig Cc: Hannes Reinecke Cc: Ming Lei Reviewed-by: Johannes Thumshirn Reviewed-by: Martin K. Petersen Signed-off-by: Bart Van Assche Signed-off-by: Jens Axboe --- block/blk-core.c | 91 +++++++++++++++++++++++++++++++++++++++++--------- block/blk-mq.c | 12 ++----- block/blk-settings.c | 6 ++-- block/blk-sysfs.c | 22 ++++-------- block/blk-timeout.c | 6 ++-- include/linux/blkdev.h | 5 +++ 6 files changed, 93 insertions(+), 49 deletions(-) (limited to 'include/linux/blkdev.h') diff --git a/block/blk-core.c b/block/blk-core.c index 241b73088617..74c6283f4509 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -71,6 +71,78 @@ struct kmem_cache *blk_requestq_cachep; */ static struct workqueue_struct *kblockd_workqueue; +/** + * blk_queue_flag_set - atomically set a queue flag + * @flag: flag to be set + * @q: request queue + */ +void blk_queue_flag_set(unsigned int flag, struct request_queue *q) +{ + unsigned long flags; + + spin_lock_irqsave(q->queue_lock, flags); + queue_flag_set(flag, q); + spin_unlock_irqrestore(q->queue_lock, flags); +} +EXPORT_SYMBOL(blk_queue_flag_set); + +/** + * blk_queue_flag_clear - atomically clear a queue flag + * @flag: flag to be cleared + * @q: request queue + */ +void blk_queue_flag_clear(unsigned int flag, struct request_queue *q) +{ + unsigned long flags; + + spin_lock_irqsave(q->queue_lock, flags); + queue_flag_clear(flag, q); + spin_unlock_irqrestore(q->queue_lock, flags); +} +EXPORT_SYMBOL(blk_queue_flag_clear); + +/** + * blk_queue_flag_test_and_set - atomically test and set a queue flag + * @flag: flag to be set + * @q: request queue + * + * Returns the previous value of @flag - 0 if the flag was not set and 1 if + * the flag was already set. + */ +bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q) +{ + unsigned long flags; + bool res; + + spin_lock_irqsave(q->queue_lock, flags); + res = queue_flag_test_and_set(flag, q); + spin_unlock_irqrestore(q->queue_lock, flags); + + return res; +} +EXPORT_SYMBOL_GPL(blk_queue_flag_test_and_set); + +/** + * blk_queue_flag_test_and_clear - atomically test and clear a queue flag + * @flag: flag to be cleared + * @q: request queue + * + * Returns the previous value of @flag - 0 if the flag was not set and 1 if + * the flag was set. + */ +bool blk_queue_flag_test_and_clear(unsigned int flag, struct request_queue *q) +{ + unsigned long flags; + bool res; + + spin_lock_irqsave(q->queue_lock, flags); + res = queue_flag_test_and_clear(flag, q); + spin_unlock_irqrestore(q->queue_lock, flags); + + return res; +} +EXPORT_SYMBOL_GPL(blk_queue_flag_test_and_clear); + static void blk_clear_congested(struct request_list *rl, int sync) { #ifdef CONFIG_CGROUP_WRITEBACK @@ -361,25 +433,14 @@ EXPORT_SYMBOL(blk_sync_queue); */ int blk_set_preempt_only(struct request_queue *q) { - unsigned long flags; - int res; - - spin_lock_irqsave(q->queue_lock, flags); - res = queue_flag_test_and_set(QUEUE_FLAG_PREEMPT_ONLY, q); - spin_unlock_irqrestore(q->queue_lock, flags); - - return res; + return blk_queue_flag_test_and_set(QUEUE_FLAG_PREEMPT_ONLY, q); } EXPORT_SYMBOL_GPL(blk_set_preempt_only); void blk_clear_preempt_only(struct request_queue *q) { - unsigned long flags; - - spin_lock_irqsave(q->queue_lock, flags); - queue_flag_clear(QUEUE_FLAG_PREEMPT_ONLY, q); + blk_queue_flag_clear(QUEUE_FLAG_PREEMPT_ONLY, q); wake_up_all(&q->mq_freeze_wq); - spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL_GPL(blk_clear_preempt_only); @@ -629,9 +690,7 @@ EXPORT_SYMBOL_GPL(blk_queue_bypass_end); void blk_set_queue_dying(struct request_queue *q) { - spin_lock_irq(q->queue_lock); - queue_flag_set(QUEUE_FLAG_DYING, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_set(QUEUE_FLAG_DYING, q); /* * When queue DYING flag is set, we need to block new req diff --git a/block/blk-mq.c b/block/blk-mq.c index e70cc7d48f58..a86899022683 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -194,11 +194,7 @@ EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); */ void blk_mq_quiesce_queue_nowait(struct request_queue *q) { - unsigned long flags; - - spin_lock_irqsave(q->queue_lock, flags); - queue_flag_set(QUEUE_FLAG_QUIESCED, q); - spin_unlock_irqrestore(q->queue_lock, flags); + blk_queue_flag_set(QUEUE_FLAG_QUIESCED, q); } EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_nowait); @@ -239,11 +235,7 @@ EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue); */ void blk_mq_unquiesce_queue(struct request_queue *q) { - unsigned long flags; - - spin_lock_irqsave(q->queue_lock, flags); - queue_flag_clear(QUEUE_FLAG_QUIESCED, q); - spin_unlock_irqrestore(q->queue_lock, flags); + blk_queue_flag_clear(QUEUE_FLAG_QUIESCED, q); /* dispatch requests which are inserted during quiescing */ blk_mq_run_hw_queues(q, true); diff --git a/block/blk-settings.c b/block/blk-settings.c index 7f719da0eadd..d1de71124656 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -859,12 +859,10 @@ EXPORT_SYMBOL(blk_queue_update_dma_alignment); void blk_queue_flush_queueable(struct request_queue *q, bool queueable) { - spin_lock_irq(q->queue_lock); if (queueable) - queue_flag_clear(QUEUE_FLAG_FLUSH_NQ, q); + blk_queue_flag_clear(QUEUE_FLAG_FLUSH_NQ, q); else - queue_flag_set(QUEUE_FLAG_FLUSH_NQ, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_set(QUEUE_FLAG_FLUSH_NQ, q); } EXPORT_SYMBOL_GPL(blk_queue_flush_queueable); diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index fd71a00c9462..d00d1b0ec109 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -276,12 +276,10 @@ queue_store_##name(struct request_queue *q, const char *page, size_t count) \ if (neg) \ val = !val; \ \ - spin_lock_irq(q->queue_lock); \ if (val) \ - queue_flag_set(QUEUE_FLAG_##flag, q); \ + blk_queue_flag_set(QUEUE_FLAG_##flag, q); \ else \ - queue_flag_clear(QUEUE_FLAG_##flag, q); \ - spin_unlock_irq(q->queue_lock); \ + blk_queue_flag_clear(QUEUE_FLAG_##flag, q); \ return ret; \ } @@ -414,12 +412,10 @@ static ssize_t queue_poll_store(struct request_queue *q, const char *page, if (ret < 0) return ret; - spin_lock_irq(q->queue_lock); if (poll_on) - queue_flag_set(QUEUE_FLAG_POLL, q); + blk_queue_flag_set(QUEUE_FLAG_POLL, q); else - queue_flag_clear(QUEUE_FLAG_POLL, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_clear(QUEUE_FLAG_POLL, q); return ret; } @@ -487,12 +483,10 @@ static ssize_t queue_wc_store(struct request_queue *q, const char *page, if (set == -1) return -EINVAL; - spin_lock_irq(q->queue_lock); if (set) - queue_flag_set(QUEUE_FLAG_WC, q); + blk_queue_flag_set(QUEUE_FLAG_WC, q); else - queue_flag_clear(QUEUE_FLAG_WC, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_clear(QUEUE_FLAG_WC, q); return count; } @@ -946,9 +940,7 @@ void blk_unregister_queue(struct gendisk *disk) */ mutex_lock(&q->sysfs_lock); - spin_lock_irq(q->queue_lock); - queue_flag_clear(QUEUE_FLAG_REGISTERED, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_clear(QUEUE_FLAG_REGISTERED, q); /* * Remove the sysfs attributes before unregistering the queue data diff --git a/block/blk-timeout.c b/block/blk-timeout.c index a05e3676d24a..34a55250f08a 100644 --- a/block/blk-timeout.c +++ b/block/blk-timeout.c @@ -57,12 +57,10 @@ ssize_t part_timeout_store(struct device *dev, struct device_attribute *attr, char *p = (char *) buf; val = simple_strtoul(p, &p, 10); - spin_lock_irq(q->queue_lock); if (val) - queue_flag_set(QUEUE_FLAG_FAIL_IO, q); + blk_queue_flag_set(QUEUE_FLAG_FAIL_IO, q); else - queue_flag_clear(QUEUE_FLAG_FAIL_IO, q); - spin_unlock_irq(q->queue_lock); + blk_queue_flag_clear(QUEUE_FLAG_FAIL_IO, q); } return count; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index c351aaec3ca7..f84b3c7887b1 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -707,6 +707,11 @@ struct request_queue { (1 << QUEUE_FLAG_SAME_COMP) | \ (1 << QUEUE_FLAG_POLL)) +void blk_queue_flag_set(unsigned int flag, struct request_queue *q); +void blk_queue_flag_clear(unsigned int flag, struct request_queue *q); +bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q); +bool blk_queue_flag_test_and_clear(unsigned int flag, struct request_queue *q); + /* * @q->queue_lock is set while a queue is being initialized. Since we know * that no other threads access the queue object before @q->queue_lock has -- cgit v1.2.3 From 1db2008b79a32db2ad41338c6c74c4735cf74f6d Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 7 Mar 2018 17:10:11 -0800 Subject: block: Complain if queue_flag_(set|clear)_unlocked() is abused Since it is not safe to use queue_flag_(set|clear)_unlocked() without holding the queue lock after the sysfs entries for a queue have been created, complain if this happens. Cc: Mike Snitzer Cc: Christoph Hellwig Cc: Hannes Reinecke Cc: Ming Lei Reviewed-by: Johannes Thumshirn Reviewed-by: Martin K. Petersen Signed-off-by: Bart Van Assche Signed-off-by: Jens Axboe --- include/linux/blkdev.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include/linux/blkdev.h') diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index f84b3c7887b1..888c9b25cb8f 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -728,12 +728,18 @@ static inline void queue_lockdep_assert_held(struct request_queue *q) static inline void queue_flag_set_unlocked(unsigned int flag, struct request_queue *q) { + if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && + kref_read(&q->kobj.kref)) + lockdep_assert_held(q->queue_lock); __set_bit(flag, &q->queue_flags); } static inline void queue_flag_clear_unlocked(unsigned int flag, struct request_queue *q) { + if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && + kref_read(&q->kobj.kref)) + lockdep_assert_held(q->queue_lock); __clear_bit(flag, &q->queue_flags); } -- cgit v1.2.3 From 8a0ac14b8da9b86cfbe7aace40c8d485ed5c5b97 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 7 Mar 2018 17:10:12 -0800 Subject: block: Move the queue_flag_*() functions from a public into a private header file This patch helps to avoid that new code gets introduced in block drivers that manipulates queue flags without holding the queue lock when that lock should be held. Cc: Christoph Hellwig Cc: Hannes Reinecke Cc: Ming Lei Reviewed-by: Johannes Thumshirn Reviewed-by: Martin K. Petersen Signed-off-by: Bart Van Assche Signed-off-by: Jens Axboe --- block/blk.h | 69 ++++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/blkdev.h | 69 -------------------------------------------------- 2 files changed, 69 insertions(+), 69 deletions(-) (limited to 'include/linux/blkdev.h') diff --git a/block/blk.h b/block/blk.h index 46db5dc83dcb..b034fd2460c4 100644 --- a/block/blk.h +++ b/block/blk.h @@ -41,6 +41,75 @@ extern struct kmem_cache *request_cachep; extern struct kobj_type blk_queue_ktype; extern struct ida blk_queue_ida; +/* + * @q->queue_lock is set while a queue is being initialized. Since we know + * that no other threads access the queue object before @q->queue_lock has + * been set, it is safe to manipulate queue flags without holding the + * queue_lock if @q->queue_lock == NULL. See also blk_alloc_queue_node() and + * blk_init_allocated_queue(). + */ +static inline void queue_lockdep_assert_held(struct request_queue *q) +{ + if (q->queue_lock) + lockdep_assert_held(q->queue_lock); +} + +static inline void queue_flag_set_unlocked(unsigned int flag, + struct request_queue *q) +{ + if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && + kref_read(&q->kobj.kref)) + lockdep_assert_held(q->queue_lock); + __set_bit(flag, &q->queue_flags); +} + +static inline void queue_flag_clear_unlocked(unsigned int flag, + struct request_queue *q) +{ + if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && + kref_read(&q->kobj.kref)) + lockdep_assert_held(q->queue_lock); + __clear_bit(flag, &q->queue_flags); +} + +static inline int queue_flag_test_and_clear(unsigned int flag, + struct request_queue *q) +{ + queue_lockdep_assert_held(q); + + if (test_bit(flag, &q->queue_flags)) { + __clear_bit(flag, &q->queue_flags); + return 1; + } + + return 0; +} + +static inline int queue_flag_test_and_set(unsigned int flag, + struct request_queue *q) +{ + queue_lockdep_assert_held(q); + + if (!test_bit(flag, &q->queue_flags)) { + __set_bit(flag, &q->queue_flags); + return 0; + } + + return 1; +} + +static inline void queue_flag_set(unsigned int flag, struct request_queue *q) +{ + queue_lockdep_assert_held(q); + __set_bit(flag, &q->queue_flags); +} + +static inline void queue_flag_clear(unsigned int flag, struct request_queue *q) +{ + queue_lockdep_assert_held(q); + __clear_bit(flag, &q->queue_flags); +} + static inline struct blk_flush_queue *blk_get_flush_queue( struct request_queue *q, struct blk_mq_ctx *ctx) { diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 888c9b25cb8f..19eaf8d89368 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -712,75 +712,6 @@ void blk_queue_flag_clear(unsigned int flag, struct request_queue *q); bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q); bool blk_queue_flag_test_and_clear(unsigned int flag, struct request_queue *q); -/* - * @q->queue_lock is set while a queue is being initialized. Since we know - * that no other threads access the queue object before @q->queue_lock has - * been set, it is safe to manipulate queue flags without holding the - * queue_lock if @q->queue_lock == NULL. See also blk_alloc_queue_node() and - * blk_init_allocated_queue(). - */ -static inline void queue_lockdep_assert_held(struct request_queue *q) -{ - if (q->queue_lock) - lockdep_assert_held(q->queue_lock); -} - -static inline void queue_flag_set_unlocked(unsigned int flag, - struct request_queue *q) -{ - if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && - kref_read(&q->kobj.kref)) - lockdep_assert_held(q->queue_lock); - __set_bit(flag, &q->queue_flags); -} - -static inline void queue_flag_clear_unlocked(unsigned int flag, - struct request_queue *q) -{ - if (test_bit(QUEUE_FLAG_INIT_DONE, &q->queue_flags) && - kref_read(&q->kobj.kref)) - lockdep_assert_held(q->queue_lock); - __clear_bit(flag, &q->queue_flags); -} - -static inline int queue_flag_test_and_clear(unsigned int flag, - struct request_queue *q) -{ - queue_lockdep_assert_held(q); - - if (test_bit(flag, &q->queue_flags)) { - __clear_bit(flag, &q->queue_flags); - return 1; - } - - return 0; -} - -static inline int queue_flag_test_and_set(unsigned int flag, - struct request_queue *q) -{ - queue_lockdep_assert_held(q); - - if (!test_bit(flag, &q->queue_flags)) { - __set_bit(flag, &q->queue_flags); - return 0; - } - - return 1; -} - -static inline void queue_flag_set(unsigned int flag, struct request_queue *q) -{ - queue_lockdep_assert_held(q); - __set_bit(flag, &q->queue_flags); -} - -static inline void queue_flag_clear(unsigned int flag, struct request_queue *q) -{ - queue_lockdep_assert_held(q); - __clear_bit(flag, &q->queue_flags); -} - #define blk_queue_tagged(q) test_bit(QUEUE_FLAG_QUEUED, &(q)->queue_flags) #define blk_queue_stopped(q) test_bit(QUEUE_FLAG_STOPPED, &(q)->queue_flags) #define blk_queue_dying(q) test_bit(QUEUE_FLAG_DYING, &(q)->queue_flags) -- cgit v1.2.3 From 233bde21aa43516baa013ef7ac33f3427056db3e Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 14 Mar 2018 15:48:06 -0700 Subject: block: Move SECTOR_SIZE and SECTOR_SHIFT definitions into It happens often while I'm preparing a patch for a block driver that I'm wondering: is a definition of SECTOR_SIZE and/or SECTOR_SHIFT available for this driver? Do I have to introduce definitions of these constants before I can use these constants? To avoid this confusion, move the existing definitions of SECTOR_SIZE and SECTOR_SHIFT into the header file such that these become available for all block drivers. Make the SECTOR_SIZE definition in the uapi msdos_fs.h header file conditional to avoid that including that header file after causes the compiler to complain about a SECTOR_SIZE redefinition. Note: the SECTOR_SIZE / SECTOR_SHIFT / SECTOR_BITS definitions have not been removed from uapi header files nor from NAND drivers in which these constants are used for another purpose than converting block layer offsets and sizes into a number of sectors. Cc: David S. Miller Cc: Mike Snitzer Cc: Dan Williams Cc: Minchan Kim Cc: Nitin Gupta Reviewed-by: Sergey Senozhatsky Reviewed-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Reviewed-by: Martin K. Petersen Signed-off-by: Bart Van Assche Signed-off-by: Jens Axboe --- arch/xtensa/platforms/iss/simdisk.c | 1 - drivers/block/brd.c | 1 - drivers/block/null_blk.c | 2 -- drivers/block/rbd.c | 9 -------- drivers/block/zram/zram_drv.h | 1 - drivers/ide/ide-cd.c | 8 +++---- drivers/ide/ide-cd.h | 6 +----- drivers/nvdimm/nd.h | 1 - drivers/scsi/gdth.h | 3 --- include/linux/blkdev.h | 42 +++++++++++++++++++++++++++---------- include/linux/device-mapper.h | 2 -- include/linux/ide.h | 1 - include/uapi/linux/msdos_fs.h | 2 ++ 13 files changed, 38 insertions(+), 41 deletions(-) (limited to 'include/linux/blkdev.h') diff --git a/arch/xtensa/platforms/iss/simdisk.c b/arch/xtensa/platforms/iss/simdisk.c index 1b6418407467..026211e7ab09 100644 --- a/arch/xtensa/platforms/iss/simdisk.c +++ b/arch/xtensa/platforms/iss/simdisk.c @@ -21,7 +21,6 @@ #include #define SIMDISK_MAJOR 240 -#define SECTOR_SHIFT 9 #define SIMDISK_MINORS 1 #define MAX_SIMDISK_COUNT 10 diff --git a/drivers/block/brd.c b/drivers/block/brd.c index deea78e485da..66cb0f857f64 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -24,7 +24,6 @@ #include -#define SECTOR_SHIFT 9 #define PAGE_SECTORS_SHIFT (PAGE_SHIFT - SECTOR_SHIFT) #define PAGE_SECTORS (1 << PAGE_SECTORS_SHIFT) diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c index 0517613afccb..a76553293a31 100644 --- a/drivers/block/null_blk.c +++ b/drivers/block/null_blk.c @@ -16,10 +16,8 @@ #include #include -#define SECTOR_SHIFT 9 #define PAGE_SECTORS_SHIFT (PAGE_SHIFT - SECTOR_SHIFT) #define PAGE_SECTORS (1 << PAGE_SECTORS_SHIFT) -#define SECTOR_SIZE (1 << SECTOR_SHIFT) #define SECTOR_MASK (PAGE_SECTORS - 1) #define FREE_BATCH 16 diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 0016170cde0a..1e03b04819c8 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -50,15 +50,6 @@ #define RBD_DEBUG /* Activate rbd_assert() calls */ -/* - * The basic unit of block I/O is a sector. It is interpreted in a - * number of contexts in Linux (blk, bio, genhd), but the default is - * universally 512 bytes. These symbols are just slightly more - * meaningful than the bare numbers they represent. - */ -#define SECTOR_SHIFT 9 -#define SECTOR_SIZE (1ULL << SECTOR_SHIFT) - /* * Increment the given counter and return its updated value. * If the counter is already 0 it will not be incremented. diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h index 31762db861e3..1e9bf65c0bfb 100644 --- a/drivers/block/zram/zram_drv.h +++ b/drivers/block/zram/zram_drv.h @@ -37,7 +37,6 @@ static const size_t max_zpage_size = PAGE_SIZE / 4 * 3; /*-- End of configurable params */ -#define SECTOR_SHIFT 9 #define SECTORS_PER_PAGE_SHIFT (PAGE_SHIFT - SECTOR_SHIFT) #define SECTORS_PER_PAGE (1 << SECTORS_PER_PAGE_SHIFT) #define ZRAM_LOGICAL_BLOCK_SHIFT 12 diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index 5613cc2d51fc..5a8e8e3c22cd 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -712,7 +712,7 @@ static ide_startstop_t cdrom_start_rw(ide_drive_t *drive, struct request *rq) struct request_queue *q = drive->queue; int write = rq_data_dir(rq) == WRITE; unsigned short sectors_per_frame = - queue_logical_block_size(q) >> SECTOR_BITS; + queue_logical_block_size(q) >> SECTOR_SHIFT; ide_debug_log(IDE_DBG_RQ, "rq->cmd[0]: 0x%x, rq->cmd_flags: 0x%x, " "secs_per_frame: %u", @@ -919,7 +919,7 @@ static int cdrom_read_capacity(ide_drive_t *drive, unsigned long *capacity, * end up being bogus. */ blocklen = be32_to_cpu(capbuf.blocklen); - blocklen = (blocklen >> SECTOR_BITS) << SECTOR_BITS; + blocklen = (blocklen >> SECTOR_SHIFT) << SECTOR_SHIFT; switch (blocklen) { case 512: case 1024: @@ -935,7 +935,7 @@ static int cdrom_read_capacity(ide_drive_t *drive, unsigned long *capacity, } *capacity = 1 + be32_to_cpu(capbuf.lba); - *sectors_per_frame = blocklen >> SECTOR_BITS; + *sectors_per_frame = blocklen >> SECTOR_SHIFT; ide_debug_log(IDE_DBG_PROBE, "cap: %lu, sectors_per_frame: %lu", *capacity, *sectors_per_frame); @@ -1012,7 +1012,7 @@ int ide_cd_read_toc(ide_drive_t *drive, struct request_sense *sense) drive->probed_capacity = toc->capacity * sectors_per_frame; blk_queue_logical_block_size(drive->queue, - sectors_per_frame << SECTOR_BITS); + sectors_per_frame << SECTOR_SHIFT); /* first read just the header, so we know how long the TOC is */ stat = cdrom_read_tocentry(drive, 0, 1, 0, (char *) &toc->hdr, diff --git a/drivers/ide/ide-cd.h b/drivers/ide/ide-cd.h index 264e822eba58..04f0f310a856 100644 --- a/drivers/ide/ide-cd.h +++ b/drivers/ide/ide-cd.h @@ -21,11 +21,7 @@ /************************************************************************/ -#define SECTOR_BITS 9 -#ifndef SECTOR_SIZE -#define SECTOR_SIZE (1 << SECTOR_BITS) -#endif -#define SECTORS_PER_FRAME (CD_FRAMESIZE >> SECTOR_BITS) +#define SECTORS_PER_FRAME (CD_FRAMESIZE >> SECTOR_SHIFT) #define SECTOR_BUFFER_SIZE (CD_FRAMESIZE * 32) /* Capabilities Page size including 8 bytes of Mode Page Header */ diff --git a/drivers/nvdimm/nd.h b/drivers/nvdimm/nd.h index 8d6375ee0fda..184e070d50a2 100644 --- a/drivers/nvdimm/nd.h +++ b/drivers/nvdimm/nd.h @@ -29,7 +29,6 @@ enum { * BTT instance */ ND_MAX_LANES = 256, - SECTOR_SHIFT = 9, INT_LBASIZE_ALIGNMENT = 64, NVDIMM_IO_ATOMIC = 1, }; diff --git a/drivers/scsi/gdth.h b/drivers/scsi/gdth.h index 95fc720c1b30..e6e5ccb1e0f3 100644 --- a/drivers/scsi/gdth.h +++ b/drivers/scsi/gdth.h @@ -178,9 +178,6 @@ #define MSG_SIZE 34 /* size of message structure */ #define MSG_REQUEST 0 /* async. event: message */ -/* cacheservice defines */ -#define SECTOR_SIZE 0x200 /* always 512 bytes per sec. */ - /* DPMEM constants */ #define DPMEM_MAGIC 0xC0FFEE11 #define IC_HEADER_BYTES 48 diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 19eaf8d89368..9af3e0f430bc 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -1021,6 +1021,19 @@ static inline struct request_queue *bdev_get_queue(struct block_device *bdev) return bdev->bd_disk->queue; /* this is never NULL */ } +/* + * The basic unit of block I/O is a sector. It is used in a number of contexts + * in Linux (blk, bio, genhd). The size of one sector is 512 = 2**9 + * bytes. Variables of type sector_t represent an offset or size that is a + * multiple of 512 bytes. Hence these two constants. + */ +#ifndef SECTOR_SHIFT +#define SECTOR_SHIFT 9 +#endif +#ifndef SECTOR_SIZE +#define SECTOR_SIZE (1 << SECTOR_SHIFT) +#endif + /* * blk_rq_pos() : the current sector * blk_rq_bytes() : bytes left in the entire request @@ -1048,12 +1061,12 @@ extern unsigned int blk_rq_err_bytes(const struct request *rq); static inline unsigned int blk_rq_sectors(const struct request *rq) { - return blk_rq_bytes(rq) >> 9; + return blk_rq_bytes(rq) >> SECTOR_SHIFT; } static inline unsigned int blk_rq_cur_sectors(const struct request *rq) { - return blk_rq_cur_bytes(rq) >> 9; + return blk_rq_cur_bytes(rq) >> SECTOR_SHIFT; } static inline unsigned int blk_rq_zone_no(struct request *rq) @@ -1083,7 +1096,8 @@ static inline unsigned int blk_queue_get_max_sectors(struct request_queue *q, int op) { if (unlikely(op == REQ_OP_DISCARD || op == REQ_OP_SECURE_ERASE)) - return min(q->limits.max_discard_sectors, UINT_MAX >> 9); + return min(q->limits.max_discard_sectors, + UINT_MAX >> SECTOR_SHIFT); if (unlikely(op == REQ_OP_WRITE_SAME)) return q->limits.max_write_same_sectors; @@ -1395,16 +1409,21 @@ extern int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, static inline int sb_issue_discard(struct super_block *sb, sector_t block, sector_t nr_blocks, gfp_t gfp_mask, unsigned long flags) { - return blkdev_issue_discard(sb->s_bdev, block << (sb->s_blocksize_bits - 9), - nr_blocks << (sb->s_blocksize_bits - 9), + return blkdev_issue_discard(sb->s_bdev, + block << (sb->s_blocksize_bits - + SECTOR_SHIFT), + nr_blocks << (sb->s_blocksize_bits - + SECTOR_SHIFT), gfp_mask, flags); } static inline int sb_issue_zeroout(struct super_block *sb, sector_t block, sector_t nr_blocks, gfp_t gfp_mask) { return blkdev_issue_zeroout(sb->s_bdev, - block << (sb->s_blocksize_bits - 9), - nr_blocks << (sb->s_blocksize_bits - 9), + block << (sb->s_blocksize_bits - + SECTOR_SHIFT), + nr_blocks << (sb->s_blocksize_bits - + SECTOR_SHIFT), gfp_mask, 0); } @@ -1511,7 +1530,8 @@ static inline int queue_alignment_offset(struct request_queue *q) static inline int queue_limit_alignment_offset(struct queue_limits *lim, sector_t sector) { unsigned int granularity = max(lim->physical_block_size, lim->io_min); - unsigned int alignment = sector_div(sector, granularity >> 9) << 9; + unsigned int alignment = sector_div(sector, granularity >> SECTOR_SHIFT) + << SECTOR_SHIFT; return (granularity + lim->alignment_offset - alignment) % granularity; } @@ -1545,8 +1565,8 @@ static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector return 0; /* Why are these in bytes, not sectors? */ - alignment = lim->discard_alignment >> 9; - granularity = lim->discard_granularity >> 9; + alignment = lim->discard_alignment >> SECTOR_SHIFT; + granularity = lim->discard_granularity >> SECTOR_SHIFT; if (!granularity) return 0; @@ -1557,7 +1577,7 @@ static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector offset = (granularity + alignment - offset) % granularity; /* Turn it back into bytes, gaah */ - return offset << 9; + return offset << SECTOR_SHIFT; } static inline int bdev_discard_alignment(struct block_device *bdev) diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index da83f64952e7..4384433b50e7 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -542,8 +542,6 @@ do { \ #define DMEMIT(x...) sz += ((sz >= maxlen) ? \ 0 : scnprintf(result + sz, maxlen - sz, x)) -#define SECTOR_SHIFT 9 - /* * Definitions of return values from target end_io function. */ diff --git a/include/linux/ide.h b/include/linux/ide.h index 771989d25ef8..0acfa62b1d44 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -165,7 +165,6 @@ struct ide_io_ports { */ #define PARTN_BITS 6 /* number of minor dev bits for partitions */ #define MAX_DRIVES 2 /* per interface; 2 assumed by lots of code */ -#define SECTOR_SIZE 512 /* * Timeouts for various operations: diff --git a/include/uapi/linux/msdos_fs.h b/include/uapi/linux/msdos_fs.h index a45d0754102e..fde753735aba 100644 --- a/include/uapi/linux/msdos_fs.h +++ b/include/uapi/linux/msdos_fs.h @@ -10,7 +10,9 @@ * The MS-DOS filesystem constants/structures */ +#ifndef SECTOR_SIZE #define SECTOR_SIZE 512 /* sector size (bytes) */ +#endif #define SECTOR_BITS 9 /* log2(SECTOR_SIZE) */ #define MSDOS_DPB (MSDOS_DPS) /* dir entries per block */ #define MSDOS_DPB_BITS 4 /* log2(MSDOS_DPB) */ -- cgit v1.2.3 From 0ce9144471de9ee09306ca0127e7cd27521ccc3f Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 18 Apr 2018 14:08:27 +1000 Subject: block: add blk_queue_fua() helper function So we can check FUA support status from the iomap direct IO code. Reviewed-by: Christoph Hellwig Signed-off-by: Dave Chinner Signed-off-by: Jens Axboe --- include/linux/blkdev.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/blkdev.h') diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 9af3e0f430bc..c362aadfe036 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -737,6 +737,7 @@ bool blk_queue_flag_test_and_clear(unsigned int flag, struct request_queue *q); #define blk_queue_quiesced(q) test_bit(QUEUE_FLAG_QUIESCED, &(q)->queue_flags) #define blk_queue_preempt_only(q) \ test_bit(QUEUE_FLAG_PREEMPT_ONLY, &(q)->queue_flags) +#define blk_queue_fua(q) test_bit(QUEUE_FLAG_FUA, &(q)->queue_flags) extern int blk_set_preempt_only(struct request_queue *q); extern void blk_clear_preempt_only(struct request_queue *q); -- cgit v1.2.3 From ccce20fc7968d546fb1e8e147bf5cdc8afc4278a Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Mon, 16 Apr 2018 18:04:41 -0700 Subject: scsi: sd_zbc: Avoid that resetting a zone fails sporadically Since SCSI scanning occurs asynchronously, since sd_revalidate_disk() is called from sd_probe_async() and since sd_revalidate_disk() calls sd_zbc_read_zones() it can happen that sd_zbc_read_zones() is called concurrently with blkdev_report_zones() and/or blkdev_reset_zones(). That can cause these functions to fail with -EIO because sd_zbc_read_zones() e.g. sets q->nr_zones to zero before restoring it to the actual value, even if no drive characteristics have changed. Avoid that this can happen by making the following changes: - Protect the code that updates zone information with blk_queue_enter() and blk_queue_exit(). - Modify sd_zbc_setup_seq_zones_bitmap() and sd_zbc_setup() such that these functions do not modify struct scsi_disk before all zone information has been obtained. Note: since commit 055f6e18e08f ("block: Make q_usage_counter also track legacy requests"; kernel v4.15) the request queue freezing mechanism also affects legacy request queues. Fixes: 89d947561077 ("sd: Implement support for ZBC devices") Signed-off-by: Bart Van Assche Cc: Jens Axboe Cc: Damien Le Moal Cc: Christoph Hellwig Cc: Hannes Reinecke Cc: stable@vger.kernel.org # v4.16 Reviewed-by: Damien Le Moal Signed-off-by: Martin K. Petersen --- drivers/scsi/sd_zbc.c | 140 +++++++++++++++++++++++++++++-------------------- include/linux/blkdev.h | 5 ++ 2 files changed, 87 insertions(+), 58 deletions(-) (limited to 'include/linux/blkdev.h') diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c index 41df75eea57b..210407cd2341 100644 --- a/drivers/scsi/sd_zbc.c +++ b/drivers/scsi/sd_zbc.c @@ -400,8 +400,10 @@ static int sd_zbc_check_capacity(struct scsi_disk *sdkp, unsigned char *buf) * * Check that all zones of the device are equal. The last zone can however * be smaller. The zone size must also be a power of two number of LBAs. + * + * Returns the zone size in bytes upon success or an error code upon failure. */ -static int sd_zbc_check_zone_size(struct scsi_disk *sdkp) +static s64 sd_zbc_check_zone_size(struct scsi_disk *sdkp) { u64 zone_blocks = 0; sector_t block = 0; @@ -412,8 +414,6 @@ static int sd_zbc_check_zone_size(struct scsi_disk *sdkp) int ret; u8 same; - sdkp->zone_blocks = 0; - /* Get a buffer */ buf = kmalloc(SD_ZBC_BUF_SIZE, GFP_KERNEL); if (!buf) @@ -445,16 +445,17 @@ static int sd_zbc_check_zone_size(struct scsi_disk *sdkp) /* Parse zone descriptors */ while (rec < buf + buf_len) { - zone_blocks = get_unaligned_be64(&rec[8]); - if (sdkp->zone_blocks == 0) { - sdkp->zone_blocks = zone_blocks; - } else if (zone_blocks != sdkp->zone_blocks && - (block + zone_blocks < sdkp->capacity - || zone_blocks > sdkp->zone_blocks)) { - zone_blocks = 0; + u64 this_zone_blocks = get_unaligned_be64(&rec[8]); + + if (zone_blocks == 0) { + zone_blocks = this_zone_blocks; + } else if (this_zone_blocks != zone_blocks && + (block + this_zone_blocks < sdkp->capacity + || this_zone_blocks > zone_blocks)) { + this_zone_blocks = 0; goto out; } - block += zone_blocks; + block += this_zone_blocks; rec += 64; } @@ -467,8 +468,6 @@ static int sd_zbc_check_zone_size(struct scsi_disk *sdkp) } while (block < sdkp->capacity); - zone_blocks = sdkp->zone_blocks; - out: if (!zone_blocks) { if (sdkp->first_scan) @@ -488,8 +487,7 @@ out: "Zone size too large\n"); ret = -ENODEV; } else { - sdkp->zone_blocks = zone_blocks; - sdkp->zone_shift = ilog2(zone_blocks); + ret = zone_blocks; } out_free: @@ -500,15 +498,14 @@ out_free: /** * sd_zbc_alloc_zone_bitmap - Allocate a zone bitmap (one bit per zone). - * @sdkp: The disk of the bitmap + * @nr_zones: Number of zones to allocate space for. + * @numa_node: NUMA node to allocate the memory from. */ -static inline unsigned long *sd_zbc_alloc_zone_bitmap(struct scsi_disk *sdkp) +static inline unsigned long * +sd_zbc_alloc_zone_bitmap(u32 nr_zones, int numa_node) { - struct request_queue *q = sdkp->disk->queue; - - return kzalloc_node(BITS_TO_LONGS(sdkp->nr_zones) - * sizeof(unsigned long), - GFP_KERNEL, q->node); + return kzalloc_node(BITS_TO_LONGS(nr_zones) * sizeof(unsigned long), + GFP_KERNEL, numa_node); } /** @@ -516,6 +513,7 @@ static inline unsigned long *sd_zbc_alloc_zone_bitmap(struct scsi_disk *sdkp) * @sdkp: disk used * @buf: report reply buffer * @buflen: length of @buf + * @zone_shift: logarithm base 2 of the number of blocks in a zone * @seq_zones_bitmap: bitmap of sequential zones to set * * Parse reported zone descriptors in @buf to identify sequential zones and @@ -525,7 +523,7 @@ static inline unsigned long *sd_zbc_alloc_zone_bitmap(struct scsi_disk *sdkp) * Return the LBA after the last zone reported. */ static sector_t sd_zbc_get_seq_zones(struct scsi_disk *sdkp, unsigned char *buf, - unsigned int buflen, + unsigned int buflen, u32 zone_shift, unsigned long *seq_zones_bitmap) { sector_t lba, next_lba = sdkp->capacity; @@ -544,7 +542,7 @@ static sector_t sd_zbc_get_seq_zones(struct scsi_disk *sdkp, unsigned char *buf, if (type != ZBC_ZONE_TYPE_CONV && cond != ZBC_ZONE_COND_READONLY && cond != ZBC_ZONE_COND_OFFLINE) - set_bit(lba >> sdkp->zone_shift, seq_zones_bitmap); + set_bit(lba >> zone_shift, seq_zones_bitmap); next_lba = lba + get_unaligned_be64(&rec[8]); rec += 64; } @@ -553,12 +551,16 @@ static sector_t sd_zbc_get_seq_zones(struct scsi_disk *sdkp, unsigned char *buf, } /** - * sd_zbc_setup_seq_zones_bitmap - Initialize the disk seq zone bitmap. + * sd_zbc_setup_seq_zones_bitmap - Initialize a seq zone bitmap. * @sdkp: target disk + * @zone_shift: logarithm base 2 of the number of blocks in a zone + * @nr_zones: number of zones to set up a seq zone bitmap for * * Allocate a zone bitmap and initialize it by identifying sequential zones. */ -static int sd_zbc_setup_seq_zones_bitmap(struct scsi_disk *sdkp) +static unsigned long * +sd_zbc_setup_seq_zones_bitmap(struct scsi_disk *sdkp, u32 zone_shift, + u32 nr_zones) { struct request_queue *q = sdkp->disk->queue; unsigned long *seq_zones_bitmap; @@ -566,9 +568,9 @@ static int sd_zbc_setup_seq_zones_bitmap(struct scsi_disk *sdkp) unsigned char *buf; int ret = -ENOMEM; - seq_zones_bitmap = sd_zbc_alloc_zone_bitmap(sdkp); + seq_zones_bitmap = sd_zbc_alloc_zone_bitmap(nr_zones, q->node); if (!seq_zones_bitmap) - return -ENOMEM; + return ERR_PTR(-ENOMEM); buf = kmalloc(SD_ZBC_BUF_SIZE, GFP_KERNEL); if (!buf) @@ -579,7 +581,7 @@ static int sd_zbc_setup_seq_zones_bitmap(struct scsi_disk *sdkp) if (ret) goto out; lba = sd_zbc_get_seq_zones(sdkp, buf, SD_ZBC_BUF_SIZE, - seq_zones_bitmap); + zone_shift, seq_zones_bitmap); } if (lba != sdkp->capacity) { @@ -591,12 +593,9 @@ out: kfree(buf); if (ret) { kfree(seq_zones_bitmap); - return ret; + return ERR_PTR(ret); } - - q->seq_zones_bitmap = seq_zones_bitmap; - - return 0; + return seq_zones_bitmap; } static void sd_zbc_cleanup(struct scsi_disk *sdkp) @@ -612,44 +611,64 @@ static void sd_zbc_cleanup(struct scsi_disk *sdkp) q->nr_zones = 0; } -static int sd_zbc_setup(struct scsi_disk *sdkp) +static int sd_zbc_setup(struct scsi_disk *sdkp, u32 zone_blocks) { struct request_queue *q = sdkp->disk->queue; + u32 zone_shift = ilog2(zone_blocks); + u32 nr_zones; int ret; - /* READ16/WRITE16 is mandatory for ZBC disks */ - sdkp->device->use_16_for_rw = 1; - sdkp->device->use_10_for_rw = 0; - /* chunk_sectors indicates the zone size */ - blk_queue_chunk_sectors(sdkp->disk->queue, - logical_to_sectors(sdkp->device, sdkp->zone_blocks)); - sdkp->nr_zones = - round_up(sdkp->capacity, sdkp->zone_blocks) >> sdkp->zone_shift; + blk_queue_chunk_sectors(q, + logical_to_sectors(sdkp->device, zone_blocks)); + nr_zones = round_up(sdkp->capacity, zone_blocks) >> zone_shift; /* * Initialize the device request queue information if the number * of zones changed. */ - if (sdkp->nr_zones != q->nr_zones) { - - sd_zbc_cleanup(sdkp); - - q->nr_zones = sdkp->nr_zones; - if (sdkp->nr_zones) { - q->seq_zones_wlock = sd_zbc_alloc_zone_bitmap(sdkp); - if (!q->seq_zones_wlock) { + if (nr_zones != sdkp->nr_zones || nr_zones != q->nr_zones) { + unsigned long *seq_zones_wlock = NULL, *seq_zones_bitmap = NULL; + size_t zone_bitmap_size; + + if (nr_zones) { + seq_zones_wlock = sd_zbc_alloc_zone_bitmap(nr_zones, + q->node); + if (!seq_zones_wlock) { ret = -ENOMEM; goto err; } - ret = sd_zbc_setup_seq_zones_bitmap(sdkp); - if (ret) { - sd_zbc_cleanup(sdkp); + seq_zones_bitmap = sd_zbc_setup_seq_zones_bitmap(sdkp, + zone_shift, nr_zones); + if (IS_ERR(seq_zones_bitmap)) { + ret = PTR_ERR(seq_zones_bitmap); + kfree(seq_zones_wlock); goto err; } } - + zone_bitmap_size = BITS_TO_LONGS(nr_zones) * + sizeof(unsigned long); + blk_mq_freeze_queue(q); + if (q->nr_zones != nr_zones) { + /* READ16/WRITE16 is mandatory for ZBC disks */ + sdkp->device->use_16_for_rw = 1; + sdkp->device->use_10_for_rw = 0; + + sdkp->zone_blocks = zone_blocks; + sdkp->zone_shift = zone_shift; + sdkp->nr_zones = nr_zones; + q->nr_zones = nr_zones; + swap(q->seq_zones_wlock, seq_zones_wlock); + swap(q->seq_zones_bitmap, seq_zones_bitmap); + } else if (memcmp(q->seq_zones_bitmap, seq_zones_bitmap, + zone_bitmap_size) != 0) { + memcpy(q->seq_zones_bitmap, seq_zones_bitmap, + zone_bitmap_size); + } + blk_mq_unfreeze_queue(q); + kfree(seq_zones_wlock); + kfree(seq_zones_bitmap); } return 0; @@ -661,6 +680,7 @@ err: int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf) { + int64_t zone_blocks; int ret; if (!sd_is_zoned(sdkp)) @@ -697,12 +717,16 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf) * Check zone size: only devices with a constant zone size (except * an eventual last runt zone) that is a power of 2 are supported. */ - ret = sd_zbc_check_zone_size(sdkp); - if (ret) + zone_blocks = sd_zbc_check_zone_size(sdkp); + ret = -EFBIG; + if (zone_blocks != (u32)zone_blocks) + goto err; + ret = zone_blocks; + if (ret < 0) goto err; /* The drive satisfies the kernel restrictions: set it up */ - ret = sd_zbc_setup(sdkp); + ret = sd_zbc_setup(sdkp, zone_blocks); if (ret) goto err; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 9af3e0f430bc..21e21f273a21 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -605,6 +605,11 @@ struct request_queue { * initialized by the low level device driver (e.g. scsi/sd.c). * Stacking drivers (device mappers) may or may not initialize * these fields. + * + * Reads of this information must be protected with blk_queue_enter() / + * blk_queue_exit(). Modifying this information is only allowed while + * no requests are being processed. See also blk_mq_freeze_queue() and + * blk_mq_unfreeze_queue(). */ unsigned int nr_zones; unsigned long *seq_zones_bitmap; -- cgit v1.2.3