linux-dev/drivers/block/null_blk_zoned.c, branch linus/master

null_blk: Move driver into its own directory

2020-12-08T00:36:04Z

Move null_blk driver code into the new sub-directory drivers/block/null_blk. Suggested-by: Bart Van Assche Signed-off-by: Damien Le Moal Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: discard zones on reset

2020-12-08T00:36:03Z

When memory backing is enabled, use null_handle_discard() to free the backing memory used by a zone when the zone is being reset. Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: Improve implicit zone close

2020-12-08T00:36:03Z

When open zone resource management is enabled, that is, when a null_blk zoned device is created with zone_max_open different than 0, implicitly or explicitly opening a zone may require implicitly closing a zone that is already implicitly open. This operation is done using the function null_close_first_imp_zone(), which search for an implicitly open zone to close starting from the first sequential zone. This implementation is simple but may result in the same being constantly implicitly closed and then implicitly reopened on write, namely, the lowest numbered zone that is being written. Avoid this by starting the search for an implicitly open zone starting from the zone following the last zone that was implicitly closed. The function null_close_first_imp_zone() is renamed null_close_imp_open_zone(). Signed-off-by: Damien Le Moal Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: improve zone locking

2020-12-08T00:36:03Z

With memory backing disabled, using a single spinlock for protecting zone information and zone resource management prevents the parallel execution on multiple queue of IO requests to different zones. Furthermore, regardless of the use of memory backing, if a null_blk device is created without limits on the number of open and active zones, accounting for zone resource management is not necessary. >From these observations, zone locking is changed as follows to improve performance: 1) the zone_lock spinlock is renamed zone_res_lock and used only if zone resource management is necessary, that is, if either zone_max_open or zone_max_active are not 0. This is indicated using the new boolean need_zone_res_mgmt in the nullb_device structure. null_zone_write() is modified to reduce the amount of code executed with the zone_res_lock spinlock held. 2) With memory backing disabled, per zone locking is changed to a spinlock per zone. 3) Introduce the structure nullb_zone to replace the use of struct blk_zone for zone information. This new structure includes a union of a spinlock and a mutex for zone locking. The spinlock is used when memory backing is disabled and the mutex is used with memory backing. With these changes, fio performance with zonemode=zbd for 4K random read and random write on a dual socket (24 cores per socket) machine using the none schedulder is as follows: before patch: write (psync x 96 jobs) = 465 KIOPS read (libaio@qd=8 x 96 jobs) = 1361 KIOPS after patch: write (psync x 96 jobs) = 456 KIOPS read (libaio@qd=8 x 96 jobs) = 4096 KIOPS Write performance remains mostly unchanged but read performance is three times higher. Performance when using the mq-deadline scheduler is not changed by this patch as mq-deadline becomes the bottleneck for a multi-queue device. Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: Fail zone append to conventional zones

2020-12-08T00:36:03Z

Conventional zones do not have a write pointer and so cannot accept zone append writes. Make sure to fail any zone append write command issued to a conventional zone. Reported-by: Naohiro Aota Fixes: e0489ed5daeb ("null_blk: Support REQ_OP_ZONE_APPEND") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: Fix zone size initialization

2020-12-08T00:36:03Z

For a null_blk device with zoned mode enabled is currently initialized with a number of zones equal to the device capacity divided by the zone size, without considering if the device capacity is a multiple of the zone size. If the zone size is not a divisor of the capacity, the zones end up not covering the entire capacity, potentially resulting is out of bounds accesses to the zone array. Fix this by adding one last smaller zone with a size equal to the remainder of the disk capacity divided by the zone size if the capacity is not a multiple of the zone size. For such smaller last zone, the zone capacity is also checked so that it does not exceed the smaller zone size. Reported-by: Naohiro Aota Fixes: ca4b2a011948 ("null_blk: add zone support") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Reviewed-by: Johannes Thumshirn Signed-off-by: Jens Axboe

null_blk: Fix scheduling in atomic with zoned mode

2020-11-06T16:36:42Z

Commit aa1c09cb65e2 ("null_blk: Fix locking in zoned mode") changed zone locking to using the potentially sleeping wait_on_bit_io() function. This is acceptable when memory backing is enabled as the device queue is in that case marked as blocking, but this triggers a scheduling while in atomic context with memory backing disabled. Fix this by relying solely on the device zone spinlock for zone information protection without temporarily releasing this lock around null_process_cmd() execution in null_zone_write(). This is OK to do since when memory backing is disabled, command processing does not block and the memory backing lock nullb->lock is unused. This solution avoids the overhead of having to mark a zoned null_blk device queue as blocking when memory backing is unused. This patch also adds comments to the zone locking code to explain the unusual locking scheme. Fixes: aa1c09cb65e2 ("null_blk: Fix locking in zoned mode") Reported-by: kernel test robot Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Cc: stable@vger.kernel.org Signed-off-by: Jens Axboe

null_blk: Fix locking in zoned mode

2020-10-29T13:27:30Z

When the zoned mode is enabled in null_blk, Serializing read, write and zone management operations for each zone is necessary to protect device level information for managing zone resources (zone open and closed counters) as well as each zone condition and write pointer position. Commit 35bc10b2eafb ("null_blk: synchronization fix for zoned device") introduced a spinlock to implement this serialization. However, when memory backing is also enabled, GFP_NOIO memory allocations are executed under the spinlock, resulting in might_sleep() warnings. Furthermore, the zone_lock spinlock is locked/unlocked using spin_lock_irq/spin_unlock_irq, similarly to the memory backing code with the nullb->lock spinlock. This nested use of irq locks wrecks the irq enabled/disabled state. Fix all this by introducing a bitmap for per-zone lock, with locking implemented using wait_on_bit_lock_io() and clear_and_wake_up_bit(). This locking mechanism allows keeping a zone locked while executing null_process_cmd(), serializing all operations to the zone while allowing to sleep during memory backing allocation with GFP_NOIO. Device level zone resource management information is protected using a spinlock which is not held while executing null_process_cmd(); Fixes: 35bc10b2eafb ("null_blk: synchronization fix for zoned device") Signed-off-by: Damien Le Moal Signed-off-by: Jens Axboe

null_blk: Fix zone reset all tracing

2020-10-29T13:27:30Z

In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect. Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty. Fixes: 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c") Signed-off-by: Damien Le Moal Cc: stable@vger.kernel.org Signed-off-by: Jens Axboe

null_blk: synchronization fix for zoned device

2020-10-28T02:40:42Z

Parallel write,read,zone-mgmt operations accessing/altering zone state and write-pointer may get into race. Avoid the situation by using a new spinlock for zoned device. Concurrent zone-appends (on a zone) returning same write-pointer issue is also avoided using this lock. Cc: stable@vger.kernel.org Fixes: e0489ed5daeb ("null_blk: Support REQ_OP_ZONE_APPEND") Signed-off-by: Kanchan Joshi Reviewed-by: Damien Le Moal Signed-off-by: Jens Axboe