user/sven/linux.git/include/linux/genhd.h, branch v5.4.90

block/diskstats: more accurate approximation of io_ticks for slow disks

2020-10-07T06:01:29Z

commit 2b8bd423614c595540eaadcfbc702afe8e155e50 upstream. Currently io_ticks is approximated by adding one at each start and end of requests if jiffies counter has changed. This works perfectly for requests shorter than a jiffy or if one of requests starts/ends at each jiffy. If disk executes just one request at a time and they are longer than two jiffies then only first and last jiffies will be accounted. Fix is simple: at the end of request add up into io_ticks jiffies passed since last update rather than just one jiffy. Example: common HDD executes random read 4k requests around 12ms. fio --name=test --filename=/dev/sdb --rw=randread --direct=1 --runtime=30 & iostat -x 10 sdb Note changes of iostat's "%util" 8,43% -> 99,99% before/after patch: Before: Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sdb 0,00 0,00 82,60 0,00 330,40 0,00 8,00 0,96 12,09 12,09 0,00 1,02 8,43 After: Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util sdb 0,00 0,00 82,50 0,00 330,00 0,00 8,00 1,00 12,10 12,10 0,00 12,12 99,99 Now io_ticks does not loose time between start and end of requests, but for queue-depth > 1 some I/O time between adjacent starts might be lost. For load estimation "%util" is not as useful as average queue length, but it clearly shows how often disk queue is completely empty. Fixes: 5b18b5a73760 ("block: delete part_round_stats and switch to less precise counting") Signed-off-by: Konstantin Khlebnikov Reviewed-by: Ming Lei Signed-off-by: Jens Axboe From: "Banerjee, Debabrata" Signed-off-by: Greg Kroah-Hartman

block: nr_sects_write(): Disable preemption on seqcount write

2020-06-24T15:50:49Z

[ Upstream commit 15b81ce5abdc4b502aa31dff2d415b79d2349d2f ] For optimized block readers not holding a mutex, the "number of sectors" 64-bit value is protected from tearing on 32-bit architectures by a sequence counter. Disable preemption before entering that sequence counter's write side critical section. Otherwise, the read side can preempt the write side section and spin for the entire scheduler tick. If the reader belongs to a real-time scheduling class, it can spin forever and the kernel will livelock. Fixes: c83f6bf98dc1 ("block: add partition resize function to blkpg ioctl") Cc: Signed-off-by: Ahmed S. Darwish Reviewed-by: Sebastian Andrzej Siewior Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: fix use-after-free on gendisk

2019-04-22T15:48:12Z

commit 2da78092dda "block: Fix dev_t minor allocation lifetime" specifically moved blk_free_devt(dev->devt) call to part_release() to avoid reallocating device number before the device is fully shutdown. However, it can cause use-after-free on gendisk in get_gendisk(). We use md device as example to show the race scenes: Process1 Worker Process2 md_free blkdev_open del_gendisk add delete_partition_work_fn() to wq __blkdev_get get_gendisk put_disk disk_release kfree(disk) find part from ext_devt_idr get_disk_and_module(disk) cause use after free delete_partition_work_fn put_device(part) part_release remove part from ext_devt_idr Before is removed from ext_devt_idr by delete_partition_work_fn(), we can find the devt and then access gendisk by hd_struct pointer. But, if we access the gendisk after it have been freed, it can cause in use-after-freeon gendisk in get_gendisk(). We fix this by adding a new helper blk_invalidate_devt() in delete_partition() and del_gendisk(). It replaces hd_struct pointer in idr with value 'NULL', and deletes the entry from idr in part_release() as we do now. Thanks to Jan Kara for providing the solution and more clear comments for the code. Fixes: 2da78092dda1 ("block: Fix dev_t minor allocation lifetime") Cc: Al Viro Reviewed-by: Bart Van Assche Reviewed-by: Keith Busch Reviewed-by: Jan Kara Suggested-by: Jan Kara Signed-off-by: Yufen Yu Signed-off-by: Jens Axboe

block: disk_events: introduce event flags

2019-04-12T19:35:24Z

Currently, an empty disk->events field tells the block layer not to forward media change events to user space. This was done in commit 7c88a168da80 ("block: don't propagate unlisted DISK_EVENTs to userland") in order to avoid events from "fringe" drivers to be forwarded to user space. By doing so, the block layer lost the information which events were supported by a particular block device, and most importantly, whether or not a given device supports media change events at all. Prepare for not interpreting the "events" field this way in the future any more. This is done by adding an additional field "event_flags" to struct gendisk, and two flag bits that can be set to have the device treated like one that had the "events" field set to a non-zero value before. This applies only to the sd and sr drivers, which are changed to set the new flags. The new flags are DISK_EVENT_FLAG_POLL to enforce polling of the device for synchronous events, and DISK_EVENT_FLAG_UEVENT to tell the blocklayer to generate udev events from kernel events. In order to add the event_flags field to struct gendisk, the events field is converted to an "unsigned short"; it doesn't need to hold values bigger than 2 anyway. This patch doesn't change behavior. Reviewed-by: Christoph Hellwig Signed-off-by: Martin Wilck Signed-off-by: Jens Axboe

block: genhd: remove async_events field

2019-04-12T19:35:22Z

The async_events field, intended to be used for drivers that support asynchronous notifications about disk events (aka media change events), isn't currently used by any driver, and apparently that has been that way for a long time (if not forever). Remove it. Reviewed-by: Hannes Reinecke Reviewed-by: Christoph Hellwig Signed-off-by: Martin Wilck Signed-off-by: Jens Axboe

block: remove CONFIG_LBDAF

2019-04-06T16:48:35Z

Currently support for 64-bit sector_t and blkcnt_t is optional on 32-bit architectures. These types are required to support block device and/or file sizes larger than 2 TiB, and have generally defaulted to on for a long time. Enabling the option only increases the i386 tinyconfig size by 145 bytes, and many data structures already always use 64-bit values for their in-core and on-disk data structures anyway, so there should not be a large change in dynamic memory usage either. Dropping this option removes a somewhat weird non-default config that has cause various bugs or compiler warnings when actually used. Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe

block: return just one value from part_in_flight

2018-12-10T15:30:38Z

The previous patches deleted all the code that needed the second value returned from part_in_flight - now the kernel only uses the first value. Consequently, part_in_flight (and blk_mq_in_flight) may be changed so that it only returns one value. This patch just refactors the code, there's no functional change. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Signed-off-by: Jens Axboe

block: switch to per-cpu in-flight counters

2018-12-10T15:30:37Z

Now when part_round_stats is gone, we can switch to per-cpu in-flight counters. We use the local-atomic type local_t, so that if part_inc_in_flight or part_dec_in_flight is reentrantly called from an interrupt, the value will be correct. The other counters could be corrupted due to reentrant interrupt, but the corruption only results in slight counter skew - the in_flight counter must be exact, so it needs local_t. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Signed-off-by: Jens Axboe

block: delete part_round_stats and switch to less precise counting

2018-12-10T15:30:37Z

We want to convert to per-cpu in_flight counters. The function part_round_stats needs the in_flight counter every jiffy, it would be too costly to sum all the percpu variables every jiffy, so it must be deleted. part_round_stats is used to calculate two counters - time_in_queue and io_ticks. time_in_queue can be calculated without part_round_stats, by adding the duration of the I/O when the I/O ends (the value is almost as exact as the previously calculated value, except that time for in-progress I/Os is not counted). io_ticks can be approximated by increasing the value when I/O is started or ended and the jiffies value has changed. If the I/Os take less than a jiffy, the value is as exact as the previously calculated value. If the I/Os take more than a jiffy, io_ticks can drift behind the previously calculated value. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Signed-off-by: Jens Axboe

block: stop passing 'cpu' to all percpu stats methods

2018-12-10T15:30:37Z

All of part_stat_* and related methods are used with preempt disabled, so there is no need to pass cpu around to allow of them. Just call smp_processor_id() as needed. Suggested-by: Jens Axboe Signed-off-by: Mike Snitzer Signed-off-by: Jens Axboe