Merge tag 'for-5.12/block-2021-02-17' of git://git.kernel.dk/linux-block

582cd91f · Linus Torvalds <[email protected]> · 2021-02-21 11:02:48 -08:00

Pull core block updates from Jens Axboe:
 "Another nice round of removing more code than what is added, mostly
  due to Christoph's relentless pursuit of tech debt removal/cleanups.
  This pull request contains:

   - Two series of BFQ improvements (Paolo, Jan, Jia)

   - Block iov_iter improvements (Pavel)

   - bsg error path fix (Pan)

   - blk-mq scheduler improvements (Jan)

   - -EBUSY discard fix (Jan)

   - bvec allocation improvements (Ming, Christoph)

   - bio allocation and init improvements (Christoph)

   - Store bdev pointer in bio instead of gendisk + partno (Christoph)

   - Block trace point cleanups (Christoph)

   - hard read-only vs read-only split (Christoph)

   - Block based swap cleanups (Christoph)

   - Zoned write granularity support (Damien)

   - Various fixes/tweaks (Chunguang, Guoqing, Lei, Lukas, Huhai)"

* tag 'for-5.12/block-2021-02-17' of git://git.kernel.dk/linux-block: (104 commits)
  mm: simplify swapdev_block
  sd_zbc: clear zone resources for non-zoned case
  block: introduce blk_queue_clear_zone_settings()
  zonefs: use zone write granularity as block size
  block: introduce zone_write_granularity limit
  block: use blk_queue_set_zoned in add_partition()
  nullb: use blk_queue_set_zoned() to setup zoned devices
  nvme: cleanup zone information initialization
  block: document zone_append_max_bytes attribute
  block: use bi_max_vecs to find the bvec pool
  md/raid10: remove dead code in reshape_request
  block: mark the bio as cloned in bio_iov_bvec_set
  block: set BIO_NO_PAGE_REF in bio_iov_bvec_set
  block: remove a layer of indentation in bio_iov_iter_get_pages
  block: turn the nr_iovecs argument to bio_alloc* into an unsigned short
  block: remove the 1 and 4 vec bvec_slabs entries
  block: streamline bvec_alloc
  block: factor out a bvec_alloc_gfp helper
  block: move struct biovec_slab to bio.c
  block: reuse BIO_INLINE_VECS for integrity bvecs
  ...

Commit: 582cd91f69de8e44857cb610ebca661dac8656b7
Parents: bd018bba · f885056a

Merge commit. Changes below are compared with the first parent: bd018bba.

Changes

148 files changed · +1343 −1608

File	Status	Add	Del	Content
Documentation/block/biovecs.rst	modified	+2	−0	View file
Documentation/block/queue-sysfs.rst	modified	+13	−0	View file
Documentation/filesystems/f2fs.rst	modified	+0	−1	View file
Documentation/filesystems/porting.rst	modified	+16	−0	View file
arch/m68k/emu/nfblock.c	modified	+1	−1	View file
arch/xtensa/platforms/iss/simdisk.c	modified	+1	−1	View file
block/bfq-iosched.c	modified	+276	−169	View file
block/bfq-iosched.h	modified	+19	−10	View file
block/bfq-wf2q.c	modified	+0	−3	View file
block/bio-integrity.c	modified	+15	−20	View file
block/bio.c	modified	+253	−318	View file
block/blk-cgroup.c	modified	+12	−10	View file
block/blk-core.c	modified	+46	−53	View file
block/blk-crypto-fallback.c	modified	+4	−2	View file
block/blk-crypto.c	modified	+1	−1	View file
block/blk-exec.c	modified	+5	−9	View file
block/blk-flush.c	modified	+6	−11	View file
block/blk-merge.c	modified	+8	−9	View file
block/blk-mq.c	modified	+61	−8	View file
block/blk-settings.c	modified	+39	−2	View file
block/blk-sysfs.c	modified	+8	−0	View file
block/blk-throttle.c	modified	+1	−1	View file
block/blk-wbt.c	modified	+2	−2	View file
block/blk-zoned.c	modified	+17	−0	View file
block/blk.h	modified	+7	−5	View file
block/bounce.c	modified	+3	−1	View file
block/bsg.c	modified	+4	−2	View file
block/genhd.c	modified	+49	−257	View file
block/kyber-iosched.c	modified	+1	−0	View file
block/mq-deadline.c	modified	+0	−6	View file
block/partitions/core.c	modified	+8	−28	View file
block/scsi_ioctl.c	modified	+3	−3	View file
drivers/block/brd.c	modified	+2	−6	View file
drivers/block/drbd/drbd_actlog.c	modified	+1	−1	View file
drivers/block/drbd/drbd_bitmap.c	modified	+1	−1	View file
drivers/block/drbd/drbd_int.h	modified	+2	−4	View file
drivers/block/drbd/drbd_main.c	modified	+0	−13	View file
drivers/block/drbd/drbd_req.c	modified	+5	−2	View file
drivers/block/drbd/drbd_req.h	modified	+0	−12	View file
drivers/block/drbd/drbd_worker.c	modified	+4	−1	View file
drivers/block/mtip32xx/mtip32xx.c	modified	+1	−1	View file
drivers/block/null_blk/main.c	modified	+1	−1	View file
drivers/block/null_blk/zoned.c	modified	+4	−4	View file
drivers/block/paride/pd.c	modified	+1	−1	View file
drivers/block/pktcdvd.c	modified	+3	−3	View file
drivers/block/ps3vram.c	modified	+1	−1	View file
drivers/block/rbd.c	modified	+0	−19	View file
drivers/block/rsxx/dev.c	modified	+1	−1	View file
drivers/block/sx8.c	modified	+2	−2	View file
drivers/block/umem.c	modified	+1	−1	View file
drivers/block/virtio_blk.c	modified	+1	−1	View file
drivers/block/zram/zram_drv.c	modified	+1	−1	View file
drivers/cdrom/cdrom.c	modified	+1	−1	View file
drivers/ide/ide-atapi.c	modified	+1	−1	View file
drivers/ide/ide-cd.c	modified	+1	−1	View file
drivers/ide/ide-cd_ioctl.c	modified	+1	−1	View file
drivers/ide/ide-devsets.c	modified	+1	−1	View file
drivers/ide/ide-disk.c	modified	+1	−1	View file
drivers/ide/ide-ioctls.c	modified	+2	−2	View file
drivers/ide/ide-park.c	modified	+1	−1	View file
drivers/ide/ide-pm.c	modified	+2	−2	View file
drivers/ide/ide-tape.c	modified	+1	−1	View file
drivers/ide/ide-taskfile.c	modified	+1	−1	View file
drivers/lightnvm/pblk-init.c	modified	+1	−1	View file
drivers/md/bcache/debug.c	modified	+1	−1	View file
drivers/md/bcache/request.c	modified	+24	−15	View file
drivers/md/bcache/super.c	modified	+1	−1	View file
drivers/md/dm-bio-record.h	modified	+3	−6	View file
drivers/md/dm-cache-metadata.c	modified	+1	−1	View file
drivers/md/dm-clone-target.c	modified	+1	−13	View file
drivers/md/dm-raid1.c	modified	+5	−5	View file
drivers/md/dm-thin-metadata.c	modified	+1	−1	View file
drivers/md/dm-zoned-metadata.c	modified	+3	−3	View file
drivers/md/dm.c	modified	+7	−7	View file
drivers/md/md-linear.c	modified	+1	−1	View file
drivers/md/md.c	modified	+29	−44	View file
drivers/md/md.h	modified	+3	−5	View file
drivers/md/raid1.c	modified	+4	−4	View file
drivers/md/raid10.c	modified	+7	−11	View file
drivers/md/raid5-ppl.c	modified	+1	−1	View file
drivers/md/raid5.c	modified	+46	−64	View file
drivers/mmc/core/block.c	modified	+5	−5	View file
drivers/nvdimm/blk.c	modified	+2	−2	View file
drivers/nvdimm/btt.c	modified	+2	−2	View file
drivers/nvdimm/pmem.c	modified	+2	−2	View file
drivers/nvme/host/core.c	modified	+16	−15	View file
drivers/nvme/host/lightnvm.c	modified	+3	−4	View file
drivers/nvme/host/multipath.c	modified	+3	−3	View file
drivers/nvme/host/pci.c	modified	+2	−2	View file
drivers/nvme/host/rdma.c	modified	+1	−1	View file
drivers/nvme/host/zns.c	modified	+3	−8	View file
drivers/nvme/target/io-cmd-bdev.c	modified	+1	−1	View file
drivers/nvme/target/passthru.c	modified	+1	−1	View file
drivers/s390/block/dasd.c	modified	+5	−21	View file
drivers/s390/block/dcssblk.c	modified	+1	−5	View file
drivers/s390/block/xpram.c	modified	+1	−1	View file
drivers/scsi/scsi_error.c	modified	+1	−1	View file
drivers/scsi/scsi_lib.c	modified	+1	−1	View file
drivers/scsi/sd_zbc.c	modified	+39	−4	View file
drivers/scsi/sg.c	modified	+1	−2	View file
drivers/scsi/st.c	modified	+1	−1	View file
drivers/target/target_core_file.c	modified	+6	−14	View file
drivers/target/target_core_pscsi.c	modified	+1	−2	View file
fs/block_dev.c	modified	+9	−11	View file
fs/btrfs/check-integrity.c	modified	+5	−5	View file
fs/btrfs/raid56.c	modified	+2	−5	View file
fs/btrfs/scrub.c	modified	+1	−1	View file
fs/btrfs/volumes.c	modified	+1	−1	View file
fs/btrfs/zoned.c	modified	+2	−2	View file
fs/direct-io.c	modified	+3	−1	View file
fs/exfat/file.c	modified	+1	−1	View file
fs/ext4/fast_commit.c	modified	+2	−2	View file
fs/ext4/fsync.c	modified	+1	−1	View file
fs/ext4/ialloc.c	modified	+1	−1	View file
fs/ext4/super.c	modified	+1	−1	View file
fs/f2fs/data.c	modified	+5	−35	View file
fs/f2fs/f2fs.h	modified	+0	−2	View file
fs/f2fs/segment.c	modified	+1	−11	View file
fs/f2fs/super.c	modified	+0	−1	View file
fs/fat/file.c	modified	+1	−1	View file
fs/hfsplus/inode.c	modified	+1	−1	View file
fs/hfsplus/super.c	modified	+1	−1	View file
fs/iomap/direct-io.c	modified	+4	−5	View file
fs/jbd2/checkpoint.c	modified	+1	−1	View file
fs/jbd2/commit.c	modified	+2	−2	View file
fs/jbd2/recovery.c	modified	+1	−1	View file
fs/libfs.c	modified	+1	−1	View file
fs/nfs/blocklayout/blocklayout.c	modified	+0	−5	View file
fs/nfsd/blocklayout.c	modified	+1	−1	View file
fs/nilfs2/segbuf.c	modified	+0	−4	View file
fs/nilfs2/the_nilfs.h	modified	+1	−1	View file
fs/ocfs2/file.c	modified	+1	−1	View file
fs/reiserfs/file.c	modified	+1	−1	View file
fs/splice.c	modified	+6	−3	View file
fs/super.c	modified	+2	−1	View file
fs/xfs/xfs_super.c	modified	+1	−1	View file
fs/zonefs/super.c	modified	+6	−7	View file
include/linux/bio.h	modified	+28	−27	View file
include/linux/blk-mq.h	modified	+2	−6	View file
include/linux/blk_types.h	modified	+3	−30	View file
include/linux/blkdev.h	modified	+25	−28	View file
include/linux/elevator.h	modified	+2	−0	View file
include/linux/genhd.h	modified	+9	−18	View file
include/linux/swap.h	modified	+0	−1	View file
kernel/trace/blktrace.c	modified	+9	−7	View file
lib/iov_iter.c	modified	+19	−2	View file
mm/page_io.c	modified	+14	−33	View file
mm/swapfile.c	modified	+3	−33	View file

Only the first 600 diff lines are shown. Download the complete patch.

Documentation/block/biovecs.rst

View file

diff --git a/Documentation/block/biovecs.rst b/Documentation/block/biovecs.rst
index 36771a1..ddb867e 100644
--- a/Documentation/block/biovecs.rst
+++ b/Documentation/block/biovecs.rst
@@ -40,6 +40,8 @@ normal code doesn't have to deal with bi_bvec_done.
    There is a lower level advance function - bvec_iter_advance() - which takes
    a pointer to a biovec, not a bio; this is used by the bio integrity code.
 
+As of 5.12 bvec segments with zero bv_len are not supported.
+
 What's all this get us?
 =======================
 

Documentation/block/queue-sysfs.rst

View file

diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index 2638d34..4dc7f0d 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -261,6 +261,12 @@ For block drivers that support REQ_OP_WRITE_ZEROES, the maximum number of
 bytes that can be zeroed at once. The value 0 means that REQ_OP_WRITE_ZEROES
 is not supported.
 
+zone_append_max_bytes (RO)
+--------------------------
+This is the maximum number of bytes that can be written to a sequential
+zone of a zoned block device using a zone append write operation
+(REQ_OP_ZONE_APPEND). This value is always 0 for regular block devices.
+
 zoned (RO)
 ----------
 This indicates if the device is a zoned block device and the zone model of the
@@ -273,4 +279,11 @@ devices are described in the ZBC (Zoned Block Commands) and ZAC
 do not support zone commands, they will be treated as regular block devices
 and zoned will report "none".
 
+zone_write_granularity (RO)
+---------------------------
+This indicates the alignment constraint, in bytes, for write operations in
+sequential zones of zoned block devices (devices with a zoned attributed
+that reports "host-managed" or "host-aware"). This value is always 0 for
+regular block devices.
+
 Jens Axboe <[email protected]>, February 2009

Documentation/filesystems/f2fs.rst

View file

diff --git a/Documentation/filesystems/f2fs.rst b/Documentation/filesystems/f2fs.rst
index 81c05ba..35ed01a 100644
--- a/Documentation/filesystems/f2fs.rst
+++ b/Documentation/filesystems/f2fs.rst
@@ -179,7 +179,6 @@ fault_type=%d		 Support configuring fault injection type, should be
 			 FAULT_KVMALLOC		  0x000000002
 			 FAULT_PAGE_ALLOC	  0x000000004
 			 FAULT_PAGE_GET		  0x000000008
-			 FAULT_ALLOC_BIO	  0x000000010
 			 FAULT_ALLOC_NID	  0x000000020
 			 FAULT_ORPHAN		  0x000000040
 			 FAULT_BLOCK		  0x000000080

Documentation/filesystems/porting.rst

View file

diff --git a/Documentation/filesystems/porting.rst b/Documentation/filesystems/porting.rst
index 867036a..1f8cf8e 100644
--- a/Documentation/filesystems/porting.rst
+++ b/Documentation/filesystems/porting.rst
@@ -865,3 +865,19 @@ no matter what.  Everything is handled by the caller.
 
 clone_private_mount() returns a longterm mount now, so the proper destructor of
 its result is kern_unmount() or kern_unmount_array().
+
+---
+
+**mandatory**
+
+zero-length bvec segments are disallowed, they must be filtered out before
+passed on to an iterator.
+
+---
+
+**mandatory**
+
+For bvec based itererators bio_iov_iter_get_pages() now doesn't copy bvecs but
+uses the one provided. Anyone issuing kiocb-I/O should ensure that the bvec and
+page references stay until I/O has completed, i.e. until ->ki_complete() has
+been called or returned with non -EIOCBQUEUED code.

arch/m68k/emu/nfblock.c

View file

diff --git a/arch/m68k/emu/nfblock.c b/arch/m68k/emu/nfblock.c
index 92d26c8..ba80854 100644
--- a/arch/m68k/emu/nfblock.c
+++ b/arch/m68k/emu/nfblock.c
@@ -61,7 +61,7 @@ struct nfhd_device {
 
 static blk_qc_t nfhd_submit_bio(struct bio *bio)
 {
-	struct nfhd_device *dev = bio->bi_disk->private_data;
+	struct nfhd_device *dev = bio->bi_bdev->bd_disk->private_data;
 	struct bio_vec bvec;
 	struct bvec_iter iter;
 	int dir, len, shift;

arch/xtensa/platforms/iss/simdisk.c

View file

diff --git a/arch/xtensa/platforms/iss/simdisk.c b/arch/xtensa/platforms/iss/simdisk.c
index 3447556..fc09be7 100644
--- a/arch/xtensa/platforms/iss/simdisk.c
+++ b/arch/xtensa/platforms/iss/simdisk.c
@@ -103,7 +103,7 @@ static void simdisk_transfer(struct simdisk *dev, unsigned long sector,
 
 static blk_qc_t simdisk_submit_bio(struct bio *bio)
 {
-	struct simdisk *dev = bio->bi_disk->private_data;
+	struct simdisk *dev = bio->bi_bdev->bd_disk->private_data;
 	struct bio_vec bvec;
 	struct bvec_iter iter;
 	sector_t sector = bio->bi_iter.bi_sector;

block/bfq-iosched.c

View file

diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index 9e81d10..b398dde 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -158,7 +158,6 @@ BFQ_BFQQ_FNS(in_large_burst);
 BFQ_BFQQ_FNS(coop);
 BFQ_BFQQ_FNS(split_coop);
 BFQ_BFQQ_FNS(softrt_update);
-BFQ_BFQQ_FNS(has_waker);
 #undef BFQ_BFQQ_FNS						\
 
 /* Expiration time of sync (0) and async (1) requests, in ns. */
@@ -1024,9 +1023,16 @@ bfq_bfqq_resume_state(struct bfq_queue *bfqq, struct bfq_data *bfqd,
 	else
 		bfq_clear_bfqq_IO_bound(bfqq);
 
+	bfqq->last_serv_time_ns = bic->saved_last_serv_time_ns;
+	bfqq->inject_limit = bic->saved_inject_limit;
+	bfqq->decrease_time_jif = bic->saved_decrease_time_jif;
+
 	bfqq->entity.new_weight = bic->saved_weight;
 	bfqq->ttime = bic->saved_ttime;
+	bfqq->io_start_time = bic->saved_io_start_time;
+	bfqq->tot_idle_time = bic->saved_tot_idle_time;
 	bfqq->wr_coeff = bic->saved_wr_coeff;
+	bfqq->service_from_wr = bic->saved_service_from_wr;
 	bfqq->wr_start_at_switch_to_srt = bic->saved_wr_start_at_switch_to_srt;
 	bfqq->last_wr_start_finish = bic->saved_last_wr_start_finish;
 	bfqq->wr_cur_max_time = bic->saved_wr_cur_max_time;
@@ -1647,6 +1653,8 @@ static bool bfq_bfqq_higher_class_or_weight(struct bfq_queue *bfqq,
 	return bfqq_weight > in_serv_weight;
 }
 
+static bool bfq_better_to_idle(struct bfq_queue *bfqq);
+
 static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 					     struct bfq_queue *bfqq,
 					     int old_wr_coeff,
@@ -1671,15 +1679,19 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 	 * - it is sync,
 	 * - it does not belong to a large burst,
 	 * - it has been idle for enough time or is soft real-time,
-	 * - is linked to a bfq_io_cq (it is not shared in any sense).
+	 * - is linked to a bfq_io_cq (it is not shared in any sense),
+	 * - has a default weight (otherwise we assume the user wanted
+	 *   to control its weight explicitly)
 	 */
 	in_burst = bfq_bfqq_in_large_burst(bfqq);
 	soft_rt = bfqd->bfq_wr_max_softrt_rate > 0 &&
 		!BFQQ_TOTALLY_SEEKY(bfqq) &&
 		!in_burst &&
 		time_is_before_jiffies(bfqq->soft_rt_next_start) &&
-		bfqq->dispatched == 0;
-	*interactive = !in_burst && idle_for_long_time;
+		bfqq->dispatched == 0 &&
+		bfqq->entity.new_weight == 40;
+	*interactive = !in_burst && idle_for_long_time &&
+		bfqq->entity.new_weight == 40;
 	wr_or_deserves_wr = bfqd->low_latency &&
 		(bfqq->wr_coeff > 1 ||
 		 (bfq_bfqq_sync(bfqq) &&
@@ -1717,17 +1729,6 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 
 	bfq_clear_bfqq_just_created(bfqq);
 
-
-	if (!bfq_bfqq_IO_bound(bfqq)) {
-		if (arrived_in_time) {
-			bfqq->requests_within_timer++;
-			if (bfqq->requests_within_timer >=
-			    bfqd->bfq_requests_within_timer)
-				bfq_mark_bfqq_IO_bound(bfqq);
-		} else
-			bfqq->requests_within_timer = 0;
-	}
-
 	if (bfqd->low_latency) {
 		if (unlikely(time_is_after_jiffies(bfqq->split_time)))
 			/* wraparound */
@@ -1755,10 +1756,10 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 	bfq_add_bfqq_busy(bfqd, bfqq);
 
 	/*
-	 * Expire in-service queue only if preemption may be needed
-	 * for guarantees. In particular, we care only about two
-	 * cases. The first is that bfqq has to recover a service
-	 * hole, as explained in the comments on
+	 * Expire in-service queue if preemption may be needed for
+	 * guarantees or throughput. As for guarantees, we care
+	 * explicitly about two cases. The first is that bfqq has to
+	 * recover a service hole, as explained in the comments on
 	 * bfq_bfqq_update_budg_for_activation(), i.e., that
 	 * bfqq_wants_to_preempt is true. However, if bfqq does not
 	 * carry time-critical I/O, then bfqq's bandwidth is less
@@ -1785,11 +1786,23 @@ static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd,
 	 * timestamps of the in-service queue would need to be
 	 * updated, and this operation is quite costly (see the
 	 * comments on bfq_bfqq_update_budg_for_activation()).
+	 *
+	 * As for throughput, we ask bfq_better_to_idle() whether we
+	 * still need to plug I/O dispatching. If bfq_better_to_idle()
+	 * says no, then plugging is not needed any longer, either to
+	 * boost throughput or to perserve service guarantees. Then
+	 * the best option is to stop plugging I/O, as not doing so
+	 * would certainly lower throughput. We may end up in this
+	 * case if: (1) upon a dispatch attempt, we detected that it
+	 * was better to plug I/O dispatch, and to wait for a new
+	 * request to arrive for the currently in-service queue, but
+	 * (2) this switch of bfqq to busy changes the scenario.
 	 */
 	if (bfqd->in_service_queue &&
 	    ((bfqq_wants_to_preempt &&
 	      bfqq->wr_coeff >= bfqd->in_service_queue->wr_coeff) ||
-	     bfq_bfqq_higher_class_or_weight(bfqq, bfqd->in_service_queue)) &&
+	     bfq_bfqq_higher_class_or_weight(bfqq, bfqd->in_service_queue) ||
+	     !bfq_better_to_idle(bfqd->in_service_queue)) &&
 	    next_queue_may_preempt(bfqd))
 		bfq_bfqq_expire(bfqd, bfqd->in_service_queue,
 				false, BFQQE_PREEMPTED);
@@ -1861,6 +1874,138 @@ static void bfq_reset_inject_limit(struct bfq_data *bfqd,
 	bfqq->decrease_time_jif = jiffies;
 }
 
+static void bfq_update_io_intensity(struct bfq_queue *bfqq, u64 now_ns)
+{
+	u64 tot_io_time = now_ns - bfqq->io_start_time;
+
+	if (RB_EMPTY_ROOT(&bfqq->sort_list) && bfqq->dispatched == 0)
+		bfqq->tot_idle_time +=
+			now_ns - bfqq->ttime.last_end_request;
+
+	if (unlikely(bfq_bfqq_just_created(bfqq)))
+		return;
+
+	/*
+	 * Must be busy for at least about 80% of the time to be
+	 * considered I/O bound.
+	 */
+	if (bfqq->tot_idle_time * 5 > tot_io_time)
+		bfq_clear_bfqq_IO_bound(bfqq);
+	else
+		bfq_mark_bfqq_IO_bound(bfqq);
+
+	/*
+	 * Keep an observation window of at most 200 ms in the past
+	 * from now.
+	 */
+	if (tot_io_time > 200 * NSEC_PER_MSEC) {
+		bfqq->io_start_time = now_ns - (tot_io_time>>1);
+		bfqq->tot_idle_time >>= 1;
+	}
+}
+
+/*
+ * Detect whether bfqq's I/O seems synchronized with that of some
+ * other queue, i.e., whether bfqq, after remaining empty, happens to
+ * receive new I/O only right after some I/O request of the other
+ * queue has been completed. We call waker queue the other queue, and
+ * we assume, for simplicity, that bfqq may have at most one waker
+ * queue.
+ *
+ * A remarkable throughput boost can be reached by unconditionally
+ * injecting the I/O of the waker queue, every time a new
+ * bfq_dispatch_request happens to be invoked while I/O is being
+ * plugged for bfqq.  In addition to boosting throughput, this
+ * unblocks bfqq's I/O, thereby improving bandwidth and latency for
+ * bfqq. Note that these same results may be achieved with the general
+ * injection mechanism, but less effectively. For details on this
+ * aspect, see the comments on the choice of the queue for injection
+ * in bfq_select_queue().
+ *
+ * Turning back to the detection of a waker queue, a queue Q is deemed
+ * as a waker queue for bfqq if, for three consecutive times, bfqq
+ * happens to become non empty right after a request of Q has been
+ * completed. In particular, on the first time, Q is tentatively set
+ * as a candidate waker queue, while on the third consecutive time
+ * that Q is detected, the field waker_bfqq is set to Q, to confirm
+ * that Q is a waker queue for bfqq. These detection steps are
+ * performed only if bfqq has a long think time, so as to make it more
+ * likely that bfqq's I/O is actually being blocked by a
+ * synchronization. This last filter, plus the above three-times
+ * requirement, make false positives less likely.
+ *
+ * NOTE
+ *
+ * The sooner a waker queue is detected, the sooner throughput can be
+ * boosted by injecting I/O from the waker queue. Fortunately,
+ * detection is likely to be actually fast, for the following
+ * reasons. While blocked by synchronization, bfqq has a long think
+ * time. This implies that bfqq's inject limit is at least equal to 1
+ * (see the comments in bfq_update_inject_limit()). So, thanks to
+ * injection, the waker queue is likely to be served during the very
+ * first I/O-plugging time interval for bfqq. This triggers the first
+ * step of the detection mechanism. Thanks again to injection, the
+ * candidate waker queue is then likely to be confirmed no later than
+ * during the next I/O-plugging interval for bfqq.
+ *
+ * ISSUE
+ *
+ * On queue merging all waker information is lost.
+ */
+static void bfq_check_waker(struct bfq_data *bfqd, struct bfq_queue *bfqq,
+			    u64 now_ns)
+{
+	if (!bfqd->last_completed_rq_bfqq ||
+	    bfqd->last_completed_rq_bfqq == bfqq ||
+	    bfq_bfqq_has_short_ttime(bfqq) ||
+	    now_ns - bfqd->last_completion >= 4 * NSEC_PER_MSEC ||
+	    bfqd->last_completed_rq_bfqq == bfqq->waker_bfqq)
+		return;
+
+	if (bfqd->last_completed_rq_bfqq !=
+	    bfqq->tentative_waker_bfqq) {
+		/*
+		 * First synchronization detected with a
+		 * candidate waker queue, or with a different
+		 * candidate waker queue from the current one.
+		 */
+		bfqq->tentative_waker_bfqq =
+			bfqd->last_completed_rq_bfqq;
+		bfqq->num_waker_detections = 1;
+	} else /* Same tentative waker queue detected again */
+		bfqq->num_waker_detections++;
+
+	if (bfqq->num_waker_detections == 3) {
+		bfqq->waker_bfqq = bfqd->last_completed_rq_bfqq;
+		bfqq->tentative_waker_bfqq = NULL;
+
+		/*
+		 * If the waker queue disappears, then
+		 * bfqq->waker_bfqq must be reset. To
+		 * this goal, we maintain in each
+		 * waker queue a list, woken_list, of
+		 * all the queues that reference the
+		 * waker queue through their
+		 * waker_bfqq pointer. When the waker
+		 * queue exits, the waker_bfqq pointer
+		 * of all the queues in the woken_list
+		 * is reset.
+		 *
+		 * In addition, if bfqq is already in
+		 * the woken_list of a waker queue,
+		 * then, before being inserted into
+		 * the woken_list of a new waker
+		 * queue, bfqq must be removed from
+		 * the woken_list of the old waker
+		 * queue.
+		 */
+		if (!hlist_unhashed(&bfqq->woken_list_node))
+			hlist_del_init(&bfqq->woken_list_node);
+		hlist_add_head(&bfqq->woken_list_node,
+			       &bfqd->last_completed_rq_bfqq->woken_list);
+	}
+}
+
 static void bfq_add_request(struct request *rq)
 {
 	struct bfq_queue *bfqq = RQ_BFQQ(rq);
@@ -1868,117 +2013,14 @@ static void bfq_add_request(struct request *rq)
 	struct request *next_rq, *prev;
 	unsigned int old_wr_coeff = bfqq->wr_coeff;
 	bool interactive = false;
+	u64 now_ns = ktime_get_ns();
 
 	bfq_log_bfqq(bfqd, bfqq, "add_request %d", rq_is_sync(rq));
 	bfqq->queued[rq_is_sync(rq)]++;
 	bfqd->queued++;
 
 	if (RB_EMPTY_ROOT(&bfqq->sort_list) && bfq_bfqq_sync(bfqq)) {
-		/*
-		 * Detect whether bfqq's I/O seems synchronized with
-		 * that of some other queue, i.e., whether bfqq, after
-		 * remaining empty, happens to receive new I/O only
-		 * right after some I/O request of the other queue has
-		 * been completed. We call waker queue the other
-		 * queue, and we assume, for simplicity, that bfqq may
-		 * have at most one waker queue.
-		 *
-		 * A remarkable throughput boost can be reached by
-		 * unconditionally injecting the I/O of the waker
-		 * queue, every time a new bfq_dispatch_request
-		 * happens to be invoked while I/O is being plugged
-		 * for bfqq.  In addition to boosting throughput, this
-		 * unblocks bfqq's I/O, thereby improving bandwidth
-		 * and latency for bfqq. Note that these same results
-		 * may be achieved with the general injection
-		 * mechanism, but less effectively. For details on
-		 * this aspect, see the comments on the choice of the
-		 * queue for injection in bfq_select_queue().
-		 *
-		 * Turning back to the detection of a waker queue, a
-		 * queue Q is deemed as a waker queue for bfqq if, for
-		 * two consecutive times, bfqq happens to become non
-		 * empty right after a request of Q has been
-		 * completed. In particular, on the first time, Q is
-		 * tentatively set as a candidate waker queue, while
-		 * on the second time, the flag
-		 * bfq_bfqq_has_waker(bfqq) is set to confirm that Q
-		 * is a waker queue for bfqq. These detection steps
-		 * are performed only if bfqq has a long think time,
-		 * so as to make it more likely that bfqq's I/O is
-		 * actually being blocked by a synchronization. This
-		 * last filter, plus the above two-times requirement,
-		 * make false positives less likely.
-		 *
-		 * NOTE
-		 *
-		 * The sooner a waker queue is detected, the sooner
-		 * throughput can be boosted by injecting I/O from the
-		 * waker queue. Fortunately, detection is likely to be
-		 * actually fast, for the following reasons. While
-		 * blocked by synchronization, bfqq has a long think
-		 * time. This implies that bfqq's inject limit is at
-		 * least equal to 1 (see the comments in
-		 * bfq_update_inject_limit()). So, thanks to
-		 * injection, the waker queue is likely to be served
-		 * during the very first I/O-plugging time interval
-		 * for bfqq. This triggers the first step of the
-		 * detection mechanism. Thanks again to injection, the
-		 * candidate waker queue is then likely to be
-		 * confirmed no later than during the next
-		 * I/O-plugging interval for bfqq.
-		 */
-		if (bfqd->last_completed_rq_bfqq &&
-		    !bfq_bfqq_has_short_ttime(bfqq) &&
-		    ktime_get_ns() - bfqd->last_completion <
-		    200 * NSEC_PER_USEC) {
-			if (bfqd->last_completed_rq_bfqq != bfqq &&
-			    bfqd->last_completed_rq_bfqq !=
-			    bfqq->waker_bfqq) {
-				/*
-				 * First synchronization detected with
-				 * a candidate waker queue, or with a
-				 * different candidate waker queue
-				 * from the current one.
-				 */
-				bfqq->waker_bfqq = bfqd->last_completed_rq_bfqq;
-
-				/*
-				 * If the waker queue disappears, then
-				 * bfqq->waker_bfqq must be reset. To
-				 * this goal, we maintain in each
-				 * waker queue a list, woken_list, of
-				 * all the queues that reference the
-				 * waker queue through their
-				 * waker_bfqq pointer. When the waker
-				 * queue exits, the waker_bfqq pointer
-				 * of all the queues in the woken_list
-				 * is reset.
-				 *
-				 * In addition, if bfqq is already in
-				 * the woken_list of a waker queue,
-				 * then, before being inserted into
-				 * the woken_list of a new waker
-				 * queue, bfqq must be removed from
-				 * the woken_list of the old waker
-				 * queue.
-				 */
-				if (!hlist_unhashed(&bfqq->woken_list_node))
-					hlist_del_init(&bfqq->woken_list_node);
-				hlist_add_head(&bfqq->woken_list_node,
-				    &bfqd->last_completed_rq_bfqq->woken_list);
-
-				bfq_clear_bfqq_has_waker(bfqq);
-			} else if (bfqd->last_completed_rq_bfqq ==
-				   bfqq->waker_bfqq &&
-				   !bfq_bfqq_has_waker(bfqq)) {
-				/*
-				 * synchronization with waker_bfqq
-				 * seen for the second time
-				 */
-				bfq_mark_bfqq_has_waker(bfqq);
-			}
-		}
+		bfq_check_waker(bfqd, bfqq, now_ns);
 
 		/*
 		 * Periodically reset inject limit, to make sure that
@@ -2047,6 +2089,9 @@ static void bfq_add_request(struct request *rq)
 		}
 	}
 
+	if (bfq_bfqq_sync(bfqq))
+		bfq_update_io_intensity(bfqq, now_ns);
+
 	elv_rb_add(&bfqq->sort_list, rq);
 
 	/*
@@ -2352,6 +2397,24 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq,
 /* Must be called with bfqq != NULL */
 static void bfq_bfqq_end_wr(struct bfq_queue *bfqq)
 {
+	/*
+	 * If bfqq has been enjoying interactive weight-raising, then
+	 * reset soft_rt_next_start. We do it for the following
+	 * reason. bfqq may have been conveying the I/O needed to load
+	 * a soft real-time application. Such an application actually
+	 * exhibits a soft real-time I/O pattern after it finishes
+	 * loading, and finally starts doing its job. But, if bfqq has
+	 * been receiving a lot of bandwidth so far (likely to happen
+	 * on a fast device), then soft_rt_next_start now contains a
+	 * high value that. So, without this reset, bfqq would be
+	 * prevented from being possibly considered as soft_rt for a
+	 * very long time.
+	 */
+
+	if (bfqq->wr_cur_max_time !=
+	    bfqq->bfqd->bfq_wr_rt_max_time)
+		bfqq->soft_rt_next_start = jiffies;
+
 	if (bfq_bfqq_busy(bfqq))
 		bfqq->bfqd->wr_busy_queues--;
 	bfqq->wr_coeff = 1;
@@ -2686,10 +2749,16 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfqq)
 	if (!bic)
 		return;
 
+	bic->saved_last_serv_time_ns = bfqq->last_serv_time_ns;
+	bic->saved_inject_limit = bfqq->inject_limit;
+	bic->saved_decrease_time_jif = bfqq->decrease_time_jif;
+
 	bic->saved_weight = bfqq->entity.orig_weight;
 	bic->saved_ttime = bfqq->ttime;
 	bic->saved_has_short_ttime = bfq_bfqq_has_short_ttime(bfqq);
 	bic->saved_IO_bound = bfq_bfqq_IO_bound(bfqq);
+	bic->saved_io_start_time = bfqq->io_start_time;
+	bic->saved_tot_idle_time = bfqq->tot_idle_time;
 	bic->saved_in_large_burst = bfq_bfqq_in_large_burst(bfqq);
 	bic->was_in_burst_list = !hlist_unhashed(&bfqq->burst_list_node);
 	if (unlikely(bfq_bfqq_just_created(bfqq) &&
@@ -2712,6 +2781,7 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfqq)
 		bic->saved_wr_coeff = bfqq->wr_coeff;
 		bic->saved_wr_start_at_switch_to_srt =
 			bfqq->wr_start_at_switch_to_srt;
+		bic->saved_service_from_wr = bfqq->service_from_wr;
 		bic->saved_last_wr_start_finish = bfqq->last_wr_start_finish;
 		bic->saved_wr_cur_max_time = bfqq->wr_cur_max_time;
 	}
@@ -2937,6 +3007,7 @@ static void __bfq_set_in_service_queue(struct bfq_data *bfqd,
 	}
 
 	bfqd->in_service_queue = bfqq;
+	bfqd->in_serv_last_pos = 0;
 }
 
 /*
@@ -3442,20 +3513,38 @@ static void bfq_dispatch_remove(struct request_queue *q, struct request *rq)
  * order until all the requests already queued in the device have been
  * served. The last sub-condition commented above somewhat mitigates
  * this problem for weight-raised queues.
+ *
+ * However, as an additional mitigation for this problem, we preserve
+ * plugging for a special symmetric case that may suddenly turn into
+ * asymmetric: the case where only bfqq is busy. In this case, not
+ * expiring bfqq does not cause any harm to any other queues in terms
+ * of service guarantees. In contrast, it avoids the following unlucky
+ * sequence of events: (1) bfqq is expired, (2) a new queue with a
+ * lower weight than bfqq becomes busy (or more queues), (3) the new
+ * queue is served until a new request arrives for bfqq, (4) when bfqq
+ * is finally served, there are so many requests of the new queue in
+ * the drive that the pending requests for bfqq take a lot of time to
+ * be served. In particular, event (2) may case even already
+ * dispatched requests of bfqq to be delayed, inside the drive. So, to
+ * avoid this series of events, the scenario is preventively declared
+ * as asymmetric also if bfqq is the only busy queues
  */
 static bool idling_needed_for_service_guarantees(struct bfq_data *bfqd,
 						 struct bfq_queue *bfqq)
 {
+	int tot_busy_queues = bfq_tot_busy_queues(bfqd);
+
 	/* No point in idling for bfqq if it won't get requests any longer */
 	if (unlikely(!bfqq_process_refs(bfqq)))
 		return false;
 
 	return (bfqq->wr_coeff > 1 &&
 		(bfqd->wr_busy_queues <
-		 bfq_tot_busy_queues(bfqd) ||
+		 tot_busy_queues ||
 		 bfqd->rq_in_driver >=
 		 bfqq->dispatched + 4)) ||
-		bfq_asymmetric_scenario(bfqd, bfqq);
+		bfq_asymmetric_scenario(bfqd, bfqq) ||
+		tot_busy_queues == 1;
 }
 
 static bool __bfq_bfqq_expire(struct bfq_data *bfqd, struct bfq_queue *bfqq,
@@ -3939,10 +4028,6 @@ void bfq_bfqq_expire(struct bfq_data *bfqd,
 	      bfq_bfqq_budget_left(bfqq) >=  entity->budget / 3)))
 		bfq_bfqq_charge_time(bfqd, bfqq, delta);
 
-	if (reason == BFQQE_TOO_IDLE &&
-	    entity->service <= 2 * entity->budget / 10)
-		bfq_clear_bfqq_IO_bound(bfqq);
-
 	if (bfqd->low_latency && bfqq->wr_coeff == 1)