From d14469ed7c00314fe8957b2841bda329e4eaf4ab Mon Sep 17 00:00:00 2001 From: Yu Kuai Date: Mon, 25 Aug 2025 17:32:05 +0800 Subject: [PATCH 1/6] loop: fix zero sized loop for block special file By default, /dev/sda is block special file from devtmpfs, getattr will return file size as zero, causing loop failed for raw block device. We can add bdev_statx() to return device size, however this may introduce changes that are not acknowledged by user. Fix this problem by reverting changes for block special file, file mapping host is set to bdev inode while opening, and use i_size_read() directly to get device size. Fixes: 47b71abd5846 ("loop: use vfs_getattr_nosec for accurate file size") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-lkp/202508200409.b2459c02-lkp@intel.com Signed-off-by: Yu Kuai Reviewed-by: Ming Lei Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20250825093205.3684121-1-yukuai1@huaweicloud.com [axboe: fix spelling error] Signed-off-by: Jens Axboe --- drivers/block/loop.c | 26 ++++++++++++++++---------- 1 file changed, 16 insertions(+), 10 deletions(-) diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 57263c273f0f..053a086d547e 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -139,20 +139,26 @@ static int part_shift; static loff_t lo_calculate_size(struct loop_device *lo, struct file *file) { - struct kstat stat; loff_t loopsize; int ret; - /* - * Get the accurate file size. This provides better results than - * cached inode data, particularly for network filesystems where - * metadata may be stale. - */ - ret = vfs_getattr_nosec(&file->f_path, &stat, STATX_SIZE, 0); - if (ret) - return 0; + if (S_ISBLK(file_inode(file)->i_mode)) { + loopsize = i_size_read(file->f_mapping->host); + } else { + struct kstat stat; + + /* + * Get the accurate file size. This provides better results than + * cached inode data, particularly for network filesystems where + * metadata may be stale. + */ + ret = vfs_getattr_nosec(&file->f_path, &stat, STATX_SIZE, 0); + if (ret) + return 0; + + loopsize = stat.size; + } - loopsize = stat.size; if (lo->lo_offset > 0) loopsize -= lo->lo_offset; /* offset is beyond i_size, weird but possible */ From 198f36f902ec7e99b645382505f74b87a4523ed9 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Mon, 25 Aug 2025 11:27:19 -0700 Subject: [PATCH 2/6] blk-zoned: Fix a lockdep complaint about recursive locking If preparing a write bio fails then blk_zone_wplug_bio_work() calls bio_endio() with zwplug->lock held. If a device mapper driver is stacked on top of the zoned block device then this results in nested locking of zwplug->lock. The resulting lockdep complaint is a false positive because this is nested locking and not recursive locking. Suppress this false positive by calling blk_zone_wplug_bio_io_error() without holding zwplug->lock. This is safe because no code in blk_zone_wplug_bio_io_error() depends on zwplug->lock being held. This patch suppresses the following lockdep complaint: WARNING: possible recursive locking detected -------------------------------------------- kworker/3:0H/46 is trying to acquire lock: ffffff882968b830 (&zwplug->lock){-...}-{2:2}, at: blk_zone_write_plug_bio_endio+0x64/0x1f0 but task is already holding lock: ffffff88315bc230 (&zwplug->lock){-...}-{2:2}, at: blk_zone_wplug_bio_work+0x8c/0x48c other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&zwplug->lock); lock(&zwplug->lock); *** DEADLOCK *** May be due to missing lock nesting notation 3 locks held by kworker/3:0H/46: #0: ffffff8809486758 ((wq_completion)sdd_zwplugs){+.+.}-{0:0}, at: process_one_work+0x1bc/0x65c #1: ffffffc085de3d70 ((work_completion)(&zwplug->bio_work)){+.+.}-{0:0}, at: process_one_work+0x1e4/0x65c #2: ffffff88315bc230 (&zwplug->lock){-...}-{2:2}, at: blk_zone_wplug_bio_work+0x8c/0x48c stack backtrace: CPU: 3 UID: 0 PID: 46 Comm: kworker/3:0H Tainted: G W OE 6.12.38-android16-5-maybe-dirty-4k #1 8b362b6f76e3645a58cd27d86982bce10d150025 Tainted: [W]=WARN, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE Hardware name: Spacecraft board based on MALIBU (DT) Workqueue: sdd_zwplugs blk_zone_wplug_bio_work Call trace: dump_backtrace+0xfc/0x17c show_stack+0x18/0x28 dump_stack_lvl+0x40/0xa0 dump_stack+0x18/0x24 print_deadlock_bug+0x38c/0x398 __lock_acquire+0x13e8/0x2e1c lock_acquire+0x134/0x2b4 _raw_spin_lock_irqsave+0x5c/0x80 blk_zone_write_plug_bio_endio+0x64/0x1f0 bio_endio+0x9c/0x240 __dm_io_complete+0x214/0x260 clone_endio+0xe8/0x214 bio_endio+0x218/0x240 blk_zone_wplug_bio_work+0x204/0x48c process_one_work+0x26c/0x65c worker_thread+0x33c/0x498 kthread+0x110/0x134 ret_from_fork+0x10/0x20 Cc: stable@vger.kernel.org Cc: Damien Le Moal Cc: Christoph Hellwig Fixes: dd291d77cc90 ("block: Introduce zone write plugging") Signed-off-by: Bart Van Assche Reviewed-by: Damien Le Moal Link: https://lore.kernel.org/r/20250825182720.1697203-1-bvanassche@acm.org Signed-off-by: Jens Axboe --- block/blk-zoned.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/block/blk-zoned.c b/block/blk-zoned.c index ef43aaca49f4..5e2a5788dc3b 100644 --- a/block/blk-zoned.c +++ b/block/blk-zoned.c @@ -1286,14 +1286,14 @@ static void blk_zone_wplug_bio_work(struct work_struct *work) struct block_device *bdev; unsigned long flags; struct bio *bio; + bool prepared; /* * Submit the next plugged BIO. If we do not have any, clear * the plugged flag. */ - spin_lock_irqsave(&zwplug->lock, flags); - again: + spin_lock_irqsave(&zwplug->lock, flags); bio = bio_list_pop(&zwplug->bio_list); if (!bio) { zwplug->flags &= ~BLK_ZONE_WPLUG_PLUGGED; @@ -1304,13 +1304,14 @@ static void blk_zone_wplug_bio_work(struct work_struct *work) trace_blk_zone_wplug_bio(zwplug->disk->queue, zwplug->zone_no, bio->bi_iter.bi_sector, bio_sectors(bio)); - if (!blk_zone_wplug_prepare_bio(zwplug, bio)) { + prepared = blk_zone_wplug_prepare_bio(zwplug, bio); + spin_unlock_irqrestore(&zwplug->lock, flags); + + if (!prepared) { blk_zone_wplug_bio_io_error(zwplug, bio); goto again; } - spin_unlock_irqrestore(&zwplug->lock, flags); - bdev = bio->bi_bdev; /* From e3ef9445cd9d90e43de0bd3cd55d437773dfd139 Mon Sep 17 00:00:00 2001 From: Nilay Shroff Date: Tue, 26 Aug 2025 22:00:32 +0530 Subject: [PATCH 3/6] block: validate QoS before calling __rq_qos_done_bio() If a bio has BIO_QOS_xxx set, it doesn't guarantee that q->rq_qos is also present at-least for stacked block devices. For instance, in case of NVMe when multipath is enabled, the bottom device may have QoS enabled but top device doesn't. So always validate QoS is enabled and q->rq_qos is present before calling __rq_qos_done_bio(). Fixes: 370ac285f23a ("block: avoid cpu_hotplug_lock depedency on freeze_lock") Reported-by: Venkat Rao Bagalkote Closes: https://lore.kernel.org/all/3a07b752-06a4-4eee-b302-f4669feb859d@linux.ibm.com/ Signed-off-by: Nilay Shroff Link: https://lore.kernel.org/r/20250826163128.1952394-1-nilay@linux.ibm.com Signed-off-by: Jens Axboe --- block/blk-rq-qos.h | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/block/blk-rq-qos.h b/block/blk-rq-qos.h index 1fe22000a379..b538f2c0febc 100644 --- a/block/blk-rq-qos.h +++ b/block/blk-rq-qos.h @@ -149,12 +149,15 @@ static inline void rq_qos_done_bio(struct bio *bio) q = bdev_get_queue(bio->bi_bdev); /* - * If a bio has BIO_QOS_xxx set, it implicitly implies that - * q->rq_qos is present. So, we skip re-checking q->rq_qos - * here as an extra optimization and directly call - * __rq_qos_done_bio(). + * A BIO may carry BIO_QOS_* flags even if the associated request_queue + * does not have rq_qos enabled. This can happen with stacked block + * devices — for example, NVMe multipath, where it's possible that the + * bottom device has QoS enabled but the top device does not. Therefore, + * always verify that q->rq_qos is present and QoS is enabled before + * calling __rq_qos_done_bio(). */ - __rq_qos_done_bio(q->rq_qos, bio); + if (test_bit(QUEUE_FLAG_QOS_ENABLED, &q->queue_flags) && q->rq_qos) + __rq_qos_done_bio(q->rq_qos, bio); } static inline void rq_qos_throttle(struct request_queue *q, struct bio *bio) From c5c5eb24ed6177fc0ef4bb75fc18d07a99c1d3f0 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 27 Aug 2025 20:15:59 +0800 Subject: [PATCH 4/6] ublk: avoid ublk_io_release() called after ublk char dev is closed When running test_stress_04.sh, the following warning is triggered: WARNING: CPU: 1 PID: 135 at drivers/block/ublk_drv.c:1933 ublk_ch_release+0x423/0x4b0 [ublk_drv] This happens when the daemon is abruptly killed: - some references may still be held, because registering IO buffer doesn't grab ublk char device reference OR - io->task_registered_buffers won't be cleared because io buffer is released from non-daemon context For zero-copy and auto buffer register modes, I/O reference crosses syscalls, so IO reference may not be dropped naturally when ublk server is killed abruptly. However, when releasing io_uring context, it is guaranteed that the reference is dropped finally, see io_sqe_buffers_unregister() from io_ring_ctx_free(). Fix this by adding ublk_drain_io_references() that: - Waits for active I/O references dropped in async way by scheduling work function, for avoiding ublk dev and io_uring file's release dependency - Reinitializes io->ref and io->task_registered_buffers to clean state This ensures the reference count state is clean when ublk_queue_reinit() is called, preventing the warning and potential use-after-free. Fixes: 1f6540e2aabb ("ublk: zc register/unregister bvec") Fixes: 1ceeedb59749 ("ublk: optimize UBLK_IO_UNREGISTER_IO_BUF on daemon task") Fixes: 8a8fe42d765b ("ublk: optimize UBLK_IO_REGISTER_IO_BUF on daemon task") Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20250827121602.2619736-2-ming.lei@redhat.com Signed-off-by: Jens Axboe --- drivers/block/ublk_drv.c | 72 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 70 insertions(+), 2 deletions(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 99abd67b708b..67d4a867aec4 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -239,6 +239,7 @@ struct ublk_device { struct mutex cancel_mutex; bool canceling; pid_t ublksrv_tgid; + struct delayed_work exit_work; }; /* header of ublk_params */ @@ -1595,12 +1596,62 @@ static void ublk_set_canceling(struct ublk_device *ub, bool canceling) ublk_get_queue(ub, i)->canceling = canceling; } -static int ublk_ch_release(struct inode *inode, struct file *filp) +static bool ublk_check_and_reset_active_ref(struct ublk_device *ub) { - struct ublk_device *ub = filp->private_data; + int i, j; + + if (!(ub->dev_info.flags & (UBLK_F_SUPPORT_ZERO_COPY | + UBLK_F_AUTO_BUF_REG))) + return false; + + for (i = 0; i < ub->dev_info.nr_hw_queues; i++) { + struct ublk_queue *ubq = ublk_get_queue(ub, i); + + for (j = 0; j < ubq->q_depth; j++) { + struct ublk_io *io = &ubq->ios[j]; + unsigned int refs = refcount_read(&io->ref) + + io->task_registered_buffers; + + /* + * UBLK_REFCOUNT_INIT or zero means no active + * reference + */ + if (refs != UBLK_REFCOUNT_INIT && refs != 0) + return true; + + /* reset to zero if the io hasn't active references */ + refcount_set(&io->ref, 0); + io->task_registered_buffers = 0; + } + } + return false; +} + +static void ublk_ch_release_work_fn(struct work_struct *work) +{ + struct ublk_device *ub = + container_of(work, struct ublk_device, exit_work.work); struct gendisk *disk; int i; + /* + * For zero-copy and auto buffer register modes, I/O references + * might not be dropped naturally when the daemon is killed, but + * io_uring guarantees that registered bvec kernel buffers are + * unregistered finally when freeing io_uring context, then the + * active references are dropped. + * + * Wait until active references are dropped for avoiding use-after-free + * + * registered buffer may be unregistered in io_ring's release hander, + * so have to wait by scheduling work function for avoiding the two + * file release dependency. + */ + if (ublk_check_and_reset_active_ref(ub)) { + schedule_delayed_work(&ub->exit_work, 1); + return; + } + /* * disk isn't attached yet, either device isn't live, or it has * been removed already, so we needn't to do anything @@ -1673,6 +1724,23 @@ static int ublk_ch_release(struct inode *inode, struct file *filp) ublk_reset_ch_dev(ub); out: clear_bit(UB_STATE_OPEN, &ub->state); + + /* put the reference grabbed in ublk_ch_release() */ + ublk_put_device(ub); +} + +static int ublk_ch_release(struct inode *inode, struct file *filp) +{ + struct ublk_device *ub = filp->private_data; + + /* + * Grab ublk device reference, so it won't be gone until we are + * really released from work function. + */ + ublk_get_device(ub); + + INIT_DELAYED_WORK(&ub->exit_work, ublk_ch_release_work_fn); + schedule_delayed_work(&ub->exit_work, 0); return 0; } From 9b2785ea8592f239836405de023c75c4f3f5ce00 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 27 Aug 2025 20:16:00 +0800 Subject: [PATCH 5/6] ublk selftests: add --no_ublk_fixed_fd for not using registered ublk char device Add a new command line option --no_ublk_fixed_fd that excludes the ublk control device (/dev/ublkcN) from io_uring's registered files array. When this option is used, only backing files are registered starting from index 1, while the ublk control device is accessed using its raw file descriptor. Add ublk_get_registered_fd() helper function that returns the appropriate file descriptor for use with io_uring operations. Key optimizations implemented: - Cache UBLKS_Q_NO_UBLK_FIXED_FD flag in ublk_queue.flags to avoid reading dev->no_ublk_fixed_fd in fast path - Cache ublk char device fd in ublk_queue.ublk_fd for fast access - Update ublk_get_registered_fd() to use ublk_queue * parameter - Update io_uring_prep_buf_register/unregister() to use ublk_queue * - Replace ublk_device * access with ublk_queue * access in fast paths Also pass --no_ublk_fixed_fd to test_stress_04.sh for covering plain ublk char device mode. Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20250827121602.2619736-3-ming.lei@redhat.com Signed-off-by: Jens Axboe --- tools/testing/selftests/ublk/file_backed.c | 10 ++--- tools/testing/selftests/ublk/kublk.c | 38 +++++++++++++--- tools/testing/selftests/ublk/kublk.h | 45 +++++++++++++------ tools/testing/selftests/ublk/null.c | 4 +- tools/testing/selftests/ublk/stripe.c | 4 +- .../testing/selftests/ublk/test_stress_04.sh | 6 +-- 6 files changed, 74 insertions(+), 33 deletions(-) diff --git a/tools/testing/selftests/ublk/file_backed.c b/tools/testing/selftests/ublk/file_backed.c index 2d93ac860bd5..cd9fe69ecce2 100644 --- a/tools/testing/selftests/ublk/file_backed.c +++ b/tools/testing/selftests/ublk/file_backed.c @@ -20,7 +20,7 @@ static int loop_queue_flush_io(struct ublk_thread *t, struct ublk_queue *q, struct io_uring_sqe *sqe[1]; ublk_io_alloc_sqes(t, sqe, 1); - io_uring_prep_fsync(sqe[0], 1 /*fds[1]*/, IORING_FSYNC_DATASYNC); + io_uring_prep_fsync(sqe[0], ublk_get_registered_fd(q, 1) /*fds[1]*/, IORING_FSYNC_DATASYNC); io_uring_sqe_set_flags(sqe[0], IOSQE_FIXED_FILE); /* bit63 marks us as tgt io */ sqe[0]->user_data = build_user_data(tag, ublk_op, 0, q->q_id, 1); @@ -42,7 +42,7 @@ static int loop_queue_tgt_rw_io(struct ublk_thread *t, struct ublk_queue *q, if (!sqe[0]) return -ENOMEM; - io_uring_prep_rw(op, sqe[0], 1 /*fds[1]*/, + io_uring_prep_rw(op, sqe[0], ublk_get_registered_fd(q, 1) /*fds[1]*/, addr, iod->nr_sectors << 9, iod->start_sector << 9); @@ -56,19 +56,19 @@ static int loop_queue_tgt_rw_io(struct ublk_thread *t, struct ublk_queue *q, ublk_io_alloc_sqes(t, sqe, 3); - io_uring_prep_buf_register(sqe[0], 0, tag, q->q_id, ublk_get_io(q, tag)->buf_index); + io_uring_prep_buf_register(sqe[0], q, tag, q->q_id, ublk_get_io(q, tag)->buf_index); sqe[0]->flags |= IOSQE_CQE_SKIP_SUCCESS | IOSQE_IO_HARDLINK; sqe[0]->user_data = build_user_data(tag, ublk_cmd_op_nr(sqe[0]->cmd_op), 0, q->q_id, 1); - io_uring_prep_rw(op, sqe[1], 1 /*fds[1]*/, 0, + io_uring_prep_rw(op, sqe[1], ublk_get_registered_fd(q, 1) /*fds[1]*/, 0, iod->nr_sectors << 9, iod->start_sector << 9); sqe[1]->buf_index = tag; sqe[1]->flags |= IOSQE_FIXED_FILE | IOSQE_IO_HARDLINK; sqe[1]->user_data = build_user_data(tag, ublk_op, 0, q->q_id, 1); - io_uring_prep_buf_unregister(sqe[2], 0, tag, q->q_id, ublk_get_io(q, tag)->buf_index); + io_uring_prep_buf_unregister(sqe[2], q, tag, q->q_id, ublk_get_io(q, tag)->buf_index); sqe[2]->user_data = build_user_data(tag, ublk_cmd_op_nr(sqe[2]->cmd_op), 0, q->q_id, 1); return 2; diff --git a/tools/testing/selftests/ublk/kublk.c b/tools/testing/selftests/ublk/kublk.c index 6512dfbdbce3..b71faba86c3b 100644 --- a/tools/testing/selftests/ublk/kublk.c +++ b/tools/testing/selftests/ublk/kublk.c @@ -432,7 +432,7 @@ static void ublk_thread_deinit(struct ublk_thread *t) } } -static int ublk_queue_init(struct ublk_queue *q, unsigned extra_flags) +static int ublk_queue_init(struct ublk_queue *q, unsigned long long extra_flags) { struct ublk_dev *dev = q->dev; int depth = dev->dev_info.queue_depth; @@ -446,6 +446,9 @@ static int ublk_queue_init(struct ublk_queue *q, unsigned extra_flags) q->flags = dev->dev_info.flags; q->flags |= extra_flags; + /* Cache fd in queue for fast path access */ + q->ublk_fd = dev->fds[0]; + cmd_buf_size = ublk_queue_cmd_buf_sz(q); off = UBLKSRV_CMD_BUF_OFFSET + q->q_id * ublk_queue_max_cmd_buf_sz(); q->io_cmd_buf = mmap(0, cmd_buf_size, PROT_READ, @@ -481,9 +484,10 @@ static int ublk_queue_init(struct ublk_queue *q, unsigned extra_flags) return -ENOMEM; } -static int ublk_thread_init(struct ublk_thread *t) +static int ublk_thread_init(struct ublk_thread *t, unsigned long long extra_flags) { struct ublk_dev *dev = t->dev; + unsigned long long flags = dev->dev_info.flags | extra_flags; int ring_depth = dev->tgt.sq_depth, cq_depth = dev->tgt.cq_depth; int ret; @@ -512,7 +516,17 @@ static int ublk_thread_init(struct ublk_thread *t) io_uring_register_ring_fd(&t->ring); - ret = io_uring_register_files(&t->ring, dev->fds, dev->nr_fds); + if (flags & UBLKS_Q_NO_UBLK_FIXED_FD) { + /* Register only backing files starting from index 1, exclude ublk control device */ + if (dev->nr_fds > 1) { + ret = io_uring_register_files(&t->ring, &dev->fds[1], dev->nr_fds - 1); + } else { + /* No backing files to register, skip file registration */ + ret = 0; + } + } else { + ret = io_uring_register_files(&t->ring, dev->fds, dev->nr_fds); + } if (ret) { ublk_err("ublk dev %d thread %d register files failed %d\n", t->dev->dev_info.dev_id, t->idx, ret); @@ -626,9 +640,12 @@ int ublk_queue_io_cmd(struct ublk_thread *t, struct ublk_io *io) /* These fields should be written once, never change */ ublk_set_sqe_cmd_op(sqe[0], cmd_op); - sqe[0]->fd = 0; /* dev->fds[0] */ + sqe[0]->fd = ublk_get_registered_fd(q, 0); /* dev->fds[0] */ sqe[0]->opcode = IORING_OP_URING_CMD; - sqe[0]->flags = IOSQE_FIXED_FILE; + if (q->flags & UBLKS_Q_NO_UBLK_FIXED_FD) + sqe[0]->flags = 0; /* Use raw FD, not fixed file */ + else + sqe[0]->flags = IOSQE_FIXED_FILE; sqe[0]->rw_flags = 0; cmd->tag = io->tag; cmd->q_id = q->q_id; @@ -832,6 +849,7 @@ struct ublk_thread_info { unsigned idx; sem_t *ready; cpu_set_t *affinity; + unsigned long long extra_flags; }; static void *ublk_io_handler_fn(void *data) @@ -844,7 +862,7 @@ static void *ublk_io_handler_fn(void *data) t->dev = info->dev; t->idx = info->idx; - ret = ublk_thread_init(t); + ret = ublk_thread_init(t, info->extra_flags); if (ret) { ublk_err("ublk dev %d thread %u init failed\n", dev_id, t->idx); @@ -934,6 +952,8 @@ static int ublk_start_daemon(const struct dev_ctx *ctx, struct ublk_dev *dev) if (ctx->auto_zc_fallback) extra_flags = UBLKS_Q_AUTO_BUF_REG_FALLBACK; + if (ctx->no_ublk_fixed_fd) + extra_flags |= UBLKS_Q_NO_UBLK_FIXED_FD; for (i = 0; i < dinfo->nr_hw_queues; i++) { dev->q[i].dev = dev; @@ -951,6 +971,7 @@ static int ublk_start_daemon(const struct dev_ctx *ctx, struct ublk_dev *dev) tinfo[i].dev = dev; tinfo[i].idx = i; tinfo[i].ready = &ready; + tinfo[i].extra_flags = extra_flags; /* * If threads are not tied 1:1 to queues, setting thread @@ -1471,7 +1492,7 @@ static void __cmd_create_help(char *exe, bool recovery) printf("%s %s -t [null|loop|stripe|fault_inject] [-q nr_queues] [-d depth] [-n dev_id]\n", exe, recovery ? "recover" : "add"); printf("\t[--foreground] [--quiet] [-z] [--auto_zc] [--auto_zc_fallback] [--debug_mask mask] [-r 0|1 ] [-g]\n"); - printf("\t[-e 0|1 ] [-i 0|1]\n"); + printf("\t[-e 0|1 ] [-i 0|1] [--no_ublk_fixed_fd]\n"); printf("\t[--nthreads threads] [--per_io_tasks]\n"); printf("\t[target options] [backfile1] [backfile2] ...\n"); printf("\tdefault: nr_queues=2(max 32), depth=128(max 1024), dev_id=-1(auto allocation)\n"); @@ -1534,6 +1555,7 @@ int main(int argc, char *argv[]) { "size", 1, NULL, 's'}, { "nthreads", 1, NULL, 0 }, { "per_io_tasks", 0, NULL, 0 }, + { "no_ublk_fixed_fd", 0, NULL, 0 }, { 0, 0, 0, 0 } }; const struct ublk_tgt_ops *ops = NULL; @@ -1613,6 +1635,8 @@ int main(int argc, char *argv[]) ctx.nthreads = strtol(optarg, NULL, 10); if (!strcmp(longopts[option_idx].name, "per_io_tasks")) ctx.per_io_tasks = 1; + if (!strcmp(longopts[option_idx].name, "no_ublk_fixed_fd")) + ctx.no_ublk_fixed_fd = 1; break; case '?': /* diff --git a/tools/testing/selftests/ublk/kublk.h b/tools/testing/selftests/ublk/kublk.h index 219233f8a053..5e55484fb0aa 100644 --- a/tools/testing/selftests/ublk/kublk.h +++ b/tools/testing/selftests/ublk/kublk.h @@ -77,6 +77,7 @@ struct dev_ctx { unsigned int recovery:1; unsigned int auto_zc_fallback:1; unsigned int per_io_tasks:1; + unsigned int no_ublk_fixed_fd:1; int _evtfd; int _shmid; @@ -166,7 +167,9 @@ struct ublk_queue { /* borrow one bit of ublk uapi flags, which may never be used */ #define UBLKS_Q_AUTO_BUF_REG_FALLBACK (1ULL << 63) +#define UBLKS_Q_NO_UBLK_FIXED_FD (1ULL << 62) __u64 flags; + int ublk_fd; /* cached ublk char device fd */ struct ublk_io ios[UBLK_QUEUE_DEPTH]; }; @@ -273,34 +276,48 @@ static inline int ublk_io_alloc_sqes(struct ublk_thread *t, return nr_sqes; } -static inline void io_uring_prep_buf_register(struct io_uring_sqe *sqe, - int dev_fd, int tag, int q_id, __u64 index) +static inline int ublk_get_registered_fd(struct ublk_queue *q, int fd_index) +{ + if (q->flags & UBLKS_Q_NO_UBLK_FIXED_FD) { + if (fd_index == 0) + /* Return the raw ublk FD for index 0 */ + return q->ublk_fd; + /* Adjust index for backing files (index 1 becomes 0, etc.) */ + return fd_index - 1; + } + return fd_index; +} + +static inline void __io_uring_prep_buf_reg_unreg(struct io_uring_sqe *sqe, + struct ublk_queue *q, int tag, int q_id, __u64 index) { struct ublksrv_io_cmd *cmd = (struct ublksrv_io_cmd *)sqe->cmd; + int dev_fd = ublk_get_registered_fd(q, 0); io_uring_prep_read(sqe, dev_fd, 0, 0, 0); sqe->opcode = IORING_OP_URING_CMD; - sqe->flags |= IOSQE_FIXED_FILE; - sqe->cmd_op = UBLK_U_IO_REGISTER_IO_BUF; + if (q->flags & UBLKS_Q_NO_UBLK_FIXED_FD) + sqe->flags &= ~IOSQE_FIXED_FILE; + else + sqe->flags |= IOSQE_FIXED_FILE; cmd->tag = tag; cmd->addr = index; cmd->q_id = q_id; } -static inline void io_uring_prep_buf_unregister(struct io_uring_sqe *sqe, - int dev_fd, int tag, int q_id, __u64 index) +static inline void io_uring_prep_buf_register(struct io_uring_sqe *sqe, + struct ublk_queue *q, int tag, int q_id, __u64 index) { - struct ublksrv_io_cmd *cmd = (struct ublksrv_io_cmd *)sqe->cmd; + __io_uring_prep_buf_reg_unreg(sqe, q, tag, q_id, index); + sqe->cmd_op = UBLK_U_IO_REGISTER_IO_BUF; +} - io_uring_prep_read(sqe, dev_fd, 0, 0, 0); - sqe->opcode = IORING_OP_URING_CMD; - sqe->flags |= IOSQE_FIXED_FILE; +static inline void io_uring_prep_buf_unregister(struct io_uring_sqe *sqe, + struct ublk_queue *q, int tag, int q_id, __u64 index) +{ + __io_uring_prep_buf_reg_unreg(sqe, q, tag, q_id, index); sqe->cmd_op = UBLK_U_IO_UNREGISTER_IO_BUF; - - cmd->tag = tag; - cmd->addr = index; - cmd->q_id = q_id; } static inline void *ublk_get_sqe_cmd(const struct io_uring_sqe *sqe) diff --git a/tools/testing/selftests/ublk/null.c b/tools/testing/selftests/ublk/null.c index f0e0003a4860..280043f6b689 100644 --- a/tools/testing/selftests/ublk/null.c +++ b/tools/testing/selftests/ublk/null.c @@ -63,7 +63,7 @@ static int null_queue_zc_io(struct ublk_thread *t, struct ublk_queue *q, ublk_io_alloc_sqes(t, sqe, 3); - io_uring_prep_buf_register(sqe[0], 0, tag, q->q_id, ublk_get_io(q, tag)->buf_index); + io_uring_prep_buf_register(sqe[0], q, tag, q->q_id, ublk_get_io(q, tag)->buf_index); sqe[0]->user_data = build_user_data(tag, ublk_cmd_op_nr(sqe[0]->cmd_op), 0, q->q_id, 1); sqe[0]->flags |= IOSQE_CQE_SKIP_SUCCESS | IOSQE_IO_HARDLINK; @@ -71,7 +71,7 @@ static int null_queue_zc_io(struct ublk_thread *t, struct ublk_queue *q, __setup_nop_io(tag, iod, sqe[1], q->q_id); sqe[1]->flags |= IOSQE_IO_HARDLINK; - io_uring_prep_buf_unregister(sqe[2], 0, tag, q->q_id, ublk_get_io(q, tag)->buf_index); + io_uring_prep_buf_unregister(sqe[2], q, tag, q->q_id, ublk_get_io(q, tag)->buf_index); sqe[2]->user_data = build_user_data(tag, ublk_cmd_op_nr(sqe[2]->cmd_op), 0, q->q_id, 1); // buf register is marked as IOSQE_CQE_SKIP_SUCCESS diff --git a/tools/testing/selftests/ublk/stripe.c b/tools/testing/selftests/ublk/stripe.c index 1fb9b7cc281b..791fa8dc1651 100644 --- a/tools/testing/selftests/ublk/stripe.c +++ b/tools/testing/selftests/ublk/stripe.c @@ -142,7 +142,7 @@ static int stripe_queue_tgt_rw_io(struct ublk_thread *t, struct ublk_queue *q, ublk_io_alloc_sqes(t, sqe, s->nr + extra); if (zc) { - io_uring_prep_buf_register(sqe[0], 0, tag, q->q_id, io->buf_index); + io_uring_prep_buf_register(sqe[0], q, tag, q->q_id, io->buf_index); sqe[0]->flags |= IOSQE_CQE_SKIP_SUCCESS | IOSQE_IO_HARDLINK; sqe[0]->user_data = build_user_data(tag, ublk_cmd_op_nr(sqe[0]->cmd_op), 0, q->q_id, 1); @@ -168,7 +168,7 @@ static int stripe_queue_tgt_rw_io(struct ublk_thread *t, struct ublk_queue *q, if (zc) { struct io_uring_sqe *unreg = sqe[s->nr + 1]; - io_uring_prep_buf_unregister(unreg, 0, tag, q->q_id, io->buf_index); + io_uring_prep_buf_unregister(unreg, q, tag, q->q_id, io->buf_index); unreg->user_data = build_user_data( tag, ublk_cmd_op_nr(unreg->cmd_op), 0, q->q_id, 1); } diff --git a/tools/testing/selftests/ublk/test_stress_04.sh b/tools/testing/selftests/ublk/test_stress_04.sh index 40d1437ca298..3f901db4d09d 100755 --- a/tools/testing/selftests/ublk/test_stress_04.sh +++ b/tools/testing/selftests/ublk/test_stress_04.sh @@ -28,14 +28,14 @@ _create_backfile 0 256M _create_backfile 1 128M _create_backfile 2 128M -ublk_io_and_kill_daemon 8G -t null -q 4 -z & -ublk_io_and_kill_daemon 256M -t loop -q 4 -z "${UBLK_BACKFILES[0]}" & +ublk_io_and_kill_daemon 8G -t null -q 4 -z --no_ublk_fixed_fd & +ublk_io_and_kill_daemon 256M -t loop -q 4 -z --no_ublk_fixed_fd "${UBLK_BACKFILES[0]}" & ublk_io_and_kill_daemon 256M -t stripe -q 4 -z "${UBLK_BACKFILES[1]}" "${UBLK_BACKFILES[2]}" & if _have_feature "AUTO_BUF_REG"; then ublk_io_and_kill_daemon 8G -t null -q 4 --auto_zc & ublk_io_and_kill_daemon 256M -t loop -q 4 --auto_zc "${UBLK_BACKFILES[0]}" & - ublk_io_and_kill_daemon 256M -t stripe -q 4 --auto_zc "${UBLK_BACKFILES[1]}" "${UBLK_BACKFILES[2]}" & + ublk_io_and_kill_daemon 256M -t stripe -q 4 --auto_zc --no_ublk_fixed_fd "${UBLK_BACKFILES[1]}" "${UBLK_BACKFILES[2]}" & ublk_io_and_kill_daemon 8G -t null -q 4 -z --auto_zc --auto_zc_fallback & fi From 95a7c5000956f939b86d8b00b8e6b8345f4a9b65 Mon Sep 17 00:00:00 2001 From: Coly Li Date: Thu, 28 Aug 2025 23:48:35 +0800 Subject: [PATCH 6/6] bcache: change maintainer's email address Change to my new email address on fnnas.com. Signed-off-by: Coly Li Link: https://lore.kernel.org/r/20250828154835.32926-1-colyli@kernel.org Signed-off-by: Jens Axboe --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index fe168477caa4..b47daf498a97 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4205,7 +4205,7 @@ W: http://www.baycom.org/~tom/ham/ham.html F: drivers/net/hamradio/baycom* BCACHE (BLOCK LAYER CACHE) -M: Coly Li +M: Coly Li M: Kent Overstreet L: linux-bcache@vger.kernel.org S: Maintained