summaryrefslogtreecommitdiffstats
path: root/drivers/block
Commit message (Collapse)AuthorAgeFilesLines
* libceph: add osd_req_op_extent_osd_data_bvecs()Ilya Dryomov2018-05-101-1/+3
| | | | | | | | ... and store num_bvecs for client code's convenience. Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
* Merge tag 'for-linus-20180425' of git://git.kernel.dk/linux-blockLinus Torvalds2018-04-264-52/+68
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull block updates from Jens Axboe: "I ended up sitting on this about a week longer than I wanted to, since we were hashing out details with a timeout change. I've now killed that patch, so we can flush the existing queue in due time. This contains: - Fix for an old regression, where entering the queue can be disturbed by a signal to the process. This can cause spurious EIO. Fix from Alan Jenkins. - cdrom information leak fix from Dan. - Trivial helper for testing queue FUA from Dave Chinner, part of his O_DIRECT FUA series. - Series of swim fixes from Finn that actually makes it work again. - Loop O_DIRECT corruption fix, which caused data corruption in production for us. From me. - BFQ crash fix from me. - bcache maintainer update. Michael no longer has the time to do it, Coly has stepped up to serve as the new maintainer. - blkcg locking fixes from Jiang Biao. - Revert of a change from this merge window from Ming, that causes an issue on some hardware. - Minor clarification doc addition from Linus Walleij" * tag 'for-linus-20180425' of git://git.kernel.dk/linux-block: (22 commits) Revert "blk-mq: remove code for dealing with remapping queue" block: mq: Add some minor doc for core structs bcache: mark Coly Li as bcache maintainer MAINTAINERS: Remove me as maintainer of bcache blkcg: init root blkcg_gq under lock blkcg: small fix on comment in blkcg_init_queue blkcg: don't hold blkcg lock when deactivating policy block: add blk_queue_fua() helper function cdrom: information leak in cdrom_ioctl_media_changed() bfq-iosched: ensure to clear bic/bfqq pointers when preparing request blk-mq: start request gstate with gen 1 block/swim: Select appropriate drive on device open block/swim: Fix IO error at end of medium block/swim: Check drive type block/swim: Rename macros to avoid inconsistent inverted logic block/swim: Don't log an error message for an invalid ioctl block/swim: Remove extra put_disk() call from error path block/swim: Fix array bounds check m68k/mac: Don't remap SWIM MMIO region loop: handle short DIO reads ...
| * block/swim: Select appropriate drive on device openFinn Thain2018-04-171-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | The driver supports internal and external FDD units so the floppy_open function must not hard-code the drive location. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Fix IO error at end of mediumFinn Thain2018-04-171-5/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Reading to the end of a 720K disk results in an IO error instead of EOF because the block layer thinks the disk has 2880 sectors. (Partly this is a result of inverted logic of the ONEMEG_MEDIA bit that's now fixed.) Initialize the density and head count in swim_add_floppy() to agree with the device size passed to set_capacity() during drive probe. Call set_capacity() again upon device open, after refreshing the density and head count values. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Check drive typeFinn Thain2018-04-171-2/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The SWIM chip is compatible with GCR-mode Sony 400K/800K drives but this driver only supports MFM mode. Therefore only Sony FDHD drives are supported. Skip incompatible drives. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Rename macros to avoid inconsistent inverted logicFinn Thain2018-04-172-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The Sony drive status bits use active-low logic. The swim_readbit() function converts that to 'C' logic for readability. Hence, the sense of the names of the status bit macros should not be inverted. Mostly they are correct. However, the TWOMEG_DRIVE, MFM_MODE and TWOMEG_MEDIA macros have inverted sense (like MkLinux). Fix this inconsistency and make the following patches less confusing. The same problem affects swim3.c so fix that too. No functional change. The FDHD drive status bits are documented in sonydriv.cpp from MAME and in swimiii.h from MkLinux. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: linuxppc-dev@lists.ozlabs.org Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Don't log an error message for an invalid ioctlFinn Thain2018-04-171-7/+2Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The 'eject' shell command may send various different ioctl commands. This leads to error messages on the console even though the FDEJECT ioctl succeeds. ~# eject floppy SWIM floppy_ioctl: unknown cmd 21257 SWIM floppy_ioctl: unknown cmd 1 Don't log an error message for an invalid ioctl, just do as the swim3 driver does and return -ENOTTY. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Remove extra put_disk() call from error pathFinn Thain2018-04-171-1/+0Star
| | | | | | | | | | | | | | | | | | | | | | | | Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Fixes: 103db8b2dfa5 ("[PATCH] swim: stop sharing request queue across multiple gendisks") Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/swim: Fix array bounds checkFinn Thain2018-04-171-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the floppy_find() function in swim.c is a call to get_disk(swd->unit[drive].disk). The actual parameter to this call can be a NULL pointer when drive == swd->floppy_count. This causes an oops in get_disk(). Data read fault at 0x00000198 in Super Data (pc=0x1be5b6) BAD KERNEL BUSERR Oops: 00000000 Modules linked in: swim_mod ipv6 mac8390 PC: [<001be5b6>] get_disk+0xc/0x76 SR: 2004 SP: 9a078bc1 a2: 0213ed90 d0: 00000000 d1: 00000000 d2: 00000000 d3: 000000ff d4: 00000002 d5: 02983590 a0: 02332e00 a1: 022dfd64 Process dd (pid: 285, task=020ab25b) Frame format=B ssw=074d isc=4a88 isb=6732 daddr=00000198 dobuf=00000000 baddr=001be5bc dibuf=bfffffff ver=f Stack from 022dfca4: 00000000 0203fc00 0213ed90 022dfcc0 02982936 00000000 00200000 022dfd08 0020f85a 00200000 022dfd64 02332e00 004040fc 00000014 001be77e 022dfd64 00334e4a 001be3f8 0800001d 022dfd64 01c04b60 01c04b70 022aba80 029828f8 02332e00 022dfd2c 001be7ac 0203fc00 00200000 022dfd64 02103a00 01c04b60 01c04b60 0200e400 022dfd68 000e191a 00200000 022dfd64 02103a00 0800001d 00000000 00000003 000b89de 00500000 02103a00 01c04b60 02103a08 01c04c2e Call Trace: [<02982936>] floppy_find+0x3e/0x4a [swim_mod] [<00200000>] uart_remove_one_port+0x1a2/0x260 [<0020f85a>] kobj_lookup+0xde/0x132 [<00200000>] uart_remove_one_port+0x1a2/0x260 [<001be77e>] get_gendisk+0x0/0x130 [<00334e4a>] mutex_lock+0x0/0x2e [<001be3f8>] disk_block_events+0x0/0x6c [<029828f8>] floppy_find+0x0/0x4a [swim_mod] [<001be7ac>] get_gendisk+0x2e/0x130 [<00200000>] uart_remove_one_port+0x1a2/0x260 [<000e191a>] __blkdev_get+0x32/0x45a [<00200000>] uart_remove_one_port+0x1a2/0x260 [<000b89de>] complete_walk+0x0/0x8a [<000e1e22>] blkdev_get+0xe0/0x29a [<000e1fdc>] blkdev_open+0x0/0xb0 [<000b89de>] complete_walk+0x0/0x8a [<000e1fdc>] blkdev_open+0x0/0xb0 [<000e01cc>] bd_acquire+0x74/0x8a [<000e205c>] blkdev_open+0x80/0xb0 [<000e1fdc>] blkdev_open+0x0/0xb0 [<000abf24>] do_dentry_open+0x1a4/0x322 [<00020000>] __do_proc_douintvec+0x22/0x27e [<000b89de>] complete_walk+0x0/0x8a [<000baa62>] link_path_walk+0x0/0x48e [<000ba3f8>] inode_permission+0x20/0x54 [<000ac0e4>] vfs_open+0x42/0x78 [<000bc372>] path_openat+0x2b2/0xeaa [<000bc0c0>] path_openat+0x0/0xeaa [<0004463e>] __irq_wake_thread+0x0/0x4e [<0003a45a>] task_tick_fair+0x18/0xc8 [<000bd00a>] do_filp_open+0xa0/0xea [<000abae0>] do_sys_open+0x11a/0x1ee [<00020000>] __do_proc_douintvec+0x22/0x27e [<000abbf4>] SyS_open+0x1e/0x22 [<00020000>] __do_proc_douintvec+0x22/0x27e [<00002b40>] syscall+0x8/0xc [<00020000>] __do_proc_douintvec+0x22/0x27e [<0000c00b>] dyadic+0x1/0x28 Code: 4e5e 4e75 4e56 fffc 2f0b 2f02 266e 0008 <206b> 0198 4a88 6732 2428 002c 661e 486b 0058 4eb9 0032 0b96 588f 4a88 672c 2008 Disabling lock debugging due to kernel taint Fix the array index bounds check to avoid this. Cc: Laurent Vivier <lvivier@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: stable@vger.kernel.org # v4.14+ Fixes: 8852ecd97488 ("[PATCH] m68k: mac - Add SWIM floppy support") Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * m68k/mac: Don't remap SWIM MMIO regionFinn Thain2018-04-171-7/+3Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For reasons I don't understand, calling ioremap() then iounmap() on the SWIM MMIO region causes a hang on 68030 (but not on 68040). ~# modprobe swim_mod SWIM floppy driver Version 0.2 (2008-10-30) SWIM device not found ! watchdog: BUG: soft lockup - CPU#0 stuck for 23s! [modprobe:285] Modules linked in: swim_mod(+) Format 00 Vector: 0064 PC: 000075aa Status: 2000 Not tainted ORIG_D0: ffffffff D0: d00c0000 A2: 007c2370 A1: 003f810c A0: 00040000 D5: d0096800 D4: d0097e00 D3: 00000001 D2: 00000003 D1: 00000000 Non-Maskable Interrupt Modules linked in: swim_mod(+) PC: [<000075ba>] __iounmap+0x24/0x10e SR: 2000 SP: 007abc48 a2: 007c2370 d0: d00c0000 d1: 000001a0 d2: 00000019 d3: 00000001 d4: d0097e00 d5: d0096800 a0: 00040000 a1: 003f810c Process modprobe (pid: 285, task=007c2370) Frame format=0 Stack from 007abc7c: ffffffed 00000000 006a4060 004712e0 007abca0 000076ea d0080000 00080000 010bb4b8 007abcd8 010ba542 d0096000 00000000 00000000 00000001 010bb59c 00000000 007abf30 010bb4b8 0047760a 0047763c 00477612 00616540 007abcec 0020a91a 00477600 0047760a 010bb4cc 007abd18 002092f2 0047760a 00333b06 007abd5c 00000000 0047760a 010bb4cc 00404f90 004776b8 00000001 007abd38 00209446 010bb4cc 0047760a 010bb4cc 0020938e 0031f8be 00616540 007abd64 Call Trace: [<000076ea>] iounmap+0x46/0x5a [<00080000>] shrink_page_list+0x7f6/0xe06 [<010ba542>] swim_probe+0xe4/0x496 [swim_mod] [<0020a91a>] platform_drv_probe+0x20/0x5e [<002092f2>] driver_probe_device+0x21c/0x2b8 [<00333b06>] mutex_lock+0x0/0x2e [<00209446>] __driver_attach+0xb8/0xce [<0020938e>] __driver_attach+0x0/0xce [<0031f8be>] klist_next+0x0/0xa0 [<00207562>] bus_for_each_dev+0x74/0xba [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<00333b06>] mutex_lock+0x0/0x2e [<00208e44>] driver_attach+0x1a/0x1e [<0020938e>] __driver_attach+0x0/0xce [<00207e26>] bus_add_driver+0x188/0x234 [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<00209894>] driver_register+0x58/0x104 [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<010bd000>] swim_init+0x0/0x2c [swim_mod] [<0020a7be>] __platform_driver_register+0x38/0x3c [<010bd028>] swim_init+0x28/0x2c [swim_mod] [<000020dc>] do_one_initcall+0x38/0x196 [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<003331cc>] mutex_unlock+0x0/0x3e [<00333b06>] mutex_lock+0x0/0x2e [<003331cc>] mutex_unlock+0x0/0x3e [<00333b06>] mutex_lock+0x0/0x2e [<003331cc>] mutex_unlock+0x0/0x3e [<00333b06>] mutex_lock+0x0/0x2e [<003331cc>] mutex_unlock+0x0/0x3e [<00333b06>] mutex_lock+0x0/0x2e [<00075008>] __free_pages+0x0/0x38 [<000045c0>] mangle_kernel_stack+0x30/0xda [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<003331cc>] mutex_unlock+0x0/0x3e [<00333b06>] mutex_lock+0x0/0x2e [<0005ced4>] do_init_module+0x42/0x266 [<010bd000>] swim_init+0x0/0x2c [swim_mod] [<000344c0>] blocking_notifier_call_chain+0x0/0x20 [<0005eda0>] load_module+0x1a30/0x1e70 [<0000465d>] mangle_kernel_stack+0xcd/0xda [<00331c64>] __generic_copy_from_user+0x0/0x46 [<0033256e>] _cond_resched+0x0/0x32 [<00331b9c>] memset+0x0/0x98 [<0033256e>] _cond_resched+0x0/0x32 [<0005f25c>] SyS_init_module+0x7c/0x112 [<00002000>] _start+0x0/0x8 [<00002000>] _start+0x0/0x8 [<00331c82>] __generic_copy_from_user+0x1e/0x46 [<0005f2b2>] SyS_init_module+0xd2/0x112 [<0000465d>] mangle_kernel_stack+0xcd/0xda [<00002b40>] syscall+0x8/0xc [<0000465d>] mangle_kernel_stack+0xcd/0xda [<0008c00c>] pcpu_balance_workfn+0xb2/0x40e Code: 2200 7419 e4a9 e589 2841 d9fc 0000 1000 <2414> 7203 c282 7602 b681 6600 0096 0242 fe00 0482 0000 0000 e9c0 11c3 ed89 2642 There's no need to call ioremap() for the SWIM address range, as it lies within the usual IO device region at 0x5000 0000, which has already been mapped by head.S. Remove the redundant ioremap() and iounmap() calls to fix the hang. Cc: Laurent Vivier <lvivier@redhat.com> Cc: stable@vger.kernel.org # v4.14+ Tested-by: Stan Johnson <userm57@yahoo.com> Signed-off-by: Finn Thain <fthain@telegraphics.com.au> Acked-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * loop: handle short DIO readsJens Axboe2018-04-151-7/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We ran into an issue with loop and btrfs, where btrfs would complain about checksum errors. It turns out that is because we don't handle short reads at all, we just zero fill the remainder. Worse than that, we don't handle the filling properly, which results in loop trying to advance a single bio by much more than its size, since it doesn't take chaining into account. Handle short reads appropriately, by simply retrying at the new correct offset. End the remainder of the request with EIO, if we get a 0 read. Fixes: bc07c10a3603 ("block: loop: support DIO & AIO") Reviewed-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * loop: remove cmd->rq memberJens Axboe2018-04-152-18/+19
| | | | | | | | | | | | | | | | We can always get at the request from the payload, no need to store a pointer to it. Reviewed-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
* | rbd: notrim map optionIlya Dryomov2018-04-161-5/+14
| | | | | | | | | | | | | | | | | | | | Add an option to turn off discard and write zeroes offload support to avoid deprovisioning a fully provisioned image. When enabled, discard requests will fail with -EOPNOTSUPP, write zeroes requests will fall back to manually zeroing. Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Tested-by: Hitoshi Kamei <hitoshi.kamei.xm@hitachi.com>
* | rbd: adjust queue limits for "fancy" stripingIlya Dryomov2018-04-161-9/+8Star
| | | | | | | | | | | | | | | | In order to take full advantage of merging in ceph_file_to_extents(), allow object set sized I/Os. If the layout is not "fancy", an object set consists of just one object. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
* | rbd: avoid Wreturn-type warningsArnd Bergmann2018-04-161-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In some configurations gcc cannot see that rbd_assert(0) leads to an unreachable code path: drivers/block/rbd.c: In function 'rbd_img_is_write': drivers/block/rbd.c:1397:1: error: control reaches end of non-void function [-Werror=return-type] drivers/block/rbd.c: In function '__rbd_obj_handle_request': drivers/block/rbd.c:2499:1: error: control reaches end of non-void function [-Werror=return-type] drivers/block/rbd.c: In function 'rbd_obj_handle_write': drivers/block/rbd.c:2471:1: error: control reaches end of non-void function [-Werror=return-type] As the rbd_assert() here shows has no extra information beyond the verbose BUG(), we can simply use BUG() directly in its place. This is reliably detected as not returning on any architecture, since it doesn't depend on the unlikely() comparison that confused gcc. Fixes: 3da691bf4366 ("rbd: new request handling code") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
* | rbd: support timeout in rbd_wait_state_locked()Dongsheng Yang2018-04-161-1/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | currently, the rbd_wait_state_locked() will wait forever if we can't get our state locked. Example: rbd map --exclusive test1 --> /dev/rbd0 rbd map test1 --> /dev/rbd1 dd if=/dev/zero of=/dev/rbd1 bs=1M count=1 --> IO blocked To avoid this problem, this patch introduce a timeout design in rbd_wait_state_locked(). Then rbd_wait_state_locked() will return error when we reach a timeout. This patch allow user to set the lock_timeout in rbd mapping. Signed-off-by: Dongsheng Yang <dongsheng.yang@easystack.cn> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
* | rbd: refactor rbd_wait_state_locked()Ilya Dryomov2018-04-161-17/+26
|/ | | | | | | In preparation for lock_timeout option, make rbd_wait_state_locked() return error codes. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
* Merge tag 'for-linus-20180413' of git://git.kernel.dk/linux-blockLinus Torvalds2018-04-141-19/+26
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull block fixes from Jens Axboe: "Followup fixes for this merge window. This contains: - Series from Ming, fixing corner cases in our CPU <-> queue mapping. This triggered repeated warnings on especially s390, but I also hit it in cpu hot plug/unplug testing while doing IO on NVMe on x86-64. - Another fix from Ming, ensuring that we always order budget and driver tag identically, avoiding a deadlock on QD=1 devices. - Loop locking regression fix from this merge window, from Omar. - Another loop locking fix, this time missing an unlock, from Tetsuo Handa. - Fix for racing IO submission with device removal from Bart. - sr reference fix from me, fixing a case where disk change or getevents can race with device removal. - Set of nvme fixes by way of Keith, from various contributors" * tag 'for-linus-20180413' of git://git.kernel.dk/linux-block: (28 commits) nvme: expand nvmf_check_if_ready checks nvme: Use admin command effects for admin commands nvmet: fix space padding in serial number nvme: check return value of init_srcu_struct function nvmet: Fix nvmet_execute_write_zeroes sector count nvme-pci: Separate IO and admin queue IRQ vectors nvme-pci: Remove unused queue parameter nvme-pci: Skip queue deletion if there are no queues nvme: target: fix buffer overflow nvme: don't send keep-alives to the discovery controller nvme: unexport nvme_start_keep_alive nvme-loop: fix kernel oops in case of unhandled command nvme: enforce 64bit offset for nvme_get_log_ext fn sr: get/drop reference to device in revalidate and check_events blk-mq: Revert "blk-mq: reimplement blk_mq_hw_queue_mapped" blk-mq: Avoid that submitting a bio concurrently with device removal triggers a crash backing: silence compiler warning using __printf blk-mq: remove code for dealing with remapping queue blk-mq: reimplement blk_mq_hw_queue_mapped blk-mq: don't check queue mapped in __blk_mq_delay_run_hw_queue() ...
| * loop: fix LOOP_GET_STATUS lock imbalanceOmar Sandoval2018-04-101-15/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | Commit 2d1d4c1e591f made loop_get_status() drop lo_ctx_mutex before returning, but the loop_get_status_old(), loop_get_status64(), and loop_get_status_compat() wrappers don't call loop_get_status() if the passed argument is NULL. The callers expect that the lock is dropped, so make sure we drop it in that case, too. Reported-by: syzbot+31e8daa8b3fc129e75f2@syzkaller.appspotmail.com Fixes: 2d1d4c1e591f ("loop: don't call into filesystem while holding lo_ctl_mutex") Signed-off-by: Omar Sandoval <osandov@fb.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
| * block/loop: fix deadlock after loop_set_statusTetsuo Handa2018-04-101-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | syzbot is reporting deadlocks at __blkdev_get() [1]. ---------------------------------------- [ 92.493919] systemd-udevd D12696 525 1 0x00000000 [ 92.495891] Call Trace: [ 92.501560] schedule+0x23/0x80 [ 92.502923] schedule_preempt_disabled+0x5/0x10 [ 92.504645] __mutex_lock+0x416/0x9e0 [ 92.510760] __blkdev_get+0x73/0x4f0 [ 92.512220] blkdev_get+0x12e/0x390 [ 92.518151] do_dentry_open+0x1c3/0x2f0 [ 92.519815] path_openat+0x5d9/0xdc0 [ 92.521437] do_filp_open+0x7d/0xf0 [ 92.527365] do_sys_open+0x1b8/0x250 [ 92.528831] do_syscall_64+0x6e/0x270 [ 92.530341] entry_SYSCALL_64_after_hwframe+0x42/0xb7 [ 92.931922] 1 lock held by systemd-udevd/525: [ 92.933642] #0: 00000000a2849e25 (&bdev->bd_mutex){+.+.}, at: __blkdev_get+0x73/0x4f0 ---------------------------------------- The reason of deadlock turned out that wait_event_interruptible() in blk_queue_enter() got stuck with bdev->bd_mutex held at __blkdev_put() due to q->mq_freeze_depth == 1. ---------------------------------------- [ 92.787172] a.out S12584 634 633 0x80000002 [ 92.789120] Call Trace: [ 92.796693] schedule+0x23/0x80 [ 92.797994] blk_queue_enter+0x3cb/0x540 [ 92.803272] generic_make_request+0xf0/0x3d0 [ 92.807970] submit_bio+0x67/0x130 [ 92.810928] submit_bh_wbc+0x15e/0x190 [ 92.812461] __block_write_full_page+0x218/0x460 [ 92.815792] __writepage+0x11/0x50 [ 92.817209] write_cache_pages+0x1ae/0x3d0 [ 92.825585] generic_writepages+0x5a/0x90 [ 92.831865] do_writepages+0x43/0xd0 [ 92.836972] __filemap_fdatawrite_range+0xc1/0x100 [ 92.838788] filemap_write_and_wait+0x24/0x70 [ 92.840491] __blkdev_put+0x69/0x1e0 [ 92.841949] blkdev_close+0x16/0x20 [ 92.843418] __fput+0xda/0x1f0 [ 92.844740] task_work_run+0x87/0xb0 [ 92.846215] do_exit+0x2f5/0xba0 [ 92.850528] do_group_exit+0x34/0xb0 [ 92.852018] SyS_exit_group+0xb/0x10 [ 92.853449] do_syscall_64+0x6e/0x270 [ 92.854944] entry_SYSCALL_64_after_hwframe+0x42/0xb7 [ 92.943530] 1 lock held by a.out/634: [ 92.945105] #0: 00000000a2849e25 (&bdev->bd_mutex){+.+.}, at: __blkdev_put+0x3c/0x1e0 ---------------------------------------- The reason of q->mq_freeze_depth == 1 turned out that loop_set_status() forgot to call blk_mq_unfreeze_queue() at error paths for info->lo_encrypt_type != NULL case. ---------------------------------------- [ 37.509497] CPU: 2 PID: 634 Comm: a.out Tainted: G W 4.16.0+ #457 [ 37.513608] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 05/19/2017 [ 37.518832] RIP: 0010:blk_freeze_queue_start+0x17/0x40 [ 37.521778] RSP: 0018:ffffb0c2013e7c60 EFLAGS: 00010246 [ 37.524078] RAX: 0000000000000000 RBX: ffff8b07b1519798 RCX: 0000000000000000 [ 37.527015] RDX: 0000000000000002 RSI: ffffb0c2013e7cc0 RDI: ffff8b07b1519798 [ 37.529934] RBP: ffffb0c2013e7cc0 R08: 0000000000000008 R09: 47a189966239b898 [ 37.532684] R10: dad78b99b278552f R11: 9332dca72259d5ef R12: ffff8b07acd73678 [ 37.535452] R13: 0000000000004c04 R14: 0000000000000000 R15: ffff8b07b841e940 [ 37.538186] FS: 00007fede33b9740(0000) GS:ffff8b07b8e80000(0000) knlGS:0000000000000000 [ 37.541168] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 37.543590] CR2: 00000000206fdf18 CR3: 0000000130b30006 CR4: 00000000000606e0 [ 37.546410] Call Trace: [ 37.547902] blk_freeze_queue+0x9/0x30 [ 37.549968] loop_set_status+0x67/0x3c0 [loop] [ 37.549975] loop_set_status64+0x3b/0x70 [loop] [ 37.549986] lo_ioctl+0x223/0x810 [loop] [ 37.549995] blkdev_ioctl+0x572/0x980 [ 37.550003] block_ioctl+0x34/0x40 [ 37.550006] do_vfs_ioctl+0xa7/0x6d0 [ 37.550017] ksys_ioctl+0x6b/0x80 [ 37.573076] SyS_ioctl+0x5/0x10 [ 37.574831] do_syscall_64+0x6e/0x270 [ 37.576769] entry_SYSCALL_64_after_hwframe+0x42/0xb7 ---------------------------------------- [1] https://syzkaller.appspot.com/bug?id=cd662bc3f6022c0979d01a262c318fab2ee9b56f Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reported-by: syzbot <bot+48594378e9851eab70bcd6f99327c7db58c5a28a@syzkaller.appspotmail.com> Fixes: ecdd09597a572513 ("block/loop: fix race between I/O and set_status") Cc: Ming Lei <tom.leiming@gmail.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: stable <stable@vger.kernel.org> Cc: Jens Axboe <axboe@fb.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
* | Merge tag 'ceph-for-4.17-rc1' of git://github.com/ceph/ceph-clientLinus Torvalds2018-04-101-1472/+980Star
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull ceph updates from Ilya Dryomov: "The big ticket items are: - support for rbd "fancy" striping (myself). The striping feature bit is now fully implemented, allowing mapping v2 images with non-default striping patterns. This completes support for --image-format 2. - CephFS quota support (Luis Henriques and Zheng Yan). This set is based on the new SnapRealm code in the upcoming v13.y.z ("Mimic") release. Quota handling will be rejected on older filesystems. - memory usage improvements in CephFS (Chengguang Xu). Directory specific bits have been split out of ceph_file_info and some effort went into improving cap reservation code to avoid OOM crashes. Also included a bunch of assorted fixes all over the place from Chengguang and others" * tag 'ceph-for-4.17-rc1' of git://github.com/ceph/ceph-client: (67 commits) ceph: quota: report root dir quota usage in statfs ceph: quota: add counter for snaprealms with quota ceph: quota: cache inode pointer in ceph_snap_realm ceph: fix root quota realm check ceph: don't check quota for snap inode ceph: quota: update MDS when max_bytes is approaching ceph: quota: support for ceph.quota.max_bytes ceph: quota: don't allow cross-quota renames ceph: quota: support for ceph.quota.max_files ceph: quota: add initial infrastructure to support cephfs quotas rbd: remove VLA usage rbd: fix spelling mistake: "reregisteration" -> "reregistration" ceph: rename function drop_leases() to a more descriptive name ceph: fix invalid point dereference for error case in mdsc destroy ceph: return proper bool type to caller instead of pointer ceph: optimize memory usage ceph: optimize mds session register libceph, ceph: add __init attribution to init funcitons ceph: filter out used flags when printing unused open flags ceph: don't wait on writeback when there is no more dirty pages ...
| * | rbd: remove VLA usageKyle Spiers2018-04-021-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As part of the effort to remove VLAs from the kernel[1], this moves the literal values into the stack array calculation instead of using a variable for the sizing. The resulting size can be found from sizeof(buf). [1] https://lkml.org/lkml/2018/3/7/621 Signed-off-by: Kyle Spiers <kyle@spiers.me> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: fix spelling mistake: "reregisteration" -> "reregistration"Colin Ian King2018-04-021-1/+1
| | | | | | | | | | | | | | | | | | | | | Trivial fix to spelling mistake in rdb_warn message text. Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: get the latest osdmap when using an existing clientIlya Dryomov2018-04-021-36/+33Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently we request the latest osdmap only if ceph_pg_poolid_by_name() fails with -ENOENT. This is effective with newly created pools, but we also want to avoid attempting to map from pools that were recently deleted and report "pool does not exist" instead. (Such an attempt eventually fails in the OSD client after map check code kicks in, but the error message is confusing.) Request the latest osdmap unconditionally after bumping a ref on an existing client in rbd_client_find(). Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: move rbd_get_client() below rbd_put_client()Ilya Dryomov2018-04-021-20/+20
| | | | | | | | | | | | | | | | | | ... to avoid a forward declaration in the next commit. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove redundant declaration of rbd_spec_put()Ilya Dryomov2018-04-021-1/+0Star
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: allow "fancy" stripingIlya Dryomov2018-04-021-27/+2Star
| | | | | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Acked-by: Jason Dillaman <dillaman@redhat.com>
| * | rbd: introduce OWN_BVECS data typeIlya Dryomov2018-04-021-7/+149
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If the layout is "fancy", we need to be able to rearrange the provided bio_vecs in stripe unit chunks to make it possible for the messenger to read/write directly from/to the provided data buffer, without employing a temporary data buffer for assembling the result. Higher level bio_vec arrays are generally immutable, so this requires copying into a private array. Only the bio_vecs themselves are shuffled around, not the actual data. OWN_BVECS doesn't own any pages. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove rbd_parent_request_{create,destroy}()Ilya Dryomov2018-04-021-68/+6Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | rbd_parent_request_create() takes a ref on obj_req for child_img_req. There is no point in doing that because child_img_req is created on behalf of obj_req -- obj_req is the initiator and can't be completed before child_img_req. Open-code the rest of rbd_parent_request_create() and remove it along with rbd_parent_request_destroy(). Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: get rid of img_req->{offset,length}Ilya Dryomov2018-04-021-18/+8Star
| | | | | | | | | | | | | | | | | | These are set, but no longer used. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove rbd_img_request_fill() and helpersIlya Dryomov2018-04-021-98/+0Star
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: switch to common striping frameworkIlya Dryomov2018-04-021-23/+168
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: create+truncate for whole-object layered discardsIlya Dryomov2018-04-021-1/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A whole-object layered discard is implemented as a truncate rather than a delete: a dummy object is needed to prevent the CoW machinery from kicking in. However, a truncate on a non-existent object is a no-op. If the object doesn't exist in HEAD, a discard request is effectively ignored, which violates our "discard zeroes data" promise and breaks REQ_OP_WRITE_ZEROES implementation. A non-exclusive create on an existing object is also a no-op, so the fix is to do a compound create+truncate instead. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: move to obj_req->img_extentsIlya Dryomov2018-04-021-52/+98
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In preparation for rbd "fancy" striping, replace obj_req->img_offset with obj_req->img_extents. A single starting offset isn't sufficient because we want only one OSD request per object and will merge adjacent object extents in ceph_file_to_extents(). The final object extent may map into multiple different byte ranges in the image. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: incorporate ceph_object_extentIlya Dryomov2018-04-021-37/+34Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | obj_req->object_no -> obj_req->ex.oe_objno obj_req->offset -> obj_req->ex.oe_off obj_req->length -> obj_req->ex.oe_len ... and use ex for linking object requests to image requests. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: store data_type in img_req instead of obj_reqIlya Dryomov2018-04-021-26/+8Star
| | | | | | | | | | | | | | | | | | | | | All object requests are associated with an image request now -- avoid duplicating the same info in each object request. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove obj_req->flags fieldIlya Dryomov2018-04-021-35/+0Star
| | | | | | | | | | | | | | | | | | There are no standalone (!IMG_DATA) object requests anymore. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove old request completion codeIlya Dryomov2018-04-021-172/+3Star
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: new request completion codeIlya Dryomov2018-04-021-13/+55
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Do away with partial request completions and all the associated complexity. Individual object requests no longer need to be completed in order -- when the last one becomes ready, we complete the entire higher level request all at once. This also wraps up the conversion to a state machine model and eliminates the recursion described in commit 6d69bb536bac ("rbd: prevent kernel stack blow up on rbd map"). Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: update rbd_img_request_submit() signatureIlya Dryomov2018-04-021-10/+3Star
| | | | | | | | | | | | | | | | | | | | | | | | It should be void now. Also, object requests are unlinked only in image request destructor, which can't run before rbd_img_request_put(), so no need for _safe. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: add img_req->op_type fieldIlya Dryomov2018-04-021-63/+12Star
| | | | | | | | | | | | | | | | | | Store op_type in its own field instead of packing it into flags. Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: simplify rbd_osd_req_create()Ilya Dryomov2018-04-021-45/+14Star
| | | | | | | | | | | | | | | | | | | | | | | | No need to pass rbd_dev and op_type to rbd_osd_req_create(): there are no standalone (!IMG_DATA) object requests anymore and osd_req->r_flags can be set in rbd_osd_req_format_{read,write}(). Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: remove old request handling codeIlya Dryomov2018-04-021-730/+4Star
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: new request handling codeIlya Dryomov2018-04-021-77/+601
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The notable changes are: - instead of explicitly stat'ing the object to see if it exists before issuing the write, send the write optimistically along with the stat in a single OSD request - zero copyup optimization - all object requests are associated with an image request and have a valid ->img_request pointer; there are no standalone (!IMG_DATA) object requests anymore - code is structured as a state machine (vs a bunch of callbacks with implicit state) Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: move from raw pages to bvec data descriptorsIlya Dryomov2018-04-021-78/+77Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In preparation for rbd "fancy" striping which requires bio_vec arrays, wire up BVECS data type and kill off PAGES data type. There is nothing wrong with using page vectors for copyup requests -- it's just less iterator boilerplate code to write for the new striping framework. Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Alex Elder <elder@linaro.org>
| * | rbd: get rid of img_req->copyup_pagesIlya Dryomov2018-04-021-34/+9Star
| | | | | | | | | | | | | | | | | | | | | | | | The initiating object request is the proper owner -- save a bit of space. Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Alex Elder <elder@linaro.org>
| * | rbd: don't (ab)use obj_req->pages for stat requestsIlya Dryomov2018-04-021-10/+5Star
| | | | | | | | | | | | | | | | | | | | | | | | obj_req->pages is for provided data buffers. stat requests are internal and should be NODATA. Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Alex Elder <elder@linaro.org>
| * | rbd: remove bio cloning helpersIlya Dryomov2018-04-021-141/+0Star
| | | | | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Alex Elder <elder@linaro.org>
| * | libceph, rbd: new bio handling code (aka don't clone bios)Ilya Dryomov2018-04-021-27/+40
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The reason we clone bios is to be able to give each object request (and consequently each ceph_osd_data/ceph_msg_data item) its own pointer to a (list of) bio(s). The messenger then initializes its cursor with cloned bio's ->bi_iter, so it knows where to start reading from/writing to. That's all the cloned bios are used for: to determine each object request's starting position in the provided data buffer. Introduce ceph_bio_iter to do exactly that -- store position within bio list (i.e. pointer to bio) + position within that bio (i.e. bvec_iter). Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
| * | rbd: start enums at 1 instead of 0Ilya Dryomov2018-04-021-2/+4
| | | | | | | | | | | | Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2024-11-19 04:00:54 +0100