summaryrefslogtreecommitdiffstats
path: root/block
AgeCommit message (Collapse)AuthorFilesLines
2015-01-08blk-mq: End unstarted requests on a dying queueKeith Busch1-1/+10
Requests that haven't been started prior to a queue dying can be ended in error without waiting for them to start and time out. Signed-off-by: Keith Busch <keith.busch@intel.com> Added code comment to explain why this is done. Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-08blk-mq: Allow requests to never expireKeith Busch2-1/+6
Some types of requests may be started that are not gauranteed to ever complete. This adds a request flag that a driver can use so mark the request as such. Signed-off-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-08blk-mq: Add helper to abort requeued requestsJens Axboe1-0/+20
Adds a helper function a driver can use to abort requeued requests in case any are pending when h/w queues are being removed. Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-08blk-mq: Let drivers cancel requeue_workKeith Busch1-0/+6
Kicking requeued requests will start h/w queues in a work_queue, which may alter the driver's requested state to temporarily stop them. This patch exports a method to cancel the q->requeue_work so a driver can be assured stopped h/w queues won't be started up before it is ready. Signed-off-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-08blk-mq: Export if requests were startedKeith Busch1-0/+6
Drivers can iterate over all allocated request tags, but their callback needs a way to know if the driver started the request in the first place. Signed-off-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-08blk-mq: Wake tasks entering queue on dyingKeith Busch1-0/+7
When the queue is set to dying, wake up tasks that are waiting on frozen queue so they realize it is dying and abandon their request. Signed-off-by: Keith Busch <keith.busch@intel.com> Modified by me to add a code comment on the need for the wakeup. Signed-off-by: Jens Axboe <axboe@fb.com>
2015-01-07blk-mq: get rid of ->cmd_size in the hardware queueJens Axboe1-1/+0
We store it in the tag set, we don't need it in the hardware queue. While removing cmd_size, place ->queue_num further down to avoid a hole on 64-bit archs. It's not used in any fast paths, so we can safely move it. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-31block: wake up waiters when a queue is marked dyingJens Axboe5-5/+42
If it's dying, we can't expect new request to complete and come in an wake up other tasks waiting for requests. So after we have marked it as dying, wake up everybody currently waiting for a request. Once they wake, they will retry their allocation and fail appropriately due to the state of the queue. Tested-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-20blk-mq: Export freeze/unfreeze functionsKeith Busch1-2/+4
Let drivers prevent entering a queue that isn't available. Signed-off-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-20blk-mq: Exit queue on alloc failureKeith Busch1-1/+3
Fixes usage counter when a request could not be allocated. Signed-off-by: Keith Busch <keith.busch@intel.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-15Revert "blk-mq: Micro-optimize bt_get()"Jens Axboe1-1/+3
This reverts commit 52f7eb945f2ba62b324bb9ae16d945326a961dcf. The optimization is only really safe for a single queue, otherwise 'bs' and 'bt' can indeed change, and if we don't do a finish_wait() for each loop, we'll potentially change the wait structure and corrupt task wait list. Reported-by: Jan Kara <jack@suse.cz>
2014-12-13Merge branch 'for-3.19/core' of git://git.kernel.dk/linux-blockLinus Torvalds10-119/+197
Pull block driver core update from Jens Axboe: "This is the pull request for the core block IO changes for 3.19. Not a huge round this time, mostly lots of little good fixes: - Fix a bug in sysfs blktrace interface causing a NULL pointer dereference, when enabled/disabled through that API. From Arianna Avanzini. - Various updates/fixes/improvements for blk-mq: - A set of updates from Bart, mostly fixing buts in the tag handling. - Cleanup/code consolidation from Christoph. - Extend queue_rq API to be able to handle batching issues of IO requests. NVMe will utilize this shortly. From me. - A few tag and request handling updates from me. - Cleanup of the preempt handling for running queues from Paolo. - Prevent running of unmapped hardware queues from Ming Lei. - Move the kdump memory limiting check to be in the correct location, from Shaohua. - Initialize all software queues at init time from Takashi. This prevents a kobject warning when CPUs are brought online that weren't online when a queue was registered. - Single writeback fix for I_DIRTY clearing from Tejun. Queued with the core IO changes, since it's just a single fix. - Version X of the __bio_add_page() segment addition retry from Maurizio. Hope the Xth time is the charm. - Documentation fixup for IO scheduler merging from Jan. - Introduce (and use) generic IO stat accounting helpers for non-rq drivers, from Gu Zheng. - Kill off artificial limiting of max sectors in a request from Christoph" * 'for-3.19/core' of git://git.kernel.dk/linux-block: (26 commits) bio: modify __bio_add_page() to accept pages that don't start a new segment blk-mq: Fix uninitialized kobject at CPU hotplugging blktrace: don't let the sysfs interface remove trace from running list blk-mq: Use all available hardware queues blk-mq: Micro-optimize bt_get() blk-mq: Fix a race between bt_clear_tag() and bt_get() blk-mq: Avoid that __bt_get_word() wraps multiple times blk-mq: Fix a use-after-free blk-mq: prevent unmapped hw queue from being scheduled blk-mq: re-check for available tags after running the hardware queue blk-mq: fix hang in bt_get() blk-mq: move the kdump check to blk_mq_alloc_tag_set blk-mq: cleanup tag free handling blk-mq: use 'nr_cpu_ids' as highest CPU ID count for hwq <-> cpu map blk: introduce generic io stat accounting help function blk-mq: handle the single queue case in blk_mq_hctx_next_cpu genhd: check for int overflow in disk_expand_part_tbl() blk-mq: add blk_mq_free_hctx_request() blk-mq: export blk_mq_free_request() blk-mq: use get_cpu/put_cpu instead of preempt_disable/preempt_enable ...
2014-12-11bio: modify __bio_add_page() to accept pages that don't start a new segmentMaurizio Lombardi1-24/+30
The original behaviour is to refuse to add a new page if the maximum number of segments has been reached, regardless of the fact the page we are going to add can be merged into the last segment or not. Unfortunately, when the system runs under heavy memory fragmentation conditions, a driver may try to add multiple pages to the last segment. The original code won't accept them and EBUSY will be reported to userspace. This patch modifies the function so it refuses to add a page only in case the latter starts a new segment and the maximum number of segments has already been reached. The bug can be easily reproduced with the st driver: 1) set CONFIG_SCSI_MPT2SAS_MAX_SGE or CONFIG_SCSI_MPT3SAS_MAX_SGE to 16 2) modprobe st buffer_kbs=1024 3) #dd if=/dev/zero of=/dev/st0 bs=1M count=10 dd: error writing `/dev/st0': Device or resource busy Signed-off-by: Maurizio Lombardi <mlombard@redhat.com> Signed-off-by: Ming Lei <ming.lei@canonical.com> Cc: Jet Chen <jet.chen@intel.com> Cc: Tomas Henzl <thenzl@redhat.com> Cc: Jens Axboe <axboe@kernel.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-10Merge tag 'pm+acpi-3.19-rc1' of ↵Linus Torvalds2-4/+4
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm Pull ACPI and power management updates from Rafael Wysocki: "This time we have some more new material than we used to have during the last couple of development cycles. The most important part of it to me is the introduction of a unified interface for accessing device properties provided by platform firmware. It works with Device Trees and ACPI in a uniform way and drivers using it need not worry about where the properties come from as long as the platform firmware (either DT or ACPI) makes them available. It covers both devices and "bare" device node objects without struct device representation as that turns out to be necessary in some cases. This has been in the works for quite a few months (and development cycles) and has been approved by all of the relevant maintainers. On top of that, some drivers are switched over to the new interface (at25, leds-gpio, gpio_keys_polled) and some additional changes are made to the core GPIO subsystem to allow device drivers to manipulate GPIOs in the "canonical" way on platforms that provide GPIO information in their ACPI tables, but don't assign names to GPIO lines (in which case the driver needs to do that on the basis of what it knows about the device in question). That also has been approved by the GPIO core maintainers and the rfkill driver is now going to use it. Second is support for hardware P-states in the intel_pstate driver. It uses CPUID to detect whether or not the feature is supported by the processor in which case it will be enabled by default. However, it can be disabled entirely from the kernel command line if necessary. Next is support for a platform firmware interface based on ACPI operation regions used by the PMIC (Power Management Integrated Circuit) chips on the Intel Baytrail-T and Baytrail-T-CR platforms. That interface is used for manipulating power resources and for thermal management: sensor temperature reporting, trip point setting and so on. Also the ACPI core is now going to support the _DEP configuration information in a limited way. Basically, _DEP it supposed to reflect off-the-hierarchy dependencies between devices which may be very indirect, like when AML for one device accesses locations in an operation region handled by another device's driver (usually, the device depended on this way is a serial bus or GPIO controller). The support added this time is sufficient to make the ACPI battery driver work on Asus T100A, but it is general enough to be able to cover some other use cases in the future. Finally, we have a new cpufreq driver for the Loongson1B processor. In addition to the above, there are fixes and cleanups all over the place as usual and a traditional ACPICA update to a recent upstream release. As far as the fixes go, the ACPI LPSS (Low-power Subsystem) driver for Intel platforms should be able to handle power management of the DMA engine correctly, the cpufreq-dt driver should interact with the thermal subsystem in a better way and the ACPI backlight driver should handle some more corner cases, among other things. On top of the ACPICA update there are fixes for race conditions in the ACPICA's interrupt handling code which might lead to some random and strange looking failures on some systems. In the cleanups department the most visible part is the series of commits targeted at getting rid of the CONFIG_PM_RUNTIME configuration option. That was triggered by a discussion regarding the generic power domains code during which we realized that trying to support certain combinations of PM config options was painful and not really worth it, because nobody would use them in production anyway. For this reason, we decided to make CONFIG_PM_SLEEP select CONFIG_PM_RUNTIME and that lead to the conclusion that the latter became redundant and CONFIG_PM could be used instead of it. The material here makes that replacement in a major part of the tree, but there will be at least one more batch of that in the second part of the merge window. Specifics: - Support for retrieving device properties information from ACPI _DSD device configuration objects and a unified device properties interface for device drivers (and subsystems) on top of that. As stated above, this works with Device Trees and ACPI and allows device drivers to be written in a platform firmware (DT or ACPI) agnostic way. The at25, leds-gpio and gpio_keys_polled drivers are now going to use this new interface and the GPIO subsystem is additionally modified to allow device drivers to assign names to GPIO resources returned by ACPI _CRS objects (in case _DSD is not present or does not provide the expected data). The changes in this set are mostly from Mika Westerberg, Rafael J Wysocki, Aaron Lu, and Darren Hart with some fixes from others (Fabio Estevam, Geert Uytterhoeven). - Support for Hardware Managed Performance States (HWP) as described in Volume 3, section 14.4, of the Intel SDM in the intel_pstate driver. CPUID is used to detect whether or not the feature is supported by the processor. If supported, it will be enabled automatically unless the intel_pstate=no_hwp switch is present in the kernel command line. From Dirk Brandewie. - New Intel Broadwell-H ID for intel_pstate (Dirk Brandewie). - Support for firmware interface based on ACPI operation regions used by the PMIC chips on the Intel Baytrail-T and Baytrail-T-CR platforms for power resource control and thermal management (Aaron Lu). - Limited support for retrieving off-the-hierarchy dependencies between devices from ACPI _DEP device configuration objects and deferred probing support for the ACPI battery driver based on the _DEP information to make that driver work on Asus T100A (Lan Tianyu). - New cpufreq driver for the Loongson1B processor (Kelvin Cheung). - ACPICA update to upstream revision 20141107 which only affects tools (Bob Moore). - Fixes for race conditions in the ACPICA's interrupt handling code and in the ACPI code related to system suspend and resume (Lv Zheng and Rafael J Wysocki). - ACPI core fix for an RCU-related issue in the ioremap() regions management code that slowed down significantly after CPUs had been allowed to enter idle states even if they'd had RCU callbakcs queued and triggered some problems in certain proprietary graphics driver (and elsewhere). The fix replaces synchronize_rcu() in that code with synchronize_rcu_expedited() which makes the issue go away. From Konstantin Khlebnikov. - ACPI LPSS (Low-Power Subsystem) driver fix to handle power management of the DMA engine included into the LPSS correctly. The problem is that the DMA engine doesn't have ACPI PM support of its own and it simply is turned off when the last LPSS device having ACPI PM support goes into D3cold. To work around that, the PM domain used by the ACPI LPSS driver is redesigned so at least one device with ACPI PM support will be on as long as the DMA engine is in use. From Andy Shevchenko. - ACPI backlight driver fix to avoid using it on "Win8-compatible" systems where it doesn't work and where it was used by default by mistake (Aaron Lu). - Assorted minor ACPI core fixes and cleanups from Tomasz Nowicki, Sudeep Holla, Huang Rui, Hanjun Guo, Fabian Frederick, and Ashwin Chaugule (mostly related to the upcoming ARM64 support). - Intel RAPL (Running Average Power Limit) power capping driver fixes and improvements including new processor IDs (Jacob Pan). - Generic power domains modification to power up domains after attaching devices to them to meet the expectations of device drivers and bus types assuming devices to be accessible at probe time (Ulf Hansson). - Preliminary support for controlling device clocks from the generic power domains core code and modifications of the ARM/shmobile platform to use that feature (Ulf Hansson). - Assorted minor fixes and cleanups of the generic power domains core code (Ulf Hansson, Geert Uytterhoeven). - Assorted minor fixes and cleanups of the device clocks control code in the PM core (Geert Uytterhoeven, Grygorii Strashko). - Consolidation of device power management Kconfig options by making CONFIG_PM_SLEEP select CONFIG_PM_RUNTIME and removing the latter which is now redundant (Rafael J Wysocki and Kevin Hilman). That is the first batch of the changes needed for this purpose. - Core device runtime power management support code cleanup related to the execution of callbacks (Andrzej Hajda). - cpuidle ARM support improvements (Lorenzo Pieralisi). - cpuidle cleanup related to the CPUIDLE_FLAG_TIME_VALID flag and a new MAINTAINERS entry for ARM Exynos cpuidle (Daniel Lezcano and Bartlomiej Zolnierkiewicz). - New cpufreq driver callback (->ready) to be executed when the cpufreq core is ready to use a given policy object and cpufreq-dt driver modification to use that callback for cooling device registration (Viresh Kumar). - cpufreq core fixes and cleanups (Viresh Kumar, Vince Hsu, James Geboski, Tomeu Vizoso). - Assorted fixes and cleanups in the cpufreq-pcc, intel_pstate, cpufreq-dt, pxa2xx cpufreq drivers (Lenny Szubowicz, Ethan Zhao, Stefan Wahren, Petr Cvek). - OPP (Operating Performance Points) framework modification to allow OPPs to be removed too and update of a few cpufreq drivers (cpufreq-dt, exynos5440, imx6q, cpufreq) to remove OPPs (added during initialization) on driver removal (Viresh Kumar). - Hibernation core fixes and cleanups (Tina Ruchandani and Markus Elfring). - PM Kconfig fix related to CPU power management (Pankaj Dubey). - cpupower tool fix (Prarit Bhargava)" * tag 'pm+acpi-3.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (120 commits) i2c-omap / PM: Drop CONFIG_PM_RUNTIME from i2c-omap.c dmaengine / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM tools: cpupower: fix return checks for sysfs_get_idlestate_count() drivers: sh / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM e1000e / igb / PM: Eliminate CONFIG_PM_RUNTIME MMC / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM MFD / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM misc / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM media / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM input / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM leds: leds-gpio: Fix multiple instances registration without 'label' property iio / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM hsi / OMAP / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM i2c-hid / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM drm / exynos / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM gpio / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM hwrandom / exynos / PM: Use CONFIG_PM in #ifdef block / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM USB / PM: Drop CONFIG_PM_RUNTIME from the USB core PM: Merge the SET*_RUNTIME_PM_OPS() macros ...
2014-12-10blk-mq: Fix uninitialized kobject at CPU hotpluggingTakashi Iwai1-5/+4
When a CPU is hotplugged, the current blk-mq spews a warning like: kobject '(null)' (ffffe8ffffc8b5d8): tried to add an uninitialized object, something is seriously wrong. CPU: 1 PID: 1386 Comm: systemd-udevd Not tainted 3.18.0-rc7-2.g088d59b-default #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_171129-lamiak 04/01/2014 0000000000000000 0000000000000002 ffffffff81605f07 ffffe8ffffc8b5d8 ffffffff8132c7a0 ffff88023341d370 0000000000000020 ffff8800bb05bd58 ffff8800bb05bd08 000000000000a0a0 000000003f441940 0000000000000007 Call Trace: [<ffffffff81005306>] dump_trace+0x86/0x330 [<ffffffff81005644>] show_stack_log_lvl+0x94/0x170 [<ffffffff81006d21>] show_stack+0x21/0x50 [<ffffffff81605f07>] dump_stack+0x41/0x51 [<ffffffff8132c7a0>] kobject_add+0xa0/0xb0 [<ffffffff8130aee1>] blk_mq_register_hctx+0x91/0xb0 [<ffffffff8130b82e>] blk_mq_sysfs_register+0x3e/0x60 [<ffffffff81309298>] blk_mq_queue_reinit_notify+0xf8/0x190 [<ffffffff8107cfdc>] notifier_call_chain+0x4c/0x70 [<ffffffff8105fd23>] cpu_notify+0x23/0x50 [<ffffffff81060037>] _cpu_up+0x157/0x170 [<ffffffff810600d9>] cpu_up+0x89/0xb0 [<ffffffff815fa5b5>] cpu_subsys_online+0x35/0x80 [<ffffffff814323cd>] device_online+0x5d/0xa0 [<ffffffff81432485>] online_store+0x75/0x80 [<ffffffff81236a5a>] kernfs_fop_write+0xda/0x150 [<ffffffff811c5532>] vfs_write+0xb2/0x1f0 [<ffffffff811c5f42>] SyS_write+0x42/0xb0 [<ffffffff8160c4ed>] system_call_fastpath+0x16/0x1b [<00007f0132fb24e0>] 0x7f0132fb24e0 This is indeed because of an uninitialized kobject for blk_mq_ctx. The blk_mq_ctx kobjects are initialized in blk_mq_sysfs_init(), but it goes loop over hctx_for_each_ctx(), i.e. it initializes only for online CPUs. Thus, when a CPU is hotplugged, the ctx for the newly onlined CPU is registered without initialization. This patch fixes the issue by initializing the all ctx kobjects belonging to each queue. Bugzilla: https://bugzilla.novell.com/show_bug.cgi?id=908794 Cc: <stable@vger.kernel.org> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-09blk-mq: Use all available hardware queuesBart Van Assche1-1/+1
Suppose that a system has two CPU sockets, three cores per socket, that it does not support hyperthreading and that four hardware queues are provided by a block driver. With the current algorithm this will lead to the following assignment of CPU cores to hardware queues: HWQ 0: 0 1 HWQ 1: 2 3 HWQ 2: 4 5 HWQ 3: (none) This patch changes the queue assignment into: HWQ 0: 0 1 HWQ 1: 2 HWQ 2: 3 4 HWQ 3: 5 In other words, this patch has the following three effects: - All four hardware queues are used instead of only three. - CPU cores are spread more evenly over hardware queues. For the above example the range of the number of CPU cores associated with a single HWQ is reduced from [0..2] to [1..2]. - If the number of HWQ's is a multiple of the number of CPU sockets it is now guaranteed that all CPU cores associated with a single HWQ reside on the same CPU socket. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Sagi Grimberg <sagig@mellanox.com> Cc: Jens Axboe <axboe@fb.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Ming Lei <ming.lei@canonical.com> Cc: Alexander Gordeev <agordeev@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-09blk-mq: Micro-optimize bt_get()Bart Van Assche1-3/+1
Remove a superfluous finish_wait() call. Convert the two bt_wait_ptr() calls into a single call. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Robert Elliott <elliott@hp.com> Cc: Ming Lei <ming.lei@canonical.com> Cc: Alexander Gordeev <agordeev@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-09blk-mq: Fix a race between bt_clear_tag() and bt_get()Bart Van Assche1-6/+5
What we need is the following two guarantees: * Any thread that observes the effect of the test_and_set_bit() by __bt_get_word() also observes the preceding addition of 'current' to the appropriate wait list. This is guaranteed by the semantics of the spin_unlock() operation performed by prepare_and_wait(). Hence the conversion of test_and_set_bit_lock() into test_and_set_bit(). * The wait lists are examined by bt_clear() after the tag bit has been cleared. clear_bit_unlock() guarantees that any thread that observes that the bit has been cleared also observes the store operations preceding clear_bit_unlock(). However, clear_bit_unlock() does not prevent that the wait lists are examined before that the tag bit is cleared. Hence the addition of a memory barrier between clear_bit() and the wait list examination. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Robert Elliott <elliott@hp.com> Cc: Ming Lei <ming.lei@canonical.com> Cc: Alexander Gordeev <agordeev@redhat.com> Cc: <stable@vger.kernel.org> # v3.13+ Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-09blk-mq: Avoid that __bt_get_word() wraps multiple timesBart Van Assche1-2/+4
If __bt_get_word() is called with last_tag != 0, if the first find_next_zero_bit() fails, if after wrap-around the test_and_set_bit() call fails and find_next_zero_bit() succeeds, if the next test_and_set_bit() call fails and subsequently find_next_zero_bit() does not find a zero bit, then another wrap-around will occur. Avoid this by introducing an additional local variable. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Robert Elliott <elliott@hp.com> Cc: Ming Lei <ming.lei@canonical.com> Cc: Alexander Gordeev <agordeev@redhat.com> Cc: <stable@vger.kernel.org> # v3.13+ Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-09blk-mq: Fix a use-after-freeBart Van Assche2-8/+7
blk-mq users are allowed to free the memory request_queue.tag_set points at after blk_cleanup_queue() has finished but before blk_release_queue() has started. This can happen e.g. in the SCSI core. The SCSI core namely embeds the tag_set structure in a SCSI host structure. The SCSI host structure is freed by scsi_host_dev_release(). This function is called after blk_cleanup_queue() finished but can be called before blk_release_queue(). This means that it is not safe to access request_queue.tag_set from inside blk_release_queue(). Hence remove the blk_sync_queue() call from blk_release_queue(). This call is not necessary - outstanding requests must have finished before blk_release_queue() is called. Additionally, move the blk_mq_free_queue() call from blk_release_queue() to blk_cleanup_queue() to avoid that struct request_queue.tag_set gets accessed after it has been freed. This patch avoids that the following kernel oops can be triggered when deleting a SCSI host for which scsi-mq was enabled: Call Trace: [<ffffffff8109a7c4>] lock_acquire+0xc4/0x270 [<ffffffff814ce111>] mutex_lock_nested+0x61/0x380 [<ffffffff812575f0>] blk_mq_free_queue+0x30/0x180 [<ffffffff8124d654>] blk_release_queue+0x84/0xd0 [<ffffffff8126c29b>] kobject_cleanup+0x7b/0x1a0 [<ffffffff8126c140>] kobject_put+0x30/0x70 [<ffffffff81245895>] blk_put_queue+0x15/0x20 [<ffffffff8125c409>] disk_release+0x99/0xd0 [<ffffffff8133d056>] device_release+0x36/0xb0 [<ffffffff8126c29b>] kobject_cleanup+0x7b/0x1a0 [<ffffffff8126c140>] kobject_put+0x30/0x70 [<ffffffff8125a78a>] put_disk+0x1a/0x20 [<ffffffff811d4cb5>] __blkdev_put+0x135/0x1b0 [<ffffffff811d56a0>] blkdev_put+0x50/0x160 [<ffffffff81199eb4>] kill_block_super+0x44/0x70 [<ffffffff8119a2a4>] deactivate_locked_super+0x44/0x60 [<ffffffff8119a87e>] deactivate_super+0x4e/0x70 [<ffffffff811b9833>] cleanup_mnt+0x43/0x90 [<ffffffff811b98d2>] __cleanup_mnt+0x12/0x20 [<ffffffff8107252c>] task_work_run+0xac/0xe0 [<ffffffff81002c01>] do_notify_resume+0x61/0xa0 [<ffffffff814d2c58>] int_signal+0x12/0x17 Signed-off-by: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Robert Elliott <elliott@hp.com> Cc: Ming Lei <ming.lei@canonical.com> Cc: Alexander Gordeev <agordeev@redhat.com> Cc: <stable@vger.kernel.org> # v3.13+ Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-08Merge tag 'scsi-for-linus' of ↵Linus Torvalds4-3/+33
git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi Pull SCSI updates from James Bottomley: "This patch is the usual mix of driver updates (srp, ipr, scsi_debug, NCR5380, fnic, 53c974, ses, wd719x, hpsa, megaraid_sas). Of those, wd7a9x is new and 53c974 is a rewrite of the old tmscsim driver and the extensive work by Finn Thain rewrites all the NCR5380 based drivers. There's also extensive infrastructure updates: a new logging infrastructure for sense information and a rewrite of the tagged command queue API and an assortment of minor updates" * tag 'scsi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (183 commits) scsi: set fmt to NULL scsi_extd_sense_format() by default libsas: remove task_collector mode wd719x: remove dma_cache_sync call scsi_debug: add Report supported opcodes+tmfs; Compare and write scsi_debug: change SCSI command parser to table driven scsi_debug: add Capacity Changed Unit Attention scsi_debug: append inject error flags onto scsi_cmnd object scsi_debug: pinpoint invalid field in sense data wd719x: Add firmware documentation wd719x: Introduce Western Digital WD7193/7197/7296 PCI SCSI card driver eeprom-93cx6: Add (read-only) support for 8-bit mode esas2r: fix an oversight in setting return value esas2r: fix an error path in esas2r_ioctl_handler esas2r: fir error handling in do_fm_api scsi: add SPC-3 command definitions scsi: rename SERVICE_ACTION_IN to SERVICE_ACTION_IN_16 scsi: remove scsi_driver owner field scsi: move scsi_dispatch_cmd to scsi_lib.c scsi: stop passing a gfp_mask argument down the command setup path scsi: remove scsi_next_command ...
2014-12-08blk-mq: prevent unmapped hw queue from being scheduledMing Lei2-2/+11
When one hardware queue has no mapped software queues, it shouldn't have been scheduled. Otherwise WARNING or OOPS can triggered. blk_mq_hw_queue_mapped() helper is introduce for fixing the problem. Signed-off-by: Ming Lei <ming.lei@canonical.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-08Merge branch 'pm-runtime'Rafael J. Wysocki2-4/+4
* pm-runtime: (25 commits) i2c-omap / PM: Drop CONFIG_PM_RUNTIME from i2c-omap.c dmaengine / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM drivers: sh / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM e1000e / igb / PM: Eliminate CONFIG_PM_RUNTIME MMC / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM MFD / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM misc / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM media / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM input / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM iio / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM hsi / OMAP / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM i2c-hid / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM drm / exynos / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM gpio / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM hwrandom / exynos / PM: Use CONFIG_PM in #ifdef block / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PM USB / PM: Drop CONFIG_PM_RUNTIME from the USB core PM: Merge the SET*_RUNTIME_PM_OPS() macros PM / Kconfig: Do not select PM directly from Kconfig files PCI / PM: Drop CONFIG_PM_RUNTIME from the PCI core ...
2014-12-08blk-mq: re-check for available tags after running the hardware queueJens Axboe1-0/+8
If we run out of tags and have to sleep, we run the hardware queue to kick pending IO into gear. During that run, we may have completed requests, so re-check if we have free tags before going to sleep. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-08blk-mq: fix hang in bt_get()Bart Van Assche1-0/+7
Avoid that if there are fewer hardware queues than CPU threads that bt_get() can hang. The symptoms of the hang were as follows: * All tags allocated for a particular hardware queue. * (nr_tags) pending commands for that hardware queue. * No pending commands for the software queues associated with that hardware queue. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-12-08Merge remote-tracking branch 'scsi-queue/core-for-3.19' into for-linusJames Bottomley4-3/+33
2014-12-04block / PM: Replace CONFIG_PM_RUNTIME with CONFIG_PMRafael J. Wysocki2-4/+4
After commit b2b49ccbdd54 (PM: Kconfig: Set PM_RUNTIME if PM_SLEEP is selected) PM_RUNTIME is always set if PM is set, so #ifdef blocks depending on CONFIG_PM_RUNTIME may now be changed to depend on CONFIG_PM. Replace CONFIG_PM_RUNTIME with CONFIG_PM in the block device core. Reviewed-by: Aaron Lu <aaron.lu@intel.com> Acked-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2014-12-02block: fix regression where bio_integrity_process uses wrong bio_vec iteratorDarrick J. Wong1-6/+7
bio integrity handling is broken on a system with LVM layered atop a DIF/DIX SCSI drive because device mapper clones the bio, modifies the clone, and sends the clone to the lower layers for processing. However, the clone bio has bi_vcnt == 0, which means that when the sd driver calls bio_integrity_process to attach DIX data, the for_each_segment_all() call (which uses bi_vcnt) returns immediately and random garbage is sent to the disk on a disk write. The disk of course returns an error. Therefore, teach bio_integrity_process() to use bio_for_each_segment() to iterate the bio_vecs, since the per-bio iterator tracks which bio_vecs are associated with that particular bio. The integrity handling code is effectively part of the "driver" (it's not the bio owner), so it must use the correct iterator function. v2: Fix a compiler warning about abandoned local variables. This patch supersedes "block: bio_integrity_process uses wrong bio_vec iterator". Patch applies against 3.18-rc6. Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Acked-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-30blk-mq: move the kdump check to blk_mq_alloc_tag_setShaohua Li1-10/+10
We call blk_mq_alloc_tag_set() first then blk_mq_init_queue(). The requests are allocated in the former function. So the kdump check should be moved to there to really save memory. Signed-off-by: Shaohua Li <shli@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-24blk-mq: cleanup tag free handlingJens Axboe1-18/+6
We only call __blk_mq_put_tag() and __blk_mq_put_reserved_tag() from blk_mq_put_tag(), so just inline the two calls instead of having them as separate functions. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-24blk-mq: use 'nr_cpu_ids' as highest CPU ID count for hwq <-> cpu mapJens Axboe1-1/+1
We currently use num_possible_cpus(), but that breaks on sparc64 where the CPU ID space is discontig. Use nr_cpu_ids as the highest CPU ID instead, so we don't end up reading from invalid memory. Cc: stable@kernel.org # 3.13+ Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-24scsi: rename SERVICE_ACTION_IN to SERVICE_ACTION_IN_16Hannes Reinecke1-1/+1
SPC-3 defines SERVICE ACTION IN(12) and SERVICE ACTION IN(16). So rename SERVICE_ACTION_IN to SERVICE_ACTION_IN_16 to be consistent with SPC and to allow for better distinction. Signed-off-by: Hannes Reinecke <hare@suse.de> Tested-by: Robert Elliott <elliott@hp.com> Reviewed-by: Robert Elliott <elliott@hp.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-11-24blk: introduce generic io stat accounting help functionGu Zheng1-0/+28
Many block drivers accounting io stat based on bio (e.g. NVMe...), the blk_account_io_start/end() which is based on request does not make sense to them, so here we introduce the similar help function named generic_start/end_io_acct base on raw sectors, and it can simplify some driver's open io accounting code. Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-24blk-mq: handle the single queue case in blk_mq_hctx_next_cpuChristoph Hellwig1-21/+10
Don't duplicate the code to handle the not cpu bounce case in the caller, do it inside blk_mq_hctx_next_cpu instead. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-19genhd: check for int overflow in disk_expand_part_tbl()Jens Axboe1-2/+9
We can get here from blkdev_ioctl() -> blkpg_ioctl() -> add_partition() with a user passed in partno value. If we pass in 0x7fffffff, the new target in disk_expand_part_tbl() overflows the 'int' and we access beyond the end of ptbl->part[] and even write to it when we do the rcu_assign_pointer() to assign the new partition. Reported-by: David Ramos <daramos@stanford.edu> Cc: stable@kernel.org Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-17blk-mq: add blk_mq_free_hctx_request()Jens Axboe1-5/+12
It's silly to use blk_mq_free_request() which in turn maps the request to the hardware queue, for places where we already know what the hardware queue is. This saves us an extra mapping of a hardware queue on request completion, if the caller knows this information already. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-17blk-mq: export blk_mq_free_request()Jens Axboe1-0/+1
Drivers that know they are blk-mq should just use this function instead of calling through blk_put_request(). Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-12scsi: add new scsi-command flag for tagged commandsChristoph Hellwig1-2/+2
Currently scsi piggy backs on the block layer to define the concept of a tagged command. But we want to be able to have block-level host-wide tags assigned even for untagged commands like the initial INQUIRY, so add a new SCSI-level flag for commands that are tagged at the scsi level, so that even commands without that set can have tags assigned to them. Note that this alredy is the case for the blk-mq code path, and this just lets the old path catch up with it. We also set this flag based upon sdev->simple_tags instead of the block queue flag, so that it is entirely independent of the block layer tagging, and thus always correct even if a driver doesn't use block level tagging yet. Also remove the old blk_rq_tagged; it was only used by SCSI drivers, and removing it forces them to look for the proper replacement. Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.de>
2014-11-12blk-mq: add blk_mq_unique_tag()Bart Van Assche2-0/+30
The queuecommand() callback functions in SCSI low-level drivers need to know which hardware context has been selected by the block layer. Since this information is not available in the request structure, and since passing the hctx pointer directly to the queuecommand callback function would require modification of all SCSI LLDs, add a function to the block layer that allows to query the hardware context index. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: Sagi Grimberg <sagig@mellanox.com> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-11-11block: blk-merge: fix blk_recount_segments()Ming Lei1-8/+11
For cloned bio, bio->bi_vcnt can't be used at all, and we have resort to bio_segments() to figure out how many segment there are in the bio. Signed-off-by: Ming Lei <tom.leiming@gmail.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-11blk-mq: use get_cpu/put_cpu instead of preempt_disable/preempt_enablePaolo Bonzini1-4/+4
blk-mq is using preempt_disable/enable in order to ensure that the queue runners are placed on the right CPU. This does not work with the RT patches, because __blk_mq_run_hw_queue takes a non-raw spinlock with the preemption-disabled region. If there is contention on the lock, this violates the rules for preemption-disabled regions. While this should be easily fixable within the RT patches just by doing migrate_disable/enable, we can do better and document _why_ this particular region runs with disabled preemption. After the previous patch, it is trivial to switch it to get/put_cpu; the RT patches then can change it to get_cpu_light, which lets virtio-blk run under RT kernels. Cc: Jens Axboe <axboe@kernel.dk> Cc: Thomas Gleixner <tglx@linutronix.de> Reported-by: Clark Williams <williams@redhat.com> Tested-by: Clark Williams <williams@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-11blk_mq: call preempt_disable/enable in blk_mq_run_hw_queue, and only if neededPaolo Bonzini1-9/+12
preempt_disable/enable surrounds every call to blk_mq_run_hw_queue, except the one in blk-flush.c. In fact that one is always asynchronous, and it does not need smp_processor_id(). We can do the same for all other calls, avoiding preempt_disable when async is true. This avoids peppering blk-mq.c with preemption-disabled regions. Cc: Jens Axboe <axboe@kernel.dk> Cc: Thomas Gleixner <tglx@linutronix.de> Reported-by: Clark Williams <williams@redhat.com> Tested-by: Clark Williams <williams@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-10scsi: Fix more error handling in SCSI_IOCTL_SEND_COMMANDTony Battersby1-3/+5
Fix an error path in SCSI_IOCTL_SEND_COMMAND that calls blk_put_request(rq) on an invalid IS_ERR(rq) pointer. Fixes: a492f075450f ("block,scsi: fixup blk_get_request dead queue scenarios") Signed-off-by: Tony Battersby <tonyb@cybernetics.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-11-04blk-mq: make mq_queue_reinit_notify() freeze queues in parallelTejun Heo1-8/+33
q->mq_usage_counter is a percpu_ref which is killed and drained when the queue is frozen. On a CPU hotplug event, blk_mq_queue_reinit() which involves freezing the queue is invoked on all existing queues. Because percpu_ref killing and draining involve a RCU grace period, doing the above on one queue after another may take a long time if there are many queues on the system. This patch splits out initiation of freezing and waiting for its completion, and updates blk_mq_queue_reinit_notify() so that the queues are frozen in parallel instead of one after another. Note that freezing and unfreezing are moved from blk_mq_queue_reinit() to blk_mq_queue_reinit_notify(). Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-31block: Fix computation of merged request priorityJan Kara1-6/+8
Priority of a merged request is computed by ioprio_best(). If one of the requests has undefined priority (IOPRIO_CLASS_NONE) and another request has priority from IOPRIO_CLASS_BE, the function will return the undefined priority which is wrong. Fix the function to properly return priority of a request with the defined priority. Fixes: d58cdfb89ce0c6bd5f81ae931a984ef298dbda20 CC: stable@vger.kernel.org Signed-off-by: Jan Kara <jack@suse.cz> Reviewed-by: Jeff Moyer <jmoyer@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-29blk-mq: add BLK_MQ_F_DEFER_ISSUE support flagJens Axboe1-1/+6
Drivers can now tell blk-mq if they take advantage of the deferred issue through 'last' or not. If they do, don't do queue-direct for sync IO. This is a preparation patch for the nvme conversion. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-29blk-mq: add a 'list' parameter to ->queue_rq()Jens Axboe1-2/+27
Since we have the notion of a 'last' request in a chain, we can use this to have the hardware optimize the issuing of requests. Add a list_head parameter to queue_rq that the driver can use to temporarily store hw commands for issue when 'last' is true. If we are doing a chain of requests, pass in a NULL list for the first request to force issue of that immediately, then batch the remainder for deferred issue until the last request has been sent. Instead of adding yet another argument to the hot ->queue_rq path, encapsulate the passed arguments in a blk_mq_queue_data structure. This is passed as a constant, and has been tested as faster than passing 4 (or even 3) args through ->queue_rq. Update drivers for the new ->queue_rq() prototype. There are no functional changes in this patch for drivers - if they don't use the passed in list, then they will just queue requests individually like before. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-23block: fix wrong error return in elevator_init()Sudip Mukherjee1-1/+3
while compiling integer err was showing as a set but unused variable. elevator_init_fn can be either cfq_init_queue or deadline_init_queue or noop_init_queue. all three of these functions are returning -ENOMEM if they fail to allocate the queue. so we should actually be returning the error code rather than returning 0 always. Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org> Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-22scsi: Fix error handling in SCSI_IOCTL_SEND_COMMANDJan Kara1-2/+1
When sg_scsi_ioctl() fails to prepare request to submit in blk_rq_map_kern() we jump to a label where we just end up copying (luckily zeroed-out) kernel buffer to userspace instead of reporting error. Fix the problem by jumping to the right label. CC: Jens Axboe <axboe@kernel.dk> CC: linux-scsi@vger.kernel.org CC: stable@vger.kernel.org Coverity-id: 1226871 Signed-off-by: Jan Kara <jack@suse.cz> Fixed up the, now unused, out label. Signed-off-by: Jens Axboe <axboe@fb.com>
2014-10-21blk-merge: recaculate segment if it isn't less than max segmentsMing Lei1-2/+3
The problem is introduced by commit 764f612c6c3c231b(blk-merge: don't compute bi_phys_segments from bi_vcnt for cloned bio), and merge is needed if number of current segment isn't less than max segments. Strictly speaking, bio->bi_vcnt shouldn't be used here since it may not be accurate in cases of both cloned bio or bio cloned from, but bio_segments() is a bit expensive, and bi_vcnt is still the biggest number, so the approach should work. Signed-off-by: Ming Lei <tom.leiming@gmail.com> Signed-off-by: Jens Axboe <axboe@fb.com>