From 07b0fdecb2477396bcb69609019aade2b22124a1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 16 Jul 2019 07:58:31 -0700 Subject: blkcg: allow blkcg_policy->pd_stat() to print non-debug info too Currently, ->pd_stat() is called only when moduleparam blkcg_debug_stats is set which prevents it from printing non-debug policy-specific statistics. Let's move debug testing down so that ->pd_stat() can print non-debug stat too. This patch doesn't cause any visible behavior change. Signed-off-by: Tejun Heo Cc: Josef Bacik Signed-off-by: Jens Axboe --- include/linux/blk-cgroup.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/blk-cgroup.h b/include/linux/blk-cgroup.h index 689a58231288..12811091fd50 100644 --- a/include/linux/blk-cgroup.h +++ b/include/linux/blk-cgroup.h @@ -181,6 +181,7 @@ struct blkcg_policy { extern struct blkcg blkcg_root; extern struct cgroup_subsys_state * const blkcg_root_css; +extern bool blkcg_debug_stats; struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg, struct request_queue *q, bool update_hint); -- cgit v1.2.3 From a6d81d30d3cd87f85bfd922358eb18b8146c4925 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 16 Jul 2019 16:19:25 -0400 Subject: wait: add wq_has_single_sleeper helper rq-qos sits in the io path so we want to take locks as sparingly as possible. To accomplish this we try not to take the waitqueue head lock unless we are sure we need to go to sleep, and we have an optimization to make sure that we don't starve out existing waiters. Since we check if there are existing waiters locklessly we need to be able to update our view of the waitqueue list after we've added ourselves to the waitqueue. Accomplish this by adding this helper to see if there is more than just ourselves on the list. Reviewed-by: Oleg Nesterov Signed-off-by: Josef Bacik Signed-off-by: Jens Axboe --- include/linux/wait.h | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'include/linux') diff --git a/include/linux/wait.h b/include/linux/wait.h index b6f77cf60dd7..30c515520fb2 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -126,6 +126,19 @@ static inline int waitqueue_active(struct wait_queue_head *wq_head) return !list_empty(&wq_head->head); } +/** + * wq_has_single_sleeper - check if there is only one sleeper + * @wq_head: wait queue head + * + * Returns true of wq_head has only one sleeper on the list. + * + * Please refer to the comment for waitqueue_active. + */ +static inline bool wq_has_single_sleeper(struct wait_queue_head *wq_head) +{ + return list_is_singular(&wq_head->head); +} + /** * wq_has_sleeper - check if there are any waiting processes * @wq_head: wait queue head -- cgit v1.2.3 From 00289cd87676e14913d2d8492d1ce05c4baafdae Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 17 Jul 2019 18:07:53 -0700 Subject: drivers/base: Introduce kill_device() The libnvdimm subsystem arranges for devices to be destroyed as a result of a sysfs operation. Since device_unregister() cannot be called from an actively running sysfs attribute of the same device libnvdimm arranges for device_unregister() to be performed in an out-of-line async context. The driver core maintains a 'dead' state for coordinating its own racing async registration / de-registration requests. Rather than add local 'dead' state tracking infrastructure to libnvdimm device objects, export the existing state tracking via a new kill_device() helper. The kill_device() helper simply marks the device as dead, i.e. that it is on its way to device_del(), or returns that the device was already dead. This can be used in advance of calling device_unregister() for subsystems like libnvdimm that might need to handle multiple user threads racing to delete a device. This refactoring does not change any behavior, but it is a pre-requisite for follow-on fixes and therefore marked for -stable. Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Fixes: 4d88a97aa9e8 ("libnvdimm, nvdimm: dimm driver and base libnvdimm device-driver...") Cc: Tested-by: Jane Chu Reviewed-by: Greg Kroah-Hartman Link: https://lore.kernel.org/r/156341207332.292348.14959761496009347574.stgit@dwillia2-desk3.amr.corp.intel.com Signed-off-by: Dan Williams --- drivers/base/core.c | 27 +++++++++++++++++++-------- include/linux/device.h | 1 + 2 files changed, 20 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/core.c b/drivers/base/core.c index fd7511e04e62..eaf3aa0cb803 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -2211,6 +2211,24 @@ void put_device(struct device *dev) } EXPORT_SYMBOL_GPL(put_device); +bool kill_device(struct device *dev) +{ + /* + * Require the device lock and set the "dead" flag to guarantee that + * the update behavior is consistent with the other bitfields near + * it and that we cannot have an asynchronous probe routine trying + * to run while we are tearing out the bus/class/sysfs from + * underneath the device. + */ + lockdep_assert_held(&dev->mutex); + + if (dev->p->dead) + return false; + dev->p->dead = true; + return true; +} +EXPORT_SYMBOL_GPL(kill_device); + /** * device_del - delete device from system. * @dev: device. @@ -2230,15 +2248,8 @@ void device_del(struct device *dev) struct kobject *glue_dir = NULL; struct class_interface *class_intf; - /* - * Hold the device lock and set the "dead" flag to guarantee that - * the update behavior is consistent with the other bitfields near - * it and that we cannot have an asynchronous probe routine trying - * to run while we are tearing out the bus/class/sysfs from - * underneath the device. - */ device_lock(dev); - dev->p->dead = true; + kill_device(dev); device_unlock(dev); /* Notify clients of device removal. This call must come diff --git a/include/linux/device.h b/include/linux/device.h index e85264fb6616..0da5c67f6be1 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -1373,6 +1373,7 @@ extern int (*platform_notify_remove)(struct device *dev); */ extern struct device *get_device(struct device *dev); extern void put_device(struct device *dev); +extern bool kill_device(struct device *dev); #ifdef CONFIG_DEVTMPFS extern int devtmpfs_create_node(struct device *dev); -- cgit v1.2.3 From 87a30e1f05d73a34e6d1895065541369131aaf1c Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 17 Jul 2019 18:08:26 -0700 Subject: driver-core, libnvdimm: Let device subsystems add local lockdep coverage For good reason, the standard device_lock() is marked lockdep_set_novalidate_class() because there is simply no sane way to describe the myriad ways the device_lock() ordered with other locks. However, that leaves subsystems that know their own local device_lock() ordering rules to find lock ordering mistakes manually. Instead, introduce an optional / additional lockdep-enabled lock that a subsystem can acquire in all the same paths that the device_lock() is acquired. A conversion of the NFIT driver and NVDIMM subsystem to a lockdep-validate device_lock() scheme is included. The debug_nvdimm_lock() implementation implements the correct lock-class and stacking order for the libnvdimm device topology hierarchy. Yes, this is a hack, but hopefully it is a useful hack for other subsystems device_lock() debug sessions. Quoting Greg: "Yeah, it feels a bit hacky but it's really up to a subsystem to mess up using it as much as anything else, so user beware :) I don't object to it if it makes things easier for you to debug." Cc: Ingo Molnar Cc: Ira Weiny Cc: Will Deacon Cc: Dave Jiang Cc: Keith Busch Cc: Peter Zijlstra Cc: Vishal Verma Cc: "Rafael J. Wysocki" Cc: Greg Kroah-Hartman Signed-off-by: Dan Williams Acked-by: Greg Kroah-Hartman Reviewed-by: Ira Weiny Link: https://lore.kernel.org/r/156341210661.292348.7014034644265455704.stgit@dwillia2-desk3.amr.corp.intel.com --- drivers/acpi/nfit/core.c | 28 ++++++++--------- drivers/acpi/nfit/nfit.h | 24 +++++++++++++++ drivers/base/core.c | 3 ++ drivers/nvdimm/btt_devs.c | 16 +++++----- drivers/nvdimm/bus.c | 28 ++++++++++------- drivers/nvdimm/core.c | 10 +++--- drivers/nvdimm/dimm_devs.c | 4 +-- drivers/nvdimm/namespace_devs.c | 36 +++++++++++----------- drivers/nvdimm/nd-core.h | 68 +++++++++++++++++++++++++++++++++++++++++ drivers/nvdimm/pfn_devs.c | 24 +++++++-------- drivers/nvdimm/pmem.c | 4 +-- drivers/nvdimm/region.c | 2 +- drivers/nvdimm/region_devs.c | 16 +++++----- include/linux/device.h | 5 +++ 14 files changed, 187 insertions(+), 81 deletions(-) (limited to 'include/linux') diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c index 23022cf20d26..f22139458ce1 100644 --- a/drivers/acpi/nfit/core.c +++ b/drivers/acpi/nfit/core.c @@ -1282,7 +1282,7 @@ static ssize_t hw_error_scrub_store(struct device *dev, if (rc) return rc; - device_lock(dev); + nfit_device_lock(dev); nd_desc = dev_get_drvdata(dev); if (nd_desc) { struct acpi_nfit_desc *acpi_desc = to_acpi_desc(nd_desc); @@ -1299,7 +1299,7 @@ static ssize_t hw_error_scrub_store(struct device *dev, break; } } - device_unlock(dev); + nfit_device_unlock(dev); if (rc) return rc; return size; @@ -1319,7 +1319,7 @@ static ssize_t scrub_show(struct device *dev, ssize_t rc = -ENXIO; bool busy; - device_lock(dev); + nfit_device_lock(dev); nd_desc = dev_get_drvdata(dev); if (!nd_desc) { device_unlock(dev); @@ -1339,7 +1339,7 @@ static ssize_t scrub_show(struct device *dev, } mutex_unlock(&acpi_desc->init_mutex); - device_unlock(dev); + nfit_device_unlock(dev); return rc; } @@ -1356,14 +1356,14 @@ static ssize_t scrub_store(struct device *dev, if (val != 1) return -EINVAL; - device_lock(dev); + nfit_device_lock(dev); nd_desc = dev_get_drvdata(dev); if (nd_desc) { struct acpi_nfit_desc *acpi_desc = to_acpi_desc(nd_desc); rc = acpi_nfit_ars_rescan(acpi_desc, ARS_REQ_LONG); } - device_unlock(dev); + nfit_device_unlock(dev); if (rc) return rc; return size; @@ -1749,9 +1749,9 @@ static void acpi_nvdimm_notify(acpi_handle handle, u32 event, void *data) struct acpi_device *adev = data; struct device *dev = &adev->dev; - device_lock(dev->parent); + nfit_device_lock(dev->parent); __acpi_nvdimm_notify(dev, event); - device_unlock(dev->parent); + nfit_device_unlock(dev->parent); } static bool acpi_nvdimm_has_method(struct acpi_device *adev, char *method) @@ -3457,8 +3457,8 @@ static int acpi_nfit_flush_probe(struct nvdimm_bus_descriptor *nd_desc) struct device *dev = acpi_desc->dev; /* Bounce the device lock to flush acpi_nfit_add / acpi_nfit_notify */ - device_lock(dev); - device_unlock(dev); + nfit_device_lock(dev); + nfit_device_unlock(dev); /* Bounce the init_mutex to complete initial registration */ mutex_lock(&acpi_desc->init_mutex); @@ -3602,8 +3602,8 @@ void acpi_nfit_shutdown(void *data) * acpi_nfit_ars_rescan() submissions have had a chance to * either submit or see ->cancel set. */ - device_lock(bus_dev); - device_unlock(bus_dev); + nfit_device_lock(bus_dev); + nfit_device_unlock(bus_dev); flush_workqueue(nfit_wq); } @@ -3746,9 +3746,9 @@ EXPORT_SYMBOL_GPL(__acpi_nfit_notify); static void acpi_nfit_notify(struct acpi_device *adev, u32 event) { - device_lock(&adev->dev); + nfit_device_lock(&adev->dev); __acpi_nfit_notify(&adev->dev, adev->handle, event); - device_unlock(&adev->dev); + nfit_device_unlock(&adev->dev); } static const struct acpi_device_id acpi_nfit_ids[] = { diff --git a/drivers/acpi/nfit/nfit.h b/drivers/acpi/nfit/nfit.h index 6ee2b02af73e..24241941181c 100644 --- a/drivers/acpi/nfit/nfit.h +++ b/drivers/acpi/nfit/nfit.h @@ -312,6 +312,30 @@ static inline struct acpi_nfit_desc *to_acpi_desc( return container_of(nd_desc, struct acpi_nfit_desc, nd_desc); } +#ifdef CONFIG_PROVE_LOCKING +static inline void nfit_device_lock(struct device *dev) +{ + device_lock(dev); + mutex_lock(&dev->lockdep_mutex); +} + +static inline void nfit_device_unlock(struct device *dev) +{ + mutex_unlock(&dev->lockdep_mutex); + device_unlock(dev); +} +#else +static inline void nfit_device_lock(struct device *dev) +{ + device_lock(dev); +} + +static inline void nfit_device_unlock(struct device *dev) +{ + device_unlock(dev); +} +#endif + const guid_t *to_nfit_uuid(enum nfit_uuids id); int acpi_nfit_init(struct acpi_nfit_desc *acpi_desc, void *nfit, acpi_size sz); void acpi_nfit_shutdown(void *data); diff --git a/drivers/base/core.c b/drivers/base/core.c index eaf3aa0cb803..4825949d6547 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -1663,6 +1663,9 @@ void device_initialize(struct device *dev) kobject_init(&dev->kobj, &device_ktype); INIT_LIST_HEAD(&dev->dma_pools); mutex_init(&dev->mutex); +#ifdef CONFIG_PROVE_LOCKING + mutex_init(&dev->lockdep_mutex); +#endif lockdep_set_novalidate_class(&dev->mutex); spin_lock_init(&dev->devres_lock); INIT_LIST_HEAD(&dev->devres_head); diff --git a/drivers/nvdimm/btt_devs.c b/drivers/nvdimm/btt_devs.c index 62d00fffa4af..3508a79110c7 100644 --- a/drivers/nvdimm/btt_devs.c +++ b/drivers/nvdimm/btt_devs.c @@ -62,14 +62,14 @@ static ssize_t sector_size_store(struct device *dev, struct nd_btt *nd_btt = to_nd_btt(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); rc = nd_size_select_store(dev, buf, &nd_btt->lbasize, btt_lbasize_supported); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -91,11 +91,11 @@ static ssize_t uuid_store(struct device *dev, struct nd_btt *nd_btt = to_nd_btt(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); rc = nd_uuid_store(dev, &nd_btt->uuid, buf, len); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -120,13 +120,13 @@ static ssize_t namespace_store(struct device *dev, struct nd_btt *nd_btt = to_nd_btt(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); rc = nd_namespace_store(dev, &nd_btt->ndns, buf, len); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -138,14 +138,14 @@ static ssize_t size_show(struct device *dev, struct nd_btt *nd_btt = to_nd_btt(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); if (dev->driver) rc = sprintf(buf, "%llu\n", nd_btt->size); else { /* no size to convey if the btt instance is disabled */ rc = -ENXIO; } - device_unlock(dev); + nd_device_unlock(dev); return rc; } diff --git a/drivers/nvdimm/bus.c b/drivers/nvdimm/bus.c index df41f3571dc9..798c5c4aea9c 100644 --- a/drivers/nvdimm/bus.c +++ b/drivers/nvdimm/bus.c @@ -26,7 +26,7 @@ int nvdimm_major; static int nvdimm_bus_major; -static struct class *nd_class; +struct class *nd_class; static DEFINE_IDA(nd_ida); static int to_nd_device_type(struct device *dev) @@ -91,7 +91,10 @@ static int nvdimm_bus_probe(struct device *dev) dev->driver->name, dev_name(dev)); nvdimm_bus_probe_start(nvdimm_bus); + debug_nvdimm_lock(dev); rc = nd_drv->probe(dev); + debug_nvdimm_unlock(dev); + if (rc == 0) nd_region_probe_success(nvdimm_bus, dev); else @@ -113,8 +116,11 @@ static int nvdimm_bus_remove(struct device *dev) struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev); int rc = 0; - if (nd_drv->remove) + if (nd_drv->remove) { + debug_nvdimm_lock(dev); rc = nd_drv->remove(dev); + debug_nvdimm_unlock(dev); + } nd_region_disable(nvdimm_bus, dev); dev_dbg(&nvdimm_bus->dev, "%s.remove(%s) = %d\n", dev->driver->name, @@ -140,7 +146,7 @@ static void nvdimm_bus_shutdown(struct device *dev) void nd_device_notify(struct device *dev, enum nvdimm_event event) { - device_lock(dev); + nd_device_lock(dev); if (dev->driver) { struct nd_device_driver *nd_drv; @@ -148,7 +154,7 @@ void nd_device_notify(struct device *dev, enum nvdimm_event event) if (nd_drv->notify) nd_drv->notify(dev, event); } - device_unlock(dev); + nd_device_unlock(dev); } EXPORT_SYMBOL(nd_device_notify); @@ -296,7 +302,7 @@ static void nvdimm_bus_release(struct device *dev) kfree(nvdimm_bus); } -static bool is_nvdimm_bus(struct device *dev) +bool is_nvdimm_bus(struct device *dev) { return dev->release == nvdimm_bus_release; } @@ -575,9 +581,9 @@ void nd_device_unregister(struct device *dev, enum nd_async_mode mode) * or otherwise let the async path handle it if the * unregistration was already queued. */ - device_lock(dev); + nd_device_lock(dev); killed = kill_device(dev); - device_unlock(dev); + nd_device_unlock(dev); if (!killed) return; @@ -888,10 +894,10 @@ void wait_nvdimm_bus_probe_idle(struct device *dev) if (nvdimm_bus->probe_active == 0) break; nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); wait_event(nvdimm_bus->wait, nvdimm_bus->probe_active == 0); - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); } while (true); } @@ -1107,7 +1113,7 @@ static int __nd_ioctl(struct nvdimm_bus *nvdimm_bus, struct nvdimm *nvdimm, goto out; } - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); rc = nd_cmd_clear_to_send(nvdimm_bus, nvdimm, func, buf); if (rc) @@ -1129,7 +1135,7 @@ static int __nd_ioctl(struct nvdimm_bus *nvdimm_bus, struct nvdimm *nvdimm, out_unlock: nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); out: kfree(in_env); kfree(out_env); diff --git a/drivers/nvdimm/core.c b/drivers/nvdimm/core.c index 5e1f060547bf..9204f1e9fd14 100644 --- a/drivers/nvdimm/core.c +++ b/drivers/nvdimm/core.c @@ -246,7 +246,7 @@ static int nd_uuid_parse(struct device *dev, u8 *uuid_out, const char *buf, * * Enforce that uuids can only be changed while the device is disabled * (driver detached) - * LOCKING: expects device_lock() is held on entry + * LOCKING: expects nd_device_lock() is held on entry */ int nd_uuid_store(struct device *dev, u8 **uuid_out, const char *buf, size_t len) @@ -347,15 +347,15 @@ static DEVICE_ATTR_RO(provider); static int flush_namespaces(struct device *dev, void *data) { - device_lock(dev); - device_unlock(dev); + nd_device_lock(dev); + nd_device_unlock(dev); return 0; } static int flush_regions_dimms(struct device *dev, void *data) { - device_lock(dev); - device_unlock(dev); + nd_device_lock(dev); + nd_device_unlock(dev); device_for_each_child(dev, NULL, flush_namespaces); return 0; } diff --git a/drivers/nvdimm/dimm_devs.c b/drivers/nvdimm/dimm_devs.c index dfecd6e17043..29a065e769ea 100644 --- a/drivers/nvdimm/dimm_devs.c +++ b/drivers/nvdimm/dimm_devs.c @@ -484,12 +484,12 @@ static ssize_t security_store(struct device *dev, * done while probing is idle and the DIMM is not in active use * in any region. */ - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); rc = __security_store(dev, buf, len); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc; } diff --git a/drivers/nvdimm/namespace_devs.c b/drivers/nvdimm/namespace_devs.c index a434a5964cb9..92cd809d7e43 100644 --- a/drivers/nvdimm/namespace_devs.c +++ b/drivers/nvdimm/namespace_devs.c @@ -410,7 +410,7 @@ static ssize_t alt_name_store(struct device *dev, struct nd_region *nd_region = to_nd_region(dev->parent); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); rc = __alt_name_store(dev, buf, len); @@ -418,7 +418,7 @@ static ssize_t alt_name_store(struct device *dev, rc = nd_namespace_label_update(nd_region, dev); dev_dbg(dev, "%s(%zd)\n", rc < 0 ? "fail " : "", rc); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc < 0 ? rc : len; } @@ -1077,7 +1077,7 @@ static ssize_t size_store(struct device *dev, if (rc) return rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); rc = __size_store(dev, val); @@ -1103,7 +1103,7 @@ static ssize_t size_store(struct device *dev, dev_dbg(dev, "%llx %s (%d)\n", val, rc < 0 ? "fail" : "success", rc); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc < 0 ? rc : len; } @@ -1286,7 +1286,7 @@ static ssize_t uuid_store(struct device *dev, } else return -ENXIO; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); if (to_ndns(dev)->claim) @@ -1302,7 +1302,7 @@ static ssize_t uuid_store(struct device *dev, dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc < 0 ? rc : len; } @@ -1376,7 +1376,7 @@ static ssize_t sector_size_store(struct device *dev, } else return -ENXIO; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); if (to_ndns(dev)->claim) rc = -EBUSY; @@ -1387,7 +1387,7 @@ static ssize_t sector_size_store(struct device *dev, dev_dbg(dev, "result: %zd %s: %s%s", rc, rc < 0 ? "tried" : "wrote", buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -1502,9 +1502,9 @@ static ssize_t holder_show(struct device *dev, struct nd_namespace_common *ndns = to_ndns(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); rc = sprintf(buf, "%s\n", ndns->claim ? dev_name(ndns->claim) : ""); - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -1541,7 +1541,7 @@ static ssize_t holder_class_store(struct device *dev, struct nd_region *nd_region = to_nd_region(dev->parent); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); rc = __holder_class_store(dev, buf); @@ -1549,7 +1549,7 @@ static ssize_t holder_class_store(struct device *dev, rc = nd_namespace_label_update(nd_region, dev); dev_dbg(dev, "%s(%zd)\n", rc < 0 ? "fail " : "", rc); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc < 0 ? rc : len; } @@ -1560,7 +1560,7 @@ static ssize_t holder_class_show(struct device *dev, struct nd_namespace_common *ndns = to_ndns(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); if (ndns->claim_class == NVDIMM_CCLASS_NONE) rc = sprintf(buf, "\n"); else if ((ndns->claim_class == NVDIMM_CCLASS_BTT) || @@ -1572,7 +1572,7 @@ static ssize_t holder_class_show(struct device *dev, rc = sprintf(buf, "dax\n"); else rc = sprintf(buf, "\n"); - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -1586,7 +1586,7 @@ static ssize_t mode_show(struct device *dev, char *mode; ssize_t rc; - device_lock(dev); + nd_device_lock(dev); claim = ndns->claim; if (claim && is_nd_btt(claim)) mode = "safe"; @@ -1599,7 +1599,7 @@ static ssize_t mode_show(struct device *dev, else mode = "raw"; rc = sprintf(buf, "%s\n", mode); - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -1703,8 +1703,8 @@ struct nd_namespace_common *nvdimm_namespace_common_probe(struct device *dev) * Flush any in-progess probes / removals in the driver * for the raw personality of this namespace. */ - device_lock(&ndns->dev); - device_unlock(&ndns->dev); + nd_device_lock(&ndns->dev); + nd_device_unlock(&ndns->dev); if (ndns->dev.driver) { dev_dbg(&ndns->dev, "is active, can't bind %s\n", dev_name(dev)); diff --git a/drivers/nvdimm/nd-core.h b/drivers/nvdimm/nd-core.h index 6cd470547106..0ac52b6eb00e 100644 --- a/drivers/nvdimm/nd-core.h +++ b/drivers/nvdimm/nd-core.h @@ -9,6 +9,7 @@ #include #include #include +#include "nd.h" extern struct list_head nvdimm_bus_list; extern struct mutex nvdimm_bus_list_mutex; @@ -182,4 +183,71 @@ ssize_t nd_namespace_store(struct device *dev, struct nd_namespace_common **_ndns, const char *buf, size_t len); struct nd_pfn *to_nd_pfn_safe(struct device *dev); +bool is_nvdimm_bus(struct device *dev); + +#ifdef CONFIG_PROVE_LOCKING +extern struct class *nd_class; + +enum { + LOCK_BUS, + LOCK_NDCTL, + LOCK_REGION, + LOCK_DIMM = LOCK_REGION, + LOCK_NAMESPACE, + LOCK_CLAIM, +}; + +static inline void debug_nvdimm_lock(struct device *dev) +{ + if (is_nd_region(dev)) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_REGION); + else if (is_nvdimm(dev)) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_DIMM); + else if (is_nd_btt(dev) || is_nd_pfn(dev) || is_nd_dax(dev)) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_CLAIM); + else if (dev->parent && (is_nd_region(dev->parent))) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_NAMESPACE); + else if (is_nvdimm_bus(dev)) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_BUS); + else if (dev->class && dev->class == nd_class) + mutex_lock_nested(&dev->lockdep_mutex, LOCK_NDCTL); + else + dev_WARN(dev, "unknown lock level\n"); +} + +static inline void debug_nvdimm_unlock(struct device *dev) +{ + mutex_unlock(&dev->lockdep_mutex); +} + +static inline void nd_device_lock(struct device *dev) +{ + device_lock(dev); + debug_nvdimm_lock(dev); +} + +static inline void nd_device_unlock(struct device *dev) +{ + debug_nvdimm_unlock(dev); + device_unlock(dev); +} +#else +static inline void nd_device_lock(struct device *dev) +{ + device_lock(dev); +} + +static inline void nd_device_unlock(struct device *dev) +{ + device_unlock(dev); +} + +static inline void debug_nvdimm_lock(struct device *dev) +{ +} + +static inline void debug_nvdimm_unlock(struct device *dev) +{ +} +#endif #endif /* __ND_CORE_H__ */ diff --git a/drivers/nvdimm/pfn_devs.c b/drivers/nvdimm/pfn_devs.c index 0f81fc56bbfd..9b09fe18e666 100644 --- a/drivers/nvdimm/pfn_devs.c +++ b/drivers/nvdimm/pfn_devs.c @@ -67,7 +67,7 @@ static ssize_t mode_store(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc = 0; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); if (dev->driver) rc = -EBUSY; @@ -89,7 +89,7 @@ static ssize_t mode_store(struct device *dev, dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -132,14 +132,14 @@ static ssize_t align_store(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); rc = nd_size_select_store(dev, buf, &nd_pfn->align, nd_pfn_supported_alignments()); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -161,11 +161,11 @@ static ssize_t uuid_store(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); rc = nd_uuid_store(dev, &nd_pfn->uuid, buf, len); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); - device_unlock(dev); + nd_device_unlock(dev); return rc ? rc : len; } @@ -190,13 +190,13 @@ static ssize_t namespace_store(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); rc = nd_namespace_store(dev, &nd_pfn->ndns, buf, len); dev_dbg(dev, "result: %zd wrote: %s%s", rc, buf, buf[len - 1] == '\n' ? "" : "\n"); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -208,7 +208,7 @@ static ssize_t resource_show(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); if (dev->driver) { struct nd_pfn_sb *pfn_sb = nd_pfn->pfn_sb; u64 offset = __le64_to_cpu(pfn_sb->dataoff); @@ -222,7 +222,7 @@ static ssize_t resource_show(struct device *dev, /* no address to convey if the pfn instance is disabled */ rc = -ENXIO; } - device_unlock(dev); + nd_device_unlock(dev); return rc; } @@ -234,7 +234,7 @@ static ssize_t size_show(struct device *dev, struct nd_pfn *nd_pfn = to_nd_pfn_safe(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); if (dev->driver) { struct nd_pfn_sb *pfn_sb = nd_pfn->pfn_sb; u64 offset = __le64_to_cpu(pfn_sb->dataoff); @@ -250,7 +250,7 @@ static ssize_t size_show(struct device *dev, /* no size to convey if the pfn instance is disabled */ rc = -ENXIO; } - device_unlock(dev); + nd_device_unlock(dev); return rc; } diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c index 28cb44c61d4a..53797e7be18a 100644 --- a/drivers/nvdimm/pmem.c +++ b/drivers/nvdimm/pmem.c @@ -520,8 +520,8 @@ static int nd_pmem_remove(struct device *dev) nvdimm_namespace_detach_btt(to_nd_btt(dev)); else { /* - * Note, this assumes device_lock() context to not race - * nd_pmem_notify() + * Note, this assumes nd_device_lock() context to not + * race nd_pmem_notify() */ sysfs_put(pmem->bb_state); pmem->bb_state = NULL; diff --git a/drivers/nvdimm/region.c b/drivers/nvdimm/region.c index 488c47ac4c4a..37bf8719a2a4 100644 --- a/drivers/nvdimm/region.c +++ b/drivers/nvdimm/region.c @@ -102,7 +102,7 @@ static int nd_region_remove(struct device *dev) nvdimm_bus_unlock(dev); /* - * Note, this assumes device_lock() context to not race + * Note, this assumes nd_device_lock() context to not race * nd_region_notify() */ sysfs_put(nd_region->bb_state); diff --git a/drivers/nvdimm/region_devs.c b/drivers/nvdimm/region_devs.c index a15276cdec7d..91b5a7ade0d5 100644 --- a/drivers/nvdimm/region_devs.c +++ b/drivers/nvdimm/region_devs.c @@ -329,7 +329,7 @@ static ssize_t set_cookie_show(struct device *dev, * the v1.1 namespace label cookie definition. To read all this * data we need to wait for probing to settle. */ - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); if (nd_region->ndr_mappings) { @@ -346,7 +346,7 @@ static ssize_t set_cookie_show(struct device *dev, } } nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); if (rc) return rc; @@ -422,12 +422,12 @@ static ssize_t available_size_show(struct device *dev, * memory nvdimm_bus_lock() is dropped, but that's userspace's * problem to not race itself. */ - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); available = nd_region_available_dpa(nd_region); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return sprintf(buf, "%llu\n", available); } @@ -439,12 +439,12 @@ static ssize_t max_available_extent_show(struct device *dev, struct nd_region *nd_region = to_nd_region(dev); unsigned long long available = 0; - device_lock(dev); + nd_device_lock(dev); nvdimm_bus_lock(dev); wait_nvdimm_bus_probe_idle(dev); available = nd_region_allocatable_dpa(nd_region); nvdimm_bus_unlock(dev); - device_unlock(dev); + nd_device_unlock(dev); return sprintf(buf, "%llu\n", available); } @@ -563,12 +563,12 @@ static ssize_t region_badblocks_show(struct device *dev, struct nd_region *nd_region = to_nd_region(dev); ssize_t rc; - device_lock(dev); + nd_device_lock(dev); if (dev->driver) rc = badblocks_show(&nd_region->bb, buf, 0); else rc = -ENXIO; - device_unlock(dev); + nd_device_unlock(dev); return rc; } diff --git a/include/linux/device.h b/include/linux/device.h index 0da5c67f6be1..9237b857b598 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -909,6 +909,8 @@ struct dev_links_info { * This identifies the device type and carries type-specific * information. * @mutex: Mutex to synchronize calls to its driver. + * @lockdep_mutex: An optional debug lock that a subsystem can use as a + * peer lock to gain localized lockdep coverage of the device_lock. * @bus: Type of bus device is on. * @driver: Which driver has allocated this * @platform_data: Platform data specific to the device. @@ -991,6 +993,9 @@ struct device { core doesn't touch it */ void *driver_data; /* Driver data, set and get with dev_set_drvdata/dev_get_drvdata */ +#ifdef CONFIG_PROVE_LOCKING + struct mutex lockdep_mutex; +#endif struct mutex mutex; /* mutex to synchronize calls to * its driver. */ -- cgit v1.2.3 From 893a1c97205a3ece0cbb3f571a3b972080f3b4c7 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 16 Jul 2019 13:55:23 -0600 Subject: blk-mq: allow REQ_NOWAIT to return an error inline By default, if a caller sets REQ_NOWAIT and we need to block, we'll return -EAGAIN through the bio->bi_end_io() callback. For some use cases, this makes it hard to use. Allow a caller to ask for inline return of errors related to blocking by also setting REQ_NOWAIT_INLINE. Signed-off-by: Jens Axboe --- block/blk-mq.c | 8 ++++++-- include/linux/blk_types.h | 5 ++++- 2 files changed, 10 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/block/blk-mq.c b/block/blk-mq.c index b038ec680e84..2bc2c0705660 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -1960,9 +1960,13 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio) rq = blk_mq_get_request(q, bio, &data); if (unlikely(!rq)) { rq_qos_cleanup(q, bio); - if (bio->bi_opf & REQ_NOWAIT) + + cookie = BLK_QC_T_NONE; + if (bio->bi_opf & REQ_NOWAIT_INLINE) + cookie = BLK_QC_T_EAGAIN; + else if (bio->bi_opf & REQ_NOWAIT) bio_wouldblock_error(bio); - return BLK_QC_T_NONE; + return cookie; } trace_block_getrq(q, bio, bio->bi_opf); diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h index feff3fe4467e..1b1fa1557e68 100644 --- a/include/linux/blk_types.h +++ b/include/linux/blk_types.h @@ -311,6 +311,7 @@ enum req_flag_bits { __REQ_RAHEAD, /* read ahead, can fail anytime */ __REQ_BACKGROUND, /* background IO */ __REQ_NOWAIT, /* Don't wait if request will block */ + __REQ_NOWAIT_INLINE, /* Return would-block error inline */ /* * When a shared kthread needs to issue a bio for a cgroup, doing * so synchronously can lead to priority inversions as the kthread @@ -345,6 +346,7 @@ enum req_flag_bits { #define REQ_RAHEAD (1ULL << __REQ_RAHEAD) #define REQ_BACKGROUND (1ULL << __REQ_BACKGROUND) #define REQ_NOWAIT (1ULL << __REQ_NOWAIT) +#define REQ_NOWAIT_INLINE (1ULL << __REQ_NOWAIT_INLINE) #define REQ_CGROUP_PUNT (1ULL << __REQ_CGROUP_PUNT) #define REQ_NOUNMAP (1ULL << __REQ_NOUNMAP) @@ -418,12 +420,13 @@ static inline int op_stat_group(unsigned int op) typedef unsigned int blk_qc_t; #define BLK_QC_T_NONE -1U +#define BLK_QC_T_EAGAIN -2U #define BLK_QC_T_SHIFT 16 #define BLK_QC_T_INTERNAL (1U << 31) static inline bool blk_qc_t_valid(blk_qc_t cookie) { - return cookie != BLK_QC_T_NONE; + return cookie != BLK_QC_T_NONE && cookie != BLK_QC_T_EAGAIN; } static inline unsigned int blk_qc_t_to_queue_num(blk_qc_t cookie) -- cgit v1.2.3 From effa467870c7612012885df4e246bdb8ffd8e44c Mon Sep 17 00:00:00 2001 From: Dmitry Safonov Date: Tue, 16 Jul 2019 22:38:05 +0100 Subject: iommu/vt-d: Don't queue_iova() if there is no flush queue Intel VT-d driver was reworked to use common deferred flushing implementation. Previously there was one global per-cpu flush queue, afterwards - one per domain. Before deferring a flush, the queue should be allocated and initialized. Currently only domains with IOMMU_DOMAIN_DMA type initialize their flush queue. It's probably worth to init it for static or unmanaged domains too, but it may be arguable - I'm leaving it to iommu folks. Prevent queuing an iova flush if the domain doesn't have a queue. The defensive check seems to be worth to keep even if queue would be initialized for all kinds of domains. And is easy backportable. On 4.19.43 stable kernel it has a user-visible effect: previously for devices in si domain there were crashes, on sata devices: BUG: spinlock bad magic on CPU#6, swapper/0/1 lock: 0xffff88844f582008, .magic: 00000000, .owner: /-1, .owner_cpu: 0 CPU: 6 PID: 1 Comm: swapper/0 Not tainted 4.19.43 #1 Call Trace: dump_stack+0x61/0x7e spin_bug+0x9d/0xa3 do_raw_spin_lock+0x22/0x8e _raw_spin_lock_irqsave+0x32/0x3a queue_iova+0x45/0x115 intel_unmap+0x107/0x113 intel_unmap_sg+0x6b/0x76 __ata_qc_complete+0x7f/0x103 ata_qc_complete+0x9b/0x26a ata_qc_complete_multiple+0xd0/0xe3 ahci_handle_port_interrupt+0x3ee/0x48a ahci_handle_port_intr+0x73/0xa9 ahci_single_level_irq_intr+0x40/0x60 __handle_irq_event_percpu+0x7f/0x19a handle_irq_event_percpu+0x32/0x72 handle_irq_event+0x38/0x56 handle_edge_irq+0x102/0x121 handle_irq+0x147/0x15c do_IRQ+0x66/0xf2 common_interrupt+0xf/0xf RIP: 0010:__do_softirq+0x8c/0x2df The same for usb devices that use ehci-pci: BUG: spinlock bad magic on CPU#0, swapper/0/1 lock: 0xffff88844f402008, .magic: 00000000, .owner: /-1, .owner_cpu: 0 CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.43 #4 Call Trace: dump_stack+0x61/0x7e spin_bug+0x9d/0xa3 do_raw_spin_lock+0x22/0x8e _raw_spin_lock_irqsave+0x32/0x3a queue_iova+0x77/0x145 intel_unmap+0x107/0x113 intel_unmap_page+0xe/0x10 usb_hcd_unmap_urb_setup_for_dma+0x53/0x9d usb_hcd_unmap_urb_for_dma+0x17/0x100 unmap_urb_for_dma+0x22/0x24 __usb_hcd_giveback_urb+0x51/0xc3 usb_giveback_urb_bh+0x97/0xde tasklet_action_common.isra.4+0x5f/0xa1 tasklet_action+0x2d/0x30 __do_softirq+0x138/0x2df irq_exit+0x7d/0x8b smp_apic_timer_interrupt+0x10f/0x151 apic_timer_interrupt+0xf/0x20 RIP: 0010:_raw_spin_unlock_irqrestore+0x17/0x39 Cc: David Woodhouse Cc: Joerg Roedel Cc: Lu Baolu Cc: iommu@lists.linux-foundation.org Cc: # 4.14+ Fixes: 13cf01744608 ("iommu/vt-d: Make use of iova deferred flushing") Signed-off-by: Dmitry Safonov Reviewed-by: Lu Baolu Signed-off-by: Joerg Roedel --- drivers/iommu/intel-iommu.c | 3 ++- drivers/iommu/iova.c | 18 ++++++++++++++---- include/linux/iova.h | 6 ++++++ 3 files changed, 22 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c index 9b1d62d03370..72c6d647bec9 100644 --- a/drivers/iommu/intel-iommu.c +++ b/drivers/iommu/intel-iommu.c @@ -3561,7 +3561,8 @@ static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size) freelist = domain_unmap(domain, start_pfn, last_pfn); - if (intel_iommu_strict || (pdev && pdev->untrusted)) { + if (intel_iommu_strict || (pdev && pdev->untrusted) || + !has_iova_flush_queue(&domain->iovad)) { iommu_flush_iotlb_psi(iommu, domain, start_pfn, nrpages, !freelist, 0); /* free iova */ diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c index d499b2621239..8413ae54904a 100644 --- a/drivers/iommu/iova.c +++ b/drivers/iommu/iova.c @@ -54,9 +54,14 @@ init_iova_domain(struct iova_domain *iovad, unsigned long granule, } EXPORT_SYMBOL_GPL(init_iova_domain); +bool has_iova_flush_queue(struct iova_domain *iovad) +{ + return !!iovad->fq; +} + static void free_iova_flush_queue(struct iova_domain *iovad) { - if (!iovad->fq) + if (!has_iova_flush_queue(iovad)) return; if (timer_pending(&iovad->fq_timer)) @@ -74,13 +79,14 @@ static void free_iova_flush_queue(struct iova_domain *iovad) int init_iova_flush_queue(struct iova_domain *iovad, iova_flush_cb flush_cb, iova_entry_dtor entry_dtor) { + struct iova_fq __percpu *queue; int cpu; atomic64_set(&iovad->fq_flush_start_cnt, 0); atomic64_set(&iovad->fq_flush_finish_cnt, 0); - iovad->fq = alloc_percpu(struct iova_fq); - if (!iovad->fq) + queue = alloc_percpu(struct iova_fq); + if (!queue) return -ENOMEM; iovad->flush_cb = flush_cb; @@ -89,13 +95,17 @@ int init_iova_flush_queue(struct iova_domain *iovad, for_each_possible_cpu(cpu) { struct iova_fq *fq; - fq = per_cpu_ptr(iovad->fq, cpu); + fq = per_cpu_ptr(queue, cpu); fq->head = 0; fq->tail = 0; spin_lock_init(&fq->lock); } + smp_wmb(); + + iovad->fq = queue; + timer_setup(&iovad->fq_timer, fq_flush_timeout, 0); atomic_set(&iovad->fq_timer_on, 0); diff --git a/include/linux/iova.h b/include/linux/iova.h index 781b96ac706f..cd0f1de901a8 100644 --- a/include/linux/iova.h +++ b/include/linux/iova.h @@ -155,6 +155,7 @@ struct iova *reserve_iova(struct iova_domain *iovad, unsigned long pfn_lo, void copy_reserved_iova(struct iova_domain *from, struct iova_domain *to); void init_iova_domain(struct iova_domain *iovad, unsigned long granule, unsigned long start_pfn); +bool has_iova_flush_queue(struct iova_domain *iovad); int init_iova_flush_queue(struct iova_domain *iovad, iova_flush_cb flush_cb, iova_entry_dtor entry_dtor); struct iova *find_iova(struct iova_domain *iovad, unsigned long pfn); @@ -235,6 +236,11 @@ static inline void init_iova_domain(struct iova_domain *iovad, { } +bool has_iova_flush_queue(struct iova_domain *iovad) +{ + return false; +} + static inline int init_iova_flush_queue(struct iova_domain *iovad, iova_flush_cb flush_cb, iova_entry_dtor entry_dtor) -- cgit v1.2.3 From 201c1db90cd643282185a00770f12f95da330eca Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 23 Jul 2019 09:51:00 +0200 Subject: iommu/iova: Fix compilation error with !CONFIG_IOMMU_IOVA The stub function for !CONFIG_IOMMU_IOVA needs to be 'static inline'. Fixes: effa467870c76 ('iommu/vt-d: Don't queue_iova() if there is no flush queue') Signed-off-by: Joerg Roedel --- include/linux/iova.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/iova.h b/include/linux/iova.h index cd0f1de901a8..a0637abffee8 100644 --- a/include/linux/iova.h +++ b/include/linux/iova.h @@ -236,7 +236,7 @@ static inline void init_iova_domain(struct iova_domain *iovad, { } -bool has_iova_flush_queue(struct iova_domain *iovad) +static inline bool has_iova_flush_queue(struct iova_domain *iovad) { return false; } -- cgit v1.2.3 From 327fe1d42b83f8a06b33ba30159582b49af5fc8e Mon Sep 17 00:00:00 2001 From: Marcos Paulo de Souza Date: Tue, 23 Jul 2019 00:27:41 -0300 Subject: block: blk-mq: Remove blk_mq_sched_started_request and started_request blk_mq_sched_completed_request is a function that checks if the elevator related to the request has started_request implemented, but currently, none of the available IO schedulers implement started_request, so remove both. Signed-off-by: Marcos Paulo de Souza Signed-off-by: Jens Axboe --- block/blk-mq-sched.h | 9 --------- block/blk-mq.c | 2 -- include/linux/elevator.h | 1 - 3 files changed, 12 deletions(-) (limited to 'include/linux') diff --git a/block/blk-mq-sched.h b/block/blk-mq-sched.h index cf22ab00fefb..126021fc3a11 100644 --- a/block/blk-mq-sched.h +++ b/block/blk-mq-sched.h @@ -61,15 +61,6 @@ static inline void blk_mq_sched_completed_request(struct request *rq, u64 now) e->type->ops.completed_request(rq, now); } -static inline void blk_mq_sched_started_request(struct request *rq) -{ - struct request_queue *q = rq->q; - struct elevator_queue *e = q->elevator; - - if (e && e->type->ops.started_request) - e->type->ops.started_request(rq); -} - static inline void blk_mq_sched_requeue_request(struct request *rq) { struct request_queue *q = rq->q; diff --git a/block/blk-mq.c b/block/blk-mq.c index 2bc2c0705660..f78d3287dd82 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -669,8 +669,6 @@ void blk_mq_start_request(struct request *rq) { struct request_queue *q = rq->q; - blk_mq_sched_started_request(rq); - trace_block_rq_issue(q, rq); if (test_bit(QUEUE_FLAG_STATS, &q->queue_flags)) { diff --git a/include/linux/elevator.h b/include/linux/elevator.h index 17cd0078377c..1dd014c9c87b 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -45,7 +45,6 @@ struct elevator_mq_ops { struct request *(*dispatch_request)(struct blk_mq_hw_ctx *); bool (*has_work)(struct blk_mq_hw_ctx *); void (*completed_request)(struct request *, u64); - void (*started_request)(struct request *); void (*requeue_request)(struct request *); struct request *(*former_request)(struct request_queue *, struct request *); struct request *(*next_request)(struct request_queue *, struct request *); -- cgit v1.2.3 From 06532750010e06dd4b6d69983773677df7fc5291 Mon Sep 17 00:00:00 2001 From: Eric Auger Date: Mon, 22 Jul 2019 18:51:49 +0200 Subject: dma-mapping: use dma_get_mask in dma_addressing_limited We currently have cases where the dma_addressing_limited() gets called with dma_mask unset. This causes a NULL pointer dereference. Use dma_get_mask() accessor to prevent the crash. Fixes: b866455423e0 ("dma-mapping: add a dma_addressing_limited helper") Signed-off-by: Eric Auger Acked-by: Michael S. Tsirkin Signed-off-by: Christoph Hellwig --- include/linux/dma-mapping.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index e11b115dd0e4..f7d1eea32c78 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -689,8 +689,8 @@ static inline int dma_coerce_mask_and_coherent(struct device *dev, u64 mask) */ static inline bool dma_addressing_limited(struct device *dev) { - return min_not_zero(*dev->dma_mask, dev->bus_dma_mask) < - dma_get_required_mask(dev); + return min_not_zero(dma_get_mask(dev), dev->bus_dma_mask) < + dma_get_required_mask(dev); } #ifdef CONFIG_ARCH_HAS_SETUP_DMA_OPS -- cgit v1.2.3 From d7852fbd0f0423937fa287a598bfde188bb68c22 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 11 Jul 2019 09:54:40 -0700 Subject: access: avoid the RCU grace period for the temporary subjective credentials It turns out that 'access()' (and 'faccessat()') can cause a lot of RCU work because it installs a temporary credential that gets allocated and freed for each system call. The allocation and freeing overhead is mostly benign, but because credentials can be accessed under the RCU read lock, the freeing involves a RCU grace period. Which is not a huge deal normally, but if you have a lot of access() calls, this causes a fair amount of seconday damage: instead of having a nice alloc/free patterns that hits in hot per-CPU slab caches, you have all those delayed free's, and on big machines with hundreds of cores, the RCU overhead can end up being enormous. But it turns out that all of this is entirely unnecessary. Exactly because access() only installs the credential as the thread-local subjective credential, the temporary cred pointer doesn't actually need to be RCU free'd at all. Once we're done using it, we can just free it synchronously and avoid all the RCU overhead. So add a 'non_rcu' flag to 'struct cred', which can be set by users that know they only use it in non-RCU context (there are other potential users for this). We can make it a union with the rcu freeing list head that we need for the RCU case, so this doesn't need any extra storage. Note that this also makes 'get_current_cred()' clear the new non_rcu flag, in case we have filesystems that take a long-term reference to the cred and then expect the RCU delayed freeing afterwards. It's not entirely clear that this is required, but it makes for clear semantics: the subjective cred remains non-RCU as long as you only access it synchronously using the thread-local accessors, but you _can_ use it as a generic cred if you want to. It is possible that we should just remove the whole RCU markings for ->cred entirely. Only ->real_cred is really supposed to be accessed through RCU, and the long-term cred copies that nfs uses might want to explicitly re-enable RCU freeing if required, rather than have get_current_cred() do it implicitly. But this is a "minimal semantic changes" change for the immediate problem. Acked-by: Peter Zijlstra (Intel) Acked-by: Eric Dumazet Acked-by: Paul E. McKenney Cc: Oleg Nesterov Cc: Jan Glauber Cc: Jiri Kosina Cc: Jayachandran Chandrasekharan Nair Cc: Greg KH Cc: Kees Cook Cc: David Howells Cc: Miklos Szeredi Cc: Al Viro Signed-off-by: Linus Torvalds --- fs/open.c | 19 +++++++++++++++++++ include/linux/cred.h | 8 +++++++- kernel/cred.c | 21 +++++++++++++++++++-- 3 files changed, 45 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/fs/open.c b/fs/open.c index b5b80469b93d..a59abe3c669a 100644 --- a/fs/open.c +++ b/fs/open.c @@ -374,6 +374,25 @@ long do_faccessat(int dfd, const char __user *filename, int mode) override_cred->cap_permitted; } + /* + * The new set of credentials can *only* be used in + * task-synchronous circumstances, and does not need + * RCU freeing, unless somebody then takes a separate + * reference to it. + * + * NOTE! This is _only_ true because this credential + * is used purely for override_creds() that installs + * it as the subjective cred. Other threads will be + * accessing ->real_cred, not the subjective cred. + * + * If somebody _does_ make a copy of this (using the + * 'get_current_cred()' function), that will clear the + * non_rcu field, because now that other user may be + * expecting RCU freeing. But normal thread-synchronous + * cred accesses will keep things non-RCY. + */ + override_cred->non_rcu = 1; + old_cred = override_creds(override_cred); retry: res = user_path_at(dfd, filename, lookup_flags, &path); diff --git a/include/linux/cred.h b/include/linux/cred.h index 7eb43a038330..f7a30e0099be 100644 --- a/include/linux/cred.h +++ b/include/linux/cred.h @@ -145,7 +145,11 @@ struct cred { struct user_struct *user; /* real user ID subscription */ struct user_namespace *user_ns; /* user_ns the caps and keyrings are relative to. */ struct group_info *group_info; /* supplementary groups for euid/fsgid */ - struct rcu_head rcu; /* RCU deletion hook */ + /* RCU deletion */ + union { + int non_rcu; /* Can we skip RCU deletion? */ + struct rcu_head rcu; /* RCU deletion hook */ + }; } __randomize_layout; extern void __put_cred(struct cred *); @@ -246,6 +250,7 @@ static inline const struct cred *get_cred(const struct cred *cred) if (!cred) return cred; validate_creds(cred); + nonconst_cred->non_rcu = 0; return get_new_cred(nonconst_cred); } @@ -257,6 +262,7 @@ static inline const struct cred *get_cred_rcu(const struct cred *cred) if (!atomic_inc_not_zero(&nonconst_cred->usage)) return NULL; validate_creds(cred); + nonconst_cred->non_rcu = 0; return cred; } diff --git a/kernel/cred.c b/kernel/cred.c index c73a87a4df13..153ae369e024 100644 --- a/kernel/cred.c +++ b/kernel/cred.c @@ -144,7 +144,10 @@ void __put_cred(struct cred *cred) BUG_ON(cred == current->cred); BUG_ON(cred == current->real_cred); - call_rcu(&cred->rcu, put_cred_rcu); + if (cred->non_rcu) + put_cred_rcu(&cred->rcu); + else + call_rcu(&cred->rcu, put_cred_rcu); } EXPORT_SYMBOL(__put_cred); @@ -256,6 +259,7 @@ struct cred *prepare_creds(void) old = task->cred; memcpy(new, old, sizeof(struct cred)); + new->non_rcu = 0; atomic_set(&new->usage, 1); set_cred_subscribers(new, 0); get_group_info(new->group_info); @@ -535,7 +539,19 @@ const struct cred *override_creds(const struct cred *new) validate_creds(old); validate_creds(new); - get_cred(new); + + /* + * NOTE! This uses 'get_new_cred()' rather than 'get_cred()'. + * + * That means that we do not clear the 'non_rcu' flag, since + * we are only installing the cred into the thread-synchronous + * '->cred' pointer, not the '->real_cred' pointer that is + * visible to other threads under RCU. + * + * Also note that we did validate_creds() manually, not depending + * on the validation in 'get_cred()'. + */ + get_new_cred((struct cred *)new); alter_cred_subscribers(new, 1); rcu_assign_pointer(current->cred, new); alter_cred_subscribers(old, -1); @@ -672,6 +688,7 @@ struct cred *prepare_kernel_cred(struct task_struct *daemon) validate_creds(old); *new = *old; + new->non_rcu = 0; atomic_set(&new->usage, 1); set_cred_subscribers(new, 0); get_uid(new->user); -- cgit v1.2.3 From 8732d85a69a0411f16a4b78df8fdc7b09c50a849 Mon Sep 17 00:00:00 2001 From: Mattias Jacobsson <2pi@mok.nu> Date: Fri, 19 Jul 2019 19:51:45 +0200 Subject: platform/x86: wmi: add missing struct parameter description Add a description for the context parameter in the struct wmi_device_id. Reported-by: kbuild test robot Fixes: a48e23385fcf ("platform/x86: wmi: add context pointer field to struct wmi_device_id") Signed-off-by: Mattias Jacobsson <2pi@mok.nu> Signed-off-by: Andy Shevchenko --- include/linux/mod_devicetable.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h index b2c1648f7e5d..5714fd35a83c 100644 --- a/include/linux/mod_devicetable.h +++ b/include/linux/mod_devicetable.h @@ -814,6 +814,7 @@ struct tee_client_device_id { /** * struct wmi_device_id - WMI device identifier * @guid_string: 36 char string of the form fa50ff2b-f2e8-45de-83fa-65417f2f49ba + * @context: pointer to driver specific data */ struct wmi_device_id { const char guid_string[UUID_STRING_LEN+1]; -- cgit v1.2.3 From f1765a1819ff3489db9500c6d464e682e6844a14 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Fri, 26 Jul 2019 12:17:44 +0200 Subject: of: Fix typo in kerneldoc "Findfrom" is not a word. Replace the function synopsis by something that makes sense. Signed-off-by: Thierry Reding Signed-off-by: Rob Herring --- include/linux/of.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/of.h b/include/linux/of.h index 0cf857012f11..844f89e1b039 100644 --- a/include/linux/of.h +++ b/include/linux/of.h @@ -1164,7 +1164,7 @@ static inline int of_property_read_string_index(const struct device_node *np, } /** - * of_property_read_bool - Findfrom a property + * of_property_read_bool - Find a property * @np: device node from which the property value is to be read. * @propname: name of the property to be searched. * -- cgit v1.2.3 From c1a280b68d4e6b6db4a65aa7865c22d8789ddf09 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 26 Jul 2019 23:19:37 +0200 Subject: sched/preempt: Use CONFIG_PREEMPTION where appropriate CONFIG_PREEMPTION is selected by CONFIG_PREEMPT and by CONFIG_PREEMPT_RT. Both PREEMPT and PREEMPT_RT require the same functionality which today depends on CONFIG_PREEMPT. Switch the preemption code, scheduler and init task over to use CONFIG_PREEMPTION. That's the first step towards RT in that area. The more complex changes are coming separately. Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Masami Hiramatsu Cc: Paolo Bonzini Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Steven Rostedt Link: http://lkml.kernel.org/r/20190726212124.117528401@linutronix.de Signed-off-by: Ingo Molnar --- include/asm-generic/preempt.h | 4 ++-- include/linux/preempt.h | 6 +++--- include/linux/sched.h | 6 +++--- init/init_task.c | 2 +- init/main.c | 2 +- kernel/sched/core.c | 14 +++++++------- kernel/sched/fair.c | 2 +- kernel/sched/sched.h | 4 ++-- 8 files changed, 20 insertions(+), 20 deletions(-) (limited to 'include/linux') diff --git a/include/asm-generic/preempt.h b/include/asm-generic/preempt.h index c3046c920063..d683f5e6d791 100644 --- a/include/asm-generic/preempt.h +++ b/include/asm-generic/preempt.h @@ -78,11 +78,11 @@ static __always_inline bool should_resched(int preempt_offset) tif_need_resched()); } -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION extern asmlinkage void preempt_schedule(void); #define __preempt_schedule() preempt_schedule() extern asmlinkage void preempt_schedule_notrace(void); #define __preempt_schedule_notrace() preempt_schedule_notrace() -#endif /* CONFIG_PREEMPT */ +#endif /* CONFIG_PREEMPTION */ #endif /* __ASM_PREEMPT_H */ diff --git a/include/linux/preempt.h b/include/linux/preempt.h index dd92b1a93919..bbb68dba37cc 100644 --- a/include/linux/preempt.h +++ b/include/linux/preempt.h @@ -182,7 +182,7 @@ do { \ #define preemptible() (preempt_count() == 0 && !irqs_disabled()) -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION #define preempt_enable() \ do { \ barrier(); \ @@ -203,7 +203,7 @@ do { \ __preempt_schedule(); \ } while (0) -#else /* !CONFIG_PREEMPT */ +#else /* !CONFIG_PREEMPTION */ #define preempt_enable() \ do { \ barrier(); \ @@ -217,7 +217,7 @@ do { \ } while (0) #define preempt_check_resched() do { } while (0) -#endif /* CONFIG_PREEMPT */ +#endif /* CONFIG_PREEMPTION */ #define preempt_disable_notrace() \ do { \ diff --git a/include/linux/sched.h b/include/linux/sched.h index 9f51932bd543..6947516a2d3e 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1767,7 +1767,7 @@ static inline int test_tsk_need_resched(struct task_struct *tsk) * value indicates whether a reschedule was done in fact. * cond_resched_lock() will drop the spinlock before scheduling, */ -#ifndef CONFIG_PREEMPT +#ifndef CONFIG_PREEMPTION extern int _cond_resched(void); #else static inline int _cond_resched(void) { return 0; } @@ -1796,12 +1796,12 @@ static inline void cond_resched_rcu(void) /* * Does a critical section need to be broken due to another - * task waiting?: (technically does not depend on CONFIG_PREEMPT, + * task waiting?: (technically does not depend on CONFIG_PREEMPTION, * but a general need for low latency) */ static inline int spin_needbreak(spinlock_t *lock) { -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION return spin_is_contended(lock); #else return 0; diff --git a/init/init_task.c b/init/init_task.c index 7ab773b9b3cd..bfe06c53b14e 100644 --- a/init/init_task.c +++ b/init/init_task.c @@ -174,7 +174,7 @@ struct task_struct init_task #ifdef CONFIG_FUNCTION_GRAPH_TRACER .ret_stack = NULL, #endif -#if defined(CONFIG_TRACING) && defined(CONFIG_PREEMPT) +#if defined(CONFIG_TRACING) && defined(CONFIG_PREEMPTION) .trace_recursion = 0, #endif #ifdef CONFIG_LIVEPATCH diff --git a/init/main.c b/init/main.c index 96f8d5af52d6..653693da8da6 100644 --- a/init/main.c +++ b/init/main.c @@ -433,7 +433,7 @@ noinline void __ref rest_init(void) /* * Enable might_sleep() and smp_processor_id() checks. - * They cannot be enabled earlier because with CONFIG_PREEMPT=y + * They cannot be enabled earlier because with CONFIG_PREEMPTION=y * kernel_thread() would trigger might_sleep() splats. With * CONFIG_PREEMPT_VOLUNTARY=y the init task might have scheduled * already, but it's stuck on the kthreadd_done completion. diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 2b037f195473..604a5e137efe 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3581,7 +3581,7 @@ static inline void sched_tick_start(int cpu) { } static inline void sched_tick_stop(int cpu) { } #endif -#if defined(CONFIG_PREEMPT) && (defined(CONFIG_DEBUG_PREEMPT) || \ +#if defined(CONFIG_PREEMPTION) && (defined(CONFIG_DEBUG_PREEMPT) || \ defined(CONFIG_TRACE_PREEMPT_TOGGLE)) /* * If the value passed in is equal to the current preempt count @@ -3782,7 +3782,7 @@ again: * task, then the wakeup sets TIF_NEED_RESCHED and schedule() gets * called on the nearest possible occasion: * - * - If the kernel is preemptible (CONFIG_PREEMPT=y): + * - If the kernel is preemptible (CONFIG_PREEMPTION=y): * * - in syscall or exception context, at the next outmost * preempt_enable(). (this might be as soon as the wake_up()'s @@ -3791,7 +3791,7 @@ again: * - in IRQ context, return from interrupt-handler to * preemptible context * - * - If the kernel is not preemptible (CONFIG_PREEMPT is not set) + * - If the kernel is not preemptible (CONFIG_PREEMPTION is not set) * then at the next: * * - cond_resched() call @@ -4033,7 +4033,7 @@ static void __sched notrace preempt_schedule_common(void) } while (need_resched()); } -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION /* * this is the entry point to schedule() from in-kernel preemption * off of preempt_enable. Kernel preemptions off return from interrupt @@ -4105,7 +4105,7 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) } EXPORT_SYMBOL_GPL(preempt_schedule_notrace); -#endif /* CONFIG_PREEMPT */ +#endif /* CONFIG_PREEMPTION */ /* * this is the entry point to schedule() from kernel preemption @@ -5416,7 +5416,7 @@ SYSCALL_DEFINE0(sched_yield) return 0; } -#ifndef CONFIG_PREEMPT +#ifndef CONFIG_PREEMPTION int __sched _cond_resched(void) { if (should_resched(0)) { @@ -5433,7 +5433,7 @@ EXPORT_SYMBOL(_cond_resched); * __cond_resched_lock() - if a reschedule is pending, drop the given lock, * call schedule, and on return reacquire the lock. * - * This works OK both with and without CONFIG_PREEMPT. We do strange low-level + * This works OK both with and without CONFIG_PREEMPTION. We do strange low-level * operations here to prevent schedule() from being called twice (once via * spin_unlock(), once by hand). */ diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index bc9cfeaac8bd..aff9d76d8d65 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -7430,7 +7430,7 @@ static int detach_tasks(struct lb_env *env) detached++; env->imbalance -= load; -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION /* * NEWIDLE balancing is a source of latency, so preemptible * kernels will stop after the first task is detached to minimize diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 802b1f3405f2..f2ce6ba1c5d5 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1943,7 +1943,7 @@ unsigned long arch_scale_freq_capacity(int cpu) #endif #ifdef CONFIG_SMP -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION static inline void double_rq_lock(struct rq *rq1, struct rq *rq2); @@ -1995,7 +1995,7 @@ static inline int _double_lock_balance(struct rq *this_rq, struct rq *busiest) return ret; } -#endif /* CONFIG_PREEMPT */ +#endif /* CONFIG_PREEMPTION */ /* * double_lock_balance - lock the busiest runqueue, this_rq is locked already. -- cgit v1.2.3 From 01b1d88b09824bea1a75b0bac04dcf50d9893875 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 26 Jul 2019 23:19:38 +0200 Subject: rcu: Use CONFIG_PREEMPTION CONFIG_PREEMPTION is selected by CONFIG_PREEMPT and by CONFIG_PREEMPT_RT. Both PREEMPT and PREEMPT_RT require the same functionality which today depends on CONFIG_PREEMPT. Switch the conditionals in RCU to use CONFIG_PREEMPTION. That's the first step towards RCU on RT. The further tweaks are work in progress. This neither touches the selftest bits which need a closer look by Paul. Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Masami Hiramatsu Cc: Paolo Bonzini Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Steven Rostedt Link: http://lkml.kernel.org/r/20190726212124.210156346@linutronix.de Signed-off-by: Ingo Molnar --- arch/Kconfig | 2 +- include/linux/rcupdate.h | 2 +- include/linux/rcutree.h | 2 +- include/linux/torture.h | 2 +- kernel/rcu/Kconfig | 8 ++++---- kernel/rcu/tree.c | 6 +++--- kernel/rcu/tree_stall.h | 6 +++--- kernel/trace/Kconfig | 2 +- 8 files changed, 15 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/arch/Kconfig b/arch/Kconfig index a7b57dd42c26..c7efbc018f4f 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -103,7 +103,7 @@ config STATIC_KEYS_SELFTEST config OPTPROBES def_bool y depends on KPROBES && HAVE_OPTPROBES - select TASKS_RCU if PREEMPT + select TASKS_RCU if PREEMPTION config KPROBES_ON_FTRACE def_bool y diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 8f7167478c1d..c4f76a310443 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -578,7 +578,7 @@ do { \ * * In non-preemptible RCU implementations (TREE_RCU and TINY_RCU), * it is illegal to block while in an RCU read-side critical section. - * In preemptible RCU implementations (PREEMPT_RCU) in CONFIG_PREEMPT + * In preemptible RCU implementations (PREEMPT_RCU) in CONFIG_PREEMPTION * kernel builds, RCU read-side critical sections may be preempted, * but explicit blocking is illegal. Finally, in preemptible RCU * implementations in real-time (with -rt patchset) kernel builds, RCU diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index 735601ac27d3..18b1ed9864b0 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -53,7 +53,7 @@ void rcu_scheduler_starting(void); extern int rcu_scheduler_active __read_mostly; void rcu_end_inkernel_boot(void); bool rcu_is_watching(void); -#ifndef CONFIG_PREEMPT +#ifndef CONFIG_PREEMPTION void rcu_all_qs(void); #endif diff --git a/include/linux/torture.h b/include/linux/torture.h index a620118385bb..6241f59e2d6f 100644 --- a/include/linux/torture.h +++ b/include/linux/torture.h @@ -86,7 +86,7 @@ void _torture_stop_kthread(char *m, struct task_struct **tp); #define torture_stop_kthread(n, tp) \ _torture_stop_kthread("Stopping " #n " task", &(tp)) -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION #define torture_preempt_schedule() preempt_schedule() #else #define torture_preempt_schedule() diff --git a/kernel/rcu/Kconfig b/kernel/rcu/Kconfig index 480edf328b51..7644eda17d62 100644 --- a/kernel/rcu/Kconfig +++ b/kernel/rcu/Kconfig @@ -7,7 +7,7 @@ menu "RCU Subsystem" config TREE_RCU bool - default y if !PREEMPT && SMP + default y if !PREEMPTION && SMP help This option selects the RCU implementation that is designed for very large SMP system with hundreds or @@ -16,7 +16,7 @@ config TREE_RCU config PREEMPT_RCU bool - default y if PREEMPT + default y if PREEMPTION help This option selects the RCU implementation that is designed for very large SMP systems with hundreds or @@ -28,7 +28,7 @@ config PREEMPT_RCU config TINY_RCU bool - default y if !PREEMPT && !SMP + default y if !PREEMPTION && !SMP help This option selects the RCU implementation that is designed for UP systems from which real-time response @@ -70,7 +70,7 @@ config TREE_SRCU This option selects the full-fledged version of SRCU. config TASKS_RCU - def_bool PREEMPT + def_bool PREEMPTION select SRCU help This option enables a task-based RCU implementation that uses diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index a14e5fbbea46..5962636502bc 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -1881,7 +1881,7 @@ rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags) struct rcu_node *rnp_p; raw_lockdep_assert_held_rcu_node(rnp); - if (WARN_ON_ONCE(!IS_ENABLED(CONFIG_PREEMPT)) || + if (WARN_ON_ONCE(!IS_ENABLED(CONFIG_PREEMPTION)) || WARN_ON_ONCE(rcu_preempt_blocked_readers_cgp(rnp)) || rnp->qsmask != 0) { raw_spin_unlock_irqrestore_rcu_node(rnp, flags); @@ -2205,7 +2205,7 @@ static void force_qs_rnp(int (*f)(struct rcu_data *rdp)) mask = 0; raw_spin_lock_irqsave_rcu_node(rnp, flags); if (rnp->qsmask == 0) { - if (!IS_ENABLED(CONFIG_PREEMPT) || + if (!IS_ENABLED(CONFIG_PREEMPTION) || rcu_preempt_blocked_readers_cgp(rnp)) { /* * No point in scanning bits because they @@ -2622,7 +2622,7 @@ static int rcu_blocking_is_gp(void) { int ret; - if (IS_ENABLED(CONFIG_PREEMPT)) + if (IS_ENABLED(CONFIG_PREEMPTION)) return rcu_scheduler_active == RCU_SCHEDULER_INACTIVE; might_sleep(); /* Check for RCU read-side critical section. */ preempt_disable(); diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 065183391f75..9b92bf18b737 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -163,7 +163,7 @@ static void rcu_iw_handler(struct irq_work *iwp) // // Printing RCU CPU stall warnings -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPTION /* * Dump detailed information for all tasks blocking the current RCU @@ -215,7 +215,7 @@ static int rcu_print_task_stall(struct rcu_node *rnp) return ndetected; } -#else /* #ifdef CONFIG_PREEMPT */ +#else /* #ifdef CONFIG_PREEMPTION */ /* * Because preemptible RCU does not exist, we never have to check for @@ -233,7 +233,7 @@ static int rcu_print_task_stall(struct rcu_node *rnp) { return 0; } -#endif /* #else #ifdef CONFIG_PREEMPT */ +#endif /* #else #ifdef CONFIG_PREEMPTION */ /* * Dump stacks of all tasks running on stalled CPUs. First try using diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 98da8998c25c..d2c1fe0b451d 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -146,7 +146,7 @@ config FUNCTION_TRACER select GENERIC_TRACER select CONTEXT_SWITCH_TRACER select GLOB - select TASKS_RCU if PREEMPT + select TASKS_RCU if PREEMPTION help Enable the kernel to trace every kernel function. This is done by using a compiler feature to insert a small, 5-byte No-Operation -- cgit v1.2.3 From 27972765bd0410fc2ef5e86a41de17c71440a2dd Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 26 Jul 2019 23:19:39 +0200 Subject: locking/spinlocks: Use CONFIG_PREEMPTION CONFIG_PREEMPTION is selected by CONFIG_PREEMPT and by CONFIG_PREEMPT_RT. Both PREEMPT and PREEMPT_RT require the same functionality which today depends on CONFIG_PREEMPT. Adjust the comments in the locking code. Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Masami Hiramatsu Cc: Paolo Bonzini Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Steven Rostedt Link: http://lkml.kernel.org/r/20190726212124.302995288@linutronix.de Signed-off-by: Ingo Molnar --- include/linux/spinlock.h | 2 +- include/linux/spinlock_api_smp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h index ed7c4d6b8235..031ce8617df8 100644 --- a/include/linux/spinlock.h +++ b/include/linux/spinlock.h @@ -214,7 +214,7 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock) /* * Define the various spin_lock methods. Note we define these - * regardless of whether CONFIG_SMP or CONFIG_PREEMPT are set. The + * regardless of whether CONFIG_SMP or CONFIG_PREEMPTION are set. The * various methods are defined as nops in the case they are not * required. */ diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h index 42dfab89e740..b762eaba4cdf 100644 --- a/include/linux/spinlock_api_smp.h +++ b/include/linux/spinlock_api_smp.h @@ -96,7 +96,7 @@ static inline int __raw_spin_trylock(raw_spinlock_t *lock) /* * If lockdep is enabled then we use the non-preemption spin-ops - * even on CONFIG_PREEMPT, because lockdep assumes that interrupts are + * even on CONFIG_PREEMPTION, because lockdep assumes that interrupts are * not re-enabled during lock-acquire (which the preempt-spin-ops do): */ #if !defined(CONFIG_GENERIC_LOCKBREAK) || defined(CONFIG_DEBUG_LOCK_ALLOC) -- cgit v1.2.3