[workqueue/driver-core PATCH 0/5] Add NUMA aware async_schedule calls

[workqueue/driver-core PATCH 0/5] Add NUMA aware async_schedule calls

[Alexander Duyck]

This patch set provides functionality that will help to improve the
locality of the async_schedule calls used to provide deferred
initialization.

This patch set originally started out with me focused on just the one call
to async_schedule_domain in the nvdimm tree that was being used to
defer the device_add call however after doing some digging I realized the
scope of this was much broader than I had originally planned. As such I
went through and reworked the underlying infrastructure down to replacing
the queue_work call itself with a function of my own and opted to try and
provide a NUMA aware solution that would work for a broader audience.

v2: Dropped nvdimm patch to submit later.
	It relies on code in libnvdimm development tree.
    Simplified queue_work_near to just convert node into a CPU.
    Split up drivers core and PM core patches.

---

Alexander Duyck (5):
      workqueue: Provide queue_work_near to queue work near a given NUMA node
      async: Add support for queueing on specific NUMA node
      driver core: Probe devices asynchronously instead of the driver
      driver core: Attach devices on CPU local to device node
      PM core: Use new async_schedule_dev command


 drivers/base/bus.c        |   23 ++------------
 drivers/base/dd.c         |   44 +++++++++++++++++++++++++++
 drivers/base/power/main.c |   12 ++++---
 include/linux/async.h     |   36 ++++++++++++++++++++--
 include/linux/workqueue.h |    2 +
 kernel/async.c            |   47 ++++++++++++++---------------
 kernel/workqueue.c        |   73 +++++++++++++++++++++++++++++++++++++++++++++
 7 files changed, 182 insertions(+), 55 deletions(-)

--

[workqueue/driver-core PATCH 1/5] workqueue: Provide queue_work_near to queue work near a given NUMA node

[Alexander Duyck]

This patch provides a new function queue_work_near which is meant to
schedule work on a "random" CPU of the requested NUMA node. The main
motivation for this is to help assist asynchronous init to better improve
boot times for devices that are local to a specific node.

For now we just default to the first CPU that is in the intersection of the
cpumask of the node and the online cpumask. The only exception is if the
CPU is local to the node we will just use the current CPU. This should work
for our purposes as we are currently only using this for unbound work so
the CPU will be translated to a node anyway instead of being directly used.

Signed-off-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
---
 include/linux/workqueue.h |    2 +
 kernel/workqueue.c        |   73 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 75 insertions(+)

diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
index 60d673e15632..1f9f0a65437b 100644
--- a/include/linux/workqueue.h
+++ b/include/linux/workqueue.h
@@ -463,6 +463,8 @@ int apply_workqueue_attrs(struct workqueue_struct *wq,
 
 extern bool queue_work_on(int cpu, struct workqueue_struct *wq,
 			struct work_struct *work);
+extern bool queue_work_near(int node, struct workqueue_struct *wq,
+			    struct work_struct *work);
 extern bool queue_delayed_work_on(int cpu, struct workqueue_struct *wq,
 			struct delayed_work *work, unsigned long delay);
 extern bool mod_delayed_work_on(int cpu, struct workqueue_struct *wq,
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 0280deac392e..1467c052d9f6 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -1492,6 +1492,79 @@ bool queue_work_on(int cpu, struct workqueue_struct *wq,
 }
 EXPORT_SYMBOL(queue_work_on);
 
+/**
+ * workqueue_select_cpu_near - Select a CPU based on NUMA node
+ * @node: NUMA node ID that we want to bind a CPU from
+ *
+ * This function will attempt to find a "random" cpu available on a given
+ * node. If there are no CPUs available on the given node it will return
+ * WORK_CPU_UNBOUND indicating that we should just schedule to any
+ * available CPU if we need to schedule this work.
+ */
+static int workqueue_select_cpu_near(int node)
+{
+	int cpu;
+
+	/* No point in doing this if NUMA isn't enabled for workqueues */
+	if (!wq_numa_enabled)
+		return WORK_CPU_UNBOUND;
+
+	/* Delay binding to CPU if node is not valid or online */
+	if (node < 0 || node >= MAX_NUMNODES || !node_online(node))
+		return WORK_CPU_UNBOUND;
+
+	/* Use local node/cpu if we are already there */
+	cpu = raw_smp_processor_id();
+	if (node == cpu_to_node(cpu))
+		return cpu;
+
+	/* Use "random" otherwise know as "first" online CPU of node */
+	cpu = cpumask_any_and(cpumask_of_node(node), cpu_online_mask);
+
+	/* If CPU is valid return that, otherwise just defer */
+	return (cpu < nr_cpu_ids) ? cpu : WORK_CPU_UNBOUND;
+}
+
+/**
+ * queue_work_near - queue work on a "random" cpu for a given NUMA node
+ * @node: NUMA node that we are targeting the work for
+ * @wq: workqueue to use
+ * @work: work to queue
+ *
+ * We queue the work to a "random" CPU within a given NUMA node. The basic
+ * idea here is to provide a way to somehow associate work with a given
+ * NUMA node.
+ *
+ * This function will only make a best effort attempt at getting this onto
+ * the right NUMA node. If no node is requested or the requested node is
+ * offline then we just fall back to standard queue_work behavior.
+ *
+ * Currently the "random" CPU ends up being the first available CPU in the
+ * intersection of cpu_online_mask and the cpumask of the node, unless we
+ * are running on the node. In that case we just use the current CPU.
+ *
+ * Return: %false if @work was already on a queue, %true otherwise.
+ */
+bool queue_work_near(int node, struct workqueue_struct *wq,
+		     struct work_struct *work)
+{
+	unsigned long flags;
+	bool ret = false;
+
+	local_irq_save(flags);
+
+	if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
+		int cpu = workqueue_select_cpu_near(node);
+
+		__queue_work(cpu, wq, work);
+		ret = true;
+	}
+
+	local_irq_restore(flags);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(queue_work_near);
+
 void delayed_work_timer_fn(struct timer_list *t)
 {
 	struct delayed_work *dwork = from_timer(dwork, t, timer);

Re: [workqueue/driver-core PATCH 1/5] workqueue: Provide queue_work_near to queue work near a given NUMA node

[Tejun Heo]

Hello,

> +static int workqueue_select_cpu_near(int node)
> +{
...
> +	/* Use "random" otherwise know as "first" online CPU of node */
> +	cpu = cpumask_any_and(cpumask_of_node(node), cpu_online_mask);

Maybe we wanna round-robin within the node?

> +bool queue_work_near(int node, struct workqueue_struct *wq,
> +		     struct work_struct *work)

Can we name it queue_work_node() to stay consistent with numa-aware
memory interface?

Thanks.

-- 
tejun

Re: [workqueue/driver-core PATCH 1/5] workqueue: Provide queue_work_near to queue work near a given NUMA node

[Alexander Duyck]

On 10/10/2018 8:30 AM, Tejun Heo wrote:
> Hello,
> 
>> +static int workqueue_select_cpu_near(int node)
>> +{
> ...
>> +	/* Use "random" otherwise know as "first" online CPU of node */
>> +	cpu = cpumask_any_and(cpumask_of_node(node), cpu_online_mask);
> 
> Maybe we wanna round-robin within the node?

I had thought about it. It adds extra complexity to do it and for 
unbound work queues it doesn't really add anything since it gets 
converted back into a node anyway.

If you wanted I could look at adding extra logic that would do the 
round-robin for non-unbound workqueues. I just wasn't sure if it was 
worth the effort since the current users are all unbound workqueues.

>> +bool queue_work_near(int node, struct workqueue_struct *wq,
>> +		     struct work_struct *work)
> 
> Can we name it queue_work_node() to stay consistent with numa-aware
> memory interface?

Yes. No problem.

> Thanks.

Thanks for the review feedback.

- Alex

Re: [workqueue/driver-core PATCH 1/5] workqueue: Provide queue_work_near to queue work near a given NUMA node

[Tejun Heo]

Hello,

On Wed, Oct 10, 2018 at 08:34:06AM -0700, Alexander Duyck wrote:
> >Maybe we wanna round-robin within the node?
> 
> I had thought about it. It adds extra complexity to do it and for
> unbound work queues it doesn't really add anything since it gets
> converted back into a node anyway.

Yeap, it only makes sense for per-cpu workqueues.

> If you wanted I could look at adding extra logic that would do the
> round-robin for non-unbound workqueues. I just wasn't sure if it was
> worth the effort since the current users are all unbound workqueues.

It doesn't have to be complex, just a per-node unsynchronized int
which gets incremented and %'s would work fine.  Alternatively, we can
trigger WARN_ON_ONCE() if the function is used on a per-cpu workqueue.

Thanks.

-- 
tejun

[workqueue/driver-core PATCH 2/5] async: Add support for queueing on specific NUMA node

[Alexander Duyck]

This patch introduces four new variants of the async_schedule_ functions
that allow scheduling on a specific NUMA node.

The first two functions are async_schedule_near and
async_schedule_near_domain which end up mapping to async_schedule and
async_schedule_domain but provide NUMA node specific functionality. They
replace the original functions which were moved to inline function
definitions that call the new functions while passing NUMA_NO_NODE.

The second two functions are async_schedule_dev and
async_schedule_dev_domain which provide NUMA specific functionality when
passing a device as the data member and that device has a NUMA node other
than NUMA_NO_NODE.

The main motivation behind this is to address the need to be able to
schedule device specific init work on specific NUMA nodes in order to
improve performance of memory initialization.

Signed-off-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
---
 include/linux/async.h |   36 +++++++++++++++++++++++++++++++++---
 kernel/async.c        |   47 ++++++++++++++++++++++-------------------------
 2 files changed, 55 insertions(+), 28 deletions(-)

diff --git a/include/linux/async.h b/include/linux/async.h
index 6b0226bdaadc..1f6cb0d50263 100644
--- a/include/linux/async.h
+++ b/include/linux/async.h
@@ -14,6 +14,8 @@
 
 #include <linux/types.h>
 #include <linux/list.h>
+#include <linux/numa.h>
+#include <linux/device.h>
 
 typedef u64 async_cookie_t;
 typedef void (*async_func_t) (void *data, async_cookie_t cookie);
@@ -37,9 +39,37 @@ struct async_domain {
 	struct async_domain _name = { .pending = LIST_HEAD_INIT(_name.pending), \
 				      .registered = 0 }
 
-extern async_cookie_t async_schedule(async_func_t func, void *data);
-extern async_cookie_t async_schedule_domain(async_func_t func, void *data,
-					    struct async_domain *domain);
+async_cookie_t async_schedule_near(async_func_t func, void *data,
+				   int node);
+async_cookie_t async_schedule_near_domain(async_func_t func, void *data,
+					  int node,
+					  struct async_domain *domain);
+
+static inline async_cookie_t async_schedule(async_func_t func, void *data)
+{
+	return async_schedule_near(func, data, NUMA_NO_NODE);
+}
+
+static inline async_cookie_t
+async_schedule_domain(async_func_t func, void *data,
+		      struct async_domain *domain)
+{
+	return async_schedule_near_domain(func, data, NUMA_NO_NODE, domain);
+}
+
+static inline async_cookie_t
+async_schedule_dev(async_func_t func, struct device *dev)
+{
+	return async_schedule_near(func, dev, dev_to_node(dev));
+}
+
+static inline async_cookie_t
+async_schedule_dev_domain(async_func_t func, struct device *dev,
+			  struct async_domain *domain)
+{
+	return async_schedule_near_domain(func, dev, dev_to_node(dev), domain);
+}
+
 void async_unregister_domain(struct async_domain *domain);
 extern void async_synchronize_full(void);
 extern void async_synchronize_full_domain(struct async_domain *domain);
diff --git a/kernel/async.c b/kernel/async.c
index a893d6170944..f53d970ffdf4 100644
--- a/kernel/async.c
+++ b/kernel/async.c
@@ -149,7 +149,21 @@ static void async_run_entry_fn(struct work_struct *work)
 	wake_up(&async_done);
 }
 
-static async_cookie_t __async_schedule(async_func_t func, void *data, struct async_domain *domain)
+/**
+ * async_schedule_near_domain - schedule a function for asynchronous execution within a certain domain
+ * @func: function to execute asynchronously
+ * @data: data pointer to pass to the function
+ * @node: NUMA node that we want to schedule this on or close to
+ * @domain: the domain
+ *
+ * Returns an async_cookie_t that may be used for checkpointing later.
+ * @domain may be used in the async_synchronize_*_domain() functions to
+ * wait within a certain synchronization domain rather than globally.  A
+ * synchronization domain is specified via @domain.  Note: This function
+ * may be called from atomic or non-atomic contexts.
+ */
+async_cookie_t async_schedule_near_domain(async_func_t func, void *data,
+					  int node, struct async_domain *domain)
 {
 	struct async_entry *entry;
 	unsigned long flags;
@@ -195,43 +209,26 @@ static async_cookie_t __async_schedule(async_func_t func, void *data, struct asy
 	current->flags |= PF_USED_ASYNC;
 
 	/* schedule for execution */
-	queue_work(system_unbound_wq, &entry->work);
+	queue_work_near(node, system_unbound_wq, &entry->work);
 
 	return newcookie;
 }
+EXPORT_SYMBOL_GPL(async_schedule_near_domain);
 
 /**
- * async_schedule - schedule a function for asynchronous execution
+ * async_schedule_near - schedule a function for asynchronous execution
  * @func: function to execute asynchronously
  * @data: data pointer to pass to the function
+ * @node: NUMA node that we want to schedule this on or close to
  *
  * Returns an async_cookie_t that may be used for checkpointing later.
  * Note: This function may be called from atomic or non-atomic contexts.
  */
-async_cookie_t async_schedule(async_func_t func, void *data)
-{
-	return __async_schedule(func, data, &async_dfl_domain);
-}
-EXPORT_SYMBOL_GPL(async_schedule);
-
-/**
- * async_schedule_domain - schedule a function for asynchronous execution within a certain domain
- * @func: function to execute asynchronously
- * @data: data pointer to pass to the function
- * @domain: the domain
- *
- * Returns an async_cookie_t that may be used for checkpointing later.
- * @domain may be used in the async_synchronize_*_domain() functions to
- * wait within a certain synchronization domain rather than globally.  A
- * synchronization domain is specified via @domain.  Note: This function
- * may be called from atomic or non-atomic contexts.
- */
-async_cookie_t async_schedule_domain(async_func_t func, void *data,
-				     struct async_domain *domain)
+async_cookie_t async_schedule_near(async_func_t func, void *data, int node)
 {
-	return __async_schedule(func, data, domain);
+	return async_schedule_near_domain(func, data, node, &async_dfl_domain);
 }
-EXPORT_SYMBOL_GPL(async_schedule_domain);
+EXPORT_SYMBOL_GPL(async_schedule_near);
 
 /**
  * async_synchronize_full - synchronize all asynchronous function calls

[workqueue/driver-core PATCH 3/5] driver core: Probe devices asynchronously instead of the driver

[Alexander Duyck]

This change makes it so that we probe devices asynchronously instead of the
driver. This results in us seeing the same behavior if the device is
registered before the driver or after. This way we can avoid serializing
the initialization should the driver not be loaded until after the devices
have already been added.

The motivation behind this is that if we have a set of devices that
take a significant amount of time to load we can greatly reduce the time to
load by processing them in parallel instead of one at a time. In addition,
each device can exist on a different node so placing a single thread on one
CPU to initialize all of the devices for a given driver can result in poor
performance on a system with multiple nodes.

One issue I can see with this patch is that I am using the
dev_set/get_drvdata functions to store the driver in the device while I am
waiting on the asynchronous init to complete. For now I am protecting it by
using the lack of a dev->driver and the device lock.

Signed-off-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
---
 drivers/base/bus.c |   23 +++--------------------
 drivers/base/dd.c  |   42 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 45 insertions(+), 20 deletions(-)

diff --git a/drivers/base/bus.c b/drivers/base/bus.c
index 8bfd27ec73d6..2a17bed657ec 100644
--- a/drivers/base/bus.c
+++ b/drivers/base/bus.c
@@ -616,17 +616,6 @@ static ssize_t uevent_store(struct device_driver *drv, const char *buf,
 }
 static DRIVER_ATTR_WO(uevent);
 
-static void driver_attach_async(void *_drv, async_cookie_t cookie)
-{
-	struct device_driver *drv = _drv;
-	int ret;
-
-	ret = driver_attach(drv);
-
-	pr_debug("bus: '%s': driver %s async attach completed: %d\n",
-		 drv->bus->name, drv->name, ret);
-}
-
 /**
  * bus_add_driver - Add a driver to the bus.
  * @drv: driver.
@@ -659,15 +648,9 @@ int bus_add_driver(struct device_driver *drv)
 
 	klist_add_tail(&priv->knode_bus, &bus->p->klist_drivers);
 	if (drv->bus->p->drivers_autoprobe) {
-		if (driver_allows_async_probing(drv)) {
-			pr_debug("bus: '%s': probing driver %s asynchronously\n",
-				drv->bus->name, drv->name);
-			async_schedule(driver_attach_async, drv);
-		} else {
-			error = driver_attach(drv);
-			if (error)
-				goto out_unregister;
-		}
+		error = driver_attach(drv);
+		if (error)
+			goto out_unregister;
 	}
 	module_add_driver(drv->owner, drv);
 
diff --git a/drivers/base/dd.c b/drivers/base/dd.c
index 169412ee4ae8..5ba366c1cb83 100644
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@ -864,6 +864,29 @@ void device_initial_probe(struct device *dev)
 	__device_attach(dev, true);
 }
 
+static void __driver_attach_async_helper(void *_dev, async_cookie_t cookie)
+{
+	struct device *dev = _dev;
+
+	if (dev->parent && dev->bus->need_parent_lock)
+		device_lock(dev->parent);
+	device_lock(dev);
+
+	if (!dev->driver) {
+		struct device_driver *drv = dev_get_drvdata(dev);
+
+		driver_probe_device(drv, dev);
+	}
+
+	dev_dbg(dev, "async probe completed\n");
+
+	device_unlock(dev);
+	if (dev->parent && dev->bus->need_parent_lock)
+		device_unlock(dev->parent);
+
+	put_device(dev);
+}
+
 static int __driver_attach(struct device *dev, void *data)
 {
 	struct device_driver *drv = data;
@@ -891,6 +914,25 @@ static int __driver_attach(struct device *dev, void *data)
 		return ret;
 	} /* ret > 0 means positive match */
 
+	if (driver_allows_async_probing(drv)) {
+		/*
+		 * Instead of probing the device synchronously we will
+		 * probe it asynchronously to allow for more parallelism.
+		 *
+		 * We only take the device lock here in order to guarantee
+		 * that the dev->driver and driver_data fields are protected
+		 */
+		dev_dbg(dev, "scheduling asynchronous probe\n");
+		device_lock(dev);
+		if (!dev->driver) {
+			get_device(dev);
+			dev_set_drvdata(dev, drv);
+			async_schedule(__driver_attach_async_helper, dev);
+		}
+		device_unlock(dev);
+		return 0;
+	}
+
 	if (dev->parent && dev->bus->need_parent_lock)
 		device_lock(dev->parent);
 	device_lock(dev);

[workqueue/driver-core PATCH 4/5] driver core: Attach devices on CPU local to device node

[Alexander Duyck]

This change makes it so that we call the asynchronous probe routines on a
CPU local to the device node. By doing this we should be able to improve
our initialization time significantly as we can avoid having to access the
device from a remote node which may introduce higher latency.

Signed-off-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
---
 drivers/base/dd.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/base/dd.c b/drivers/base/dd.c
index 5ba366c1cb83..81472dc44a70 100644
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@ -826,7 +826,7 @@ static int __device_attach(struct device *dev, bool allow_async)
 			 */
 			dev_dbg(dev, "scheduling asynchronous probe\n");
 			get_device(dev);
-			async_schedule(__device_attach_async_helper, dev);
+			async_schedule_dev(__device_attach_async_helper, dev);
 		} else {
 			pm_request_idle(dev);
 		}
@@ -927,7 +927,7 @@ static int __driver_attach(struct device *dev, void *data)
 		if (!dev->driver) {
 			get_device(dev);
 			dev_set_drvdata(dev, drv);
-			async_schedule(__driver_attach_async_helper, dev);
+			async_schedule_dev(__driver_attach_async_helper, dev);
 		}
 		device_unlock(dev);
 		return 0;

[workqueue/driver-core PATCH 5/5] PM core: Use new async_schedule_dev command

[Alexander Duyck]

This change makes it so that we use the device specific version of the
async_schedule commands to defer various tasks related to power management.
By doing this we should see a slight improvement in performance as any
device that is sensitive to latency/locality in the setup will now be
initializing on the node closest to the device.

Signed-off-by: Alexander Duyck <alexander.h.duyck@linux.intel.com>
---
 drivers/base/power/main.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 3f68e2919dc5..8495d9b1e9d0 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -726,7 +726,7 @@ void dpm_noirq_resume_devices(pm_message_t state)
 		reinit_completion(&dev->power.completion);
 		if (is_async(dev)) {
 			get_device(dev);
-			async_schedule(async_resume_noirq, dev);
+			async_schedule_dev(async_resume_noirq, dev);
 		}
 	}
 
@@ -883,7 +883,7 @@ void dpm_resume_early(pm_message_t state)
 		reinit_completion(&dev->power.completion);
 		if (is_async(dev)) {
 			get_device(dev);
-			async_schedule(async_resume_early, dev);
+			async_schedule_dev(async_resume_early, dev);
 		}
 	}
 
@@ -1047,7 +1047,7 @@ void dpm_resume(pm_message_t state)
 		reinit_completion(&dev->power.completion);
 		if (is_async(dev)) {
 			get_device(dev);
-			async_schedule(async_resume, dev);
+			async_schedule_dev(async_resume, dev);
 		}
 	}
 
@@ -1366,7 +1366,7 @@ static int device_suspend_noirq(struct device *dev)
 
 	if (is_async(dev)) {
 		get_device(dev);
-		async_schedule(async_suspend_noirq, dev);
+		async_schedule_dev(async_suspend_noirq, dev);
 		return 0;
 	}
 	return __device_suspend_noirq(dev, pm_transition, false);
@@ -1569,7 +1569,7 @@ static int device_suspend_late(struct device *dev)
 
 	if (is_async(dev)) {
 		get_device(dev);
-		async_schedule(async_suspend_late, dev);
+		async_schedule_dev(async_suspend_late, dev);
 		return 0;
 	}
 
@@ -1830,7 +1830,7 @@ static int device_suspend(struct device *dev)
 
 	if (is_async(dev)) {
 		get_device(dev);
-		async_schedule(async_suspend, dev);
+		async_schedule_dev(async_suspend, dev);
 		return 0;
 	}