Message-ID: <e4fb0c39ec024d60587e5e1e70b171b99eb537f4.camel@redhat.com>
Date:   Fri, 27 Mar 2020 16:29:39 -0400
From:   Lyude Paul <lyude@...hat.com>
To:     Daniel Vetter <daniel@...ll.ch>, Tejun Heo <tj@...nel.org>
Cc:     nouveau@...ts.freedesktop.org, dri-devel@...ts.freedesktop.org,
        Ville Syrjälä 
        <ville.syrjala@...ux.intel.com>,
        Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
        Maxime Ripard <mripard@...nel.org>,
        Thomas Zimmermann <tzimmermann@...e.de>,
        David Airlie <airlied@...ux.ie>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/9] drm/vblank: Add vblank works

Adding Tejun to this thread per-Daniel's suggestion on IRC.

Hi Tejun! So, I don't know what your experience with modesetting related stuff
is so I'll quickly explain some important concepts here regarding scanlines,
vblanks, etc. If you already understand this, feel free to skip this next
paragraph.

>From the perspective of a computer, every time a computer monitor displays a
new frame it's done by "scanning out" the display image from top to bottom,
one row of pixels at a time. which row of pixels we're on is referred to as
the scanline. Additionally, there's usually a couple of extra scanlines which
we scan out, but aren't actually displayed on the screen (these sometimes get
used by HDMI audio and friends, but that's another story). The period where
we're on these scanlines is referred to as the vblank, this is important.

On a lot of display hardware, programming needs to take effect during the
vertical blanking period so that settings like gamma, what frame we're
scanning out, etc. can be safely changed without showing visual tearing on the
screen. In some unforgiving hardware, some of this programming has to both
start and end in the same vblank. This is apparently the case on the majority
of Intel GPUs in the wild right now, most notably with gamma updates which
involve mashing over 2KiB of registers during the vblank. Other drivers have
very similar needs to this (nouveau in particular is why I'm here, we need it
for CRC related stuff) so we figured it'd be a good idea to add a set of
helpers for performing realtime hardware programming that's synchronized to
vblank intervals. In particular, this is aimed at hardware programming that
would be a bit too awkward to try to pull off entirely in interrupt context.
We call these vblank workers.

We first tried doing this using plain old kthreads, as you can see in the
patch below, since we could schedule them as realtime. Additionally, the plan
was to use this in i915 combined with pm_qos to get rid of cstate latency when
handling the original vblank interrupt. Note, our time constraints are a bit
more forgiving in nouveau so you won't see any pm_qos mentions in this series.

In an effort to try to avoid reinventing parts of the kernel's worker
infrastructure though, we tried to see if we could implement these with simple
work_structs and workqueues with HIGH_PRI | UNBOUNDED,[1]. But, it would seem
that this work_struct approach is quite unreliable and we still usually fail
to start the register programming in time for Intel's vblank worker usecase.

So, so far using plain old RT kthreads seems to make things about as reliable
as I think we're going to get. Note - even using kthreads we _still_ sometimes
miss the vblank period and end up tearing a bit on screen, but it happens
significantly less often then with work_structs and is basically going to be
as fast as we can get (in the future, Intel wants to try fixing this by doing
this hardware programming outside of the CPU, but for now we're stuck with
this).

With all of this being said - we'd still like to avoid having to reinvent
workers if possible, so we were wondering if there was any kind of realtime
worker that could be used for this instead? I haven't been able to find any
way of scheduling workers to be realtime, and I'm not sure if implementing
this in Linux's workqueues would be realistic either. It'd be nice to avoid
having to use plain old kthreads if possible :). What do you think Tejun?

On Wed, 2020-03-18 at 14:46 +0100, Daniel Vetter wrote:
> On Tue, Mar 17, 2020 at 08:40:58PM -0400, Lyude Paul wrote:
> > From: Ville Syrjälä <ville.syrjala@...ux.intel.com>
> > 
> > Add some kind of vblank workers. The interface is similar to regular
> > delayed works, and also allows for re-scheduling.
> > 
> > Whatever hardware programming we do in the work must be fast
> > (must at least complete during the vblank, sometimes during
> > the first few scanlines of vblank), so we'll fire up a per-crtc
> > high priority thread for this.
> > 
> > [based off patches from Ville Syrjälä <ville.syrjala@...ux.intel.com>,
> > change below to signoff later]
> > 
> > Cc: Ville Syrjälä <ville.syrjala@...ux.intel.com>
> > Signed-off-by: Lyude Paul <lyude@...hat.com>
> 
> Hm not really sold on the idea that we have should reinvent our own worker
> infrastructure here. Imo a vblank_work should look like a delayed work,
> i.e. using struct work_struct as the base class, and wrapping the vblank
> thing around it (instead of the timer). That alos would allow drivers to
> schedule works on their own work queues, allowing for easier flushing and
> all that stuff.
> 
> Also if we do this I think we should try to follow the delayed work abi as
> closely as possible (e.g. INIT_VBLANK_WORK, queue_vblank_work,
> mod_vblank_work, ...). Delayed workers (whether timer or vblank) have a
> bunch of edges cases where consistently would be really great to avoid
> surprises and bugs.
> -Daniel
> 
> > ---
> >  drivers/gpu/drm/drm_vblank.c | 322 +++++++++++++++++++++++++++++++++++
> >  include/drm/drm_vblank.h     |  34 ++++
> >  2 files changed, 356 insertions(+)
> > 
> > diff --git a/drivers/gpu/drm/drm_vblank.c b/drivers/gpu/drm/drm_vblank.c
> > index da7b0b0c1090..06c796b6c381 100644
> > --- a/drivers/gpu/drm/drm_vblank.c
> > +++ b/drivers/gpu/drm/drm_vblank.c
> > @@ -25,7 +25,9 @@
> >   */
> >  
> >  #include <linux/export.h>
> > +#include <linux/kthread.h>
> >  #include <linux/moduleparam.h>
> > +#include <uapi/linux/sched/types.h>
> >  
> >  #include <drm/drm_crtc.h>
> >  #include <drm/drm_drv.h>
> > @@ -91,6 +93,7 @@
> >  static bool
> >  drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
> >  			  ktime_t *tvblank, bool in_vblank_irq);
> > +static int drm_vblank_get(struct drm_device *dev, unsigned int pipe);
> >  
> >  static unsigned int drm_timestamp_precision = 20;  /* Default to 20
> > usecs. */
> >  
> > @@ -440,6 +443,9 @@ void drm_vblank_cleanup(struct drm_device *dev)
> >  			drm_core_check_feature(dev, DRIVER_MODESET));
> >  
> >  		del_timer_sync(&vblank->disable_timer);
> > +
> > +		wake_up_all(&vblank->vblank_work.work_wait);
> > +		kthread_stop(vblank->vblank_work.thread);
> >  	}
> >  
> >  	kfree(dev->vblank);
> > @@ -447,6 +453,108 @@ void drm_vblank_cleanup(struct drm_device *dev)
> >  	dev->num_crtcs = 0;
> >  }
> >  
> > +static int vblank_work_thread(void *data)
> > +{
> > +	struct drm_vblank_crtc *vblank = data;
> > +
> > +	while (!kthread_should_stop()) {
> > +		struct drm_vblank_work *work, *next;
> > +		LIST_HEAD(list);
> > +		u64 count;
> > +		int ret;
> > +
> > +		spin_lock_irq(&vblank->dev->event_lock);
> > +
> > +		ret = wait_event_interruptible_lock_irq(vblank->queue,
> > +							kthread_should_stop()
> > ||
> > +							!list_empty(&vblank-
> > >vblank_work.work_list),
> > +							vblank->dev-
> > >event_lock);
> > +
> > +		WARN_ON(ret && !kthread_should_stop() &&
> > +			list_empty(&vblank->vblank_work.irq_list) &&
> > +			list_empty(&vblank->vblank_work.work_list));
> > +
> > +		list_for_each_entry_safe(work, next,
> > +					 &vblank->vblank_work.work_list,
> > +					 list) {
> > +			list_move_tail(&work->list, &list);
> > +			work->state = DRM_VBL_WORK_RUNNING;
> > +		}
> > +
> > +		spin_unlock_irq(&vblank->dev->event_lock);
> > +
> > +		if (list_empty(&list))
> > +			continue;
> > +
> > +		count = atomic64_read(&vblank->count);
> > +		list_for_each_entry(work, &list, list)
> > +			work->func(work, count);
> > +
> > +		spin_lock_irq(&vblank->dev->event_lock);
> > +
> > +		list_for_each_entry_safe(work, next, &list, list) {
> > +			if (work->reschedule) {
> > +				list_move_tail(&work->list,
> > +					       &vblank->vblank_work.irq_list);
> > +				drm_vblank_get(vblank->dev, vblank->pipe);
> > +				work->reschedule = false;
> > +				work->state = DRM_VBL_WORK_WAITING;
> > +			} else {
> > +				list_del_init(&work->list);
> > +				work->cancel = false;
> > +				work->state = DRM_VBL_WORK_IDLE;
> > +			}
> > +		}
> > +
> > +		spin_unlock_irq(&vblank->dev->event_lock);
> > +
> > +		wake_up_all(&vblank->vblank_work.work_wait);
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> > +static void vblank_work_init(struct drm_vblank_crtc *vblank)
> > +{
> > +	struct sched_param param = {
> > +		.sched_priority = MAX_RT_PRIO - 1,
> > +	};
> > +	int ret;
> > +
> > +	INIT_LIST_HEAD(&vblank->vblank_work.irq_list);
> > +	INIT_LIST_HEAD(&vblank->vblank_work.work_list);
> > +	init_waitqueue_head(&vblank->vblank_work.work_wait);
> > +
> > +	vblank->vblank_work.thread =
> > +		kthread_run(vblank_work_thread, vblank, "card %d crtc %d",
> > +			    vblank->dev->primary->index, vblank->pipe);
> > +
> > +	ret = sched_setscheduler(vblank->vblank_work.thread,
> > +				 SCHED_FIFO, &param);
> > +	WARN_ON(ret);
> > +}
> > +
> > +/**
> > + * drm_vblank_work_init - initialize a vblank work item
> > + * @work: vblank work item
> > + * @crtc: CRTC whose vblank will trigger the work execution
> > + * @func: work function to be executed
> > + *
> > + * Initialize a vblank work item for a specific crtc.
> > + */
> > +void drm_vblank_work_init(struct drm_vblank_work *work, struct drm_crtc
> > *crtc,
> > +			  void (*func)(struct drm_vblank_work *work, u64
> > count))
> > +{
> > +	struct drm_device *dev = crtc->dev;
> > +	struct drm_vblank_crtc *vblank = &dev->vblank[drm_crtc_index(crtc)];
> > +
> > +	work->vblank = vblank;
> > +	work->state = DRM_VBL_WORK_IDLE;
> > +	work->func = func;
> > +	INIT_LIST_HEAD(&work->list);
> > +}
> > +EXPORT_SYMBOL(drm_vblank_work_init);
> > +
> >  /**
> >   * drm_vblank_init - initialize vblank support
> >   * @dev: DRM device
> > @@ -481,6 +589,8 @@ int drm_vblank_init(struct drm_device *dev, unsigned
> > int num_crtcs)
> >  		init_waitqueue_head(&vblank->queue);
> >  		timer_setup(&vblank->disable_timer, vblank_disable_fn, 0);
> >  		seqlock_init(&vblank->seqlock);
> > +
> > +		vblank_work_init(vblank);
> >  	}
> >  
> >  	DRM_INFO("Supports vblank timestamp caching Rev 2 (21.10.2013).\n");
> > @@ -1825,6 +1935,22 @@ static void drm_handle_vblank_events(struct
> > drm_device *dev, unsigned int pipe)
> >  	trace_drm_vblank_event(pipe, seq, now, high_prec);
> >  }
> >  
> > +static void drm_handle_vblank_works(struct drm_vblank_crtc *vblank)
> > +{
> > +	struct drm_vblank_work *work, *next;
> > +	u64 count = atomic64_read(&vblank->count);
> > +
> > +	list_for_each_entry_safe(work, next, &vblank->vblank_work.irq_list,
> > +				 list) {
> > +		if (!vblank_passed(count, work->count))
> > +			continue;
> > +
> > +		drm_vblank_put(vblank->dev, vblank->pipe);
> > +		list_move_tail(&work->list, &vblank->vblank_work.work_list);
> > +		work->state = DRM_VBL_WORK_SCHEDULED;
> > +	}
> > +}
> > +
> >  /**
> >   * drm_handle_vblank - handle a vblank event
> >   * @dev: DRM device
> > @@ -1866,6 +1992,7 @@ bool drm_handle_vblank(struct drm_device *dev,
> > unsigned int pipe)
> >  
> >  	spin_unlock(&dev->vblank_time_lock);
> >  
> > +	drm_handle_vblank_works(vblank);
> >  	wake_up(&vblank->queue);
> >  
> >  	/* With instant-off, we defer disabling the interrupt until after
> > @@ -2076,3 +2203,198 @@ int drm_crtc_queue_sequence_ioctl(struct
> > drm_device *dev, void *data,
> >  	kfree(e);
> >  	return ret;
> >  }
> > +
> > +/**
> > + * drm_vblank_work_schedule - schedule a vblank work
> > + * @work: vblank work to schedule
> > + * @count: target vblank count
> > + * @nextonmiss: defer until the next vblank if target vblank was missed
> > + *
> > + * Schedule @work for execution once the crtc vblank count reaches
> > @count.
> > + *
> > + * If the crtc vblank count has already reached @count and @nextonmiss is
> > + * %false the work starts to execute immediately.
> > + *
> > + * If the crtc vblank count has already reached @count and @nextonmiss is
> > + * %true the work is deferred until the next vblank (as if @count has
> > been
> > + * specified as crtc vblank count + 1).
> > + *
> > + * If @work is already scheduled, this function will reschedule said work
> > + * using the new @count.
> > + *
> > + * Returns:
> > + * 0 on success, error code on failure.
> > + */
> > +int drm_vblank_work_schedule(struct drm_vblank_work *work,
> > +			     u64 count, bool nextonmiss)
> > +{
> > +	struct drm_vblank_crtc *vblank = work->vblank;
> > +	unsigned long irqflags;
> > +	u64 cur_vbl;
> > +	int ret = 0;
> > +	bool rescheduling = false;
> > +	bool passed;
> > +
> > +	spin_lock_irqsave(&vblank->dev->event_lock, irqflags);
> > +
> > +	if (work->cancel)
> > +		goto out;
> > +
> > +	if (work->state == DRM_VBL_WORK_RUNNING) {
> > +		work->reschedule = true;
> > +		work->count = count;
> > +		goto out;
> > +	} else if (work->state != DRM_VBL_WORK_IDLE) {
> > +		if (work->count == count)
> > +			goto out;
> > +		rescheduling = true;
> > +	}
> > +
> > +	if (work->state != DRM_VBL_WORK_WAITING) {
> > +		ret = drm_vblank_get(vblank->dev, vblank->pipe);
> > +		if (ret)
> > +			goto out;
> > +	}
> > +
> > +	work->count = count;
> > +
> > +	cur_vbl = atomic64_read(&vblank->count);
> > +	passed = vblank_passed(cur_vbl, count);
> > +	if (passed)
> > +		DRM_ERROR("crtc %d vblank %llu already passed (current
> > %llu)\n",
> > +			  vblank->pipe, count, cur_vbl);
> > +
> > +	if (!nextonmiss && passed) {
> > +		drm_vblank_put(vblank->dev, vblank->pipe);
> > +		if (rescheduling)
> > +			list_move_tail(&work->list,
> > +				       &vblank->vblank_work.work_list);
> > +		else
> > +			list_add_tail(&work->list,
> > +				      &vblank->vblank_work.work_list);
> > +		work->state = DRM_VBL_WORK_SCHEDULED;
> > +		wake_up_all(&vblank->queue);
> > +	} else {
> > +		if (rescheduling)
> > +			list_move_tail(&work->list,
> > +				       &vblank->vblank_work.irq_list);
> > +		else
> > +			list_add_tail(&work->list,
> > +				      &vblank->vblank_work.irq_list);
> > +		work->state = DRM_VBL_WORK_WAITING;
> > +	}
> > +
> > + out:
> > +	spin_unlock_irqrestore(&vblank->dev->event_lock, irqflags);
> > +
> > +	return ret;
> > +}
> > +EXPORT_SYMBOL(drm_vblank_work_schedule);
> > +
> > +static bool vblank_work_cancel(struct drm_vblank_work *work)
> > +{
> > +	struct drm_vblank_crtc *vblank = work->vblank;
> > +
> > +	switch (work->state) {
> > +	case DRM_VBL_WORK_RUNNING:
> > +		work->cancel = true;
> > +		work->reschedule = false;
> > +		/* fall through */
> > +	default:
> > +	case DRM_VBL_WORK_IDLE:
> > +		return false;
> > +	case DRM_VBL_WORK_WAITING:
> > +		drm_vblank_put(vblank->dev, vblank->pipe);
> > +		/* fall through */
> > +	case DRM_VBL_WORK_SCHEDULED:
> > +		list_del_init(&work->list);
> > +		work->state = DRM_VBL_WORK_IDLE;
> > +		return true;
> > +	}
> > +}
> > +
> > +/**
> > + * drm_vblank_work_cancel - cancel a vblank work
> > + * @work: vblank work to cancel
> > + *
> > + * Cancel an already scheduled vblank work.
> > + *
> > + * On return @work may still be executing, unless the return
> > + * value is %true.
> > + *
> > + * Returns:
> > + * True if the work was cancelled before it started to excute, false
> > otherwise.
> > + */
> > +bool drm_vblank_work_cancel(struct drm_vblank_work *work)
> > +{
> > +	struct drm_vblank_crtc *vblank = work->vblank;
> > +	bool cancelled;
> > +
> > +	spin_lock_irq(&vblank->dev->event_lock);
> > +
> > +	cancelled = vblank_work_cancel(work);
> > +
> > +	spin_unlock_irq(&vblank->dev->event_lock);
> > +
> > +	return cancelled;
> > +}
> > +EXPORT_SYMBOL(drm_vblank_work_cancel);
> > +
> > +/**
> > + * drm_vblank_work_cancel_sync - cancel a vblank work and wait for it to
> > finish executing
> > + * @work: vblank work to cancel
> > + *
> > + * Cancel an already scheduled vblank work and wait for its
> > + * execution to finish.
> > + *
> > + * On return @work is no longer guaraneed to be executing.
> > + *
> > + * Returns:
> > + * True if the work was cancelled before it started to excute, false
> > otherwise.
> > + */
> > +bool drm_vblank_work_cancel_sync(struct drm_vblank_work *work)
> > +{
> > +	struct drm_vblank_crtc *vblank = work->vblank;
> > +	bool cancelled;
> > +	long ret;
> > +
> > +	spin_lock_irq(&vblank->dev->event_lock);
> > +
> > +	cancelled = vblank_work_cancel(work);
> > +
> > +	ret = wait_event_lock_irq_timeout(vblank->vblank_work.work_wait,
> > +					  work->state == DRM_VBL_WORK_IDLE,
> > +					  vblank->dev->event_lock,
> > +					  10 * HZ);
> > +
> > +	spin_unlock_irq(&vblank->dev->event_lock);
> > +
> > +	WARN(!ret, "crtc %d vblank work timed out\n", vblank->pipe);
> > +
> > +	return cancelled;
> > +}
> > +EXPORT_SYMBOL(drm_vblank_work_cancel_sync);
> > +
> > +/**
> > + * drm_vblank_work_flush - wait for a scheduled vblank work to finish
> > excuting
> > + * @work: vblank work to flush
> > + *
> > + * Wait until @work has finished executing.
> > + */
> > +void drm_vblank_work_flush(struct drm_vblank_work *work)
> > +{
> > +	struct drm_vblank_crtc *vblank = work->vblank;
> > +	long ret;
> > +
> > +	spin_lock_irq(&vblank->dev->event_lock);
> > +
> > +	ret = wait_event_lock_irq_timeout(vblank->vblank_work.work_wait,
> > +					  work->state == DRM_VBL_WORK_IDLE,
> > +					  vblank->dev->event_lock,
> > +					  10 * HZ);
> > +
> > +	spin_unlock_irq(&vblank->dev->event_lock);
> > +
> > +	WARN(!ret, "crtc %d vblank work timed out\n", vblank->pipe);
> > +}
> > +EXPORT_SYMBOL(drm_vblank_work_flush);
> > diff --git a/include/drm/drm_vblank.h b/include/drm/drm_vblank.h
> > index dd9f5b9e56e4..ac9130f419af 100644
> > --- a/include/drm/drm_vblank.h
> > +++ b/include/drm/drm_vblank.h
> > @@ -203,8 +203,42 @@ struct drm_vblank_crtc {
> >  	 * disabling functions multiple times.
> >  	 */
> >  	bool enabled;
> > +
> > +	struct {
> > +		struct task_struct *thread;
> > +		struct list_head irq_list, work_list;
> > +		wait_queue_head_t work_wait;
> > +	} vblank_work;
> > +};
> > +
> > +struct drm_vblank_work {
> > +	u64 count;
> > +	struct drm_vblank_crtc *vblank;
> > +	void (*func)(struct drm_vblank_work *work, u64 count);
> > +	struct list_head list;
> > +	enum {
> > +		DRM_VBL_WORK_IDLE,
> > +		DRM_VBL_WORK_WAITING,
> > +		DRM_VBL_WORK_SCHEDULED,
> > +		DRM_VBL_WORK_RUNNING,
> > +	} state;
> > +	bool cancel : 1;
> > +	bool reschedule : 1;
> >  };
> >  
> > +int drm_vblank_work_schedule(struct drm_vblank_work *work,
> > +			     u64 count, bool nextonmiss);
> > +void drm_vblank_work_init(struct drm_vblank_work *work, struct drm_crtc
> > *crtc,
> > +			  void (*func)(struct drm_vblank_work *work, u64
> > count));
> > +bool drm_vblank_work_cancel(struct drm_vblank_work *work);
> > +bool drm_vblank_work_cancel_sync(struct drm_vblank_work *work);
> > +void drm_vblank_work_flush(struct drm_vblank_work *work);
> > +
> > +static inline bool drm_vblank_work_pending(struct drm_vblank_work *work)
> > +{
> > +	return work->state != DRM_VBL_WORK_IDLE;
> > +}
> > +
> >  int drm_vblank_init(struct drm_device *dev, unsigned int num_crtcs);
> >  bool drm_dev_has_vblank(const struct drm_device *dev);
> >  u64 drm_crtc_vblank_count(struct drm_crtc *crtc);
> > -- 
> > 2.24.1
> > 
-- 
Cheers,
	Lyude Paul (she/her)
	Associate Software Engineer at Red Hat

Powered by blists - more mailing lists