Merge tag 'sched-fifo-2020-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git...

Pull sched/fifo updates from Ingo Molnar:
 "This adds the sched_set_fifo*() encapsulation APIs to remove static
  priority level knowledge from non-scheduler code.

  The three APIs for non-scheduler code to set SCHED_FIFO are:

   - sched_set_fifo()
   - sched_set_fifo_low()
   - sched_set_normal()

  These are two FIFO priority levels: default (high), and a 'low'
  priority level, plus sched_set_normal() to set the policy back to
  non-SCHED_FIFO.

  Since the changes affect a lot of non-scheduler code, we kept this in
  a separate tree"

* tag 'sched-fifo-2020-08-04' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (24 commits)
  sched,tracing: Convert to sched_set_fifo()
  sched: Remove sched_set_*() return value
  sched: Remove sched_setscheduler*() EXPORTs
  sched,psi: Convert to sched_set_fifo_low()
  sched,rcutorture: Convert to sched_set_fifo_low()
  sched,rcuperf: Convert to sched_set_fifo_low()
  sched,locktorture: Convert to sched_set_fifo()
  sched,irq: Convert to sched_set_fifo()
  sched,watchdog: Convert to sched_set_fifo()
  sched,serial: Convert to sched_set_fifo()
  sched,powerclamp: Convert to sched_set_fifo()
  sched,ion: Convert to sched_set_normal()
  sched,powercap: Convert to sched_set_fifo*()
  sched,spi: Convert to sched_set_fifo*()
  sched,mmc: Convert to sched_set_fifo*()
  sched,ivtv: Convert to sched_set_fifo*()
  sched,drm/scheduler: Convert to sched_set_fifo*()
  sched,msm: Convert to sched_set_fifo*()
  sched,psci: Convert to sched_set_fifo*()
  sched,drbd: Convert to sched_set_fifo*()
  ...

diff --cc drivers/acpi/acpi_pad.c
Simple merge

diff --cc drivers/block/drbd/drbd_receiver.c
Simple merge

diff --cc drivers/firmware/psci/psci_checker.c
Simple merge

diff --cc drivers/gpu/drm/drm_vblank_work.c
index 7ac0fc0,0000000..bd481fd
mode 100644,000000..100644
--- 1/drivers/gpu/drm/drm_vblank_work.c
--- /dev/null
+++ b/drivers/gpu/drm/drm_vblank_work.c
@@@ -1,267 -1,0 +1,265 @@@
-       struct sched_param param = {
-               .sched_priority = MAX_RT_PRIO - 1,
-       };
 +// SPDX-License-Identifier: MIT
 +
 +#include <uapi/linux/sched/types.h>
 +
 +#include <drm/drm_print.h>
 +#include <drm/drm_vblank.h>
 +#include <drm/drm_vblank_work.h>
 +#include <drm/drm_crtc.h>
 +
 +#include "drm_internal.h"
 +
 +/**
 + * DOC: vblank works
 + *
 + * Many DRM drivers need to program hardware in a time-sensitive manner, many
 + * times with a deadline of starting and finishing within a certain region of
 + * the scanout. Most of the time the safest way to accomplish this is to
 + * simply do said time-sensitive programming in the driver's IRQ handler,
 + * which allows drivers to avoid being preempted during these critical
 + * regions. Or even better, the hardware may even handle applying such
 + * time-critical programming independently of the CPU.
 + *
 + * While there's a decent amount of hardware that's designed so that the CPU
 + * doesn't need to be concerned with extremely time-sensitive programming,
 + * there's a few situations where it can't be helped. Some unforgiving
 + * hardware may require that certain time-sensitive programming be handled
 + * completely by the CPU, and said programming may even take too long to
 + * handle in an IRQ handler. Another such situation would be where the driver
 + * needs to perform a task that needs to complete within a specific scanout
 + * period, but might possibly block and thus cannot be handled in an IRQ
 + * context. Both of these situations can't be solved perfectly in Linux since
 + * we're not a realtime kernel, and thus the scheduler may cause us to miss
 + * our deadline if it decides to preempt us. But for some drivers, it's good
 + * enough if we can lower our chance of being preempted to an absolute
 + * minimum.
 + *
 + * This is where &drm_vblank_work comes in. &drm_vblank_work provides a simple
 + * generic delayed work implementation which delays work execution until a
 + * particular vblank has passed, and then executes the work at realtime
 + * priority. This provides the best possible chance at performing
 + * time-sensitive hardware programming on time, even when the system is under
 + * heavy load. &drm_vblank_work also supports rescheduling, so that self
 + * re-arming work items can be easily implemented.
 + */
 +
 +void drm_handle_vblank_works(struct drm_vblank_crtc *vblank)
 +{
 +      struct drm_vblank_work *work, *next;
 +      u64 count = atomic64_read(&vblank->count);
 +      bool wake = false;
 +
 +      assert_spin_locked(&vblank->dev->event_lock);
 +
 +      list_for_each_entry_safe(work, next, &vblank->pending_work, node) {
 +              if (!drm_vblank_passed(count, work->count))
 +                      continue;
 +
 +              list_del_init(&work->node);
 +              drm_vblank_put(vblank->dev, vblank->pipe);
 +              kthread_queue_work(vblank->worker, &work->base);
 +              wake = true;
 +      }
 +      if (wake)
 +              wake_up_all(&vblank->work_wait_queue);
 +}
 +
 +/* Handle cancelling any pending vblank work items and drop respective vblank
 + * references in response to vblank interrupts being disabled.
 + */
 +void drm_vblank_cancel_pending_works(struct drm_vblank_crtc *vblank)
 +{
 +      struct drm_vblank_work *work, *next;
 +
 +      assert_spin_locked(&vblank->dev->event_lock);
 +
 +      list_for_each_entry_safe(work, next, &vblank->pending_work, node) {
 +              list_del_init(&work->node);
 +              drm_vblank_put(vblank->dev, vblank->pipe);
 +      }
 +
 +      wake_up_all(&vblank->work_wait_queue);
 +}
 +
 +/**
 + * 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. This can be used for self-rearming work items.
 + *
 + * Returns:
 + * %1 if @work was successfully (re)scheduled, %0 if it was either already
 + * scheduled or cancelled, or a negative 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;
 +      struct drm_device *dev = vblank->dev;
 +      u64 cur_vbl;
 +      unsigned long irqflags;
 +      bool passed, inmodeset, rescheduling = false, wake = false;
 +      int ret = 0;
 +
 +      spin_lock_irqsave(&dev->event_lock, irqflags);
 +      if (work->cancelling)
 +              goto out;
 +
 +      spin_lock(&dev->vbl_lock);
 +      inmodeset = vblank->inmodeset;
 +      spin_unlock(&dev->vbl_lock);
 +      if (inmodeset)
 +              goto out;
 +
 +      if (list_empty(&work->node)) {
 +              ret = drm_vblank_get(dev, vblank->pipe);
 +              if (ret < 0)
 +                      goto out;
 +      } else if (work->count == count) {
 +              /* Already scheduled w/ same vbl count */
 +              goto out;
 +      } else {
 +              rescheduling = true;
 +      }
 +
 +      work->count = count;
 +      cur_vbl = drm_vblank_count(dev, vblank->pipe);
 +      passed = drm_vblank_passed(cur_vbl, count);
 +      if (passed)
 +              drm_dbg_core(dev,
 +                           "crtc %d vblank %llu already passed (current %llu)\n",
 +                           vblank->pipe, count, cur_vbl);
 +
 +      if (!nextonmiss && passed) {
 +              drm_vblank_put(dev, vblank->pipe);
 +              ret = kthread_queue_work(vblank->worker, &work->base);
 +
 +              if (rescheduling) {
 +                      list_del_init(&work->node);
 +                      wake = true;
 +              }
 +      } else {
 +              if (!rescheduling)
 +                      list_add_tail(&work->node, &vblank->pending_work);
 +              ret = true;
 +      }
 +
 +out:
 +      spin_unlock_irqrestore(&dev->event_lock, irqflags);
 +      if (wake)
 +              wake_up_all(&vblank->work_wait_queue);
 +      return ret;
 +}
 +EXPORT_SYMBOL(drm_vblank_work_schedule);
 +
 +/**
 + * 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 guaranteed to no longer be scheduled or running, even
 + * if it's self-arming.
 + *
 + * Returns:
 + * %True if the work was cancelled before it started to execute, %false
 + * otherwise.
 + */
 +bool drm_vblank_work_cancel_sync(struct drm_vblank_work *work)
 +{
 +      struct drm_vblank_crtc *vblank = work->vblank;
 +      struct drm_device *dev = vblank->dev;
 +      bool ret = false;
 +
 +      spin_lock_irq(&dev->event_lock);
 +      if (!list_empty(&work->node)) {
 +              list_del_init(&work->node);
 +              drm_vblank_put(vblank->dev, vblank->pipe);
 +              ret = true;
 +      }
 +
 +      work->cancelling++;
 +      spin_unlock_irq(&dev->event_lock);
 +
 +      wake_up_all(&vblank->work_wait_queue);
 +
 +      if (kthread_cancel_work_sync(&work->base))
 +              ret = true;
 +
 +      spin_lock_irq(&dev->event_lock);
 +      work->cancelling--;
 +      spin_unlock_irq(&dev->event_lock);
 +
 +      return ret;
 +}
 +EXPORT_SYMBOL(drm_vblank_work_cancel_sync);
 +
 +/**
 + * drm_vblank_work_flush - wait for a scheduled vblank work to finish
 + * executing
 + * @work: vblank work to flush
 + *
 + * Wait until @work has finished executing once.
 + */
 +void drm_vblank_work_flush(struct drm_vblank_work *work)
 +{
 +      struct drm_vblank_crtc *vblank = work->vblank;
 +      struct drm_device *dev = vblank->dev;
 +
 +      spin_lock_irq(&dev->event_lock);
 +      wait_event_lock_irq(vblank->work_wait_queue, list_empty(&work->node),
 +                          dev->event_lock);
 +      spin_unlock_irq(&dev->event_lock);
 +
 +      kthread_flush_work(&work->base);
 +}
 +EXPORT_SYMBOL(drm_vblank_work_flush);
 +
 +/**
 + * 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 kthread_work *work))
 +{
 +      kthread_init_work(&work->base, func);
 +      INIT_LIST_HEAD(&work->node);
 +      work->vblank = &crtc->dev->vblank[drm_crtc_index(crtc)];
 +}
 +EXPORT_SYMBOL(drm_vblank_work_init);
 +
 +int drm_vblank_worker_init(struct drm_vblank_crtc *vblank)
 +{
-       return sched_setscheduler(vblank->worker->task, SCHED_FIFO, &param);
 +      struct kthread_worker *worker;
 +
 +      INIT_LIST_HEAD(&vblank->pending_work);
 +      init_waitqueue_head(&vblank->work_wait_queue);
 +      worker = kthread_create_worker(0, "card%d-crtc%d",
 +                                     vblank->dev->primary->index,
 +                                     vblank->pipe);
 +      if (IS_ERR(worker))
 +              return PTR_ERR(worker);
 +
 +      vblank->worker = worker;
 +
++      sched_set_fifo(worker->task);
++      return 0;
 +}

diff --cc drivers/gpu/drm/msm/msm_drv.c
index 36d98d4,556cca3..7d641c7
--- 1/drivers/gpu/drm/msm/msm_drv.c
--- 2/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@@ -524,11 -508,8 +517,7 @@@ static int msm_drm_init(struct device *
                        goto err_msm_uninit;
                }
  
-               ret = sched_setscheduler(priv->event_thread[i].worker->task,
-                                        SCHED_FIFO, &param);
-               if (ret)
-                       dev_warn(dev, "event_thread set priority failed:%d\n",
-                                ret);
 -              sched_set_fifo(priv->event_thread[i].thread);
++              sched_set_fifo(priv->event_thread[i].worker->task);
        }
  
        ret = drm_vblank_init(ddev, priv->num_crtcs);

diff --cc drivers/gpu/drm/scheduler/sched_main.c
Simple merge

diff --cc drivers/powercap/idle_inject.c
Simple merge

diff --cc drivers/spi/spi.c
index 0b26048,5a4f0bf..6626587
--- 1/drivers/spi/spi.c
--- 2/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@@ -1626,11 -1592,9 +1626,9 @@@ EXPORT_SYMBOL_GPL(spi_take_timestamp_po
   */
  static void spi_set_thread_rt(struct spi_controller *ctlr)
  {
-       struct sched_param param = { .sched_priority = MAX_RT_PRIO / 2 };
- 
        dev_info(&ctlr->dev,
                "will run message pump with realtime priority\n");
-       sched_setscheduler(ctlr->kworker->task, SCHED_FIFO, &param);
 -      sched_set_fifo(ctlr->kworker_task);
++      sched_set_fifo(ctlr->kworker->task);
  }
  
  static int spi_init_queue(struct spi_controller *ctlr)

diff --cc include/linux/sched.h
Simple merge

diff --cc kernel/irq/manage.c
Simple merge

diff --cc kernel/locking/locktorture.c
Simple merge

diff --cc kernel/rcu/rcuperf.c
Simple merge

diff --cc kernel/rcu/rcutorture.c
Simple merge

diff --cc kernel/sched/core.c
Simple merge