drm/vc4: Enable runtime PM.

This may actually get us a feature that the closed driver didn't have:
turning off the GPU in between rendering jobs, while the V3D device is
still opened by the client.

There may be some tuning to be applied here to use autosuspend so that
we don't bounce the device's power so much, but in steady-state
GPU-bound rendering we keep the power on (since we keep multiple jobs
outstanding) and even if we power cycle on every job we can still
manage at least 680 fps.

More importantly, though, runtime PM will allow us to power off the
device to do a GPU reset.

v2: Switch #ifdef to CONFIG_PM not CONFIG_PM_SLEEP (caught by kbuild
    test robot)

Signed-off-by: Eric Anholt <eric@anholt.net>

3 files changed, 47 insertions, 0 deletions

diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h
index b6ccf8181643..8ac3788846ce 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.h
+++ b/drivers/gpu/drm/vc4/vc4_drv.h
@@ -141,6 +141,7 @@ struct vc4_seqno_cb {
 };
 
 struct vc4_v3d {
+	struct vc4_dev *vc4;
 	struct platform_device *pdev;
 	void __iomem *regs;
 };
diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c
index 1a819dd826f8..4e0391cbf9d0 100644
--- a/drivers/gpu/drm/vc4/vc4_gem.c
+++ b/drivers/gpu/drm/vc4/vc4_gem.c
@@ -23,6 +23,7 @@
 
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
 #include <linux/device.h>
 #include <linux/io.h>
 
@@ -621,6 +622,7 @@ fail:
 static void
 vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
 {
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	unsigned i;
 
 	/* Need the struct lock for drm_gem_object_unreference(). */
@@ -639,6 +641,8 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
 	}
 	mutex_unlock(&dev->struct_mutex);
 
+	pm_runtime_put(&vc4->v3d->pdev->dev);
+
 	kfree(exec);
 }
 
@@ -792,6 +796,12 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 		return -ENOMEM;
 	}
 
+	ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
+	if (ret < 0) {
+		kfree(exec);
+		return ret;
+	}
+
 	exec->args = args;
 	INIT_LIST_HEAD(&exec->unref_list);
 
diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c
index 314ff71db978..cb38b6bf4119 100644
--- a/drivers/gpu/drm/vc4/vc4_v3d.c
+++ b/drivers/gpu/drm/vc4/vc4_v3d.c
@@ -17,6 +17,7 @@
  */
 
 #include "linux/component.h"
+#include "linux/pm_runtime.h"
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
@@ -167,6 +168,29 @@ static void vc4_v3d_init_hw(struct drm_device *dev)
 	V3D_WRITE(V3D_VPMBASE, 0);
 }
 
+#ifdef CONFIG_PM
+static int vc4_v3d_runtime_suspend(struct device *dev)
+{
+	struct vc4_v3d *v3d = dev_get_drvdata(dev);
+	struct vc4_dev *vc4 = v3d->vc4;
+
+	vc4_irq_uninstall(vc4->dev);
+
+	return 0;
+}
+
+static int vc4_v3d_runtime_resume(struct device *dev)
+{
+	struct vc4_v3d *v3d = dev_get_drvdata(dev);
+	struct vc4_dev *vc4 = v3d->vc4;
+
+	vc4_v3d_init_hw(vc4->dev);
+	vc4_irq_postinstall(vc4->dev);
+
+	return 0;
+}
+#endif
+
 static int vc4_v3d_bind(struct device *dev, struct device *master, void *data)
 {
 	struct platform_device *pdev = to_platform_device(dev);
@@ -179,6 +203,8 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data)
 	if (!v3d)
 		return -ENOMEM;
 
+	dev_set_drvdata(dev, v3d);
+
 	v3d->pdev = pdev;
 
 	v3d->regs = vc4_ioremap_regs(pdev, 0);
@@ -186,6 +212,7 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data)
 		return PTR_ERR(v3d->regs);
 
 	vc4->v3d = v3d;
+	v3d->vc4 = vc4;
 
 	if (V3D_READ(V3D_IDENT0) != V3D_EXPECTED_IDENT0) {
 		DRM_ERROR("V3D_IDENT0 read 0x%08x instead of 0x%08x\n",
@@ -207,6 +234,8 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data)
 		return ret;
 	}
 
+	pm_runtime_enable(dev);
+
 	return 0;
 }
 
@@ -216,6 +245,8 @@ static void vc4_v3d_unbind(struct device *dev, struct device *master,
 	struct drm_device *drm = dev_get_drvdata(master);
 	struct vc4_dev *vc4 = to_vc4_dev(drm);
 
+	pm_runtime_disable(dev);
+
 	drm_irq_uninstall(drm);
 
 	/* Disable the binner's overflow memory address, so the next
@@ -228,6 +259,10 @@ static void vc4_v3d_unbind(struct device *dev, struct device *master,
 	vc4->v3d = NULL;
 }
 
+static const struct dev_pm_ops vc4_v3d_pm_ops = {
+	SET_RUNTIME_PM_OPS(vc4_v3d_runtime_suspend, vc4_v3d_runtime_resume, NULL)
+};
+
 static const struct component_ops vc4_v3d_ops = {
 	.bind   = vc4_v3d_bind,
 	.unbind = vc4_v3d_unbind,
@@ -255,5 +290,6 @@ struct platform_driver vc4_v3d_driver = {
 	.driver = {
 		.name = "vc4_v3d",
 		.of_match_table = vc4_v3d_dt_match,
+		.pm = &vc4_v3d_pm_ops,
 	},
 };