drm/panfrost: Expose performance counters through unstable ioctls

Expose performance counters through 2 driver specific ioctls: one to
enable/disable the perfcnt block, and one to dump the counter values.

There are discussions to expose global performance monitors (those
counters that can't be retrieved on a per-job basis) in a consistent
way, but this is likely to take time to settle on something that works
for various HW/users.
The ioctls are marked unstable so we can get rid of them when the time
comes. We initally went for a debugfs-based interface, but this was
making the transition to per-FD address space more complicated (we need
to specify the namespace the GPU has to use when dumping the perf
counters), hence the decision to switch back to driver specific ioctls
which are passed the FD they operate on and thus will have a dedicated
address space attached to them.

Other than that, the implementation is pretty simple: it basically dumps
all counters and copy the values to a userspace buffer. The parsing is
left to userspace which has to know the specific layout that's used
by the GPU (layout differs on a per-revision basis).

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Acked-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190618081648.17297-5-boris.brezillon@collabora.com

1 files changed, 329 insertions, 0 deletions

diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
new file mode 100644
index 000000000000..83c57d325ca8
--- /dev/null
+++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
@@ -0,0 +1,329 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright 2019 Collabora Ltd */
+
+#include <drm/drm_file.h>
+#include <drm/drm_gem_shmem_helper.h>
+#include <drm/panfrost_drm.h>
+#include <linux/completion.h>
+#include <linux/iopoll.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+
+#include "panfrost_device.h"
+#include "panfrost_features.h"
+#include "panfrost_gem.h"
+#include "panfrost_issues.h"
+#include "panfrost_job.h"
+#include "panfrost_mmu.h"
+#include "panfrost_regs.h"
+
+#define COUNTERS_PER_BLOCK		64
+#define BYTES_PER_COUNTER		4
+#define BLOCKS_PER_COREGROUP		8
+#define V4_SHADERS_PER_COREGROUP	4
+
+struct panfrost_perfcnt {
+	struct panfrost_gem_object *bo;
+	size_t bosize;
+	void *buf;
+	struct panfrost_file_priv *user;
+	struct mutex lock;
+	struct completion dump_comp;
+};
+
+void panfrost_perfcnt_clean_cache_done(struct panfrost_device *pfdev)
+{
+	complete(&pfdev->perfcnt->dump_comp);
+}
+
+void panfrost_perfcnt_sample_done(struct panfrost_device *pfdev)
+{
+	gpu_write(pfdev, GPU_CMD, GPU_CMD_CLEAN_CACHES);
+}
+
+static int panfrost_perfcnt_dump_locked(struct panfrost_device *pfdev)
+{
+	u64 gpuva;
+	int ret;
+
+	reinit_completion(&pfdev->perfcnt->dump_comp);
+	gpuva = pfdev->perfcnt->bo->node.start << PAGE_SHIFT;
+	gpu_write(pfdev, GPU_PERFCNT_BASE_LO, gpuva);
+	gpu_write(pfdev, GPU_PERFCNT_BASE_HI, gpuva >> 32);
+	gpu_write(pfdev, GPU_INT_CLEAR,
+		  GPU_IRQ_CLEAN_CACHES_COMPLETED |
+		  GPU_IRQ_PERFCNT_SAMPLE_COMPLETED);
+	gpu_write(pfdev, GPU_CMD, GPU_CMD_PERFCNT_SAMPLE);
+	ret = wait_for_completion_interruptible_timeout(&pfdev->perfcnt->dump_comp,
+							msecs_to_jiffies(1000));
+	if (!ret)
+		ret = -ETIMEDOUT;
+	else if (ret > 0)
+		ret = 0;
+
+	return ret;
+}
+
+static int panfrost_perfcnt_enable_locked(struct panfrost_device *pfdev,
+					  struct panfrost_file_priv *user,
+					  unsigned int counterset)
+{
+	struct panfrost_perfcnt *perfcnt = pfdev->perfcnt;
+	struct drm_gem_shmem_object *bo;
+	u32 cfg;
+	int ret;
+
+	if (user == perfcnt->user)
+		return 0;
+	else if (perfcnt->user)
+		return -EBUSY;
+
+	ret = pm_runtime_get_sync(pfdev->dev);
+	if (ret < 0)
+		return ret;
+
+	bo = drm_gem_shmem_create(pfdev->ddev, perfcnt->bosize);
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	perfcnt->bo = to_panfrost_bo(&bo->base);
+
+	/* Map the perfcnt buf in the address space attached to file_priv. */
+	ret = panfrost_mmu_map(perfcnt->bo);
+	if (ret)
+		goto err_put_bo;
+
+	perfcnt->buf = drm_gem_shmem_vmap(&bo->base);
+	if (IS_ERR(perfcnt->buf)) {
+		ret = PTR_ERR(perfcnt->buf);
+		goto err_put_bo;
+	}
+
+	/*
+	 * Invalidate the cache and clear the counters to start from a fresh
+	 * state.
+	 */
+	reinit_completion(&pfdev->perfcnt->dump_comp);
+	gpu_write(pfdev, GPU_INT_CLEAR,
+		  GPU_IRQ_CLEAN_CACHES_COMPLETED |
+		  GPU_IRQ_PERFCNT_SAMPLE_COMPLETED);
+	gpu_write(pfdev, GPU_CMD, GPU_CMD_PERFCNT_CLEAR);
+	gpu_write(pfdev, GPU_CMD, GPU_CMD_CLEAN_INV_CACHES);
+	ret = wait_for_completion_timeout(&pfdev->perfcnt->dump_comp,
+					  msecs_to_jiffies(1000));
+	if (!ret) {
+		ret = -ETIMEDOUT;
+		goto err_vunmap;
+	}
+
+	perfcnt->user = user;
+
+	/*
+	 * Always use address space 0 for now.
+	 * FIXME: this needs to be updated when we start using different
+	 * address space.
+	 */
+	cfg = GPU_PERFCNT_CFG_AS(0) |
+	      GPU_PERFCNT_CFG_MODE(GPU_PERFCNT_CFG_MODE_MANUAL);
+
+	/*
+	 * Bifrost GPUs have 2 set of counters, but we're only interested by
+	 * the first one for now.
+	 */
+	if (panfrost_model_is_bifrost(pfdev))
+		cfg |= GPU_PERFCNT_CFG_SETSEL(counterset);
+
+	gpu_write(pfdev, GPU_PRFCNT_JM_EN, 0xffffffff);
+	gpu_write(pfdev, GPU_PRFCNT_SHADER_EN, 0xffffffff);
+	gpu_write(pfdev, GPU_PRFCNT_MMU_L2_EN, 0xffffffff);
+
+	/*
+	 * Due to PRLAM-8186 we need to disable the Tiler before we enable HW
+	 * counters.
+	 */
+	if (panfrost_has_hw_issue(pfdev, HW_ISSUE_8186))
+		gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
+	else
+		gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0xffffffff);
+
+	gpu_write(pfdev, GPU_PERFCNT_CFG, cfg);
+
+	if (panfrost_has_hw_issue(pfdev, HW_ISSUE_8186))
+		gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0xffffffff);
+
+	return 0;
+
+err_vunmap:
+	drm_gem_shmem_vunmap(&perfcnt->bo->base.base, perfcnt->buf);
+err_put_bo:
+	drm_gem_object_put_unlocked(&bo->base);
+	return ret;
+}
+
+static int panfrost_perfcnt_disable_locked(struct panfrost_device *pfdev,
+					   struct panfrost_file_priv *user)
+{
+	struct panfrost_perfcnt *perfcnt = pfdev->perfcnt;
+
+	if (user != perfcnt->user)
+		return -EINVAL;
+
+	gpu_write(pfdev, GPU_PRFCNT_JM_EN, 0x0);
+	gpu_write(pfdev, GPU_PRFCNT_SHADER_EN, 0x0);
+	gpu_write(pfdev, GPU_PRFCNT_MMU_L2_EN, 0x0);
+	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
+	gpu_write(pfdev, GPU_PERFCNT_CFG,
+		  GPU_PERFCNT_CFG_MODE(GPU_PERFCNT_CFG_MODE_OFF));
+
+	perfcnt->user = NULL;
+	drm_gem_shmem_vunmap(&perfcnt->bo->base.base, perfcnt->buf);
+	perfcnt->buf = NULL;
+	drm_gem_object_put_unlocked(&perfcnt->bo->base.base);
+	perfcnt->bo = NULL;
+	pm_runtime_mark_last_busy(pfdev->dev);
+	pm_runtime_put_autosuspend(pfdev->dev);
+
+	return 0;
+}
+
+int panfrost_ioctl_perfcnt_enable(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv)
+{
+	struct panfrost_file_priv *pfile = file_priv->driver_priv;
+	struct panfrost_device *pfdev = dev->dev_private;
+	struct panfrost_perfcnt *perfcnt = pfdev->perfcnt;
+	struct drm_panfrost_perfcnt_enable *req = data;
+	int ret;
+
+	ret = panfrost_unstable_ioctl_check();
+	if (ret)
+		return ret;
+
+	/* Only Bifrost GPUs have 2 set of counters. */
+	if (req->counterset > (panfrost_model_is_bifrost(pfdev) ? 1 : 0))
+		return -EINVAL;
+
+	mutex_lock(&perfcnt->lock);
+	if (req->enable)
+		ret = panfrost_perfcnt_enable_locked(pfdev, pfile,
+						     req->counterset);
+	else
+		ret = panfrost_perfcnt_disable_locked(pfdev, pfile);
+	mutex_unlock(&perfcnt->lock);
+
+	return ret;
+}
+
+int panfrost_ioctl_perfcnt_dump(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	struct panfrost_device *pfdev = dev->dev_private;
+	struct panfrost_perfcnt *perfcnt = pfdev->perfcnt;
+	struct drm_panfrost_perfcnt_dump *req = data;
+	void __user *user_ptr = (void __user *)(uintptr_t)req->buf_ptr;
+	int ret;
+
+	ret = panfrost_unstable_ioctl_check();
+	if (ret)
+		return ret;
+
+	mutex_lock(&perfcnt->lock);
+	if (perfcnt->user != file_priv->driver_priv) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ret = panfrost_perfcnt_dump_locked(pfdev);
+	if (ret)
+		goto out;
+
+	if (copy_to_user(user_ptr, perfcnt->buf, perfcnt->bosize))
+		ret = -EFAULT;
+
+out:
+	mutex_unlock(&perfcnt->lock);
+
+	return ret;
+}
+
+void panfrost_perfcnt_close(struct panfrost_file_priv *pfile)
+{
+	struct panfrost_device *pfdev = pfile->pfdev;
+	struct panfrost_perfcnt *perfcnt = pfdev->perfcnt;
+
+	pm_runtime_get_sync(pfdev->dev);
+	mutex_lock(&perfcnt->lock);
+	if (perfcnt->user == pfile)
+		panfrost_perfcnt_disable_locked(pfdev, pfile);
+	mutex_unlock(&perfcnt->lock);
+	pm_runtime_mark_last_busy(pfdev->dev);
+	pm_runtime_put_autosuspend(pfdev->dev);
+}
+
+int panfrost_perfcnt_init(struct panfrost_device *pfdev)
+{
+	struct panfrost_perfcnt *perfcnt;
+	size_t size;
+
+	if (panfrost_has_hw_feature(pfdev, HW_FEATURE_V4)) {
+		unsigned int ncoregroups;
+
+		ncoregroups = hweight64(pfdev->features.l2_present);
+		size = ncoregroups * BLOCKS_PER_COREGROUP *
+		       COUNTERS_PER_BLOCK * BYTES_PER_COUNTER;
+	} else {
+		unsigned int nl2c, ncores;
+
+		/*
+		 * TODO: define a macro to extract the number of l2 caches from
+		 * mem_features.
+		 */
+		nl2c = ((pfdev->features.mem_features >> 8) & GENMASK(3, 0)) + 1;
+
+		/*
+		 * shader_present might be sparse, but the counters layout
+		 * forces to dump unused regions too, hence the fls64() call
+		 * instead of hweight64().
+		 */
+		ncores = fls64(pfdev->features.shader_present);
+
+		/*
+		 * There's always one JM and one Tiler block, hence the '+ 2'
+		 * here.
+		 */
+		size = (nl2c + ncores + 2) *
+		       COUNTERS_PER_BLOCK * BYTES_PER_COUNTER;
+	}
+
+	perfcnt = devm_kzalloc(pfdev->dev, sizeof(*perfcnt), GFP_KERNEL);
+	if (!perfcnt)
+		return -ENOMEM;
+
+	perfcnt->bosize = size;
+
+	/* Start with everything disabled. */
+	gpu_write(pfdev, GPU_PERFCNT_CFG,
+		  GPU_PERFCNT_CFG_MODE(GPU_PERFCNT_CFG_MODE_OFF));
+	gpu_write(pfdev, GPU_PRFCNT_JM_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_SHADER_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_MMU_L2_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
+
+	init_completion(&perfcnt->dump_comp);
+	mutex_init(&perfcnt->lock);
+	pfdev->perfcnt = perfcnt;
+
+	return 0;
+}
+
+void panfrost_perfcnt_fini(struct panfrost_device *pfdev)
+{
+	/* Disable everything before leaving. */
+	gpu_write(pfdev, GPU_PERFCNT_CFG,
+		  GPU_PERFCNT_CFG_MODE(GPU_PERFCNT_CFG_MODE_OFF));
+	gpu_write(pfdev, GPU_PRFCNT_JM_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_SHADER_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_MMU_L2_EN, 0);
+	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
+}