diff options
author | Robert Bragg <robert@sixbynine.org> | 2016-11-07 19:49:47 +0000 |
---|---|---|
committer | Daniel Vetter <daniel.vetter@ffwll.ch> | 2016-11-22 14:27:18 +0100 |
commit | eec688e1420da584afb36ffa5f0cad75f53cf286 (patch) | |
tree | 94e93e2a08fc355f5ab4365b652e20eb6f740350 /drivers/gpu/drm/i915/i915_perf.c | |
parent | bc1d53c64773e89175766e8c77cc96aa0763de4a (diff) |
drm/i915: Add i915 perf infrastructure
Adds base i915 perf infrastructure for Gen performance metrics.
This adds a DRM_IOCTL_I915_PERF_OPEN ioctl that takes an array of uint64
properties to configure a stream of metrics and returns a new fd usable
with standard VFS system calls including read() to read typed and sized
records; ioctl() to enable or disable capture and poll() to wait for
data.
A stream is opened something like:
uint64_t properties[] = {
/* Single context sampling */
DRM_I915_PERF_PROP_CTX_HANDLE, ctx_handle,
/* Include OA reports in samples */
DRM_I915_PERF_PROP_SAMPLE_OA, true,
/* OA unit configuration */
DRM_I915_PERF_PROP_OA_METRICS_SET, metrics_set_id,
DRM_I915_PERF_PROP_OA_FORMAT, report_format,
DRM_I915_PERF_PROP_OA_EXPONENT, period_exponent,
};
struct drm_i915_perf_open_param parm = {
.flags = I915_PERF_FLAG_FD_CLOEXEC |
I915_PERF_FLAG_FD_NONBLOCK |
I915_PERF_FLAG_DISABLED,
.properties_ptr = (uint64_t)properties,
.num_properties = sizeof(properties) / 16,
};
int fd = drmIoctl(drm_fd, DRM_IOCTL_I915_PERF_OPEN, ¶m);
Records read all start with a common { type, size } header with
DRM_I915_PERF_RECORD_SAMPLE being of most interest. Sample records
contain an extensible number of fields and it's the
DRM_I915_PERF_PROP_SAMPLE_xyz properties given when opening that
determine what's included in every sample.
No specific streams are supported yet so any attempt to open a stream
will return an error.
v2:
use i915_gem_context_get() - Chris Wilson
v3:
update read() interface to avoid passing state struct - Chris Wilson
fix some rebase fallout, with i915-perf init/deinit
v4:
s/DRM_IORW/DRM_IOW/ - Emil Velikov
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Sourab Gupta <sourab.gupta@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/20161107194957.3385-2-robert@sixbynine.org
Diffstat (limited to 'drivers/gpu/drm/i915/i915_perf.c')
-rw-r--r-- | drivers/gpu/drm/i915/i915_perf.c | 449 |
1 files changed, 449 insertions, 0 deletions
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c new file mode 100644 index 000000000000..777ce65f910d --- /dev/null +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -0,0 +1,449 @@ +/* + * Copyright © 2015-2016 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + * Authors: + * Robert Bragg <robert@sixbynine.org> + */ + +#include <linux/anon_inodes.h> + +#include "i915_drv.h" + +struct perf_open_properties { + u32 sample_flags; + + u64 single_context:1; + u64 ctx_handle; +}; + +static ssize_t i915_perf_read_locked(struct i915_perf_stream *stream, + struct file *file, + char __user *buf, + size_t count, + loff_t *ppos) +{ + /* Note we keep the offset (aka bytes read) separate from any + * error status so that the final check for whether we return + * the bytes read with a higher precedence than any error (see + * comment below) doesn't need to be handled/duplicated in + * stream->ops->read() implementations. + */ + size_t offset = 0; + int ret = stream->ops->read(stream, buf, count, &offset); + + /* If we've successfully copied any data then reporting that + * takes precedence over any internal error status, so the + * data isn't lost. + * + * For example ret will be -ENOSPC whenever there is more + * buffered data than can be copied to userspace, but that's + * only interesting if we weren't able to copy some data + * because it implies the userspace buffer is too small to + * receive a single record (and we never split records). + * + * Another case with ret == -EFAULT is more of a grey area + * since it would seem like bad form for userspace to ask us + * to overrun its buffer, but the user knows best: + * + * http://yarchive.net/comp/linux/partial_reads_writes.html + */ + return offset ?: (ret ?: -EAGAIN); +} + +static ssize_t i915_perf_read(struct file *file, + char __user *buf, + size_t count, + loff_t *ppos) +{ + struct i915_perf_stream *stream = file->private_data; + struct drm_i915_private *dev_priv = stream->dev_priv; + ssize_t ret; + + if (!(file->f_flags & O_NONBLOCK)) { + /* Allow false positives from stream->ops->wait_unlocked. + */ + do { + ret = stream->ops->wait_unlocked(stream); + if (ret) + return ret; + + mutex_lock(&dev_priv->perf.lock); + ret = i915_perf_read_locked(stream, file, + buf, count, ppos); + mutex_unlock(&dev_priv->perf.lock); + } while (ret == -EAGAIN); + } else { + mutex_lock(&dev_priv->perf.lock); + ret = i915_perf_read_locked(stream, file, buf, count, ppos); + mutex_unlock(&dev_priv->perf.lock); + } + + return ret; +} + +static unsigned int i915_perf_poll_locked(struct i915_perf_stream *stream, + struct file *file, + poll_table *wait) +{ + unsigned int streams = 0; + + stream->ops->poll_wait(stream, file, wait); + + if (stream->ops->can_read(stream)) + streams |= POLLIN; + + return streams; +} + +static unsigned int i915_perf_poll(struct file *file, poll_table *wait) +{ + struct i915_perf_stream *stream = file->private_data; + struct drm_i915_private *dev_priv = stream->dev_priv; + int ret; + + mutex_lock(&dev_priv->perf.lock); + ret = i915_perf_poll_locked(stream, file, wait); + mutex_unlock(&dev_priv->perf.lock); + + return ret; +} + +static void i915_perf_enable_locked(struct i915_perf_stream *stream) +{ + if (stream->enabled) + return; + + /* Allow stream->ops->enable() to refer to this */ + stream->enabled = true; + + if (stream->ops->enable) + stream->ops->enable(stream); +} + +static void i915_perf_disable_locked(struct i915_perf_stream *stream) +{ + if (!stream->enabled) + return; + + /* Allow stream->ops->disable() to refer to this */ + stream->enabled = false; + + if (stream->ops->disable) + stream->ops->disable(stream); +} + +static long i915_perf_ioctl_locked(struct i915_perf_stream *stream, + unsigned int cmd, + unsigned long arg) +{ + switch (cmd) { + case I915_PERF_IOCTL_ENABLE: + i915_perf_enable_locked(stream); + return 0; + case I915_PERF_IOCTL_DISABLE: + i915_perf_disable_locked(stream); + return 0; + } + + return -EINVAL; +} + +static long i915_perf_ioctl(struct file *file, + unsigned int cmd, + unsigned long arg) +{ + struct i915_perf_stream *stream = file->private_data; + struct drm_i915_private *dev_priv = stream->dev_priv; + long ret; + + mutex_lock(&dev_priv->perf.lock); + ret = i915_perf_ioctl_locked(stream, cmd, arg); + mutex_unlock(&dev_priv->perf.lock); + + return ret; +} + +static void i915_perf_destroy_locked(struct i915_perf_stream *stream) +{ + struct drm_i915_private *dev_priv = stream->dev_priv; + + if (stream->enabled) + i915_perf_disable_locked(stream); + + if (stream->ops->destroy) + stream->ops->destroy(stream); + + list_del(&stream->link); + + if (stream->ctx) { + mutex_lock(&dev_priv->drm.struct_mutex); + i915_gem_context_put(stream->ctx); + mutex_unlock(&dev_priv->drm.struct_mutex); + } + + kfree(stream); +} + +static int i915_perf_release(struct inode *inode, struct file *file) +{ + struct i915_perf_stream *stream = file->private_data; + struct drm_i915_private *dev_priv = stream->dev_priv; + + mutex_lock(&dev_priv->perf.lock); + i915_perf_destroy_locked(stream); + mutex_unlock(&dev_priv->perf.lock); + + return 0; +} + + +static const struct file_operations fops = { + .owner = THIS_MODULE, + .llseek = no_llseek, + .release = i915_perf_release, + .poll = i915_perf_poll, + .read = i915_perf_read, + .unlocked_ioctl = i915_perf_ioctl, +}; + + +static struct i915_gem_context * +lookup_context(struct drm_i915_private *dev_priv, + struct drm_i915_file_private *file_priv, + u32 ctx_user_handle) +{ + struct i915_gem_context *ctx; + int ret; + + ret = i915_mutex_lock_interruptible(&dev_priv->drm); + if (ret) + return ERR_PTR(ret); + + ctx = i915_gem_context_lookup(file_priv, ctx_user_handle); + if (!IS_ERR(ctx)) + i915_gem_context_get(ctx); + + mutex_unlock(&dev_priv->drm.struct_mutex); + + return ctx; +} + +static int +i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv, + struct drm_i915_perf_open_param *param, + struct perf_open_properties *props, + struct drm_file *file) +{ + struct i915_gem_context *specific_ctx = NULL; + struct i915_perf_stream *stream = NULL; + unsigned long f_flags = 0; + int stream_fd; + int ret; + + if (props->single_context) { + u32 ctx_handle = props->ctx_handle; + struct drm_i915_file_private *file_priv = file->driver_priv; + + specific_ctx = lookup_context(dev_priv, file_priv, ctx_handle); + if (IS_ERR(specific_ctx)) { + ret = PTR_ERR(specific_ctx); + if (ret != -EINTR) + DRM_ERROR("Failed to look up context with ID %u for opening perf stream\n", + ctx_handle); + goto err; + } + } + + if (!specific_ctx && !capable(CAP_SYS_ADMIN)) { + DRM_ERROR("Insufficient privileges to open system-wide i915 perf stream\n"); + ret = -EACCES; + goto err_ctx; + } + + stream = kzalloc(sizeof(*stream), GFP_KERNEL); + if (!stream) { + ret = -ENOMEM; + goto err_ctx; + } + + stream->sample_flags = props->sample_flags; + stream->dev_priv = dev_priv; + stream->ctx = specific_ctx; + + /* + * TODO: support sampling something + * + * For now this is as far as we can go. + */ + DRM_ERROR("Unsupported i915 perf stream configuration\n"); + ret = -EINVAL; + goto err_alloc; + + list_add(&stream->link, &dev_priv->perf.streams); + + if (param->flags & I915_PERF_FLAG_FD_CLOEXEC) + f_flags |= O_CLOEXEC; + if (param->flags & I915_PERF_FLAG_FD_NONBLOCK) + f_flags |= O_NONBLOCK; + + stream_fd = anon_inode_getfd("[i915_perf]", &fops, stream, f_flags); + if (stream_fd < 0) { + ret = stream_fd; + goto err_open; + } + + if (!(param->flags & I915_PERF_FLAG_DISABLED)) + i915_perf_enable_locked(stream); + + return stream_fd; + +err_open: + list_del(&stream->link); + if (stream->ops->destroy) + stream->ops->destroy(stream); +err_alloc: + kfree(stream); +err_ctx: + if (specific_ctx) { + mutex_lock(&dev_priv->drm.struct_mutex); + i915_gem_context_put(specific_ctx); + mutex_unlock(&dev_priv->drm.struct_mutex); + } +err: + return ret; +} + +/* Note we copy the properties from userspace outside of the i915 perf + * mutex to avoid an awkward lockdep with mmap_sem. + * + * Note this function only validates properties in isolation it doesn't + * validate that the combination of properties makes sense or that all + * properties necessary for a particular kind of stream have been set. + */ +static int read_properties_unlocked(struct drm_i915_private *dev_priv, + u64 __user *uprops, + u32 n_props, + struct perf_open_properties *props) +{ + u64 __user *uprop = uprops; + int i; + + memset(props, 0, sizeof(struct perf_open_properties)); + + if (!n_props) { + DRM_ERROR("No i915 perf properties given"); + return -EINVAL; + } + + /* Considering that ID = 0 is reserved and assuming that we don't + * (currently) expect any configurations to ever specify duplicate + * values for a particular property ID then the last _PROP_MAX value is + * one greater than the maximum number of properties we expect to get + * from userspace. + */ + if (n_props >= DRM_I915_PERF_PROP_MAX) { + DRM_ERROR("More i915 perf properties specified than exist"); + return -EINVAL; + } + + for (i = 0; i < n_props; i++) { + u64 id, value; + int ret; + + ret = get_user(id, uprop); + if (ret) + return ret; + + ret = get_user(value, uprop + 1); + if (ret) + return ret; + + switch ((enum drm_i915_perf_property_id)id) { + case DRM_I915_PERF_PROP_CTX_HANDLE: + props->single_context = 1; + props->ctx_handle = value; + break; + default: + MISSING_CASE(id); + DRM_ERROR("Unknown i915 perf property ID"); + return -EINVAL; + } + + uprop += 2; + } + + return 0; +} + +int i915_perf_open_ioctl(struct drm_device *dev, void *data, + struct drm_file *file) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_perf_open_param *param = data; + struct perf_open_properties props; + u32 known_open_flags; + int ret; + + if (!dev_priv->perf.initialized) { + DRM_ERROR("i915 perf interface not available for this system"); + return -ENOTSUPP; + } + + known_open_flags = I915_PERF_FLAG_FD_CLOEXEC | + I915_PERF_FLAG_FD_NONBLOCK | + I915_PERF_FLAG_DISABLED; + if (param->flags & ~known_open_flags) { + DRM_ERROR("Unknown drm_i915_perf_open_param flag\n"); + return -EINVAL; + } + + ret = read_properties_unlocked(dev_priv, + u64_to_user_ptr(param->properties_ptr), + param->num_properties, + &props); + if (ret) + return ret; + + mutex_lock(&dev_priv->perf.lock); + ret = i915_perf_open_ioctl_locked(dev_priv, param, &props, file); + mutex_unlock(&dev_priv->perf.lock); + + return ret; +} + +void i915_perf_init(struct drm_i915_private *dev_priv) +{ + INIT_LIST_HEAD(&dev_priv->perf.streams); + mutex_init(&dev_priv->perf.lock); + + dev_priv->perf.initialized = true; +} + +void i915_perf_fini(struct drm_i915_private *dev_priv) +{ + if (!dev_priv->perf.initialized) + return; + + /* Currently nothing to clean up */ + + dev_priv->perf.initialized = false; +} |