#include <linux/pci.h>
#include <linux/dma-buf.h>
-static __must_check int i915_gem_object_flush_gpu_write_domain(struct drm_i915_gem_object *obj);
static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj);
static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj);
static __must_check int i915_gem_object_bind_to_gtt(struct drm_i915_gem_object *obj,
if (!atomic_read(&dev_priv->mm.wedged))
return 0;
- ret = wait_for_completion_interruptible(x);
- if (ret)
+ /*
+ * Only wait 10 seconds for the gpu reset to complete to avoid hanging
+ * userspace. If it takes that long something really bad is going on and
+ * we should simply try to bail out and fail as gracefully as possible.
+ */
+ ret = wait_for_completion_interruptible_timeout(x, 10*HZ);
+ if (ret == 0) {
+ DRM_ERROR("Timed out waiting for the gpu reset to complete\n");
+ return -EIO;
+ } else if (ret < 0) {
return ret;
+ }
if (atomic_read(&dev_priv->mm.wedged)) {
/* GPU is hung, bump the completion count to account for
obj->fault_mappable = true;
- pfn = ((dev->agp->base + obj->gtt_offset) >> PAGE_SHIFT) +
+ pfn = ((dev_priv->mm.gtt_base_addr + obj->gtt_offset) >> PAGE_SHIFT) +
page_offset;
/* Finally, remap it using the new GTT offset */
out:
switch (ret) {
case -EIO:
+ /* If this -EIO is due to a gpu hang, give the reset code a
+ * chance to clean up the mess. Otherwise return the proper
+ * SIGBUS. */
+ if (!atomic_read(&dev_priv->mm.wedged))
+ return VM_FAULT_SIGBUS;
case -EAGAIN:
/* Give the error handler a chance to run and move the
* objects off the GPU active list. Next time we service the
list_move_tail(&obj->mm_list, &dev_priv->mm.active_list);
list_move_tail(&obj->ring_list, &ring->active_list);
- obj->last_rendering_seqno = seqno;
+ obj->last_read_seqno = seqno;
if (obj->fenced_gpu_access) {
obj->last_fenced_seqno = seqno;
}
}
-static void
-i915_gem_object_move_off_active(struct drm_i915_gem_object *obj)
-{
- list_del_init(&obj->ring_list);
- obj->last_rendering_seqno = 0;
- obj->last_fenced_seqno = 0;
-}
-
-static void
-i915_gem_object_move_to_flushing(struct drm_i915_gem_object *obj)
-{
- struct drm_device *dev = obj->base.dev;
- drm_i915_private_t *dev_priv = dev->dev_private;
-
- BUG_ON(!obj->active);
- list_move_tail(&obj->mm_list, &dev_priv->mm.flushing_list);
-
- i915_gem_object_move_off_active(obj);
-}
-
static void
i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
{
list_move_tail(&obj->mm_list, &dev_priv->mm.inactive_list);
- BUG_ON(!list_empty(&obj->gpu_write_list));
+ BUG_ON(obj->base.write_domain & ~I915_GEM_GPU_DOMAINS);
BUG_ON(!obj->active);
+
+ list_del_init(&obj->ring_list);
obj->ring = NULL;
- i915_gem_object_move_off_active(obj);
+ obj->last_read_seqno = 0;
+ obj->last_write_seqno = 0;
+ obj->base.write_domain = 0;
+
+ obj->last_fenced_seqno = 0;
obj->fenced_gpu_access = false;
obj->active = 0;
- obj->pending_gpu_write = false;
drm_gem_object_unreference(&obj->base);
WARN_ON(i915_verify_lists(dev));
return obj->madv == I915_MADV_DONTNEED;
}
-static void
-i915_gem_process_flushing_list(struct intel_ring_buffer *ring,
- uint32_t flush_domains)
-{
- struct drm_i915_gem_object *obj, *next;
-
- list_for_each_entry_safe(obj, next,
- &ring->gpu_write_list,
- gpu_write_list) {
- if (obj->base.write_domain & flush_domains) {
- uint32_t old_write_domain = obj->base.write_domain;
-
- obj->base.write_domain = 0;
- list_del_init(&obj->gpu_write_list);
- i915_gem_object_move_to_active(obj, ring,
- i915_gem_next_request_seqno(ring));
-
- trace_i915_gem_object_change_domain(obj,
- obj->base.read_domains,
- old_write_domain);
- }
- }
-}
-
static u32
i915_gem_get_seqno(struct drm_device *dev)
{
int was_empty;
int ret;
- BUG_ON(request == NULL);
+ /*
+ * Emit any outstanding flushes - execbuf can fail to emit the flush
+ * after having emitted the batchbuffer command. Hence we need to fix
+ * things up similar to emitting the lazy request. The difference here
+ * is that the flush _must_ happen before the next request, no matter
+ * what.
+ */
+ if (ring->gpu_caches_dirty) {
+ ret = i915_gem_flush_ring(ring, 0, I915_GEM_GPU_DOMAINS);
+ if (ret)
+ return ret;
+
+ ring->gpu_caches_dirty = false;
+ }
+
+ if (request == NULL) {
+ request = kmalloc(sizeof(*request), GFP_KERNEL);
+ if (request == NULL)
+ return -ENOMEM;
+ }
+
seqno = i915_gem_next_request_seqno(ring);
/* Record the position of the start of the request so that
request_ring_position = intel_ring_get_tail(ring);
ret = ring->add_request(ring, &seqno);
- if (ret)
- return ret;
+ if (ret) {
+ kfree(request);
+ return ret;
+ }
trace_i915_gem_request_add(ring, seqno);
request->emitted_jiffies = jiffies;
was_empty = list_empty(&ring->request_list);
list_add_tail(&request->list, &ring->request_list);
+ request->file_priv = NULL;
if (file) {
struct drm_i915_file_private *file_priv = file->driver_priv;
queue_delayed_work(dev_priv->wq,
&dev_priv->mm.retire_work, HZ);
}
+
return 0;
}
struct drm_i915_gem_object,
ring_list);
- obj->base.write_domain = 0;
- list_del_init(&obj->gpu_write_list);
i915_gem_object_move_to_inactive(obj);
}
}
for_each_ring(ring, dev_priv, i)
i915_gem_reset_ring_lists(dev_priv, ring);
- /* Remove anything from the flushing lists. The GPU cache is likely
- * to be lost on reset along with the data, so simply move the
- * lost bo to the inactive list.
- */
- while (!list_empty(&dev_priv->mm.flushing_list)) {
- obj = list_first_entry(&dev_priv->mm.flushing_list,
- struct drm_i915_gem_object,
- mm_list);
-
- obj->base.write_domain = 0;
- list_del_init(&obj->gpu_write_list);
- i915_gem_object_move_to_inactive(obj);
- }
-
/* Move everything out of the GPU domains to ensure we do any
* necessary invalidation upon reuse.
*/
struct drm_i915_gem_object,
ring_list);
- if (!i915_seqno_passed(seqno, obj->last_rendering_seqno))
+ if (!i915_seqno_passed(seqno, obj->last_read_seqno))
break;
- if (obj->base.write_domain != 0)
- i915_gem_object_move_to_flushing(obj);
- else
- i915_gem_object_move_to_inactive(obj);
+ i915_gem_object_move_to_inactive(obj);
}
if (unlikely(ring->trace_irq_seqno &&
*/
idle = true;
for_each_ring(ring, dev_priv, i) {
- if (!list_empty(&ring->gpu_write_list)) {
- struct drm_i915_gem_request *request;
- int ret;
-
- ret = i915_gem_flush_ring(ring,
- 0, I915_GEM_GPU_DOMAINS);
- request = kzalloc(sizeof(*request), GFP_KERNEL);
- if (ret || request == NULL ||
- i915_add_request(ring, NULL, request))
- kfree(request);
- }
+ if (ring->gpu_caches_dirty)
+ i915_add_request(ring, NULL, NULL);
idle &= list_empty(&ring->request_list);
}
mutex_unlock(&dev->struct_mutex);
}
-static int
-i915_gem_check_wedge(struct drm_i915_private *dev_priv)
+int
+i915_gem_check_wedge(struct drm_i915_private *dev_priv,
+ bool interruptible)
{
- BUG_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
-
if (atomic_read(&dev_priv->mm.wedged)) {
struct completion *x = &dev_priv->error_completion;
bool recovery_complete;
recovery_complete = x->done > 0;
spin_unlock_irqrestore(&x->wait.lock, flags);
- return recovery_complete ? -EIO : -EAGAIN;
+ /* Non-interruptible callers can't handle -EAGAIN, hence return
+ * -EIO unconditionally for these. */
+ if (!interruptible)
+ return -EIO;
+
+ /* Recovery complete, but still wedged means reset failure. */
+ if (recovery_complete)
+ return -EIO;
+
+ return -EAGAIN;
}
return 0;
static int
i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno)
{
- int ret = 0;
+ int ret;
BUG_ON(!mutex_is_locked(&ring->dev->struct_mutex));
- if (seqno == ring->outstanding_lazy_request) {
- struct drm_i915_gem_request *request;
-
- request = kzalloc(sizeof(*request), GFP_KERNEL);
- if (request == NULL)
- return -ENOMEM;
-
- ret = i915_add_request(ring, NULL, request);
- if (ret) {
- kfree(request);
- return ret;
- }
-
- BUG_ON(seqno != request->seqno);
- }
+ ret = 0;
+ if (seqno == ring->outstanding_lazy_request)
+ ret = i915_add_request(ring, NULL, NULL);
return ret;
}
+/**
+ * __wait_seqno - wait until execution of seqno has finished
+ * @ring: the ring expected to report seqno
+ * @seqno: duh!
+ * @interruptible: do an interruptible wait (normally yes)
+ * @timeout: in - how long to wait (NULL forever); out - how much time remaining
+ *
+ * Returns 0 if the seqno was found within the alloted time. Else returns the
+ * errno with remaining time filled in timeout argument.
+ */
static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
- bool interruptible)
+ bool interruptible, struct timespec *timeout)
{
drm_i915_private_t *dev_priv = ring->dev->dev_private;
- int ret = 0;
+ struct timespec before, now, wait_time={1,0};
+ unsigned long timeout_jiffies;
+ long end;
+ bool wait_forever = true;
+ int ret;
if (i915_seqno_passed(ring->get_seqno(ring), seqno))
return 0;
trace_i915_gem_request_wait_begin(ring, seqno);
+
+ if (timeout != NULL) {
+ wait_time = *timeout;
+ wait_forever = false;
+ }
+
+ timeout_jiffies = timespec_to_jiffies(&wait_time);
+
if (WARN_ON(!ring->irq_get(ring)))
return -ENODEV;
+ /* Record current time in case interrupted by signal, or wedged * */
+ getrawmonotonic(&before);
+
#define EXIT_COND \
(i915_seqno_passed(ring->get_seqno(ring), seqno) || \
atomic_read(&dev_priv->mm.wedged))
+ do {
+ if (interruptible)
+ end = wait_event_interruptible_timeout(ring->irq_queue,
+ EXIT_COND,
+ timeout_jiffies);
+ else
+ end = wait_event_timeout(ring->irq_queue, EXIT_COND,
+ timeout_jiffies);
- if (interruptible)
- ret = wait_event_interruptible(ring->irq_queue,
- EXIT_COND);
- else
- wait_event(ring->irq_queue, EXIT_COND);
+ ret = i915_gem_check_wedge(dev_priv, interruptible);
+ if (ret)
+ end = ret;
+ } while (end == 0 && wait_forever);
+
+ getrawmonotonic(&now);
ring->irq_put(ring);
trace_i915_gem_request_wait_end(ring, seqno);
#undef EXIT_COND
- return ret;
+ if (timeout) {
+ struct timespec sleep_time = timespec_sub(now, before);
+ *timeout = timespec_sub(*timeout, sleep_time);
+ }
+
+ switch (end) {
+ case -EIO:
+ case -EAGAIN: /* Wedged */
+ case -ERESTARTSYS: /* Signal */
+ return (int)end;
+ case 0: /* Timeout */
+ if (timeout)
+ set_normalized_timespec(timeout, 0, 0);
+ return -ETIME;
+ default: /* Completed */
+ WARN_ON(end < 0); /* We're not aware of other errors */
+ return 0;
+ }
}
/**
* request and object lists appropriately for that event.
*/
int
-i915_wait_request(struct intel_ring_buffer *ring,
- uint32_t seqno)
+i915_wait_seqno(struct intel_ring_buffer *ring, uint32_t seqno)
{
drm_i915_private_t *dev_priv = ring->dev->dev_private;
int ret = 0;
BUG_ON(seqno == 0);
- ret = i915_gem_check_wedge(dev_priv);
+ ret = i915_gem_check_wedge(dev_priv, dev_priv->mm.interruptible);
if (ret)
return ret;
if (ret)
return ret;
- ret = __wait_seqno(ring, seqno, dev_priv->mm.interruptible);
- if (atomic_read(&dev_priv->mm.wedged))
- ret = -EAGAIN;
+ ret = __wait_seqno(ring, seqno, dev_priv->mm.interruptible, NULL);
return ret;
}
* Ensures that all rendering to the object has completed and the object is
* safe to unbind from the GTT or access from the CPU.
*/
-int
-i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj)
+static __must_check int
+i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
+ bool readonly)
{
+ u32 seqno;
int ret;
- /* This function only exists to support waiting for existing rendering,
- * not for emitting required flushes.
- */
- BUG_ON((obj->base.write_domain & I915_GEM_GPU_DOMAINS) != 0);
-
/* If there is rendering queued on the buffer being evicted, wait for
* it.
*/
+ if (readonly)
+ seqno = obj->last_write_seqno;
+ else
+ seqno = obj->last_read_seqno;
+ if (seqno == 0)
+ return 0;
+
+ ret = i915_wait_seqno(obj->ring, seqno);
+ if (ret)
+ return ret;
+
+ /* Manually manage the write flush as we may have not yet retired
+ * the buffer.
+ */
+ if (obj->last_write_seqno &&
+ i915_seqno_passed(seqno, obj->last_write_seqno)) {
+ obj->last_write_seqno = 0;
+ obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
+ }
+
+ i915_gem_retire_requests_ring(obj->ring);
+ return 0;
+}
+
+/**
+ * Ensures that an object will eventually get non-busy by flushing any required
+ * write domains, emitting any outstanding lazy request and retiring and
+ * completed requests.
+ */
+static int
+i915_gem_object_flush_active(struct drm_i915_gem_object *obj)
+{
+ int ret;
+
if (obj->active) {
- ret = i915_wait_request(obj->ring, obj->last_rendering_seqno);
+ ret = i915_gem_check_olr(obj->ring, obj->last_read_seqno);
if (ret)
return ret;
+
i915_gem_retire_requests_ring(obj->ring);
}
return 0;
}
+/**
+ * i915_gem_wait_ioctl - implements DRM_IOCTL_I915_GEM_WAIT
+ * @DRM_IOCTL_ARGS: standard ioctl arguments
+ *
+ * Returns 0 if successful, else an error is returned with the remaining time in
+ * the timeout parameter.
+ * -ETIME: object is still busy after timeout
+ * -ERESTARTSYS: signal interrupted the wait
+ * -ENONENT: object doesn't exist
+ * Also possible, but rare:
+ * -EAGAIN: GPU wedged
+ * -ENOMEM: damn
+ * -ENODEV: Internal IRQ fail
+ * -E?: The add request failed
+ *
+ * The wait ioctl with a timeout of 0 reimplements the busy ioctl. With any
+ * non-zero timeout parameter the wait ioctl will wait for the given number of
+ * nanoseconds on an object becoming unbusy. Since the wait itself does so
+ * without holding struct_mutex the object may become re-busied before this
+ * function completes. A similar but shorter * race condition exists in the busy
+ * ioctl
+ */
+int
+i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
+{
+ struct drm_i915_gem_wait *args = data;
+ struct drm_i915_gem_object *obj;
+ struct intel_ring_buffer *ring = NULL;
+ struct timespec timeout_stack, *timeout = NULL;
+ u32 seqno = 0;
+ int ret = 0;
+
+ if (args->timeout_ns >= 0) {
+ timeout_stack = ns_to_timespec(args->timeout_ns);
+ timeout = &timeout_stack;
+ }
+
+ ret = i915_mutex_lock_interruptible(dev);
+ if (ret)
+ return ret;
+
+ obj = to_intel_bo(drm_gem_object_lookup(dev, file, args->bo_handle));
+ if (&obj->base == NULL) {
+ mutex_unlock(&dev->struct_mutex);
+ return -ENOENT;
+ }
+
+ /* Need to make sure the object gets inactive eventually. */
+ ret = i915_gem_object_flush_active(obj);
+ if (ret)
+ goto out;
+
+ if (obj->active) {
+ seqno = obj->last_read_seqno;
+ ring = obj->ring;
+ }
+
+ if (seqno == 0)
+ goto out;
+
+ /* Do this after OLR check to make sure we make forward progress polling
+ * on this IOCTL with a 0 timeout (like busy ioctl)
+ */
+ if (!args->timeout_ns) {
+ ret = -ETIME;
+ goto out;
+ }
+
+ drm_gem_object_unreference(&obj->base);
+ mutex_unlock(&dev->struct_mutex);
+
+ ret = __wait_seqno(ring, seqno, true, timeout);
+ if (timeout) {
+ WARN_ON(!timespec_valid(timeout));
+ args->timeout_ns = timespec_to_ns(timeout);
+ }
+ return ret;
+
+out:
+ drm_gem_object_unreference(&obj->base);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+}
+
/**
* i915_gem_object_sync - sync an object to a ring.
*
return 0;
if (to == NULL || !i915_semaphore_is_enabled(obj->base.dev))
- return i915_gem_object_wait_rendering(obj);
+ return i915_gem_object_wait_rendering(obj, false);
idx = intel_ring_sync_index(from, to);
- seqno = obj->last_rendering_seqno;
+ seqno = obj->last_read_seqno;
if (seqno <= from->sync_seqno[idx])
return 0;
if (ret)
return ret;
- if (flush_domains & I915_GEM_GPU_DOMAINS)
- i915_gem_process_flushing_list(ring, flush_domains);
-
return 0;
}
static int i915_ring_idle(struct intel_ring_buffer *ring)
{
- int ret;
-
- if (list_empty(&ring->gpu_write_list) && list_empty(&ring->active_list))
+ if (list_empty(&ring->active_list))
return 0;
- if (!list_empty(&ring->gpu_write_list)) {
- ret = i915_gem_flush_ring(ring,
- I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
- if (ret)
- return ret;
- }
-
- return i915_wait_request(ring, i915_gem_next_request_seqno(ring));
+ return i915_wait_seqno(ring, i915_gem_next_request_seqno(ring));
}
int i915_gpu_idle(struct drm_device *dev)
if (ret)
return ret;
- /* Is the device fubar? */
- if (WARN_ON(!list_empty(&ring->gpu_write_list)))
- return -EBUSY;
+ ret = i915_switch_context(ring, NULL, DEFAULT_CONTEXT_ID);
+ if (ret)
+ return ret;
}
return 0;
static int
i915_gem_object_flush_fence(struct drm_i915_gem_object *obj)
{
- int ret;
-
- if (obj->fenced_gpu_access) {
- if (obj->base.write_domain & I915_GEM_GPU_DOMAINS) {
- ret = i915_gem_flush_ring(obj->ring,
- 0, obj->base.write_domain);
- if (ret)
- return ret;
- }
-
- obj->fenced_gpu_access = false;
- }
-
if (obj->last_fenced_seqno) {
- ret = i915_wait_request(obj->ring, obj->last_fenced_seqno);
+ int ret = i915_wait_seqno(obj->ring, obj->last_fenced_seqno);
if (ret)
return ret;
if (obj->base.read_domains & I915_GEM_DOMAIN_GTT)
mb();
+ obj->fenced_gpu_access = false;
return 0;
}
if (map_and_fenceable)
free_space =
drm_mm_search_free_in_range(&dev_priv->mm.gtt_space,
- size, alignment, 0,
- dev_priv->mm.gtt_mappable_end,
+ size, alignment,
+ 0, dev_priv->mm.gtt_mappable_end,
0);
else
free_space = drm_mm_search_free(&dev_priv->mm.gtt_space,
obj->gtt_space =
drm_mm_get_block_range_generic(free_space,
size, alignment, 0,
- dev_priv->mm.gtt_mappable_end,
+ 0, dev_priv->mm.gtt_mappable_end,
0);
else
obj->gtt_space =
drm_clflush_pages(obj->pages, obj->base.size / PAGE_SIZE);
}
-/** Flushes any GPU write domain for the object if it's dirty. */
-static int
-i915_gem_object_flush_gpu_write_domain(struct drm_i915_gem_object *obj)
-{
- if ((obj->base.write_domain & I915_GEM_GPU_DOMAINS) == 0)
- return 0;
-
- /* Queue the GPU write cache flushing we need. */
- return i915_gem_flush_ring(obj->ring, 0, obj->base.write_domain);
-}
-
/** Flushes the GTT write domain for the object if it's dirty. */
static void
i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj)
if (obj->base.write_domain == I915_GEM_DOMAIN_GTT)
return 0;
- ret = i915_gem_object_flush_gpu_write_domain(obj);
+ ret = i915_gem_object_wait_rendering(obj, !write);
if (ret)
return ret;
- if (obj->pending_gpu_write || write) {
- ret = i915_gem_object_wait_rendering(obj);
- if (ret)
- return ret;
- }
-
i915_gem_object_flush_cpu_write_domain(obj);
old_write_domain = obj->base.write_domain;
u32 old_read_domains, old_write_domain;
int ret;
- ret = i915_gem_object_flush_gpu_write_domain(obj);
- if (ret)
- return ret;
-
if (pipelined != obj->ring) {
ret = i915_gem_object_sync(obj, pipelined);
if (ret)
/* It should now be out of any other write domains, and we can update
* the domain values for our changes.
*/
- BUG_ON((obj->base.write_domain & ~I915_GEM_DOMAIN_GTT) != 0);
+ obj->base.write_domain = 0;
obj->base.read_domains |= I915_GEM_DOMAIN_GTT;
trace_i915_gem_object_change_domain(obj,
if ((obj->base.read_domains & I915_GEM_GPU_DOMAINS) == 0)
return 0;
- if (obj->base.write_domain & I915_GEM_GPU_DOMAINS) {
- ret = i915_gem_flush_ring(obj->ring, 0, obj->base.write_domain);
- if (ret)
- return ret;
- }
-
- ret = i915_gem_object_wait_rendering(obj);
+ ret = i915_gem_object_wait_rendering(obj, false);
if (ret)
return ret;
if (obj->base.write_domain == I915_GEM_DOMAIN_CPU)
return 0;
- ret = i915_gem_object_flush_gpu_write_domain(obj);
+ ret = i915_gem_object_wait_rendering(obj, !write);
if (ret)
return ret;
- if (write || obj->pending_gpu_write) {
- ret = i915_gem_object_wait_rendering(obj);
- if (ret)
- return ret;
- }
-
i915_gem_object_flush_gtt_write_domain(obj);
old_write_domain = obj->base.write_domain;
if (seqno == 0)
return 0;
- ret = __wait_seqno(ring, seqno, true);
+ ret = __wait_seqno(ring, seqno, true, NULL);
if (ret == 0)
queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0);
* become non-busy without any further actions, therefore emit any
* necessary flushes here.
*/
- args->busy = obj->active;
- if (args->busy) {
- /* Unconditionally flush objects, even when the gpu still uses this
- * object. Userspace calling this function indicates that it wants to
- * use this buffer rather sooner than later, so issuing the required
- * flush earlier is beneficial.
- */
- if (obj->base.write_domain & I915_GEM_GPU_DOMAINS) {
- ret = i915_gem_flush_ring(obj->ring,
- 0, obj->base.write_domain);
- } else {
- ret = i915_gem_check_olr(obj->ring,
- obj->last_rendering_seqno);
- }
-
- /* Update the active list for the hardware's current position.
- * Otherwise this only updates on a delayed timer or when irqs
- * are actually unmasked, and our working set ends up being
- * larger than required.
- */
- i915_gem_retire_requests_ring(obj->ring);
+ ret = i915_gem_object_flush_active(obj);
- args->busy = obj->active;
+ args->busy = obj->active;
+ if (obj->ring) {
+ BUILD_BUG_ON(I915_NUM_RINGS > 16);
+ args->busy |= intel_ring_flag(obj->ring) << 16;
}
drm_gem_object_unreference(&obj->base);
INIT_LIST_HEAD(&obj->gtt_list);
INIT_LIST_HEAD(&obj->ring_list);
INIT_LIST_HEAD(&obj->exec_list);
- INIT_LIST_HEAD(&obj->gpu_write_list);
obj->madv = I915_MADV_WILLNEED;
/* Avoid an unnecessary call to unbind on the first bind. */
obj->map_and_fenceable = true;
return 0;
}
+void i915_gem_l3_remap(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ u32 misccpctl;
+ int i;
+
+ if (!IS_IVYBRIDGE(dev))
+ return;
+
+ if (!dev_priv->mm.l3_remap_info)
+ return;
+
+ misccpctl = I915_READ(GEN7_MISCCPCTL);
+ I915_WRITE(GEN7_MISCCPCTL, misccpctl & ~GEN7_DOP_CLOCK_GATE_ENABLE);
+ POSTING_READ(GEN7_MISCCPCTL);
+
+ for (i = 0; i < GEN7_L3LOG_SIZE; i += 4) {
+ u32 remap = I915_READ(GEN7_L3LOG_BASE + i);
+ if (remap && remap != dev_priv->mm.l3_remap_info[i/4])
+ DRM_DEBUG("0x%x was already programmed to %x\n",
+ GEN7_L3LOG_BASE + i, remap);
+ if (remap && !dev_priv->mm.l3_remap_info[i/4])
+ DRM_DEBUG_DRIVER("Clearing remapped register\n");
+ I915_WRITE(GEN7_L3LOG_BASE + i, dev_priv->mm.l3_remap_info[i/4]);
+ }
+
+ /* Make sure all the writes land before disabling dop clock gating */
+ POSTING_READ(GEN7_L3LOG_BASE);
+
+ I915_WRITE(GEN7_MISCCPCTL, misccpctl);
+}
+
void i915_gem_init_swizzling(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = dev->dev_private;
}
}
+static bool
+intel_enable_blt(struct drm_device *dev)
+{
+ if (!HAS_BLT(dev))
+ return false;
+
+ /* The blitter was dysfunctional on early prototypes */
+ if (IS_GEN6(dev) && dev->pdev->revision < 8) {
+ DRM_INFO("BLT not supported on this pre-production hardware;"
+ " graphics performance will be degraded.\n");
+ return false;
+ }
+
+ return true;
+}
+
int
i915_gem_init_hw(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = dev->dev_private;
int ret;
+ if (!intel_enable_gtt())
+ return -EIO;
+
+ i915_gem_l3_remap(dev);
+
i915_gem_init_swizzling(dev);
ret = intel_init_render_ring_buffer(dev);
goto cleanup_render_ring;
}
- if (HAS_BLT(dev)) {
+ if (intel_enable_blt(dev)) {
ret = intel_init_blt_ring_buffer(dev);
if (ret)
goto cleanup_bsd_ring;
dev_priv->next_seqno = 1;
+ /*
+ * XXX: There was some w/a described somewhere suggesting loading
+ * contexts before PPGTT.
+ */
+ i915_gem_context_init(dev);
i915_gem_init_ppgtt(dev);
return 0;
}
BUG_ON(!list_empty(&dev_priv->mm.active_list));
- BUG_ON(!list_empty(&dev_priv->mm.flushing_list));
BUG_ON(!list_empty(&dev_priv->mm.inactive_list));
mutex_unlock(&dev->struct_mutex);
{
INIT_LIST_HEAD(&ring->active_list);
INIT_LIST_HEAD(&ring->request_list);
- INIT_LIST_HEAD(&ring->gpu_write_list);
}
void
drm_i915_private_t *dev_priv = dev->dev_private;
INIT_LIST_HEAD(&dev_priv->mm.active_list);
- INIT_LIST_HEAD(&dev_priv->mm.flushing_list);
INIT_LIST_HEAD(&dev_priv->mm.inactive_list);
INIT_LIST_HEAD(&dev_priv->mm.fence_list);
INIT_LIST_HEAD(&dev_priv->mm.gtt_list);
i915_gpu_is_active(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = dev->dev_private;
- int lists_empty;
-
- lists_empty = list_empty(&dev_priv->mm.flushing_list) &&
- list_empty(&dev_priv->mm.active_list);
-
- return !lists_empty;
+ return !list_empty(&dev_priv->mm.active_list);
}
static int