mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-02 19:36:42 +07:00
drm/i915: Combine seqno + tracking into a global timeline struct
Our timelines are more than just a seqno. They also provide an ordered list of requests to be executed. Due to the restriction of handling individual address spaces, we are limited to a timeline per address space but we use a fence context per engine within. Our first step to introducing independent timelines per context (i.e. to allow each context to have a queue of requests to execute that have a defined set of dependencies on other requests) is to provide a timeline abstraction for the global execution queue. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-23-chris@chris-wilson.co.uk
This commit is contained in:
Chris Wilson
parent
c004a90b72
commit
73cb97010d
@ -42,6 +42,7 @@ i915-y += i915_cmd_parser.o \
|
||||
i915_gem_shrinker.o \
|
||||
i915_gem_stolen.o \
|
||||
i915_gem_tiling.o \
|
||||
i915_gem_timeline.o \
|
||||
i915_gem_userptr.o \
|
||||
i915_trace_points.o \
|
||||
intel_breadcrumbs.o \
|
||||
|
||||
@ -552,7 +552,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
|
||||
seq_printf(m, "Flip queued on %s at seqno %x, next seqno %x [current breadcrumb %x], completed? %d\n",
|
||||
engine->name,
|
||||
i915_gem_request_get_seqno(work->flip_queued_req),
|
||||
dev_priv->next_seqno,
|
||||
dev_priv->gt.global_timeline.next_seqno,
|
||||
intel_engine_get_seqno(engine),
|
||||
i915_gem_request_completed(work->flip_queued_req));
|
||||
} else
|
||||
@ -662,13 +662,13 @@ static int i915_gem_request_info(struct seq_file *m, void *data)
|
||||
int count;
|
||||
|
||||
count = 0;
|
||||
list_for_each_entry(req, &engine->request_list, link)
|
||||
list_for_each_entry(req, &engine->timeline->requests, link)
|
||||
count++;
|
||||
if (count == 0)
|
||||
continue;
|
||||
|
||||
seq_printf(m, "%s requests: %d\n", engine->name, count);
|
||||
list_for_each_entry(req, &engine->request_list, link)
|
||||
list_for_each_entry(req, &engine->timeline->requests, link)
|
||||
print_request(m, req, " ");
|
||||
|
||||
any++;
|
||||
@ -1052,15 +1052,8 @@ static int
|
||||
i915_next_seqno_get(void *data, u64 *val)
|
||||
{
|
||||
struct drm_i915_private *dev_priv = data;
|
||||
int ret;
|
||||
|
||||
ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex);
|
||||
if (ret)
|
||||
return ret;
|
||||
|
||||
*val = dev_priv->next_seqno;
|
||||
mutex_unlock(&dev_priv->drm.struct_mutex);
|
||||
|
||||
*val = READ_ONCE(dev_priv->gt.global_timeline.next_seqno);
|
||||
return 0;
|
||||
}
|
||||
|
||||
@ -1075,7 +1068,7 @@ i915_next_seqno_set(void *data, u64 val)
|
||||
if (ret)
|
||||
return ret;
|
||||
|
||||
ret = i915_gem_set_seqno(dev, val);
|
||||
ret = i915_gem_set_global_seqno(dev, val);
|
||||
mutex_unlock(&dev->struct_mutex);
|
||||
|
||||
return ret;
|
||||
@ -1364,7 +1357,7 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
|
||||
seq_printf(m, "\tseqno = %x [current %x, last %x]\n",
|
||||
engine->hangcheck.seqno,
|
||||
seqno[id],
|
||||
engine->last_submitted_seqno);
|
||||
engine->timeline->last_submitted_seqno);
|
||||
seq_printf(m, "\twaiters? %s, fake irq active? %s\n",
|
||||
yesno(intel_engine_has_waiter(engine)),
|
||||
yesno(test_bit(engine->id,
|
||||
@ -3181,7 +3174,7 @@ static int i915_engine_info(struct seq_file *m, void *unused)
|
||||
seq_printf(m, "%s\n", engine->name);
|
||||
seq_printf(m, "\tcurrent seqno %x, last %x, hangcheck %x [score %d]\n",
|
||||
intel_engine_get_seqno(engine),
|
||||
engine->last_submitted_seqno,
|
||||
engine->timeline->last_submitted_seqno,
|
||||
engine->hangcheck.seqno,
|
||||
engine->hangcheck.score);
|
||||
|
||||
@ -3189,14 +3182,14 @@ static int i915_engine_info(struct seq_file *m, void *unused)
|
||||
|
||||
seq_printf(m, "\tRequests:\n");
|
||||
|
||||
rq = list_first_entry(&engine->request_list,
|
||||
struct drm_i915_gem_request, link);
|
||||
if (&rq->link != &engine->request_list)
|
||||
rq = list_first_entry(&engine->timeline->requests,
|
||||
struct drm_i915_gem_request, link);
|
||||
if (&rq->link != &engine->timeline->requests)
|
||||
print_request(m, rq, "\t\tfirst ");
|
||||
|
||||
rq = list_last_entry(&engine->request_list,
|
||||
struct drm_i915_gem_request, link);
|
||||
if (&rq->link != &engine->request_list)
|
||||
rq = list_last_entry(&engine->timeline->requests,
|
||||
struct drm_i915_gem_request, link);
|
||||
if (&rq->link != &engine->timeline->requests)
|
||||
print_request(m, rq, "\t\tlast ");
|
||||
|
||||
rq = i915_gem_find_active_request(engine);
|
||||
|
||||
@ -831,7 +831,9 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
|
||||
intel_init_display_hooks(dev_priv);
|
||||
intel_init_clock_gating_hooks(dev_priv);
|
||||
intel_init_audio_hooks(dev_priv);
|
||||
i915_gem_load_init(&dev_priv->drm);
|
||||
ret = i915_gem_load_init(&dev_priv->drm);
|
||||
if (ret < 0)
|
||||
goto err_gvt;
|
||||
|
||||
intel_display_crc_init(dev_priv);
|
||||
|
||||
@ -841,6 +843,8 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
|
||||
|
||||
return 0;
|
||||
|
||||
err_gvt:
|
||||
intel_gvt_cleanup(dev_priv);
|
||||
err_workqueues:
|
||||
i915_workqueues_cleanup(dev_priv);
|
||||
return ret;
|
||||
|
||||
@ -63,6 +63,7 @@
|
||||
#include "i915_gem_gtt.h"
|
||||
#include "i915_gem_render_state.h"
|
||||
#include "i915_gem_request.h"
|
||||
#include "i915_gem_timeline.h"
|
||||
|
||||
#include "intel_gvt.h"
|
||||
|
||||
@ -1830,7 +1831,6 @@ struct drm_i915_private {
|
||||
struct i915_gem_context *kernel_context;
|
||||
struct intel_engine_cs *engine[I915_NUM_ENGINES];
|
||||
struct i915_vma *semaphore;
|
||||
u32 next_seqno;
|
||||
|
||||
struct drm_dma_handle *status_page_dmah;
|
||||
struct resource mch_res;
|
||||
@ -2090,6 +2090,9 @@ struct drm_i915_private {
|
||||
void (*resume)(struct drm_i915_private *);
|
||||
void (*cleanup_engine)(struct intel_engine_cs *engine);
|
||||
|
||||
struct list_head timelines;
|
||||
struct i915_gem_timeline global_timeline;
|
||||
|
||||
/**
|
||||
* Is the GPU currently considered idle, or busy executing
|
||||
* userspace requests? Whilst idle, we allow runtime power
|
||||
@ -3175,7 +3178,7 @@ int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
|
||||
struct drm_file *file_priv);
|
||||
int i915_gem_wait_ioctl(struct drm_device *dev, void *data,
|
||||
struct drm_file *file_priv);
|
||||
void i915_gem_load_init(struct drm_device *dev);
|
||||
int i915_gem_load_init(struct drm_device *dev);
|
||||
void i915_gem_load_cleanup(struct drm_device *dev);
|
||||
void i915_gem_load_init_fences(struct drm_i915_private *dev_priv);
|
||||
int i915_gem_freeze(struct drm_i915_private *dev_priv);
|
||||
@ -3347,7 +3350,7 @@ void i915_gem_track_fb(struct drm_i915_gem_object *old,
|
||||
struct drm_i915_gem_object *new,
|
||||
unsigned frontbuffer_bits);
|
||||
|
||||
int __must_check i915_gem_set_seqno(struct drm_device *dev, u32 seqno);
|
||||
int __must_check i915_gem_set_global_seqno(struct drm_device *dev, u32 seqno);
|
||||
|
||||
struct drm_i915_gem_request *
|
||||
i915_gem_find_active_request(struct intel_engine_cs *engine);
|
||||
|
||||
@ -371,7 +371,7 @@ i915_gem_object_wait_fence(struct dma_fence *fence,
|
||||
if (flags & I915_WAIT_LOCKED && i915_gem_request_completed(rq))
|
||||
i915_gem_request_retire_upto(rq);
|
||||
|
||||
if (rps && rq->fence.seqno == rq->engine->last_submitted_seqno) {
|
||||
if (rps && rq->fence.seqno == rq->timeline->last_submitted_seqno) {
|
||||
/* The GPU is now idle and this client has stalled.
|
||||
* Since no other client has submitted a request in the
|
||||
* meantime, assume that this client is the only one
|
||||
@ -2563,7 +2563,7 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
|
||||
* extra delay for a recent interrupt is pointless. Hence, we do
|
||||
* not need an engine->irq_seqno_barrier() before the seqno reads.
|
||||
*/
|
||||
list_for_each_entry(request, &engine->request_list, link) {
|
||||
list_for_each_entry(request, &engine->timeline->requests, link) {
|
||||
if (i915_gem_request_completed(request))
|
||||
continue;
|
||||
|
||||
@ -2632,7 +2632,7 @@ static void i915_gem_reset_engine(struct intel_engine_cs *engine)
|
||||
if (i915_gem_context_is_default(incomplete_ctx))
|
||||
return;
|
||||
|
||||
list_for_each_entry_continue(request, &engine->request_list, link)
|
||||
list_for_each_entry_continue(request, &engine->timeline->requests, link)
|
||||
if (request->ctx == incomplete_ctx)
|
||||
reset_request(request);
|
||||
}
|
||||
@ -2671,7 +2671,8 @@ static void i915_gem_cleanup_engine(struct intel_engine_cs *engine)
|
||||
* (lockless) lookup doesn't try and wait upon the request as we
|
||||
* reset it.
|
||||
*/
|
||||
intel_engine_init_seqno(engine, engine->last_submitted_seqno);
|
||||
intel_engine_init_global_seqno(engine,
|
||||
engine->timeline->last_submitted_seqno);
|
||||
|
||||
/*
|
||||
* Clear the execlists queue up before freeing the requests, as those
|
||||
@ -2979,18 +2980,26 @@ int i915_vma_unbind(struct i915_vma *vma)
|
||||
return 0;
|
||||
}
|
||||
|
||||
int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv,
|
||||
unsigned int flags)
|
||||
static int wait_for_timeline(struct i915_gem_timeline *tl, unsigned int flags)
|
||||
{
|
||||
struct intel_engine_cs *engine;
|
||||
enum intel_engine_id id;
|
||||
int ret, i;
|
||||
|
||||
for (i = 0; i < ARRAY_SIZE(tl->engine); i++) {
|
||||
ret = i915_gem_active_wait(&tl->engine[i].last_request, flags);
|
||||
if (ret)
|
||||
return ret;
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags)
|
||||
{
|
||||
struct i915_gem_timeline *tl;
|
||||
int ret;
|
||||
|
||||
for_each_engine(engine, dev_priv, id) {
|
||||
if (engine->last_context == NULL)
|
||||
continue;
|
||||
|
||||
ret = intel_engine_idle(engine, flags);
|
||||
list_for_each_entry(tl, &i915->gt.timelines, link) {
|
||||
ret = wait_for_timeline(tl, flags);
|
||||
if (ret)
|
||||
return ret;
|
||||
}
|
||||
@ -4680,21 +4689,32 @@ i915_gem_load_init_fences(struct drm_i915_private *dev_priv)
|
||||
i915_gem_detect_bit_6_swizzle(dev);
|
||||
}
|
||||
|
||||
void
|
||||
int
|
||||
i915_gem_load_init(struct drm_device *dev)
|
||||
{
|
||||
struct drm_i915_private *dev_priv = to_i915(dev);
|
||||
int err;
|
||||
|
||||
dev_priv->objects =
|
||||
kmem_cache_create("i915_gem_object",
|
||||
sizeof(struct drm_i915_gem_object), 0,
|
||||
SLAB_HWCACHE_ALIGN,
|
||||
NULL);
|
||||
if (!dev_priv->objects) {
|
||||
err = -ENOMEM;
|
||||
goto err_out;
|
||||
}
|
||||
|
||||
dev_priv->vmas =
|
||||
kmem_cache_create("i915_gem_vma",
|
||||
sizeof(struct i915_vma), 0,
|
||||
SLAB_HWCACHE_ALIGN,
|
||||
NULL);
|
||||
if (!dev_priv->vmas) {
|
||||
err = -ENOMEM;
|
||||
goto err_objects;
|
||||
}
|
||||
|
||||
dev_priv->requests =
|
||||
kmem_cache_create("i915_gem_request",
|
||||
sizeof(struct drm_i915_gem_request), 0,
|
||||
@ -4702,6 +4722,19 @@ i915_gem_load_init(struct drm_device *dev)
|
||||
SLAB_RECLAIM_ACCOUNT |
|
||||
SLAB_DESTROY_BY_RCU,
|
||||
NULL);
|
||||
if (!dev_priv->requests) {
|
||||
err = -ENOMEM;
|
||||
goto err_vmas;
|
||||
}
|
||||
|
||||
mutex_lock(&dev_priv->drm.struct_mutex);
|
||||
INIT_LIST_HEAD(&dev_priv->gt.timelines);
|
||||
err = i915_gem_timeline_init(dev_priv,
|
||||
&dev_priv->gt.global_timeline,
|
||||
"[execution]");
|
||||
mutex_unlock(&dev_priv->drm.struct_mutex);
|
||||
if (err)
|
||||
goto err_requests;
|
||||
|
||||
INIT_LIST_HEAD(&dev_priv->context_list);
|
||||
INIT_WORK(&dev_priv->mm.free_work, __i915_gem_free_work);
|
||||
@ -4726,6 +4759,17 @@ i915_gem_load_init(struct drm_device *dev)
|
||||
atomic_set(&dev_priv->mm.bsd_engine_dispatch_index, 0);
|
||||
|
||||
spin_lock_init(&dev_priv->fb_tracking.lock);
|
||||
|
||||
return 0;
|
||||
|
||||
err_requests:
|
||||
kmem_cache_destroy(dev_priv->requests);
|
||||
err_vmas:
|
||||
kmem_cache_destroy(dev_priv->vmas);
|
||||
err_objects:
|
||||
kmem_cache_destroy(dev_priv->objects);
|
||||
err_out:
|
||||
return err;
|
||||
}
|
||||
|
||||
void i915_gem_load_cleanup(struct drm_device *dev)
|
||||
|
||||
@ -31,4 +31,6 @@
|
||||
#define GEM_BUG_ON(expr)
|
||||
#endif
|
||||
|
||||
#define I915_NUM_ENGINES 5
|
||||
|
||||
#endif /* __I915_GEM_H__ */
|
||||
|
||||
@ -40,7 +40,7 @@ static const char *i915_fence_get_timeline_name(struct dma_fence *fence)
|
||||
* multiple execution contexts (fence contexts) as we allow
|
||||
* engines within a single timeline to execute in parallel.
|
||||
*/
|
||||
return "global";
|
||||
return to_request(fence)->timeline->common->name;
|
||||
}
|
||||
|
||||
static bool i915_fence_signaled(struct dma_fence *fence)
|
||||
@ -211,7 +211,7 @@ void i915_gem_request_retire_upto(struct drm_i915_gem_request *req)
|
||||
return;
|
||||
|
||||
do {
|
||||
tmp = list_first_entry(&engine->request_list,
|
||||
tmp = list_first_entry(&engine->timeline->requests,
|
||||
typeof(*tmp), link);
|
||||
|
||||
i915_gem_request_retire(tmp);
|
||||
@ -238,37 +238,39 @@ static int i915_gem_check_wedge(struct drm_i915_private *dev_priv)
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int i915_gem_init_seqno(struct drm_i915_private *dev_priv, u32 seqno)
|
||||
static int i915_gem_init_global_seqno(struct drm_i915_private *dev_priv,
|
||||
u32 seqno)
|
||||
{
|
||||
struct i915_gem_timeline *timeline = &dev_priv->gt.global_timeline;
|
||||
struct intel_engine_cs *engine;
|
||||
enum intel_engine_id id;
|
||||
int ret;
|
||||
|
||||
/* Carefully retire all requests without writing to the rings */
|
||||
for_each_engine(engine, dev_priv, id) {
|
||||
ret = intel_engine_idle(engine,
|
||||
I915_WAIT_INTERRUPTIBLE |
|
||||
I915_WAIT_LOCKED);
|
||||
if (ret)
|
||||
return ret;
|
||||
}
|
||||
ret = i915_gem_wait_for_idle(dev_priv,
|
||||
I915_WAIT_INTERRUPTIBLE |
|
||||
I915_WAIT_LOCKED);
|
||||
if (ret)
|
||||
return ret;
|
||||
|
||||
i915_gem_retire_requests(dev_priv);
|
||||
|
||||
/* If the seqno wraps around, we need to clear the breadcrumb rbtree */
|
||||
if (!i915_seqno_passed(seqno, dev_priv->next_seqno)) {
|
||||
if (!i915_seqno_passed(seqno, timeline->next_seqno)) {
|
||||
while (intel_kick_waiters(dev_priv) ||
|
||||
intel_kick_signalers(dev_priv))
|
||||
yield();
|
||||
yield();
|
||||
}
|
||||
|
||||
/* Finally reset hw state */
|
||||
for_each_engine(engine, dev_priv, id)
|
||||
intel_engine_init_seqno(engine, seqno);
|
||||
intel_engine_init_global_seqno(engine, seqno);
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
int i915_gem_set_seqno(struct drm_device *dev, u32 seqno)
|
||||
int i915_gem_set_global_seqno(struct drm_device *dev, u32 seqno)
|
||||
{
|
||||
struct drm_i915_private *dev_priv = to_i915(dev);
|
||||
int ret;
|
||||
@ -281,28 +283,31 @@ int i915_gem_set_seqno(struct drm_device *dev, u32 seqno)
|
||||
/* HWS page needs to be set less than what we
|
||||
* will inject to ring
|
||||
*/
|
||||
ret = i915_gem_init_seqno(dev_priv, seqno - 1);
|
||||
ret = i915_gem_init_global_seqno(dev_priv, seqno - 1);
|
||||
if (ret)
|
||||
return ret;
|
||||
|
||||
dev_priv->next_seqno = seqno;
|
||||
dev_priv->gt.global_timeline.next_seqno = seqno;
|
||||
return 0;
|
||||
}
|
||||
|
||||
static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno)
|
||||
static int i915_gem_get_global_seqno(struct drm_i915_private *dev_priv,
|
||||
u32 *seqno)
|
||||
{
|
||||
struct i915_gem_timeline *tl = &dev_priv->gt.global_timeline;
|
||||
|
||||
/* reserve 0 for non-seqno */
|
||||
if (unlikely(dev_priv->next_seqno == 0)) {
|
||||
if (unlikely(tl->next_seqno == 0)) {
|
||||
int ret;
|
||||
|
||||
ret = i915_gem_init_seqno(dev_priv, 0);
|
||||
ret = i915_gem_init_global_seqno(dev_priv, 0);
|
||||
if (ret)
|
||||
return ret;
|
||||
|
||||
dev_priv->next_seqno = 1;
|
||||
tl->next_seqno = 1;
|
||||
}
|
||||
|
||||
*seqno = dev_priv->next_seqno++;
|
||||
*seqno = tl->next_seqno++;
|
||||
return 0;
|
||||
}
|
||||
|
||||
@ -311,13 +316,14 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
|
||||
{
|
||||
struct drm_i915_gem_request *request =
|
||||
container_of(fence, typeof(*request), submit);
|
||||
struct intel_engine_cs *engine = request->engine;
|
||||
|
||||
/* Will be called from irq-context when using foreign DMA fences */
|
||||
|
||||
switch (state) {
|
||||
case FENCE_COMPLETE:
|
||||
request->engine->last_submitted_seqno = request->fence.seqno;
|
||||
request->engine->submit_request(request);
|
||||
engine->timeline->last_submitted_seqno = request->fence.seqno;
|
||||
engine->submit_request(request);
|
||||
break;
|
||||
|
||||
case FENCE_FREE:
|
||||
@ -357,7 +363,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
|
||||
return ERR_PTR(ret);
|
||||
|
||||
/* Move the oldest request to the slab-cache (if not in use!) */
|
||||
req = list_first_entry_or_null(&engine->request_list,
|
||||
req = list_first_entry_or_null(&engine->timeline->requests,
|
||||
typeof(*req), link);
|
||||
if (req && i915_gem_request_completed(req))
|
||||
i915_gem_request_retire(req);
|
||||
@ -394,15 +400,17 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
|
||||
if (!req)
|
||||
return ERR_PTR(-ENOMEM);
|
||||
|
||||
ret = i915_gem_get_seqno(dev_priv, &seqno);
|
||||
ret = i915_gem_get_global_seqno(dev_priv, &seqno);
|
||||
if (ret)
|
||||
goto err;
|
||||
|
||||
req->timeline = engine->timeline;
|
||||
|
||||
spin_lock_init(&req->lock);
|
||||
dma_fence_init(&req->fence,
|
||||
&i915_fence_ops,
|
||||
&req->lock,
|
||||
engine->fence_context,
|
||||
req->timeline->fence_context,
|
||||
seqno);
|
||||
|
||||
i915_sw_fence_init(&req->submit, submit_notify);
|
||||
@ -457,9 +465,16 @@ i915_gem_request_await_request(struct drm_i915_gem_request *to,
|
||||
|
||||
GEM_BUG_ON(to == from);
|
||||
|
||||
if (to->engine == from->engine)
|
||||
if (to->timeline == from->timeline)
|
||||
return 0;
|
||||
|
||||
if (to->engine == from->engine) {
|
||||
ret = i915_sw_fence_await_sw_fence_gfp(&to->submit,
|
||||
&from->submit,
|
||||
GFP_KERNEL);
|
||||
return ret < 0 ? ret : 0;
|
||||
}
|
||||
|
||||
idx = intel_engine_sync_index(from->engine, to->engine);
|
||||
if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx])
|
||||
return 0;
|
||||
@ -622,6 +637,7 @@ void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches)
|
||||
{
|
||||
struct intel_engine_cs *engine = request->engine;
|
||||
struct intel_ring *ring = request->ring;
|
||||
struct intel_timeline *timeline = request->timeline;
|
||||
struct drm_i915_gem_request *prev;
|
||||
u32 request_start;
|
||||
u32 reserved_tail;
|
||||
@ -679,17 +695,17 @@ void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches)
|
||||
* see a more recent value in the hws than we are tracking.
|
||||
*/
|
||||
|
||||
prev = i915_gem_active_raw(&engine->last_request,
|
||||
prev = i915_gem_active_raw(&timeline->last_request,
|
||||
&request->i915->drm.struct_mutex);
|
||||
if (prev)
|
||||
i915_sw_fence_await_sw_fence(&request->submit, &prev->submit,
|
||||
&request->submitq);
|
||||
|
||||
request->emitted_jiffies = jiffies;
|
||||
request->previous_seqno = engine->last_pending_seqno;
|
||||
engine->last_pending_seqno = request->fence.seqno;
|
||||
i915_gem_active_set(&engine->last_request, request);
|
||||
list_add_tail(&request->link, &engine->request_list);
|
||||
request->previous_seqno = timeline->last_pending_seqno;
|
||||
timeline->last_pending_seqno = request->fence.seqno;
|
||||
i915_gem_active_set(&timeline->last_request, request);
|
||||
list_add_tail(&request->link, &timeline->requests);
|
||||
list_add_tail(&request->ring_link, &ring->request_list);
|
||||
|
||||
i915_gem_mark_busy(engine);
|
||||
@ -899,7 +915,8 @@ static bool engine_retire_requests(struct intel_engine_cs *engine)
|
||||
{
|
||||
struct drm_i915_gem_request *request, *next;
|
||||
|
||||
list_for_each_entry_safe(request, next, &engine->request_list, link) {
|
||||
list_for_each_entry_safe(request, next,
|
||||
&engine->timeline->requests, link) {
|
||||
if (!i915_gem_request_completed(request))
|
||||
return false;
|
||||
|
||||
|
||||
@ -81,6 +81,7 @@ struct drm_i915_gem_request {
|
||||
struct i915_gem_context *ctx;
|
||||
struct intel_engine_cs *engine;
|
||||
struct intel_ring *ring;
|
||||
struct intel_timeline *timeline;
|
||||
struct intel_signal_node signaling;
|
||||
|
||||
struct i915_sw_fence submit;
|
||||
|
||||
64
drivers/gpu/drm/i915/i915_gem_timeline.c
Normal file
64
drivers/gpu/drm/i915/i915_gem_timeline.c
Normal file
@ -0,0 +1,64 @@
|
||||
/*
|
||||
* Copyright © 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.
|
||||
*
|
||||
*/
|
||||
|
||||
#include "i915_drv.h"
|
||||
|
||||
int i915_gem_timeline_init(struct drm_i915_private *i915,
|
||||
struct i915_gem_timeline *timeline,
|
||||
const char *name)
|
||||
{
|
||||
unsigned int i;
|
||||
u64 fences;
|
||||
|
||||
lockdep_assert_held(&i915->drm.struct_mutex);
|
||||
|
||||
timeline->i915 = i915;
|
||||
timeline->name = kstrdup(name ?: "[kernel]", GFP_KERNEL);
|
||||
if (!timeline->name)
|
||||
return -ENOMEM;
|
||||
|
||||
list_add(&timeline->link, &i915->gt.timelines);
|
||||
|
||||
/* Called during early_init before we know how many engines there are */
|
||||
fences = dma_fence_context_alloc(ARRAY_SIZE(timeline->engine));
|
||||
for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
|
||||
struct intel_timeline *tl = &timeline->engine[i];
|
||||
|
||||
tl->fence_context = fences++;
|
||||
tl->common = timeline;
|
||||
|
||||
init_request_active(&tl->last_request, NULL);
|
||||
INIT_LIST_HEAD(&tl->requests);
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
void i915_gem_timeline_fini(struct i915_gem_timeline *tl)
|
||||
{
|
||||
lockdep_assert_held(&tl->i915->drm.struct_mutex);
|
||||
|
||||
list_del(&tl->link);
|
||||
kfree(tl->name);
|
||||
}
|
||||
70
drivers/gpu/drm/i915/i915_gem_timeline.h
Normal file
70
drivers/gpu/drm/i915/i915_gem_timeline.h
Normal file
@ -0,0 +1,70 @@
|
||||
/*
|
||||
* Copyright © 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.
|
||||
*
|
||||
*/
|
||||
|
||||
#ifndef I915_GEM_TIMELINE_H
|
||||
#define I915_GEM_TIMELINE_H
|
||||
|
||||
#include <linux/list.h>
|
||||
|
||||
#include "i915_gem_request.h"
|
||||
|
||||
struct i915_gem_timeline;
|
||||
|
||||
struct intel_timeline {
|
||||
u64 fence_context;
|
||||
u32 last_submitted_seqno;
|
||||
u32 last_pending_seqno;
|
||||
|
||||
/**
|
||||
* List of breadcrumbs associated with GPU requests currently
|
||||
* outstanding.
|
||||
*/
|
||||
struct list_head requests;
|
||||
|
||||
/* Contains an RCU guarded pointer to the last request. No reference is
|
||||
* held to the request, users must carefully acquire a reference to
|
||||
* the request using i915_gem_active_get_request_rcu(), or hold the
|
||||
* struct_mutex.
|
||||
*/
|
||||
struct i915_gem_active last_request;
|
||||
|
||||
struct i915_gem_timeline *common;
|
||||
};
|
||||
|
||||
struct i915_gem_timeline {
|
||||
struct list_head link;
|
||||
u32 next_seqno;
|
||||
|
||||
struct drm_i915_private *i915;
|
||||
const char *name;
|
||||
|
||||
struct intel_timeline engine[I915_NUM_ENGINES];
|
||||
};
|
||||
|
||||
int i915_gem_timeline_init(struct drm_i915_private *i915,
|
||||
struct i915_gem_timeline *tl,
|
||||
const char *name);
|
||||
void i915_gem_timeline_fini(struct i915_gem_timeline *tl);
|
||||
|
||||
#endif
|
||||
@ -1110,7 +1110,7 @@ static void error_record_engine_registers(struct drm_i915_error_state *error,
|
||||
ee->instpm = I915_READ(RING_INSTPM(engine->mmio_base));
|
||||
ee->acthd = intel_engine_get_active_head(engine);
|
||||
ee->seqno = intel_engine_get_seqno(engine);
|
||||
ee->last_seqno = engine->last_submitted_seqno;
|
||||
ee->last_seqno = engine->timeline->last_submitted_seqno;
|
||||
ee->start = I915_READ_START(engine);
|
||||
ee->head = I915_READ_HEAD(engine);
|
||||
ee->tail = I915_READ_TAIL(engine);
|
||||
@ -1195,7 +1195,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
|
||||
|
||||
count = 0;
|
||||
request = first;
|
||||
list_for_each_entry_from(request, &engine->request_list, link)
|
||||
list_for_each_entry_from(request, &engine->timeline->requests, link)
|
||||
count++;
|
||||
if (!count)
|
||||
return;
|
||||
@ -1208,7 +1208,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
|
||||
|
||||
count = 0;
|
||||
request = first;
|
||||
list_for_each_entry_from(request, &engine->request_list, link) {
|
||||
list_for_each_entry_from(request, &engine->timeline->requests, link) {
|
||||
if (count >= ee->num_requests) {
|
||||
/*
|
||||
* If the ring request list was changed in
|
||||
|
||||
@ -1522,7 +1522,8 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
|
||||
engine->submit_request = i915_guc_submit;
|
||||
|
||||
/* Replay the current set of previously submitted requests */
|
||||
list_for_each_entry(request, &engine->request_list, link) {
|
||||
list_for_each_entry(request,
|
||||
&engine->timeline->requests, link) {
|
||||
client->wq_rsvd += sizeof(struct guc_wq_item);
|
||||
if (i915_sw_fence_done(&request->submit))
|
||||
i915_guc_submit(request);
|
||||
|
||||
@ -3168,7 +3168,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
|
||||
|
||||
acthd = intel_engine_get_active_head(engine);
|
||||
seqno = intel_engine_get_seqno(engine);
|
||||
submit = READ_ONCE(engine->last_submitted_seqno);
|
||||
submit = READ_ONCE(engine->timeline->last_submitted_seqno);
|
||||
|
||||
if (engine->hangcheck.seqno == seqno) {
|
||||
if (i915_seqno_passed(seqno, submit)) {
|
||||
|
||||
@ -174,7 +174,7 @@ int intel_engines_init(struct drm_device *dev)
|
||||
return ret;
|
||||
}
|
||||
|
||||
void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
|
||||
void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno)
|
||||
{
|
||||
struct drm_i915_private *dev_priv = engine->i915;
|
||||
|
||||
@ -210,7 +210,9 @@ void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
|
||||
intel_write_status_page(engine, I915_GEM_HWS_INDEX, seqno);
|
||||
if (engine->irq_seqno_barrier)
|
||||
engine->irq_seqno_barrier(engine);
|
||||
engine->last_submitted_seqno = seqno;
|
||||
|
||||
GEM_BUG_ON(i915_gem_active_isset(&engine->timeline->last_request));
|
||||
engine->timeline->last_submitted_seqno = seqno;
|
||||
|
||||
engine->hangcheck.seqno = seqno;
|
||||
|
||||
@ -225,10 +227,9 @@ void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
|
||||
memset(&engine->hangcheck, 0, sizeof(engine->hangcheck));
|
||||
}
|
||||
|
||||
static void intel_engine_init_requests(struct intel_engine_cs *engine)
|
||||
static void intel_engine_init_timeline(struct intel_engine_cs *engine)
|
||||
{
|
||||
init_request_active(&engine->last_request, NULL);
|
||||
INIT_LIST_HEAD(&engine->request_list);
|
||||
engine->timeline = &engine->i915->gt.global_timeline.engine[engine->id];
|
||||
}
|
||||
|
||||
/**
|
||||
@ -245,9 +246,7 @@ void intel_engine_setup_common(struct intel_engine_cs *engine)
|
||||
INIT_LIST_HEAD(&engine->execlist_queue);
|
||||
spin_lock_init(&engine->execlist_lock);
|
||||
|
||||
engine->fence_context = dma_fence_context_alloc(1);
|
||||
|
||||
intel_engine_init_requests(engine);
|
||||
intel_engine_init_timeline(engine);
|
||||
intel_engine_init_hangcheck(engine);
|
||||
i915_gem_batch_pool_init(engine, &engine->batch_pool);
|
||||
|
||||
|
||||
@ -4,6 +4,7 @@
|
||||
#include <linux/hashtable.h>
|
||||
#include "i915_gem_batch_pool.h"
|
||||
#include "i915_gem_request.h"
|
||||
#include "i915_gem_timeline.h"
|
||||
|
||||
#define I915_CMD_HASH_ORDER 9
|
||||
|
||||
@ -169,7 +170,6 @@ struct intel_engine_cs {
|
||||
VCS2, /* Keep instances of the same type engine together. */
|
||||
VECS
|
||||
} id;
|
||||
#define I915_NUM_ENGINES 5
|
||||
#define _VCS(n) (VCS + (n))
|
||||
unsigned int exec_id;
|
||||
enum intel_engine_hw_id {
|
||||
@ -180,10 +180,10 @@ struct intel_engine_cs {
|
||||
VCS2_HW
|
||||
} hw_id;
|
||||
enum intel_engine_hw_id guc_id; /* XXX same as hw_id? */
|
||||
u64 fence_context;
|
||||
u32 mmio_base;
|
||||
unsigned int irq_shift;
|
||||
struct intel_ring *buffer;
|
||||
struct intel_timeline *timeline;
|
||||
|
||||
struct intel_render_state *render_state;
|
||||
|
||||
@ -346,27 +346,6 @@ struct intel_engine_cs {
|
||||
bool preempt_wa;
|
||||
u32 ctx_desc_template;
|
||||
|
||||
/**
|
||||
* List of breadcrumbs associated with GPU requests currently
|
||||
* outstanding.
|
||||
*/
|
||||
struct list_head request_list;
|
||||
|
||||
/**
|
||||
* Seqno of request most recently submitted to request_list.
|
||||
* Used exclusively by hang checker to avoid grabbing lock while
|
||||
* inspecting request list.
|
||||
*/
|
||||
u32 last_submitted_seqno;
|
||||
u32 last_pending_seqno;
|
||||
|
||||
/* An RCU guarded pointer to the last request. No reference is
|
||||
* held to the request, users must carefully acquire a reference to
|
||||
* the request using i915_gem_active_get_rcu(), or hold the
|
||||
* struct_mutex.
|
||||
*/
|
||||
struct i915_gem_active last_request;
|
||||
|
||||
struct i915_gem_context *last_context;
|
||||
|
||||
struct intel_engine_hangcheck hangcheck;
|
||||
@ -516,20 +495,13 @@ static inline u32 intel_ring_offset(struct intel_ring *ring, u32 value)
|
||||
int __intel_ring_space(int head, int tail, int size);
|
||||
void intel_ring_update_space(struct intel_ring *ring);
|
||||
|
||||
void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno);
|
||||
void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno);
|
||||
|
||||
void intel_engine_setup_common(struct intel_engine_cs *engine);
|
||||
int intel_engine_init_common(struct intel_engine_cs *engine);
|
||||
int intel_engine_create_scratch(struct intel_engine_cs *engine, int size);
|
||||
void intel_engine_cleanup_common(struct intel_engine_cs *engine);
|
||||
|
||||
static inline int intel_engine_idle(struct intel_engine_cs *engine,
|
||||
unsigned int flags)
|
||||
{
|
||||
/* Wait upon the last request to be completed */
|
||||
return i915_gem_active_wait(&engine->last_request, flags);
|
||||
}
|
||||
|
||||
int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
|
||||
int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine);
|
||||
int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine);
|
||||
@ -619,7 +591,7 @@ unsigned int intel_kick_signalers(struct drm_i915_private *i915);
|
||||
|
||||
static inline bool intel_engine_is_active(struct intel_engine_cs *engine)
|
||||
{
|
||||
return i915_gem_active_isset(&engine->last_request);
|
||||
return i915_gem_active_isset(&engine->timeline->last_request);
|
||||
}
|
||||
|
||||
#endif /* _INTEL_RINGBUFFER_H_ */
|
||||
|
||||
Loading…
Reference in New Issue
Block a user