summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/gt/intel_engine_cs.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915/gt/intel_engine_cs.c')
-rw-r--r--drivers/gpu/drm/i915/gt/intel_engine_cs.c73
1 files changed, 63 insertions, 10 deletions
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
index 82630db0394b..80fd072ac719 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
@@ -680,6 +680,8 @@ static int measure_breadcrumb_dw(struct intel_engine_cs *engine)
engine->status_page.vma))
goto out_frame;
+ mutex_lock(&frame->timeline.mutex);
+
frame->ring.vaddr = frame->cs;
frame->ring.size = sizeof(frame->cs);
frame->ring.effective_size = frame->ring.size;
@@ -688,18 +690,22 @@ static int measure_breadcrumb_dw(struct intel_engine_cs *engine)
frame->rq.i915 = engine->i915;
frame->rq.engine = engine;
frame->rq.ring = &frame->ring;
- frame->rq.timeline = &frame->timeline;
+ rcu_assign_pointer(frame->rq.timeline, &frame->timeline);
dw = intel_timeline_pin(&frame->timeline);
if (dw < 0)
goto out_timeline;
+ spin_lock_irq(&engine->active.lock);
dw = engine->emit_fini_breadcrumb(&frame->rq, frame->cs) - frame->cs;
+ spin_unlock_irq(&engine->active.lock);
+
GEM_BUG_ON(dw & 1); /* RING_TAIL must be qword aligned */
intel_timeline_unpin(&frame->timeline);
out_timeline:
+ mutex_unlock(&frame->timeline.mutex);
intel_timeline_fini(&frame->timeline);
out_frame:
kfree(frame);
@@ -814,8 +820,10 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
if (engine->default_state)
i915_gem_object_put(engine->default_state);
- intel_context_unpin(engine->kernel_context);
- intel_context_put(engine->kernel_context);
+ if (engine->kernel_context) {
+ intel_context_unpin(engine->kernel_context);
+ intel_context_put(engine->kernel_context);
+ }
GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
intel_wa_list_free(&engine->ctx_wa_list);
@@ -948,6 +956,7 @@ void intel_engine_get_instdone(struct intel_engine_cs *engine,
struct intel_instdone *instdone)
{
struct drm_i915_private *i915 = engine->i915;
+ const struct sseu_dev_info *sseu = &RUNTIME_INFO(i915)->sseu;
struct intel_uncore *uncore = engine->uncore;
u32 mmio_base = engine->mmio_base;
int slice;
@@ -965,7 +974,7 @@ void intel_engine_get_instdone(struct intel_engine_cs *engine,
instdone->slice_common =
intel_uncore_read(uncore, GEN7_SC_INSTDONE);
- for_each_instdone_slice_subslice(i915, slice, subslice) {
+ for_each_instdone_slice_subslice(i915, sseu, slice, subslice) {
instdone->sampler[slice][subslice] =
read_subslice_reg(engine, slice, subslice,
GEN7_SAMPLER_INSTDONE);
@@ -1118,6 +1127,8 @@ bool intel_engine_can_store_dword(struct intel_engine_cs *engine)
case 3:
/* maybe only uses physical not virtual addresses */
return !(IS_I915G(engine->i915) || IS_I915GM(engine->i915));
+ case 4:
+ return !IS_I965G(engine->i915); /* who knows! */
case 6:
return engine->class != VIDEO_DECODE_CLASS; /* b0rked */
default:
@@ -1193,6 +1204,27 @@ static void hexdump(struct drm_printer *m, const void *buf, size_t len)
}
}
+static struct intel_timeline *get_timeline(struct i915_request *rq)
+{
+ struct intel_timeline *tl;
+
+ /*
+ * Even though we are holding the engine->active.lock here, there
+ * is no control over the submission queue per-se and we are
+ * inspecting the active state at a random point in time, with an
+ * unknown queue. Play safe and make sure the timeline remains valid.
+ * (Only being used for pretty printing, one extra kref shouldn't
+ * cause a camel stampede!)
+ */
+ rcu_read_lock();
+ tl = rcu_dereference(rq->timeline);
+ if (!kref_get_unless_zero(&tl->kref))
+ tl = NULL;
+ rcu_read_unlock();
+
+ return tl;
+}
+
static void intel_engine_print_registers(struct intel_engine_cs *engine,
struct drm_printer *m)
{
@@ -1287,27 +1319,37 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine,
int len;
len = snprintf(hdr, sizeof(hdr),
- "\t\tActive[%d: ",
+ "\t\tActive[%d]: ",
(int)(port - execlists->active));
- if (!i915_request_signaled(rq))
+ if (!i915_request_signaled(rq)) {
+ struct intel_timeline *tl = get_timeline(rq);
+
len += snprintf(hdr + len, sizeof(hdr) - len,
"ring:{start:%08x, hwsp:%08x, seqno:%08x}, ",
i915_ggtt_offset(rq->ring->vma),
- rq->timeline->hwsp_offset,
+ tl ? tl->hwsp_offset : 0,
hwsp_seqno(rq));
+
+ if (tl)
+ intel_timeline_put(tl);
+ }
snprintf(hdr + len, sizeof(hdr) - len, "rq: ");
print_request(m, rq, hdr);
}
for (port = execlists->pending; (rq = *port); port++) {
+ struct intel_timeline *tl = get_timeline(rq);
char hdr[80];
snprintf(hdr, sizeof(hdr),
"\t\tPending[%d] ring:{start:%08x, hwsp:%08x, seqno:%08x}, rq: ",
(int)(port - execlists->pending),
i915_ggtt_offset(rq->ring->vma),
- rq->timeline->hwsp_offset,
+ tl ? tl->hwsp_offset : 0,
hwsp_seqno(rq));
print_request(m, rq, hdr);
+
+ if (tl)
+ intel_timeline_put(tl);
}
spin_unlock_irqrestore(&engine->active.lock, flags);
} else if (INTEL_GEN(dev_priv) > 6) {
@@ -1385,6 +1427,8 @@ void intel_engine_dump(struct intel_engine_cs *engine,
spin_lock_irqsave(&engine->active.lock, flags);
rq = intel_engine_find_active_request(engine);
if (rq) {
+ struct intel_timeline *tl = get_timeline(rq);
+
print_request(m, rq, "\t\tactive ");
drm_printf(m, "\t\tring->start: 0x%08x\n",
@@ -1397,10 +1441,19 @@ void intel_engine_dump(struct intel_engine_cs *engine,
rq->ring->emit);
drm_printf(m, "\t\tring->space: 0x%08x\n",
rq->ring->space);
- drm_printf(m, "\t\tring->hwsp: 0x%08x\n",
- rq->timeline->hwsp_offset);
+
+ if (tl) {
+ drm_printf(m, "\t\tring->hwsp: 0x%08x\n",
+ tl->hwsp_offset);
+ intel_timeline_put(tl);
+ }
print_request_ring(m, rq);
+
+ if (rq->hw_context->lrc_reg_state) {
+ drm_printf(m, "Logical Ring Context:\n");
+ hexdump(m, rq->hw_context->lrc_reg_state, PAGE_SIZE);
+ }
}
spin_unlock_irqrestore(&engine->active.lock, flags);
generated by cgit v1.2.3 (git 2.39.0) at 2025-09-09 09:10:12 +0300