Re: [Intel-gfx] [PATCH 24/33] drm/i915/guc: Implement banned contexts for GuC submission

On Wed, Aug 25, 2021 at 11:39:10AM +0100, Tvrtko Ursulin wrote:

On 27/07/2021 01:23, Matthew Brost wrote:

...

When using GuC submission, if a context gets banned disable scheduling and mark all inflight requests as complete.

Cc: John Harrison John.C.Harrison@Intel.com Signed-off-by: Matthew Brost matthew.brost@intel.com Reviewed-by: John Harrison John.C.Harrison@Intel.com

---

drivers/gpu/drm/i915/gem/i915_gem_context.c | 2 +- drivers/gpu/drm/i915/gt/intel_context.h | 13 ++ drivers/gpu/drm/i915/gt/intel_context_types.h | 2 + drivers/gpu/drm/i915/gt/intel_reset.c | 32 +--- .../gpu/drm/i915/gt/intel_ring_submission.c | 20 +++ drivers/gpu/drm/i915/gt/uc/intel_guc.h | 2 + .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 151 ++++++++++++++++-- drivers/gpu/drm/i915/i915_trace.h | 10 ++ 8 files changed, 195 insertions(+), 37 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index e3df01a201d7..05c3ee191710 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -1084,7 +1084,7 @@ static void kill_engines(struct i915_gem_engines *engines, bool ban) for_each_gem_engine(ce, engines, it) { struct intel_engine_cs *engine;

• if (ban && intel_context_set_banned(ce))
  

• if (ban && intel_context_ban(ce, NULL))
  continue;
  

  /*

diff --git a/drivers/gpu/drm/i915/gt/intel_context.h b/drivers/gpu/drm/i915/gt/intel_context.h index 2ed9bf5f91a5..814d9277096a 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.h +++ b/drivers/gpu/drm/i915/gt/intel_context.h @@ -16,6 +16,7 @@ #include "intel_engine_types.h" #include "intel_ring_types.h" #include "intel_timeline_types.h" +#include "i915_trace.h" #define CE_TRACE(ce, fmt, ...) do { \ const struct intel_context *ce__ = (ce); \ @@ -243,6 +244,18 @@ static inline bool intel_context_set_banned(struct intel_context *ce) return test_and_set_bit(CONTEXT_BANNED, &ce->flags); } +static inline bool intel_context_ban(struct intel_context *ce,

• 		     struct i915_request *rq)
  

+{

• bool ret = intel_context_set_banned(ce);
• trace_intel_context_ban(ce);
• if (ce->ops->ban)

• ce->ops->ban(ce, rq);
  

• return ret;

+}

• static inline bool intel_context_force_single_submission(const struct intel_context *ce) {

diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h index 035108c10b2c..57c19ee3e313 100644 --- a/drivers/gpu/drm/i915/gt/intel_context_types.h +++ b/drivers/gpu/drm/i915/gt/intel_context_types.h @@ -35,6 +35,8 @@ struct intel_context_ops { int (*alloc)(struct intel_context *ce);

• void (*ban)(struct intel_context *ce, struct i915_request *rq);
• int (*pre_pin)(struct intel_context *ce, struct i915_gem_ww_ctx *ww, void **vaddr); int (*pin)(struct intel_context *ce, void *vaddr); void (*unpin)(struct intel_context *ce);

diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c index 4d281bc8a38c..91200c43951f 100644 --- a/drivers/gpu/drm/i915/gt/intel_reset.c +++ b/drivers/gpu/drm/i915/gt/intel_reset.c @@ -22,7 +22,6 @@ #include "intel_reset.h" #include "uc/intel_guc.h" -#include "uc/intel_guc_submission.h" #define RESET_MAX_RETRIES 3 @@ -39,21 +38,6 @@ static void rmw_clear_fw(struct intel_uncore *uncore, i915_reg_t reg, u32 clr) intel_uncore_rmw_fw(uncore, reg, clr, 0); } -static void skip_context(struct i915_request *rq) -{

• struct intel_context *hung_ctx = rq->context;
• list_for_each_entry_from_rcu(rq, &hung_ctx->timeline->requests, link) {

• if (!i915_request_is_active(rq))
  

• 	return;
  

• if (rq->context == hung_ctx) {
  

• 	i915_request_set_error_once(rq, -EIO);
  

• 	__i915_request_skip(rq);
  

• }
  

• }

-}

• static void client_mark_guilty(struct i915_gem_context *ctx, bool banned) { struct drm_i915_file_private *file_priv = ctx->file_priv;

@@ -88,10 +72,8 @@ static bool mark_guilty(struct i915_request *rq) bool banned; int i;

• if (intel_context_is_closed(rq->context)) {

• intel_context_set_banned(rq->context);
  

• if (intel_context_is_closed(rq->context)) return true;

• } rcu_read_lock(); ctx = rcu_dereference(rq->context->gem_context);

@@ -123,11 +105,9 @@ static bool mark_guilty(struct i915_request *rq) banned = !i915_gem_context_is_recoverable(ctx); if (time_before(jiffies, prev_hang + CONTEXT_FAST_HANG_JIFFIES)) banned = true;

• if (banned) {

• if (banned) drm_dbg(&ctx->i915->drm, "context %s: guilty %d, banned\n", ctx->name, atomic_read(&ctx->guilty_count));

• intel_context_set_banned(rq->context);
  

• } client_mark_guilty(ctx, banned);

@@ -149,6 +129,8 @@ static void mark_innocent(struct i915_request *rq) void __i915_request_reset(struct i915_request *rq, bool guilty) {

• bool banned = false;
• RQ_TRACE(rq, "guilty? %s\n", yesno(guilty)); GEM_BUG_ON(__i915_request_is_complete(rq));

@@ -156,13 +138,15 @@ void __i915_request_reset(struct i915_request *rq, bool guilty) if (guilty) { i915_request_set_error_once(rq, -EIO); __i915_request_skip(rq);

• if (mark_guilty(rq) && !intel_engine_uses_guc(rq->engine))
  

• 	skip_context(rq);
  

• banned = mark_guilty(rq);
  

  } else { i915_request_set_error_once(rq, -EAGAIN); mark_innocent(rq); } rcu_read_unlock();
• if (banned)

• intel_context_ban(rq->context, rq);
  

  } static bool i915_in_reset(struct pci_dev *pdev)

diff --git a/drivers/gpu/drm/i915/gt/intel_ring_submission.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c index a5404c7b600f..05bb9f449df1 100644 --- a/drivers/gpu/drm/i915/gt/intel_ring_submission.c +++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c @@ -586,9 +586,29 @@ static void ring_context_reset(struct intel_context *ce) clear_bit(CONTEXT_VALID_BIT, &ce->flags); } +static void ring_context_ban(struct intel_context *ce,

• 	     struct i915_request *rq)
  

+{

• struct intel_engine_cs *engine;
• if (!rq || !i915_request_is_active(rq))

• return;
  

When this gets called from context close, via intel_context_ban, rq will be always NULL - so the below loop which skips a line of executing requests never gets to run.

This appears a functional change versus previous state of the codebase, where skip_context() would run in any case.

Hmm, not so sure about this being a functional change. When called from __i915_request_reset, intel_context_ban (previously skip_context there) always has a non-NULL request argument. When called from kill_engines the request is NULL (short circuiting ring function), but we never hooked into the ring backend before.

I had this observation from a patch I am working on (https://intel-gfx-ci.01.org/tree/drm-tip/Trybot_7950/shard-snb6/igt@gem_ctx_...

• so a long line of executing requests which did not get zapped post reset),

but may be wrong. Maybe I am missing something since I don't yet understand why would I be first to hit this issue. So take it with a grain of salt for now.

CI was green on my series but CI doesn't always catch everything...

Do you have link to your series that I can look at?

Ah.. maybe the key is that in my patch I made intel_context_ban not call ce->ops->"ban" (I renamed it to revoke) unconditionally. Hence there may be a path there intel_context_ban is first called wo/ a rq, then from within __i915_request_reset it gets called with rq, which now fails to call the vfunc. Hm that's clunky and fragile which ever way I look at it. I'll trybot one more experiment..

I'm open to suggestions on how to change this, to make it makes sense to call into the function unconditionally with a correct arguments and the backend handles the rest (i.e. the way it is currently).

Matt

Regards,

Tvrtko

...

• engine = rq->engine;
• lockdep_assert_held(&engine->sched_engine->lock);
• list_for_each_entry_continue(rq, &engine->sched_engine->requests,

• 		     sched.link)
  

• if (rq->context == ce) {
  

• 	i915_request_set_error_once(rq, -EIO);
  

• 	__i915_request_skip(rq);
  

• }
  

+}

• static const struct intel_context_ops ring_context_ops = { .alloc = ring_context_alloc,
• .ban = ring_context_ban,
• .pre_pin = ring_context_pre_pin, .pin = ring_context_pin, .unpin = ring_context_unpin,

diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc.h index 1875303c3bca..8ab70a2223b0 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.h @@ -281,6 +281,8 @@ void intel_guc_find_hung_context(struct intel_engine_cs *engine); int intel_guc_global_policies_update(struct intel_guc *guc); +void intel_guc_context_ban(struct intel_context *ce, struct i915_request *rq);

• void intel_guc_submission_reset_prepare(struct intel_guc *guc); void intel_guc_submission_reset(struct intel_guc *guc, bool stalled); void intel_guc_submission_reset_finish(struct intel_guc *guc);

diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index cea3e3073a71..ad9a38a861df 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -125,6 +125,7 @@ static inline void clr_context_pending_enable(struct intel_context *ce) #define SCHED_STATE_WAIT_FOR_DEREGISTER_TO_REGISTER BIT(0) #define SCHED_STATE_DESTROYED BIT(1) #define SCHED_STATE_PENDING_DISABLE BIT(2) +#define SCHED_STATE_BANNED BIT(3) static inline void init_sched_state(struct intel_context *ce) { /* Only should be called from guc_lrc_desc_pin() */ @@ -185,6 +186,23 @@ static inline void clr_context_pending_disable(struct intel_context *ce) ce->guc_state.sched_state &= ~SCHED_STATE_PENDING_DISABLE; } +static inline bool context_banned(struct intel_context *ce) +{

• return ce->guc_state.sched_state & SCHED_STATE_BANNED;

+}

+static inline void set_context_banned(struct intel_context *ce) +{

• lockdep_assert_held(&ce->guc_state.lock);
• ce->guc_state.sched_state |= SCHED_STATE_BANNED;

+}

+static inline void clr_context_banned(struct intel_context *ce) +{

• lockdep_assert_held(&ce->guc_state.lock);
• ce->guc_state.sched_state &= ~SCHED_STATE_BANNED;

+}

• static inline bool context_guc_id_invalid(struct intel_context *ce) { return ce->guc_id == GUC_INVALID_LRC_ID;

@@ -357,13 +375,23 @@ static int guc_lrc_desc_pin(struct intel_context *ce, bool loop); static int guc_add_request(struct intel_guc *guc, struct i915_request *rq) {

• int err;

• int err = 0; struct intel_context *ce = rq->context; u32 action[3]; int len = 0; u32 g2h_len_dw = 0; bool enabled;
• /*

• * Corner case where requests were sitting in the priority list or a
  

• * request resubmitted after the context was banned.
  

• */
  

• if (unlikely(intel_context_is_banned(ce))) {

• i915_request_put(i915_request_mark_eio(rq));
  

• intel_engine_signal_breadcrumbs(ce->engine);
  

• goto out;
  

• }
• GEM_BUG_ON(!atomic_read(&ce->guc_id_ref)); GEM_BUG_ON(context_guc_id_invalid(ce));

@@ -399,6 +427,8 @@ static int guc_add_request(struct intel_guc *guc, struct i915_request *rq) clr_context_pending_enable(ce); intel_context_put(ce); }

• if (likely(!err))

• trace_i915_request_guc_submit(rq);
  

  out: return err;

@@ -463,7 +493,6 @@ static int guc_dequeue_one_context(struct intel_guc *guc) guc->stalled_request = last; return false; }

• trace_i915_request_guc_submit(last);
  

  } guc->stalled_request = NULL;

@@ -502,12 +531,13 @@ static void cs_irq_handler(struct intel_engine_cs *engine, u16 iir) static void __guc_context_destroy(struct intel_context *ce); static void release_guc_id(struct intel_guc *guc, struct intel_context *ce); static void guc_signal_context_fence(struct intel_context *ce); +static void guc_cancel_context_requests(struct intel_context *ce); static void scrub_guc_desc_for_outstanding_g2h(struct intel_guc *guc) { struct intel_context *ce; unsigned long index, flags;

• bool pending_disable, pending_enable, deregister, destroyed;

• bool pending_disable, pending_enable, deregister, destroyed, banned; xa_for_each(&guc->context_lookup, index, ce) { /* Flush context */

@@ -525,6 +555,7 @@ static void scrub_guc_desc_for_outstanding_g2h(struct intel_guc *guc) pending_enable = context_pending_enable(ce); pending_disable = context_pending_disable(ce); deregister = context_wait_for_deregister_to_register(ce);

• banned = context_banned(ce);
  

  init_sched_state(ce); if (pending_enable || destroyed || deregister) {

@@ -542,6 +573,10 @@ static void scrub_guc_desc_for_outstanding_g2h(struct intel_guc *guc) /* Not mutualy exclusive with above if statement. */ if (pending_disable) { guc_signal_context_fence(ce);

• 	if (banned) {
  

• 		guc_cancel_context_requests(ce);
  

• 		intel_engine_signal_breadcrumbs(ce->engine);
  

• 	}
  intel_context_sched_disable_unpin(ce);
  atomic_dec(&guc->outstanding_submission_g2h);
  intel_context_put(ce);
  

@@ -661,6 +696,9 @@ static void guc_reset_state(struct intel_context *ce, u32 head, bool scrub) { struct intel_engine_cs *engine = __context_to_physical_engine(ce);

• if (intel_context_is_banned(ce))

• return;
  

• GEM_BUG_ON(!intel_context_is_pinned(ce)); /*

@@ -731,6 +769,8 @@ static void __guc_reset_context(struct intel_context *ce, bool stalled) struct i915_request *rq; u32 head;

• intel_context_get(ce);
• /*
  • GuC will implicitly mark the context as non-schedulable
  • when it sends the reset notification. Make sure our state

@@ -756,6 +796,7 @@ static void __guc_reset_context(struct intel_context *ce, bool stalled) out_replay: guc_reset_state(ce, head, stalled); __unwind_incomplete_requests(ce);

• intel_context_put(ce); } void intel_guc_submission_reset(struct intel_guc *guc, bool stalled)

@@ -940,8 +981,6 @@ static int guc_bypass_tasklet_submit(struct intel_guc *guc, ret = guc_add_request(guc, rq); if (ret == -EBUSY) guc->stalled_request = rq;

• else

• trace_i915_request_guc_submit(rq);
  

  if (unlikely(ret == -EPIPE)) disable_submission(guc);

@@ -1344,13 +1383,77 @@ static u16 prep_context_pending_disable(struct intel_context *ce) return ce->guc_id; } +static void __guc_context_set_preemption_timeout(struct intel_guc *guc,

• 				 u16 guc_id,
  

• 				 u32 preemption_timeout)
  

+{

• u32 action[] = {

• INTEL_GUC_ACTION_SET_CONTEXT_PREEMPTION_TIMEOUT,
  

• guc_id,
  

• preemption_timeout
  

• };
• intel_guc_send_busy_loop(guc, action, ARRAY_SIZE(action), 0, true);

+}

+static void guc_context_ban(struct intel_context *ce, struct i915_request *rq) +{

• struct intel_guc *guc = ce_to_guc(ce);
• struct intel_runtime_pm *runtime_pm =

• &ce->engine->gt->i915->runtime_pm;
  

• intel_wakeref_t wakeref;
• unsigned long flags;
• guc_flush_submissions(guc);
• spin_lock_irqsave(&ce->guc_state.lock, flags);
• set_context_banned(ce);
• if (submission_disabled(guc) ||

•    (!context_enabled(ce) && !context_pending_disable(ce))) {
  

• spin_unlock_irqrestore(&ce->guc_state.lock, flags);
  

• guc_cancel_context_requests(ce);
  

• intel_engine_signal_breadcrumbs(ce->engine);
  

• } else if (!context_pending_disable(ce)) {

• u16 guc_id;
  

• /*
  

•  * We add +2 here as the schedule disable complete CTB handler
  

•  * calls intel_context_sched_disable_unpin (-2 to pin_count).
  

•  */
  

• atomic_add(2, &ce->pin_count);
  

• guc_id = prep_context_pending_disable(ce);
  

• spin_unlock_irqrestore(&ce->guc_state.lock, flags);
  

• /*
  

•  * In addition to disabling scheduling, set the preemption
  

•  * timeout to the minimum value (1 us) so the banned context
  

•  * gets kicked off the HW ASAP.
  

•  */
  

• with_intel_runtime_pm(runtime_pm, wakeref) {
  

• 	__guc_context_set_preemption_timeout(guc, guc_id, 1);
  

• 	__guc_context_sched_disable(guc, ce, guc_id);
  

• }
  

• } else {

• if (!context_guc_id_invalid(ce))
  

• 	with_intel_runtime_pm(runtime_pm, wakeref)
  

• 		__guc_context_set_preemption_timeout(guc,
  

• 						     ce->guc_id,
  

• 						     1);
  

• spin_unlock_irqrestore(&ce->guc_state.lock, flags);
  

• }

+}

• static void guc_context_sched_disable(struct intel_context *ce) { struct intel_guc *guc = ce_to_guc(ce);

• struct intel_runtime_pm *runtime_pm = &ce->engine->gt->i915->runtime_pm; unsigned long flags;
• u16 guc_id;

• struct intel_runtime_pm *runtime_pm = &ce->engine->gt->i915->runtime_pm; intel_wakeref_t wakeref;
• u16 guc_id;
• bool enabled; if (submission_disabled(guc) || context_guc_id_invalid(ce) || !lrc_desc_registered(guc, ce->guc_id)) {

@@ -1364,14 +1467,22 @@ static void guc_context_sched_disable(struct intel_context *ce) spin_lock_irqsave(&ce->guc_state.lock, flags); /*

• * We have to check if the context has been pinned again as another pin
  

• * operation is allowed to pass this function. Checking the pin count,
  

• * within ce->guc_state.lock, synchronizes this function with
  

• * We have to check if the context has been disabled by another thread.
  

• * We also have to check if the context has been pinned again as another
  

• * pin operation is allowed to pass this function. Checking the pin
  

• * count, within ce->guc_state.lock, synchronizes this function with
  

  • guc_request_alloc ensuring a request doesn't slip through the
  • 'context_pending_disable' fence. Checking within the spin lock (can't
  • sleep) ensures another process doesn't pin this context and generate
  • a request before we set the 'context_pending_disable' flag here.
  */
• enabled = context_enabled(ce);
• if (unlikely(!enabled || submission_disabled(guc))) {

• if (enabled)
  

• 	clr_context_enabled(ce);
  

• spin_unlock_irqrestore(&ce->guc_state.lock, flags);
  

• goto unpin;
  

• } if (unlikely(atomic_add_unless(&ce->pin_count, -2, 2))) { spin_unlock_irqrestore(&ce->guc_state.lock, flags); return;

@@ -1529,6 +1640,8 @@ static const struct intel_context_ops guc_context_ops = { .unpin = guc_context_unpin, .post_unpin = guc_context_post_unpin,

• .ban = guc_context_ban,
• .enter = intel_context_enter_engine, .exit = intel_context_exit_engine,

@@ -1722,6 +1835,8 @@ static const struct intel_context_ops virtual_guc_context_ops = { .unpin = guc_context_unpin, .post_unpin = guc_context_post_unpin,

• .ban = guc_context_ban,
• .enter = guc_virtual_context_enter, .exit = guc_virtual_context_exit,

@@ -2164,6 +2279,8 @@ int intel_guc_sched_done_process_msg(struct intel_guc *guc, if (context_pending_enable(ce)) { clr_context_pending_enable(ce); } else if (context_pending_disable(ce)) {

• bool banned;
  

• /*
  • Unpin must be done before __guc_signal_context_fence,
  • otherwise a race exists between the requests getting

@@ -2174,9 +2291,16 @@ int intel_guc_sched_done_process_msg(struct intel_guc *guc, intel_context_sched_disable_unpin(ce); spin_lock_irqsave(&ce->guc_state.lock, flags);

• banned = context_banned(ce);
  

• clr_context_banned(ce);
  

  clr_context_pending_disable(ce); __guc_signal_context_fence(ce); spin_unlock_irqrestore(&ce->guc_state.lock, flags);

• if (banned) {
  

• 	guc_cancel_context_requests(ce);
  

• 	intel_engine_signal_breadcrumbs(ce->engine);
  

• }
  

  } decr_outstanding_submission_g2h(guc);

@@ -2211,8 +2335,11 @@ static void guc_handle_context_reset(struct intel_guc *guc, struct intel_context *ce) { trace_intel_context_reset(ce);

• capture_error_state(guc, ce);
• guc_context_replay(ce);

• if (likely(!intel_context_is_banned(ce))) {

• capture_error_state(guc, ce);
  

• guc_context_replay(ce);
  

• } } int intel_guc_context_reset_process_msg(struct intel_guc *guc,

diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h index 3f43d904f043..9613a7c19661 100644 --- a/drivers/gpu/drm/i915/i915_trace.h +++ b/drivers/gpu/drm/i915/i915_trace.h @@ -925,6 +925,11 @@ DEFINE_EVENT(intel_context, intel_context_reset, TP_ARGS(ce) ); +DEFINE_EVENT(intel_context, intel_context_ban,

•     TP_PROTO(struct intel_context *ce),
  

•     TP_ARGS(ce)
  

+);

• DEFINE_EVENT(intel_context, intel_context_register, TP_PROTO(struct intel_context *ce), TP_ARGS(ce)

@@ -1017,6 +1022,11 @@ trace_intel_context_reset(struct intel_context *ce) { } +static inline void +trace_intel_context_ban(struct intel_context *ce) +{ +}

• static inline void trace_intel_context_register(struct intel_context *ce) {