Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Jonathan Cavitt <jonathan.cavitt@intel.com>
To: intel-gfx@lists.freedesktop.org
Cc: janusz.krzysztofik@intel.com, andi.shyti@intel.com,
	matthew.d.roper@intel.com, jonathan.cavitt@intel.com,
	saurabhg.gupta@intel.com, chris.p.wilson@linux.intel.com,
	nirmoy.das@intel.com
Subject: [Intel-gfx] [PATCH v17 5/7] drm/i915: No TLB invalidation on wedged GT
Date: Tue, 17 Oct 2023 11:08:04 -0700	[thread overview]
Message-ID: <20231017180806.3054290-6-jonathan.cavitt@intel.com> (raw)
In-Reply-To: <20231017180806.3054290-1-jonathan.cavitt@intel.com>

It is not an error for GuC TLB invalidations to fail when the GT is
wedged or disabled, so do not process a wait failure as one in
guc_send_invalidate_tlb.

Signed-off-by: Fei Yang <fei.yang@intel.com>
Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
CC: John Harrison <john.c.harrison@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Acked-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Acked-by: Nirmoy Das <nirmoy.das@intel.com>
---
 .../gpu/drm/i915/gt/uc/intel_guc_submission.c | 32 ++++++++++++++++++-
 1 file changed, 31 insertions(+), 1 deletion(-)

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 1b04b1692e48d..d37698bd6b91a 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
@@ -32,6 +32,7 @@
 
 #include "i915_drv.h"
 #include "i915_reg.h"
+#include "i915_irq.h"
 #include "i915_trace.h"
 
 /**
@@ -1935,6 +1936,12 @@ void intel_guc_submission_cancel_requests(struct intel_guc *guc)
 
 	/* GuC is blown away, drop all references to contexts */
 	xa_destroy(&guc->context_lookup);
+
+	/*
+	 * Wedged GT won't respond to any TLB invalidation request. Simply
+	 * release all the blocked waiters.
+	 */
+	wake_up_all_tlb_invalidate(guc);
 }
 
 void intel_guc_submission_reset_finish(struct intel_guc *guc)
@@ -4749,10 +4756,19 @@ static long must_wait_woken(struct wait_queue_entry *wq_entry, long timeout)
 	return timeout;
 }
 
+static bool intel_gt_is_enabled(const struct intel_gt *gt)
+{
+	/* Check if GT is wedged or suspended */
+	if (intel_gt_is_wedged(gt) || !intel_irqs_enabled(gt->i915))
+		return false;
+	return true;
+}
+
 static int guc_send_invalidate_tlb(struct intel_guc *guc,
 				   enum intel_guc_tlb_invalidation_type type)
 {
 	struct intel_guc_tlb_wait _wq, *wq = &_wq;
+	struct intel_gt *gt = guc_to_gt(guc);
 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
 	int err;
 	u32 seqno;
@@ -4766,6 +4782,13 @@ static int guc_send_invalidate_tlb(struct intel_guc *guc,
 	};
 	u32 size = ARRAY_SIZE(action);
 
+	/*
+	 * Early guard against GT enablement.  TLB invalidation should not be
+	 * attempted if the GT is disabled due to suspend/wedge.
+	 */
+	if (!intel_gt_is_enabled(gt))
+		return -EINVAL;
+
 	init_waitqueue_head(&_wq.wq);
 
 	if (xa_alloc_cyclic_irq(&guc->tlb_lookup, &seqno, wq,
@@ -4798,7 +4821,14 @@ static int guc_send_invalidate_tlb(struct intel_guc *guc,
 	if (err)
 		goto out;
 
-	if (!must_wait_woken(&wait, intel_guc_ct_max_queue_time_jiffies())) {
+	/*
+	 * Late guard against GT enablement.  It is not an error for the TLB
+	 * invalidation to time out if the GT is disabled during the process
+	 * due to suspend/wedge.  In fact, the TLB invalidation is cancelled
+	 * in this case.
+	 */
+	if (!must_wait_woken(&wait, intel_guc_ct_max_queue_time_jiffies()) &&
+	    intel_gt_is_enabled(gt)) {
 		guc_err(guc,
 			"TLB invalidation response timed out for seqno %u\n", seqno);
 		err = -ETIME;
-- 
2.25.1


  parent reply	other threads:[~2023-10-17 18:19 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-17 18:07 [Intel-gfx] [PATCH v17 0/7] drm/i915: Define and use GuC and CTB TLB invalidation routines Jonathan Cavitt
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 1/7] drm/i915: Add GuC TLB Invalidation device info flags Jonathan Cavitt
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 2/7] drm/i915/guc: Add CT size delay helper Jonathan Cavitt
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 3/7] drm/i915: Define and use GuC and CTB TLB invalidation routines Jonathan Cavitt
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 4/7] drm/i915: No TLB invalidation on suspended GT Jonathan Cavitt
2023-10-17 18:08 ` Jonathan Cavitt [this message]
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 6/7] drm/i915/gt: Increase sleep in gt_tlb selftest sanitycheck Jonathan Cavitt
2023-10-17 18:08 ` [Intel-gfx] [PATCH v17 7/7] drm/i915: Enable GuC TLB invalidations for MTL Jonathan Cavitt
2023-10-17 19:07 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915: Define and use GuC and CTB TLB invalidation routines Patchwork
2023-10-17 19:07 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2023-10-17 19:25 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork
2023-10-17 20:02 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915: Define and use GuC and CTB TLB invalidation routines (rev2) Patchwork
2023-10-17 20:02 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2023-10-17 20:20 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork
2023-10-17 21:02 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915: Define and use GuC and CTB TLB invalidation routines (rev3) Patchwork
2023-10-17 21:02 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2023-10-17 21:07 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2023-10-18  2:40 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork
2023-10-18  3:28   ` Andi Shyti
2023-10-18  3:51     ` Andi Shyti
2023-10-18  4:18 ` [Intel-gfx] [PATCH v17 0/7] drm/i915: Define and use GuC and CTB TLB invalidation routines Andi Shyti

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20231017180806.3054290-6-jonathan.cavitt@intel.com \
    --to=jonathan.cavitt@intel.com \
    --cc=andi.shyti@intel.com \
    --cc=chris.p.wilson@linux.intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=janusz.krzysztofik@intel.com \
    --cc=matthew.d.roper@intel.com \
    --cc=nirmoy.das@intel.com \
    --cc=saurabhg.gupta@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox