Re: [Intel-gfx] [PATCH v11 5/7] drm/i915: No TLB invalidation on wedged GT

[Nirmoy Das]

On 10/11/2023 2:02 AM, Jonathan Cavitt wrote:

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.y...@intel.com>
Signed-off-by: Jonathan Cavitt <jonathan.cav...@intel.com>
CC: John Harrison <john.c.harri...@intel.com>
Reviewed-by: Andi Shyti <andi.sh...@linux.intel.com>

Acked-by: Nirmoy Das <nirmoy....@intel.com>

---
  .../gpu/drm/i915/gt/uc/intel_guc_submission.c  | 18 +++++++++++++++++-
  1 file changed, 17 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 9d5f8cccaa592..1914cba5f48dd 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"

/**

@@ -1941,6 +1942,12 @@ 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)

@@ -4740,6 +4747,14 @@ 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)
  {
@@ -4789,7 +4804,8 @@ 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())) {

+       if (intel_gt_is_enabled(guc_to_gt(guc)) &&
+           !must_wait_woken(&wait, intel_guc_ct_max_queue_time_jiffies())) {
                guc_err(guc,
                        "TLB invalidation response timed out for seqno %u\n", 
seqno);
                err = -ETIME;