[Intel-gfx] [PATCH v3 9/9] drm/i915/error: Capture WA ctx batch in error state

[Arun Siluvery]

>From Gen8 onwards we apply ctx workarounds using special batch buffers that
execute during save/restore, good to have them in error state.

v2: use wa_ctx->size and print only size values (Mika)

v3: simplify conditions when recording and freeing object (Chris)

Cc: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala at intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala at intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery at linux.intel.com>
---

Applied r-b received for v2, this version just simplifies conditions so I
am assuming it still holds.

 drivers/gpu/drm/i915/i915_drv.h       |  2 +-
 drivers/gpu/drm/i915/i915_gpu_error.c | 27 +++++++++++++++++++++++++++
 2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index e8a47af..b1ba039 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -568,7 +568,7 @@ struct drm_i915_error_state {
 			bool is_ppgtt;
 			int page_count;
 			u32 *pages[0];
-		} *req_ringbuffer, *hw_ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page;
+		} *req_ringbuffer, *hw_ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page, *wa_ctx;
 
 		struct drm_i915_error_request {
 			u64 ctx_desc;
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index f426538..f48788e 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -659,6 +659,27 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			}
 		}
 
+		if ((obj = error->ring[i].wa_ctx)) {
+			u64 wa_ctx_offset = obj->gtt_offset;
+			u32 *wa_ctx_page = &obj->pages[0][0];
+			struct intel_engine_cs *ring = &dev_priv->ring[RCS];
+			u32 wa_ctx_size = (ring->wa_ctx.indirect_ctx.size +
+					   ring->wa_ctx.per_ctx.size);
+
+			err_printf(m, "%s --- WA Ctx batch buffer = 0x%08llx\n",
+				   dev_priv->ring[i].name, wa_ctx_offset);
+			offset = 0;
+			for (elt = 0; elt < wa_ctx_size; elt += 4) {
+				err_printf(m, "[%04x] %08x %08x %08x %08x\n",
+					   offset,
+					   wa_ctx_page[elt],
+					   wa_ctx_page[elt+1],
+					   wa_ctx_page[elt+2],
+					   wa_ctx_page[elt+3]);
+				offset += 16;
+			}
+		}
+
 		if ((obj = error->ring[i].ctx)) {
 			err_printf(m, "%s --- HW Context = 0x%08x\n",
 				   dev_priv->ring[i].name,
@@ -752,6 +773,7 @@ static void i915_error_state_free(struct kref *error_ref)
 		i915_error_object_free(error->ring[i].hws_page);
 		i915_error_object_free(error->ring[i].ctx);
 		kfree(error->ring[i].requests);
+		i915_error_object_free(error->ring[i].wa_ctx);
 	}
 
 	i915_error_object_free(error->semaphore_obj);
@@ -1267,6 +1289,11 @@ static void i915_gem_record_rings(struct drm_device *dev,
 		error->ring[i].hws_page =
 			i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
 
+		if (ring->wa_ctx.obj) {
+			error->ring[i].wa_ctx =
+				i915_error_ggtt_object_create(dev_priv, ring->wa_ctx.obj);
+		}
+
 		i915_gem_record_active_context(ring, error, &error->ring[i]);
 
 		count = 0;
-- 
1.9.1