[Intel-gfx] [PATCH 2/2] drm/i915: Extend reset modparam to domain resets

Stuart Summers stuart.summers at intel.com
Tue May 14 16:46:53 UTC 2019


In the event a platform does not properly implement reset,
do not go through reset flows for engine domains to avoid
an unlikely situation where writes are accepted but register
values are never cleared, as this can result in GPU wedges
in these cases.

Signed-off-by: Stuart Summers <stuart.summers at intel.com>
---
 drivers/gpu/drm/i915/gt/intel_reset.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c
index 464369bc55ad..81f9f9f73b1c 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset.c
+++ b/drivers/gpu/drm/i915/gt/intel_reset.c
@@ -309,6 +309,12 @@ static int gen6_hw_domain_reset(struct drm_i915_private *i915,
 	struct intel_uncore *uncore = &i915->uncore;
 	int err;
 
+	if (!i915_modparams.reset) {
+		DRM_DEBUG_DRIVER("Skipping 0x%08x engines reset\n",
+				 hw_domain_mask);
+		return 0;
+	}
+
 	/*
 	 * GEN6_GDRST is not in the gt power well, no need to check
 	 * for fifo space for the write or forcewake the chip for
@@ -517,6 +523,13 @@ static int gen8_engine_reset_prepare(struct intel_engine_cs *engine)
 		return 0;
 	}
 
+	if (!i915_modparams.reset) {
+		DRM_DEBUG_DRIVER("Skipping %s reset request {request: %08x, RESET_CTL: %08x}\n",
+				 engine->name, request,
+				 intel_uncore_read_fw(uncore, reg));
+		return 0;
+	}
+
 	intel_uncore_write_fw(uncore, reg, _MASKED_BIT_ENABLE(request));
 	ret = __intel_wait_for_register_fw(uncore, reg, mask, ack,
 					   700, 0, NULL);
-- 
2.21.0.5.gaeb582a983



More information about the Intel-gfx mailing list