Tim Van Patten c585341552 intel/ds: Skip expensive timestamp query until necessary ... The Xe ioctl DRM_XE_DEVICE_QUERY_ENGINE_CYCLES provides accurate timestamps correlated between the CPU and GPU. However, it is slow and impacts performance while collecting Perfetto traces. Instead, use Perfetto's GetBootTimeNs() to track when to emit the BUILTIN_CLOCK_BOOTTIME clock sync event so it only occurs every 1 second. This reduces the impact of recording gpu.renderstages from -8% to -4%. More concretely, FPS measurements when tracing Unity BoatAttack demo on an Intel ADL device: * gpu.renderstages disabled: 48.044293667 * gpu.renderstages enabled: 38.119778333 (-20.66%) * gpu.renderstages enabeled + this fix: 42.641818333 (-11.24%) Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37095>		2025-08-29 21:34:43 +00:00
..
.clang-format
intel_driver_ds.cc	intel/ds: Skip expensive timestamp query until necessary	2025-08-29 21:34:43 +00:00
intel_driver_ds.h	anv/ds: Forward VkDebugUtilsObjectNameInfoEXT to perfetto.	2025-05-08 01:21:25 +00:00
intel_pps_driver.cc	intel/perf: Implement intel_perf_query_result_accumulate() for gfx 20+	2024-07-03 17:15:05 +00:00
intel_pps_driver.h	intel/perf: Replace drm_i915_perf_record_header by intel_perf_record_header	2024-05-27 19:34:06 +00:00
intel_pps_perf.cc	intel/ds: Nuke ralloc_ctx and ralloc_cfg	2024-05-07 21:44:34 +00:00
intel_pps_perf.h	intel/ds: Nuke ralloc_ctx and ralloc_cfg	2024-05-07 21:44:34 +00:00
intel_pps_priv.h	intel/ds: use a per GPU clock ID	2022-01-14 20:17:44 +00:00
intel_tracepoints.py	anv/ds: Associate the VkCommandBuffer some anv-only renderstage events.	2025-05-08 01:21:25 +00:00
meson.build	build: pass licensing information in SPDX form	2024-06-29 12:42:49 -07:00