blob: 254489b667d98c155a1c1039804f5c8b17632685 [file] [log] [blame]
-- Gather the `nop_work_func` slices and the CPU they each ran on and use that
-- information to generate a metric that displays just the Trusty workqueue
-- events grouped by CPU.
DROP VIEW IF EXISTS android_trusty_workqueues_event;
CREATE VIEW android_trusty_workqueues_event AS
SELECT
'slice' AS track_type,
name AS slice_name,
ts,
dur,
'Cpu ' || EXTRACT_ARG(arg_set_id, 'cpu') AS track_name,
'Trusty Workqueues' AS group_name
FROM slice
WHERE slice.name GLOB 'nop_work_func*';
-- Generate the final metric output. This is empty because we're only using the
-- metric to generate custom tracks, and so don't have any aggregate data to
-- generate.
DROP VIEW IF EXISTS android_trusty_workqueues_output;
CREATE VIEW android_trusty_workqueues_output AS
SELECT AndroidTrustyWorkqueues();