perf: Optimize the output of global metrics to reduce atomic operations
This commit is contained in:
@@ -175,7 +175,7 @@ struct sce_ctx *sce_ctx_create(const char *profile)
|
||||
}
|
||||
|
||||
sce_ctx->ts = timestamp_new(sce_ctx->ts_update_interval_ms);
|
||||
sce_ctx->metrics = global_metrics_create(profile);
|
||||
sce_ctx->metrics = global_metrics_create(profile, sce_ctx->nr_worker_threads);
|
||||
if (sce_ctx->metrics == NULL)
|
||||
{
|
||||
goto error_out;
|
||||
|
||||
Reference in New Issue
Block a user