e030f808ec
Currently native clusterization simply groups all benchmarks by the opcode of key instruction, but that is suboptimal in certain cases, e.g. where we can already tell that the particular instructions already resolve into different sched classes. |
||
|---|---|---|
| .. | ||
| inverse_throughput | ||
| latency | ||
| lbr | ||
| uops | ||
| analysis-cluster-stabilization-config.test | ||
| analysis-cluster-stabilization.test | ||
| analysis-clustering-algorithms.test | ||
| analysis-epsilons.test | ||
| analysis-inconsistencies-uops-backwards.test | ||
| analysis-inconsistencies-uops.test | ||
| analysis-naive-cluster-stabilization.test | ||
| analysis-naive-clusterization-same-opcode-different-sched-class.test | ||
| analysis-naive-clusterization.test | ||
| analysis-noise.test | ||
| analysis-same-cluster-for-ops-in-different-sched-clusters.test | ||
| analysis-uops-backwards.test | ||
| analysis-uops-variant.test | ||
| analysis-uops.test | ||
| lit.local.cfg | ||