github-actions[bot]
commented
1 month ago Valgrind Results:
==4041== Cachegrind, a cache and branch-prediction profiler
==4041== Copyright (C) 2002-2017, and GNU GPL'd, by Nicholas Nethercote et al.
==4041== Using Valgrind-3.18.1 and LibVEX; rerun with -h for copyright info
==4041== Command: target/release/snailtracer
==4041==
--4041-- warning: L3 cache found, using its data for the LL simulation.
Running snailtracer example!
elapsed: 1.809849425s
==4041==
==4041== I refs: 427,439,038
==4041== I1 misses: 3,171
==4041== LLi misses: 2,854
==4041== I1 miss rate: 0.00%
==4041== LLi miss rate: 0.00%
==4041==
==4041== D refs: 203,986,865 (132,175,939 rd + 71,810,926 wr)
==4041== D1 misses: 342,605 ( 202,697 rd + 139,908 wr)
==4041== LLd misses: 137,710 ( 4,397 rd + 133,313 wr)
==4041== D1 miss rate: 0.2% ( 0.2% + 0.2% )
==4041== LLd miss rate: 0.1% ( 0.0% + 0.2% )
==4041==
==4041== LL refs: 345,776 ( 205,868 rd + 139,908 wr)
==4041== LL misses: 140,564 ( 7,251 rd + 133,313 wr)
==4041== LL miss rate: 0.0% ( 0.0% + 0.2% )
This is the easiest way for test purposes to handle EOF and Prague differences in spec.
to have EOF disable for devnets but enabled in tests