Segmentation fault while running cachesim

[EduardFedorenkov]

Hello!
I have an issue while running cachesim. Let me introduce the workflow:

1. download 202210_kv_traces_all_sort.csv.oracleGeneral.zst
2. convert 202210_kv_traces_all_sort.csv.oracleGeneral.zst -> 202210_kv_traces_all_sort.csv.oracleGeneral by
   zstd -d 202210_kv_traces_all_sort.csv.oracleGeneral.zst
3. run: ./_build/bin/cachesim ./data/202210_kv_traces_all_sort.csv.oracleGeneral oracleGeneral lru 1mb --num-thread=1

the output after step 3:

[INFO]  10-14-2025 15:14:14 cli_parser.c:559  (tid=281473220710464): trace path: ./data/202210_kv_traces_all_sort.csv.oracleGeneral, trace_type ORACLE_GENERAL_TRACE, ofilepath result/202210_kv_traces_all_sort.csv.oracleGeneral.cachesim, 1 threads, warmup -1 sec, total 1 algo x 1 size = 1 caches, lru
[DEBUG] 10-14-2025 15:14:14 request.h:125  (tid=281473220710464): req clock_time 564726470, id 798057486, size 62, op nop, valid 1
[DEBUG] 10-14-2025 15:14:14 request.h:125  (tid=281473220710464): req clock_time 564726470, id 798057486, size 62, op nop, valid 1
Segmentation fault (core dumped)

Can you help me with testing cachesim? I also tested cachesim on data from your repo, but miss ratio is always 1.0000 or 0.0000. Can I expect something between 0.000 and 1.000?

[1a1a11a]

Hi @EduardFedorenkov Thank you for your interest!
We are no longer maintaining this repo, can you try https://github.com/1a1a11a/libCacheSIm?
BTW, 1mb is too small, you can use 0.01 and 0.1, which will automatically calculates the cache size based on working set.

[EduardFedorenkov]

thank you I'll try the new libCacheSim.