Move long running o1vm test to nightly, run tests for small programs

[martyall]

I'm pretty sure I'm using the on.schedule correctly, but I'm unsure how/when it reports to the current PR. The main concern would be that get notified of a failure in the right place

bad news: it takes about the same amount of time as the current job running for 3M instructions (if I understand the current job's --stop-at =3000000 arg). Do I remember that we were going to bump this to use the preimage oracle? In that case maybe its worth it.

I don't think running in parallel will help here as even one program is utilizing all cores on my machine (there are a fair number of parallel commitments being computed)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 76.87%. Comparing base (82bbc48) to head (4888f66).
Report is 10 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #3010      +/-   ##
==========================================
+ Coverage   76.83%   76.87%   +0.03%     
==========================================
  Files         262      262              
  Lines       62028    62114      +86     
==========================================
+ Hits        47662    47750      +88     
+ Misses      14366    14364       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dannywillems]

Do I remember that we were going to bump this to use the preimage oracle? In that case maybe its worth it.

Yes, and we will need to load a previous state. Discussion on Slack:

re o1vm preimage test in CI: running on my machine gives me a hit of the preimage instruction around 56M and 57M. I would suggest to run the CI from something around ~55M instructions to ~60M instructions.

2025-02-03T12:57:20Z INFO  o1vm::interpreters::mips::witness] processing step=55000000 pc=000023951c insn=00000a1280 ips=91872.79 pages=9167 mem=35.8 MiB name=n/a
[2025-02-03T12:57:21Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:21Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:22Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:23Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:23Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:24Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:24Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:25Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:25Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:26Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:27Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:27Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:28Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:28Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:29Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk
[2025-02-03T12:57:29Z DEBUG o1vm::preimage_oracle] Reading response
[2025-02-03T12:57:29Z DEBUG o1vm::preimage_oracle] Extracting contents
[2025-02-03T12:57:29Z DEBUG o1vm::preimage_oracle] Got preimage of length 32
     dcc3b3317a858eb537c93cc1591da0a1e34dd942610c825e6095fb2b4225a7b0
[2025-02-03T12:57:29Z DEBUG o1vm::interpreters::mips::witness] Preimage has been read entirely, triggering Keccak process
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Reading response
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Extracting contents
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Got preimage of length 32
     7ebd8da4e399db97ac1b854a7a74d6d3bbda59f0dfbb2e0ee024cb4fb80e439e
[2025-02-03T12:57:30Z DEBUG o1vm::interpreters::mips::witness] Preimage has been read entirely, triggering Keccak process
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Reading response
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Extracting contents
[2025-02-03T12:57:30Z DEBUG o1vm::preimage_oracle] Got preimage of length 32
     cac628b3d4961278fb0461074c1f7df2041887b10511cb6a9099e42a195577e7
[2025-02-03T12:57:30Z DEBUG o1vm::interpreters::mips::witness] Preimage has been read entirely, triggering Keccak process
[2025-02-03T12:57:30Z DEBUG pickles_o1vm] Limit of 65536 reached. We make a proof, verify it (for testing) and start with a new chunk

[dannywillems]

As discussed online, I would parallelise it, using an equivalent strategy than in o1js.

[dannywillems]

Refusing for now as it takes > 60m "O1VM MIPS Prover CI / Run o1vm prover on all mips tests (1.74) (pull_request) Successful in 65m" to run the new job. Suggested to parallelize.