1. Jan 16, 2024
    • Andrei Eres's avatar
      subsystem-bench: cache misses profiling (#2893) · ec7bfae0
      Andrei Eres authored
      ## Why we need it
      To provide another level of understanding to why polkadot's subsystems
      may perform slower than expected. Cache misses occur when processing
      large amounts of data, such as during availability recovery.
      
      ## Why Cachegrind
      Cachegrind has many drawbacks: it is slow, it uses its own cache
      simulation, which is very basic. But unlike `perf`, which is a great
      tool, Cachegrind can run in a virtual machine. This means we can easily
      run it in remote installations and even use it in CI/CD to catch
      possible regressions.
      
      Why Cachegrind and not Callgrind, another part of Valgrind? It is simply
      empirically proven that profiling runs faster with Cachegrind.
      
      ## First results
      First results have been obtained while testing of the approach. Here is
      an example.
      
      ```
      $ target/testnet/subsystem-bench --n-cores 10 --cache-misses data-availability-read
      $ cat cachegrind_report.txt
      I refs:        64,622,081,485
      I1  misses:         3,018,168
      LLi misses:           437,654
      I1  miss rate:           0.00%
      LLi miss rate:           0.00%
      
      D refs:        12,161,833,115  (9,868,356,364 rd   + 2,293,476,751 wr)
      D1  misses:       167,940,701  (   71,060,073 rd   +    96,880,628 wr)
      LLd misses:        33,550,018  (   16,685,853 rd   +    16,864,165 wr)
      D1  miss rate:            1.4% (          0.7%     +           4.2%  )
      LLd miss rate:            0.3% (          0.2%     +           0.7%  )
      
      LL refs:          170,958,869  (   74,078,241 rd   +    96,880,628 wr)
      LL misses:         33,987,672  (   17,123,507 rd   +    16,864,165 wr)
      LL miss rate:             0.0% (          0.0%     +           0.7%  )
      ```
      
      The CLI output shows that 1.4% of the L1 data cache missed, which is not
      so bad, given that the last-level cache had that data most of the time
      missing only 0.3%. Instruction data of the L1 has 0.00% misses of the
      time. Looking at an output file with `cg_annotate` shows that most of
      the misses occur during reed-solomon, which is expected.
      ec7bfae0
    • Dónal Murray's avatar
      Bump rococo relay and coretime-rococo to 1.6 (#2913) · a42a47f8
      Dónal Murray authored
      Co-authored-by: command-bot <>
      a42a47f8
  2. Jan 15, 2024
  3. Jan 12, 2024
  4. Jan 11, 2024
  5. Jan 10, 2024
  6. Jan 09, 2024
  7. Jan 08, 2024
  8. Jan 07, 2024
  9. Jan 06, 2024
  10. Jan 05, 2024
  11. Jan 04, 2024
  12. Jan 02, 2024
  13. Dec 29, 2023
  14. Dec 28, 2023
  15. Dec 27, 2023
  16. Dec 26, 2023
  17. Dec 22, 2023
  18. Dec 21, 2023
  19. Dec 20, 2023
    • Dónal Murray's avatar
      Fix clippy lints behind feature gates and add new CI step all features (#2569) · d68868f6
      Dónal Murray authored
      
      
      Many clippy lints usually enforced by `-Dcomplexity` and `-Dcorrectness`
      are not caught by CI as they are gated by `features`, like
      `runtime-benchmarks`, while the clippy CI job runs with only the default
      features for all targets.
      
      This PR also adds a CI step to run clippy with `--all-features` to
      ensure the code quality is maintained behind feature gates from now on.
      
      To improve local development, clippy lints are downgraded to warnings,
      but they still will result in an error at CI due to the `-Dwarnings`
      rustflag.
      
      ---------
      
      Co-authored-by: default avatarLiam Aharon <[email protected]>
      d68868f6