mark time = 0

[d-netto]

Happened when running tree_mutable.jl on JuliaLang/julia#45639.

threads = 2
┌─────────┬────────────┬─────────┬───────────┬────────────┬──────────────┬───────────────────┬──────────┬────────────┐
│         │ total time │ gc time │ mark time │ sweep time │ max GC pause │ time to safepoint │ max heap │ percent gc │
│         │         ms │      ms │        ms │         ms │           ms │                us │       MB │          % │
├─────────┼────────────┼─────────┼───────────┼────────────┼──────────────┼───────────────────┼──────────┼────────────┤
│ minimum │      17269 │    7583 │         0 │         31 │          227 │                55 │      612 │         44 │
│  median │      17269 │    7583 │         0 │         31 │          227 │                55 │      612 │         44 │
│ maximum │      17269 │    7583 │         0 │         31 │          227 │                55 │      612 │         44 │
└─────────┴────────────┴─────────┴───────────┴────────────┴──────────────┴───────────────────┴──────────┴────────────┘

[d-netto]

Seems to be reproducible on master as well:

bench = "tree_mutable.jl"
┌─────────┬────────────┬─────────┬───────────┬────────────┬──────────────┬───────────────────┬──────────┬────────────┐
│         │ total time │ gc time │ mark time │ sweep time │ max GC pause │ time to safepoint │ max heap │ percent gc │
│         │         ms │      ms │        ms │         ms │           ms │                us │       MB │          % │
├─────────┼────────────┼─────────┼───────────┼────────────┼──────────────┼───────────────────┼──────────┼────────────┤
│ minimum │      19326 │    8624 │         0 │         33 │          273 │                56 │      603 │         44 │
│  median │      19984 │    9278 │         0 │         39 │          273 │                57 │      603 │         46 │
│ maximum │      22215 │   10962 │         0 │         47 │          276 │                74 │      610 │         49 │
└─────────┴────────────┴─────────┴───────────┴────────────┴──────────────┴───────────────────┴──────────┴────────────┘

[oscardssmith]

@chflood are the times in the wrong places?

[d-netto]

Also, mark + sweep times seem to be discrepant from gc time. I wouldn't expect them to add up, but would expect them to be in the same order of magnitude.

[chflood]

Yeah, we were reporting the last gc's mark and sweep time, and not the total mark and sweep time. Fix is coming.

…

On Wed, Jun 29, 2022 at 6:27 PM Diogo Netto ***@***.***> wrote: Also, mark + sweep times seem to be discrepant from gc time. I wouldn't expect them to add up, but would expect them to be in the same order of magnitude. — Reply to this email directly, view it on GitHub <#34 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AE3CLBE3I46GSMALRXSO5JLVRTEVXANCNFSM52HJOGOQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[d-netto]

Resolved a few commits ago.