As part of the dirstate refactor, fsmonitor was updated to directly access the
inner map of the dirstatemap object.
Dirstatemap reimplementations may not use a map like this, so only access it if
it is there.
As part of the dirstate refactor, fsmonitor was updated to directly access the
inner map of the dirstatemap object.
Dirstatemap reimplementations may not use a map like this, so only access it if
it is there.
Automatic diff as part of commit; lint not applicable. |
Automatic diff as part of commit; unit tests not applicable. |
hgext/fsmonitor/__init__.py | ||
---|---|---|
279 | Do we do this for perf reasons? If so, add a comment. (Maybe check and see if this shows up on perfdirstate or similar - if it does, document that in the commit message with numbers, and then you don't even need a comment because we'll find that with blame.) |
hgext/fsmonitor/__init__.py | ||
---|---|---|
279 | Not sure if I understand the perfwalk output. Without this direct access of dirstate._map._map I get: ! result: 1485 ! wall 0.004521 comb 0.000000 user 0.000000 sys 0.000000 (best of 589) with the direct access: ! result: 1485 ! wall 0.004203 comb 0.000000 user 0.000000 sys 0.000000 (best of 628) I'm guessing the wall clock time is important, so this is a ~7% improvement. Is that right? I will add a comment here in any case. |
Do we do this for perf reasons? If so, add a comment.
(Maybe check and see if this shows up on perfdirstate or similar - if it does, document that in the commit message with numbers, and then you don't even need a comment because we'll find that with blame.)