-
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Optionally use NewPM for optimization #46176
Conversation
4981f50
to
690e882
Compare
283d735
to
78dabb2
Compare
e9a65a4
to
1083828
Compare
@nanosoldier |
@nanosoldier |
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. |
With the performance regression that nanosoldier sees it might be better to have a flag that turns this on? |
There is a flag (JL_USE_NEW_PM) that can be used to disable new PM. |
I was thinking a runtime flag not a compile time flag so that's easy to switch, but maybe we can just expose it in Make.inc so that it is easy to set |
Your package evaluation job has completed - possible new issues were detected. A full report can be found here. |
Seems:
is a common cause for |
In the interest of getting the changes in and then enabling it after ironing out the bugs, I'm disabling newpm by default (it's still accessible via |
a8c3db1
to
1e562cd
Compare
1e562cd
to
86ac297
Compare
If the tests come back clean I will merge the PR since every nontrivial change is underneath an ifdef. |
The win64 and win32 failures are network errors, and are unrelated to this change. |
This continues from #46175 and enables NewPM on all of our optimization pipelines, unless ASAN is active. This is because the combination of ASAN, our JIT, and NewPM results in a recursive lookup during codegen, which triggers an assert for our memory manager.
This is the other half of #44365.