Message359225
> (slowdowns because releasing/acquiring the GIL is not free, particularly when contended)
Yes, it's relatively high. We shouldn't release the GIL only for ~0.5ms. That's why 1MB~ seems nice threshold.
> If the threshold is too low then users can always choose not to use multi-threading (and in general one doesn't expect much from it in Python)
I don't like this idea. We shouldn't force users to change their program from multi threaded to single threaded. So we need to avoid large performance regression.
> but if the threshold is too high then users have no recourse. That being said, 65536 does still seem a bit low based on the results available.
But they can get the benefit when input is too large. 65536 seems "too low" to me. |
|
Date |
User |
Action |
Args |
2020-01-03 02:42:34 | methane | set | recipients:
+ methane, pitrou, vstinner, serhiy.storchaka, josh.r, remi.lapeyre, BTaskaya, bmerry |
2020-01-03 02:42:34 | methane | set | messageid: <1578019354.05.0.979386570877.issue36051@roundup.psfhosted.org> |
2020-01-03 02:42:34 | methane | link | issue36051 messages |
2020-01-03 02:42:33 | methane | create | |
|