Message351335
I have compiled both versions with optimizations on, looks like the gap gets a bit smaller (percentage-wise) after that:
benchmark | master (ns) | PR 15633 (ns) | Δ (%)
------------------------------------+---------------+-----------------+----------
call only (datetime) | 73 (±3) | 92.3 (±7) | 26
constructor + call (datetime) | 228 (±9) | 260 (±16) | 14
timedelta + call (datetime) | 108 (±5) | 128 (±9) | 18
If this were something fundamental like a performance regression in building a tuple or constructing a dictionary or something I'd be concerned, but this just reinforces my feeling that, on balance, this is worth it, and that we are probably not going to need a "fast path" version of this. |
|
Date |
User |
Action |
Args |
2019-09-08 18:16:47 | p-ganssle | set | recipients:
+ p-ganssle, tim.peters, rhettinger, belopolsky, vstinner, taleinat, berker.peksag, serhiy.storchaka, bmispelon, corona10 |
2019-09-08 18:16:47 | p-ganssle | set | messageid: <1567966607.65.0.736480634271.issue24416@roundup.psfhosted.org> |
2019-09-08 18:16:47 | p-ganssle | link | issue24416 messages |
2019-09-08 18:16:47 | p-ganssle | create | |
|