Message157916
FWIW, doing a linear probe only within a cache line (changing the 1's bit) before applying perturb might also be useful -- and the results may change if the size of a dictentry were reduced. (Mark Shannon's now-integrated patch doesn't actually do that for the keys, but it would be doable.) |
|
Date |
User |
Action |
Args |
2012-04-09 22:38:10 | Jim.Jewett | set | recipients:
+ Jim.Jewett, tim.peters, rhettinger, mark.dickinson, pitrou, neologix, jcon, serhiy.storchaka |
2012-04-09 22:38:10 | Jim.Jewett | set | messageid: <1334011090.29.0.185496220618.issue10408@psf.upfronthosting.co.za> |
2012-04-09 22:38:09 | Jim.Jewett | link | issue10408 messages |
2012-04-09 22:38:09 | Jim.Jewett | create | |
|