Message205594
I saw a small regression over 4k when using a 64k buffer on one of my machines (dual core amd64 linux). With 32k everything (amd64 linux, armv7l 32-bit linux, 64-bit os x 10.6) showed a dramatic improvement on the microbenchmark. approaching 50% less cpu use in many cases.
i doubt applications will notice as much as they're likely to be dominated by their own application code rather than the subprocess internals.
re: 3.3 or not, true, but since it doesn't change any APIs and is minor I did it anyways. If you think it doesn't belong there, leave it to the release manager to back out. This and the #19506 change should be invisible to users. |
|
Date |
User |
Action |
Args |
2013-12-08 19:05:42 | gregory.p.smith | set | recipients:
+ gregory.p.smith, pitrou, vstinner, neologix, python-dev, sbt, serhiy.storchaka |
2013-12-08 19:05:42 | gregory.p.smith | set | messageid: <1386529542.24.0.534353452244.issue19929@psf.upfronthosting.co.za> |
2013-12-08 19:05:42 | gregory.p.smith | link | issue19929 messages |
2013-12-08 19:05:41 | gregory.p.smith | create | |
|