There's also a leak in compressed streams though it's probably less large... See https://github.com/apache/arrow/pull/4049#pullrequestreview-219550038
Regards Antoine. Le 27/03/2019 à 18:54, Wes McKinney a écrit : > Since we aren't testing CUDA extensions in CI yet I am not sure we should > block over issues with them. It seems like this (plus CI on physical > hardware) can be resolved for 0.14. > > On Wed, Mar 27, 2019, 3:42 PM Antoine Pitrou <solip...@pitrou.net> wrote: > >> On Wed, 27 Mar 2019 20:36:14 +0900 (JST) >> Kouhei Sutou <k...@clear-code.com> wrote: >>> Hi, >>> >>> I would like to propose the following release candidate (RC3) of Apache >>> Arrow version 0.13.0. This is a release consiting of 584 >>> resolved JIRA issues[1]. >>> >>> This release candidate is based on commit: >>> 6de758a9e5f74d50ec3458b2af17b7d2d892c573 [2] >>> >>> The source release rc3 is hosted at [3]. >>> The binary artifacts are hosted at [4][5][6][7]. >>> The changelog is located at [8]. >>> >>> Please download, verify checksums and signatures, run the unit tests, >>> and vote on the release. See [9] for how to validate a release candidate. >>> >>> The vote will be open for at least 72 hours. >>> >>> [ ] +1 Release this as Apache Arrow 0.13.0 >>> [ ] +0 >>> [ ] -1 Do not release this as Apache Arrow 0.13.0 because... >> >> I'm leaning towards -1 because there are unforeseen CUDA memory leaks >> in release mode. Related to >> https://issues.apache.org/jira/browse/ARROW-5029. >> >> Regards >> >> Antoine. >> >> >> >
