RFR: 8223051: support loops with long (64b) trip counts

Roland Westrelin roland at openjdk.java.net
Thu Sep 24 06:58:17 UTC 2020


Last webrev was:

https://cr.openjdk.java.net/~roland/8223051/webrev.03/

This PR includes a few minor changes:

- The change in callnode.cpp that Vladimir requested in:
  https://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2020-August/039551.html

- Extra comments that John requested in:
  https://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2020-August/039621.html

- A couple extra counters to collect more detailed statistics

8252696 (Loop unswitching may cause out of bound array load to be
executed) was the only bug that was uncovered by extended testing and
it's fixed now.

This was previously reviewed by Tobias, Vladimir and John. Given the
last changes were either requested by reviewers or a straighforward
improvement to statistics, and unless someone objects, I intend to
push this in the next few days with the reviewer list I just
mentioned.

-------------

Commit messages:
 - trailing whitespaces
 - long counted loops

Changes: https://git.openjdk.java.net/jdk/pull/318/files
 Webrev: https://webrevs.openjdk.java.net/?repo=jdk&pr=318&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8223051
  Stats: 902 lines in 11 files changed: 823 ins; 63 del; 16 mod
  Patch: https://git.openjdk.java.net/jdk/pull/318.diff
  Fetch: git fetch https://git.openjdk.java.net/jdk pull/318/head:pull/318

PR: https://git.openjdk.java.net/jdk/pull/318


More information about the hotspot-compiler-dev mailing list