-
Bug
-
Resolution: Fixed
-
P4
-
17, 21, 22
-
b06
-
generic
-
generic
Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
---|---|---|---|---|---|---|
JDK-8316550 | 21.0.2 | Victor Rudometov | P4 | Resolved | Fixed | b02 |
JDK-8318213 | 17.0.10-oracle | Adithya Haradi Gopal | P4 | Resolved | Fixed | b04 |
JDK-8316625 | 17.0.10 | Victor Rudometov | P4 | Resolved | Fixed | b01 |
It is written like this:
metaSpaceRegexp = Pattern.compile(".*Metaspace: ([0-9]+).*->([0-9]+).*");
For the following log,
[gc,metaspace ] GC(0) Metaspace: 11895K(14208K)->11895K(14208K) NonClass: 10552K(12544K)->10552K(12544K) Class: 1343K(1664K)->1343K(1664K)
It intends to capture the metaspace size before and after the GC, which is the first 11895 and second 11895 separately. But as this regex is greedy, it will capture the first 11895 and the last 1343.
This mistake might cause the test always to succeed.
It might be changed to:
metaSpaceRegexp = Pattern.compile(".*Metaspace: ([0-9]+)K\\([0-9]+K\\)->([0-9]+)K\\([0-9]+K\\).*");
- backported by
-
JDK-8316550 Incorrect regex in TestMetaSpaceLog.java
- Resolved
-
JDK-8316625 Incorrect regex in TestMetaSpaceLog.java
- Resolved
-
JDK-8318213 Incorrect regex in TestMetaSpaceLog.java
- Resolved
- relates to
-
JDK-8211123 GC Metaspace printing after full gc
- Resolved
- links to
-
Commit openjdk/jdk17u-dev/4ebd8b33
-
Commit openjdk/jdk21u/46f210c4
-
Commit openjdk/jdk/9084b6c9
-
Review openjdk/jdk17u-dev/1737
-
Review openjdk/jdk21u/159
-
Review openjdk/jdk/14782