RFR: JDK-8312525: New test runtime/os/TestTrimNative.java#trimNative is failing: did not see the expected RSS reduction

[David Holmes]

On Sat, 22 Jul 2023 06:57:43 GMT, Thomas Stuefe <stuefe at openjdk.org> wrote:

> The test causes intermittent rare errors, so far only seen on Oracle 7.9 with glibc 2.17.  
> 
> This PR disables, for automated testing, the RSS evaluation part of the test. It retains the mechanical part (testing whether trims happen at all, in the expected number). The stricter variant of the test is kept as manual tests (usable with jtreg without -a).

Seems reasonable. Not 100% sure about the extra logging ... one query in relation to that.

Thanks.

src/hotspot/os/linux/os_linux.cpp line 4649:

> 4647: 
> 4648:   // Log also:
> 4649:   // - LD_PRELOAD, since this is the most popular mechanism to divert libc functions

What about the content of /etc/ld.so.preload? (We show that in hs_err file.)

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

Marked as reviewed by dholmes (Reviewer).

PR Review: https://git.openjdk.org/jdk/pull/14984#pullrequestreview-1546882402
PR Review Comment: https://git.openjdk.org/jdk/pull/14984#discussion_r1274411924