RFR - JDK-8209517: com/sun/jdi/BreakpointWithFullGC.java fails with timeout

[Mikael Vidstedt]

Thanks for fixing this Alex! I’ve skimmed through the changes and AFAICT it all looks good.

One *very* minor thing (feel free to completely ignore this) - there’s a mix of “debugee’ (one ‘g’) and “debuggee” (two ‘g’:s). :)

Cheers,
Mikael

> On Aug 15, 2018, at 3:42 PM, Alex Menkov <alexey.menkov at oracle.com> wrote:
> 
> Hi all,
> 
> please review a fix for
> https://bugs.openjdk.java.net/browse/JDK-8209517
> webrev:
> http://cr.openjdk.java.net/~amenkov/sh2java/step1_regression/webrev/
> 
> Cause of the BreakpointWithFullGC failures is a mess of jdb and debuggee outputs (the test runs debuggee by using CommandLineLaunch connector, so jdb redirects debuggee stdout to its own stdout).
> To solve it test framework was updated to launch debuggee first (redirecting its output) and then connecting jdb to existing process.
> The approach allow to drop "simple prompt" logic (jdb mode when debugee is not yet running).
> 
> Mach5 passed 400 runs ("4 std platforms" x "--test-repeat 100") without any issues.
> 
> --alex