On Fri, 14 Oct 2022 09:53:23 GMT, Kevin Walls <[email protected]> wrote:
> Hi, are you sure you want to back this out? I'll click approve as backout
> change I think looks OK.
>
> I see test/hotspot/jtreg/serviceability/dcmd/framework/VMVersionTest.java
> failing SOMETIMES for me. I think you already noted that
> src/jdk.internal.jvmstat/share/classes/sun/jvmstat/perfdata/monitor/protocol/local/PerfDataBuffer.java
> no longer handles IllegalArgumentException
>
> If we handle that then for me the test that failed frequently, runs OK 20
> times in a row.
>
> ```
> bash-4.2$ git diff
> diff --git
> a/src/jdk.internal.jvmstat/share/classes/sun/jvmstat/perfdata/monitor/protocol/local/PerfDataBuffer.java
>
> b/src/jdk.internal.jvmstat/share/classes/sun/jvmstat/perfdata/monitor/protocol/local/PerfDataBuffer.java
> index 3ad35e350e4..056671d7841 100644
> ---
> a/src/jdk.internal.jvmstat/share/classes/sun/jvmstat/perfdata/monitor/protocol/local/PerfDataBuffer.java
> +++
> b/src/jdk.internal.jvmstat/share/classes/sun/jvmstat/perfdata/monitor/protocol/local/PerfDataBuffer.java
> @@ -63,7 +63,7 @@ public class PerfDataBuffer extends AbstractPerfDataBuffer {
> try {
> ByteBuffer bb = perf.attach(vmid.getLocalVmId());
> createPerfDataBuffer(bb, vmid.getLocalVmId());
> - } catch (IOException e) {
> + } catch (IOException | IllegalArgumentException e) {
> throw new MonitorException("Could not attach to "
> + vmid.getLocalVmId(), e);
> }
> ```
I want to do more testing and then make a REDO PR for the 1.4.1 code removal. I
also want to understand why these tests would sometimes get an
IllegalArgumentException.
-------------
PR: https://git.openjdk.org/jdk/pull/10711
