RFR: 8371420: Still sporadic failures of gc/TestAlwaysPreTouchBehavior.java#<gcname> on Linux after JDK-8359104 [v2]

Matthias Baesken mbaesken at openjdk.org
Thu Dec 18 17:04:44 UTC 2025


On Tue, 16 Dec 2025 09:32:22 GMT, Albert Mingkun Yang <ayang at openjdk.org> wrote:

> > ... where do you want to see such a SkippedException added ?
> 
> The `else` branch of `if (available > requiredAvailable) {` is empty. Shouldn't there be a `SkippedException` in the `else` branch?

We would continue the for - loop 

        // RSS values we get are sometimes somewhat delayed or inaccurate
        for (int iter=0; iter < maxIter; iter++) {

in this case, without a SkippedException.  In the original test before this change, we did not throw a SkippedException as well .
On the other hand , if we really run into the else - branch  we must have got a small   hostAvailableMemory value, so skipping the test might be a good idea because most likely the hostAvailableMemory  will not grow in the for loop,

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

PR Comment: https://git.openjdk.org/jdk/pull/28735#issuecomment-3671239322


More information about the hotspot-gc-dev mailing list