RFR (S/M): 8169703: G1 crashes with guarantee(pretouch_gang != NULL) failed: No pretouch gang specified

Dmitry Dmitriev dmitry.dmitriev at oracle.com
Fri Nov 18 12:24:24 UTC 2016


Hi Thomas,

Can comment only new test test/gc/g1/TestSharedArchiveWithPreTouch.java:
1) Can you please use different name for archive to avoid intersection 
with existing tests? E.g. "SharedArchiveWithPreTouch.jsa" instead of 
"CDSCompressedKPtrs.jsa".
2) Just interesting: Is it intended behavior of the test to use 
UseCompressedClassPointers and UseCompressedOops?

Thank you!

Dmitry

On 18.11.2016 11:47, Thomas Schatzl wrote:
> Hi all,
>
>    can I have reviews for this change that fixes use of AlwaysPreTouch
> with CDS?
>
> Basically, AlwaysPreTouch is apparently not used very often, so some
> use cases were missed, causing a guarantee to fail.
>
> The patch adds support for:
>   - parallel pretouch during (potentially concurrent) humongous object
> allocation.
>   - allocating CDS/archive regions during archive creation. In this case
> the code uses only one thread because archive regions are always
> allocated region by region anyway.
>   - mapping in a memory range covered by a CDS archive
>
> All these cases are covered by new regression tests.
>
> CR:
> https://bugs.openjdk.java.net/browse/JDK-8169703
> Webrev:
> http://cr.openjdk.java.net/~tschatzl/8169703/webrev/
> Testing:
> runtime tier0-5 tests with -XX:+AlwaysPreTouch, new tests
>
> Thanks,
>    Thomas
>




More information about the hotspot-gc-dev mailing list