Questions around codeheaps, codeblobs and code locality

[keita abdoul-kader]

Hello dean,
We did try -XX:+UseTransparentHugePages in production and saw no
performance improvements (although i check if there was a reduction in iTLB
misses).
Also i should probably note that we are still using java 8 and as such
still using the unify/non segmented code cache.

On Mon, Jun 3, 2019 at 2:55 PM <dean.long at oracle.com> wrote:

> I wonder if something like -XX:+UseTransparentHugePages would help with
> the iTLB issue.
>
> dl
>
> On 6/3/19 12:14 PM, keita abdoul-kader wrote:
> > Hello,
> >
> > I have been looking into the code cache layout and ways to reduce the
> high
> > amount of iTLB we are observing on some of our workloads. It seems that
> > after the works on segmented codeHeaps, the plan was to add support for
> non
> > contiguous codeBlobs as the next step to improve code locality in the
> code
> > cache. is the work  on non contiguous codeBlobs still active ?
> >
> > Also, does the hotspot has support for any form of code compaction ?
> > Compacting the code heap is mentioned in one of the comment in
> > relocInfo.hpp, but browsing the source code it seems that once  the
> > codeBlobs is filled with the content of a codeBuffer, they are never
> moved
> > in a way to reduce codeheap fragmentation.
> >
> >   In general, is there an umbrella project or a set of open task around
> > improving code locality ?  Please note that most of my observation
> driving
> > this come from spark-sql workloads. Since spark-sql generates dynamically
> > generate a lots of code, i might simply be observing a pathological case.
> >
> > Thanks
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/attachments/20190603/f0ddb2e1/attachment.html>