RFR 8005698 : Handle Frequent HashMap Collisions with Balanced Trees
Alan.Bateman at oracle.com
Fri May 31 14:10:15 UTC 2013
On 29/05/2013 21:47, Brent Christian wrote:
> Updated webrev is here:
> It contains the following changes from Mike's review:
> * HashMap.comparableClassFor(): corrected reference to TreeBin docs
> * fixed @run tag in InPlaceOpsCollisions.java
> * Hashtable & HashMap: hashSeed made final, initialized in
> constructor, UNSAFE restored and used in readObject().
I've read through the latest webrev, overall very good work.
I had to read splitTreeBin a few times to convince myself as to how the
keys re-hash. I wonder if would be helpful to beef up the comment on
this method. It's good to see a test that specifically targets this.
Otherwise I didn't see anything obviously wrong. Minor alignment issue
with the comments on Holder.
Assuming Doug and Mike are okay with this then I suggest we try to get
it into jdk8/tl soon so that it has a few days bake time before Lana
grabs the changes from jdk8/tl for b94. I think Mike is going to sponsor
More information about the core-libs-dev