[10?] RFR: 8193128: Reduce number of implementation classes returned by List/Set/Map.of() (original) (raw)

Claes Redestad claes.redestad at oracle.com
Wed Dec 6 20:21:55 UTC 2017

Hi,

please help review this patch to consolidate the number of implementation classes returned by the static collection factories:

http://cr.openjdk.java.net/~redestad/8193128/open.00/

I set out to explore options for addressing small inefficiencies we've been running into, the latest after replacing use of @Stable arrays in java.lang.invoke with List.of() (JDK-8184777):

Microbenchmark: http://cr.openjdk.java.net/~redestad/scratch/less_immutables/ListMorphism.java http://cr.openjdk.java.net/~redestad/scratch/less_immutables/benchmarks.jar

The benchmark explores how call-sites behave when performing some naive operations on a few different Lists.

For every benchmark using List.of() there's a variant using ArrayList for comparison:

Baseline:

Benchmark                             Mode  Cnt    Score Error   Units ListMorphism.finalGetFromArrayList   thrpt   25   92.659 ±  3.058 ops/us ListMorphism.finalGetFromList        thrpt   25  335.245 ±  0.244 ops/us

3.6x

ListMorphism.finalSumSizesArrayList  thrpt   25  245.020 ±  0.106 ops/us ListMorphism.finalSumSizesList       thrpt   25  335.107 ±  0.439 ops/us

1.4x

ListMorphism.getFromArrayList        thrpt   25   70.343 ±  0.972 ops/us ListMorphism.getFromList             thrpt   25   37.121 ±  0.135 ops/us

0.53x

ListMorphism.getFromArrayList12      thrpt   25 109.890 ±  0.286  ops/us ListMorphism.getFromList12           thrpt   25  109.552 ± 6.972  ops/us

1.0x

ListMorphism.sumSizesArrayList       thrpt   25  131.467 ±  4.672 ops/us ListMorphism.sumSizesList            thrpt   25   57.877 ±  0.102 ops/us

0.45x

ListMorphism.sumSizesArrayList12     thrpt   25 208.652 ± 11.294  ops/us ListMorphism.sumSizesList12          thrpt   25  227.269 ± 0.961  ops/us

1.1x

The good: When dealing with List literals (the final* benchmarks), List.of() allows really nice speed-ups compared to ArrayList.

The bad: When not used as a literal, List.of() implementations at call-sites can cause a substantial slowdown (megamorphic dispatch)

The ugly:

After some explorations[1], I narrowed in on the following experiment: http://cr.openjdk.java.net/~redestad/scratch/less_immutables/webrev/

Basically: Merge List1 and List2 into a single class, List12, merge List0 into ListN (List0 has a singleton instance, so might as well be an instance of ListN). Same for Set0,1,2,N. Map0 is merged into MapN.

This strikes a balance between throughput, footprint and slightly better startup/warmup behavior.

According to jol estimates[2] the size of List12 is the same as both List1 and List2 in the current JDK implementation. Set12 is footprint neutral compared to Set2 on all platforms but lose a few bytes on 64-bit VMs compared to Set1.

Benchmark                             Mode  Cnt    Score   Error Units ListMorphism.finalGetFromArrayList   thrpt   25   93.046 ± 0.569 ops/us ListMorphism.finalGetFromList        thrpt   25  335.280 ± 0.154 ops/us

3.6x

ListMorphism.finalSumSizesArrayList  thrpt   25  244.595 ± 1.085 ops/us ListMorphism.finalSumSizesList       thrpt   25  303.351 ± 0.182 ops/us

1.2x

ListMorphism.getFromArrayList        thrpt   25   70.631 ± 0.070 ops/us ListMorphism.getFromList             thrpt   25   73.258 ± 2.955 ops/us

1.04x

ListMorphism.getFromArrayList12      thrpt   25 109.921 ± 0.096  ops/us ListMorphism.getFromList12           thrpt   25  127.392 ± 0.088  ops/us

1.16x

ListMorphism.sumSizesArrayList       thrpt   25  131.393 ± 4.882 ops/us ListMorphism.sumSizesList            thrpt   25  107.686 ± 5.286 ops/us

0.82x

ListMorphism.sumSizesArrayList12     thrpt   25  212.350 ± 0.134  ops/us ListMorphism.sumSizesList12          thrpt   25  198.778 ± 0.479  ops/us

0.94x

The experiment has a flag to change number of specialized List/Set/Map classes (-Djdk.ImmutableCollections.specializations=0|1|2, default=2).

1 specialization (List1 + ListN, Set1 + SetN) is more or less the same as 2, some ~1-2% improvements, mainly in sumSizes micros.

0 specializations (List-, Set, MapN only) achieves a small increase in throughput on some micros by ensuring callsites stay monomorphic, but it's not very substantial by my measures (~5%, but mostly in sumSizes micros).

Keeping the footprint more or less the same, while evening out a few rough edges and improving startup and static footprint seems like the overall best option. An alternative would be to keep the experimental flag, but I don't think a 5% gain on micros warrants the extra maintenance burden and testing that entails.

The proposed patch is more or less this experiment with 2 specializations, but having removed the flag and code movement needed to support it removed (along with a fix in the writeReplace methods of List12/Set12)

Thanks!

/Claes

[1] Older experiments: http://cr.openjdk.java.net/~redestad/immutable/list12N.0/  -- simply merge L0 into LN (still have L1, L2 and LN) - nothing really changed, though

http://cr.openjdk.java.net/~redestad/immutable/list1N.0/  -- L0 merged into LN, drop L2. Substantial performance boost on micros. Footprint drop for 2-element lists.

http://cr.openjdk.java.net/~redestad/immutable/listNdouble.0/  -- L0+L1+L2+LN merged into one implementation. Same footprint with a single class, but loses a lot on throughput in micros.

http://cr.openjdk.java.net/~redestad/immutable/listNsingle.0/  -- L0+L1+LN merged, drop L2. Simplification of the previous. Like the list1N.0 experiment in footprint, but a loss in throughput on all measures.

No approach seemed a win across the board here: we either had to accept a footprint reduction, or see throughput suffer dramatically.

[2] http://openjdk.java.net/projects/code-tools/jol/

More information about the core-libs-dev mailing list