[SPARK-16196][SQL] Codegen in-memory scan with ColumnarBatches#13899

Closed

andrewor14 wants to merge 22 commits intoapache:masterfrom

andrewor14:speedup-cache

Commits on Jun 17, 2016

Move it
Andrew Or
committed
Add benchmark code
Andrew Or
committed

Commits on Jun 20, 2016

backup
Andrew Or
committed
Narrow benchmarked code + add back old scan code
Andrew Or
committed

Commits on Jun 21, 2016

Fix benchmark to time only the read path
Andrew Or
committed
First working impl. of ColumnarBatch based caching
Andrew Or
committed

Commits on Jun 22, 2016

Always enable codegen and vectorized hashmap
Andrew Or
committed
Don't benchmark aggregate
Andrew Or
committed
Codegen memory scan using ColumnarBatches
Andrew Or
committed
Clean up the code a little
Andrew Or
committed
Merge branch 'master' of github.com:apache/spark into speedup-cache
Andrew Or
committed
Clean up a little more
Andrew Or
committed

Commits on Jun 23, 2016

Generate code for write path to support other types
Andrew Or
committed
Merge branch 'master' of github.com:apache/spark into speedup-cache
Andrew Or
committed
Move cache benchmark to new file
Andrew Or
committed
Abstract codegen code into ColumnarBatchScan
Andrew Or
committed

Commits on Jun 24, 2016

Introduce CACHE_CODEGEN config to reduce dup code
Andrew Or
committed
Add some tests for InMemoryRelation
Andrew Or
committed
Add some tests for InMemoryRelation
Andrew Or
committed
Fix InMemoryColumnarQuerySuite
Andrew Or
committed
Clean up code: abstract CachedBatch and ColumnarBatch
Andrew Or
committed
Add end-to-end benchmark, including write path
Andrew Or
committed