[IMPALA-3360] Unroll loops / replace types in filter logic in PHJ::ProcessBuildBatch()

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: Impala 2.6.0
• Fix Version/s: Impala 2.11.0
• Component/s: Backend
• Labels: codegen
• Target Version: Product Backlog
• Assignee: Thomas Tauber-Marshall
• Reporter: Henry Robinson
• Created: 14/Apr/16 20:09
• Updated: 26/Sep/17 17:24
• Resolved: 26/Sep/17 17:24

Description

This code can be optimized for codegen:

for (const FilterContext& ctx: filters_) {
        if (ctx.local_bloom_filter == NULL) continue;
        void* e = ctx.expr->GetValue(build_row);
        uint32_t filter_hash = RawValue::GetHashValue(e, ctx.expr->root()->type(),
            RuntimeFilterBank::DefaultHashSeed());
        ctx.local_bloom_filter->Insert(filter_hash);
      }

Note also that we know ahead of time whether ctx.local_bloom_filter will be NULL, so we could prepare a list of non-null filters and only iterate over that.

Issue Links

• blocks: IMPALA-3838 Codegen HdfsParquetScanner::EvalRuntimeFilters()
• is related to: IMPALA-3837 Filtering at scan node using bloom filters is 2x slower than filtering at join