Jim Apple has posted comments on this change. Change subject: IMPALA-2809: Improve ByteSwap with builtin function or SSE or AVX2. ......................................................................
Patch Set 15: (2 comments) http://gerrit.cloudera.org:8080/#/c/3081/15/be/src/benchmarks/bswap-benchmark.cc File be/src/benchmarks/bswap-benchmark.cc: Please also add tests to bit-util-test.cc Line 124: void TestSSE42Swap(int batch_size, void* d) { When I asked earlier if we care about FIXED_LEN_SIZE, I wasn't trying to be Socratic - I was actually asking. It sounds like you weren't sure either. After some digging, it looks to me like we sometimes do care about unusual FIXED_LEN_SIZE. See parquet-common.h, which includes the comment: /// fixed_len_size can be less than sizeof(Decimal*Value) for space savings. This means /// that the value in the in-memory format has leading zeros or negative 1's. /// For example, precision 2 fits in 1 byte. All decimals stored as Decimal4Value /// will have 3 bytes of leading zeros, we will only store the interesting byte. This is oddly worded, and I wouldn't take it as gospel that "All decimals stored as Decimal4Value ..." That might not be how decimals work in decimal-value.h. It might be the case that we do actually need FIXED_LEN_SIZE different than 4,8,16. Please investigate. -- To view, visit http://gerrit.cloudera.org:8080/3081 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I392ed5a8d5683f30f161282c228c1aedd7b648c1 Gerrit-PatchSet: 15 Gerrit-Project: Impala Gerrit-Branch: cdh5-trunk Gerrit-Owner: Youwei Wang <[email protected]> Gerrit-Reviewer: Alex Behm <[email protected]> Gerrit-Reviewer: Jim Apple <[email protected]> Gerrit-Reviewer: Marcel Kornacker <[email protected]> Gerrit-Reviewer: Mostafa Mokhtar <[email protected]> Gerrit-Reviewer: Tim Armstrong <[email protected]> Gerrit-Reviewer: Youwei Wang <[email protected]> Gerrit-HasComments: Yes
