mapleFU commented on code in PR #39403:
URL: https://github.com/apache/arrow/pull/39403#discussion_r1439050003
##########
cpp/src/arrow/util/bit_util.h:
##########
@@ -167,10 +167,8 @@ constexpr int64_t CoveringBytes(int64_t offset, int64_t
length) {
// Returns the 'num_bits' least-significant bits of 'v'.
static inline uint64_t TrailingBits(uint64_t v, int num_bits) {
- if (ARROW_PREDICT_FALSE(num_bits == 0)) return 0;
Review Comment:
> and i think that is the main reason for performance increase
As I benchmark. Without `-march=skylake`, removing the
`ARROW_PREDICT_FALSE(num_bits == 0)` would not make it faster or slower. And
as mentioned https://github.com/apache/arrow/pull/39403#issuecomment-1873378808
. It might affect the command generation later. Would you mind testing that?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
