Bug #72993 | my_hash_sort_utf8 can be optimized if table has many fields | ||
---|---|---|---|
Submitted: | 13 Jun 2014 1:03 | Modified: | 11 Aug 2014 14:37 |
Reporter: | Hao Liu | Email Updates: | |
Status: | Not a Bug | Impact on me: | |
Category: | MySQL Server | Severity: | S3 (Non-critical) |
Version: | 5.6 5.5 | OS: | Any |
Assigned to: | CPU Architecture: | Any |
[13 Jun 2014 1:03]
Hao Liu
[13 Jun 2014 1:12]
Hao Liu
attached patch. I think it can be optimized further.
Attachment: bug#72993.diff (application/octet-stream, text), 4.22 KiB.
[11 Aug 2014 14:37]
MySQL Verification Team
murmu's function is basically a hash function that's operating on a block of bytes. And that is all of its functionality. On the other hand, my_hash_sort_utf8() is doing a collation aware string hashing. So it's comparing apples to apple seeds. We strongly rely on collation-aware hashing as we have to support hundreds of different collations. But, thank you for trying to improve a performance of our server.