Re: plpgsql function startup-time improvements
Pavel Stehule <pavel.stehule@gmail.com>
From: Pavel Stehule <pavel.stehule@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2018-01-25T18:58:03Z
Lists: pgsql-hackers
Hi 2018-01-25 0:16 GMT+01:00 Tom Lane <tgl@sss.pgh.pa.us>: > Pavel Stehule <pavel.stehule@gmail.com> writes: > > please, can you rebase all three patches necessary for patching? > > Done. These now need to be applied over > https://www.postgresql.org/message-id/833.1516834367@sss.pgh.pa.us Thank you I checked it 1. there are no problem with patching 2. all tests passed 3. I can confirm so some best case speedup is about 15-20%. Worst case is hard to test - but these changes should not be slower than current master. 4. faster-plpgsql-datum-setup-2.patch is simple patch with few lines of new code 5. plpgsql-promises-2.patch is little bit more complex, but still it is simple 6. the code is well formatted and well commented 7. there are not new test and code, what is not problem - these patches doesn't add any new feature I'll mark these patches as ready for commiter Regards Pavel > > > regards, tom lane > >
Commits
-
Speed up plpgsql trigger startup by introducing "promises".
- fd333bc763ea 11.0 landed
-
Speed up plpgsql function startup by doing fewer pallocs.
- 40301c1c8bcb 11.0 landed