Joshua Tolley <eggyknap(at)gmail(dot)com> writes:
> Makes sense on both counts. Thanks for the help. How does the attached look?
Applied with minor corrections, mainly around the state save/restore
logic. I also put in some code to fix the memory leak noted by Tim Bunce,
but am waiting for some confirmation that it's right before
back-patching the pre-existing bug of the same ilk.
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Hitoshi Harada||Date: 2009-11-29 04:55:11|
|Subject: Frame level memory in WindowAgg?|
|Previous:||From: David Rowley||Date: 2009-11-29 02:54:20|
|Subject: Re: patch - per-tablespace random_page_cost/seq_page_cost|