Here's an updated version of the patch I posted here:
Version 2 has the following changes:
1. It has been tested and debugged on Windows.
2. As suggested by Robert Haas, I have renamed the BKI scripts to
genbki.pl and Gen_fmgrtab.pl in the interest of not changing anything
without good reason.
3. Some cleanups.
4. Document dependency on Perl 5.8 for building from CVS, as suggested
by Tom Lane here:
I saw that there was recent discussion on moving some fields out of
pg_attribute in the thread entitled "Stats for inheritance trees". I
believe my patch, were it applied first, would make that easier, since
it generates BKI data for pg_attribute.
I'll add this to the commitfest soon, if what I've described is acceptable.
pgsql-hackers by date
|Next:||From: Kevin Grittner||Date: 2009-12-31 21:53:41|
|Subject: Re: A third lock method|
|Previous:||From: Kevin Grittner||Date: 2009-12-31 21:44:02|
|Subject: Re: Serializable Isolation without blocking|