| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Andres Freund <andres(at)2ndquadrant(dot)com> |
| Cc: | Michael Paquier <michael(dot)paquier(at)gmail(dot)com>, Fujii Masao <masao(dot)fujii(at)gmail(dot)com>, Fujii Masao <fujii(at)postgresql(dot)org>, pgsql-committers <pgsql-committers(at)postgresql(dot)org> |
| Subject: | Re: pgsql: Move pg_lzcompress.c to src/common. |
| Date: | 2014-12-25 18:31:52 |
| Message-ID: | 25733.1419532312@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-committers |
Andres Freund <andres(at)2ndquadrant(dot)com> writes:
> On 2014-12-25 23:46:52 +0900, Michael Paquier wrote:
>> Urgh.. Would that mean keeping a local copy of SET_VARSIZE_4B_C and
>> VARSIZE_4B if this hack is not used? Looking at the git history, I am
>> seeing similar things in 2008 where pg_crc stuff was moved to src/port
>> (5c9c08d).
> Surely not. It seems like a much better idea to not have lzcompress deal
> with varlena at all but pass that responsibility one layer upwards. That
> shouldn't be very hard.
All files in src/common should be using the coding pattern
#ifndef FRONTEND
#include "postgres.h"
#else
#include "postgres_fe.h"
#endif
and as noted, it's going to take additional refactoring to make it
possible to do that in pg_lzcompress. In view of all the potentially
unportable stuff Andres committed today, we need the buildfarm to be
at full strength; so I've reverted this commit until it can be worked
out a bit better.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2014-12-25 18:49:01 | pgsql: Blindly fix a dtrace probe in lwlock.c for a removed local varia |
| Previous Message | Tom Lane | 2014-12-25 18:25:27 | pgsql: Temporarily revert "Move pg_lzcompress.c to src/common." |