Re: adding new field - Mailing list pgsql-hackers

From ranbeer makin
Subject Re: adding new field
Date
Msg-id a0eedc000606060914h5b24928eu9c050df30f789fd2@mail.gmail.com
Whole thread Raw
In response to Re: adding new field  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: adding new field  (Martijn van Oosterhout <kleptog@svana.org>)
List pgsql-hackers
hey,
let me clarify something.....

1. Whenever you run a query, its result or some other thing gets cached. I want to flush this cache? Also let me know when this cache gets flushed automatically.

2. After adding a new field in Resdom structure and making necessary changes in outfuncs.c, createfuncs. etc files, I *sometimes* get garbage value for the field added. The field is initialiazed to zero in makefuns.c....

3. **IMP*** Is there something in postgres that, only for N number of fields, of a structure, the memory is allocated, i.e., do I need to modify this "N" somewhere to reflect the addtion of a new field.

And that *weird* thing is:

1. I run some query, didn't get results.
2. Then, I commented the part which I modified [ i.e., commented that new field in Resdom structure], again run the same query, got results.
3. Next, I uncommented that part, and ran the same query again. I GOT THE RESULTS.

That's how my prog is behaving, behaving in a *weird* way. It seems like some memory probs somewhere. I can't figure it out. Yes, I'm re-compiling and running initdb whenever I make some modifications in the code.

I'm using release 8.03 and due to some reasons can't shift to some other release.

Regds,
Ranbeer



On 6/6/06, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Martijn van Oosterhout <kleptog@svana.org> writes:
> On Mon, Jun 05, 2006 at 11:24:09PM +0530, ranbeer makin wrote:
>> 2. I have added a new field in RESDOM structure, made necessary
>> modifications in outfuncs.c, copy, read, equalfuncs.c ....but my prog is
>> behaving in a weird way, seems like some memory probs...What other
>> modifications needs to be done to reflect the addition of this new field?

> Did you remember to recompile *everything* affected? (--enable-depend
> is useful for this). You also have to initdb again.

Also, if you're working on a patch you hope to someday contribute, you
should be starting from CVS HEAD or some reasonable approximation of it.
Resdom disappeared more than a year ago:
http://archives.postgresql.org/pgsql-committers/2005-04/msg00060.php

                        regards, tom lane

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: AIX check in datetime.h
Next
From: Tom Lane
Date:
Subject: Re: AIX check in datetime.h