From: | Hannu Krosing <hannu(at)trust(dot)ee> |
---|---|
To: | "Thomas G(dot) Lockhart" <lockhart(at)alumni(dot)caltech(dot)edu> |
Cc: | Jan Wieck <jwieck(at)debis(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, jose(at)sferacarta(dot)com, pgsql-hackers(at)postgreSQL(dot)org |
Subject: | Re: [HACKERS] PL/pgSQL a great procedural language for PostgreSQL |
Date: | 1998-11-06 09:20:23 |
Message-ID: | 3642BF57.AFC4C48E@trust.ee |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Thomas G. Lockhart wrote:
>
> > For arguments passed by value it
> > is totally impossible to know if it's a NULL by looking at
> > the value itself.
>
> Is this a sufficient reason to move toward having all user-oriented data
> types be pass-by-reference? If we don't do that, then we would need to
> pass a null flag for every parameter, or an array of flags, or have some
> global array which contains the null flags (which we might get away with
> since Postgres is pretty much single threaded and non-reentrant in the
> backend code). What other options might there be?
Perhaps make the isNull flag a bitmap ?
In case it is an int (I haven't looked), we get room for flagging 32
arguments
Also I hope that most current code will run unchanged, in case it tests
for
0 / not 0
----------------
Hannu
From | Date | Subject | |
---|---|---|---|
Next Message | Zsolt Varga | 1998-11-06 09:33:48 | linux libc6 & pgsql 6.4 |
Previous Message | Oleg Broytmann | 1998-11-06 09:17:45 | v6.4 - duplicate entries in announce |