On Sat, 24 Jul 2004 17:28:42 -0700, Byron Servies <bservies@pacang.com> wrote:
> 
> 
> 
> On Jul 24, 2004, at 7:50 AM, Andy Ruhl wrote:
> 
> > On Fri, 23 Jul 2004 21:24:54 -0700, Byron Servies
> > <bservies@pacang.com> wrote:
> >> On Jul 23, 2004, at 12:54 PM, Andy Ruhl wrote:
> >>
> >>> o postgresql from pkgsrc won't compile due to some CPU limitation. I
> >>> set an configure option to disable spinlocks, and it would finish
> >>> building without errors, but it would not actually build the
> >>> postmaster executable. Weird. Seems like 2 problems: 1. It won't
> >>> build
> >>> right. 2. The package build should fail. I compiled it from source,
> >>> but it complains when I try to do an initdb. MySQL built just fine
> >>> but
> >>> I didn't try to use it. But I REALLY don't want to use MySQL.
> >>
> >> See
> >>
> >> <http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=25473>
> >>
> >> I have been running postgres with these changes for months.  It even
> >> passes most of the tests.
> >
> > Hopefully I'm not being dense here, but I'm not finding a file called
> > dynaloader.c. I'll do some grepping to see if I can find out where
> > this stuff is handled and see if I can make the change.
> >
> 
> Sorry.  It's dynloader.c.  No 'a'.
> 
> /usr/pkgsrc/databases/postgresql-server/work/postgresql-7.3.4/src/
> backend/port/dynloader/netbsd.c
> 
> During a build this gets copied into src/backend/port/dynloader.c

Thanks, I'll give it a look I guess this means I'm going to extract
the source in the /usr/pkgsrc/distfiles directory, make the change,
and then turn off checksum checking or something... Either that or
I'll just manage my own source tree and not deal with it with pkgsrc.
Or I'm sure there is some make target I'm missing here that I can use
(because there are a ton of them).

Andy