Thread: Warp
It runs on OS/2 Warp 4 too (Postgres V8 RC3). -- / /
Eric Caloone wrote: > It runs on OS/2 Warp 4 too (Postgres V8 RC3). That is pretty hard to believe, since the code we distribute contains neither shared library support nor spinlock support nor any of the other port specific tweaks for OS/2. And someone else is currently working on porting to OS/2. What's up with that? -- Peter Eisentraut http://developer.postgresql.org/~petere/
Peter Eisentraut wrote:
> Eric Caloone wrote:
> > It runs on OS/2 Warp 4 too (Postgres V8 RC3).
>
> That is pretty hard to believe, since the code we distribute contains
> neither shared library support nor spinlock support nor any of the
> other port specific tweaks for OS/2. And someone else is currently
> working on porting to OS/2. What's up with that?
Someone posted an 8.0rc4 binary and it is mentioned in the current FAQ,
so yea, I think it works.
http://hobbes.nmsu.edu/cgi-bin/h-search?sh=1&button=Search&key=postgreSQL&stype=all&sort=type&dir=%2F
I assume the source is included.
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce Momjian <pgman@candle.pha.pa.us> writes:
> Peter Eisentraut wrote:
>> Eric Caloone wrote:
>>> It runs on OS/2 Warp 4 too (Postgres V8 RC3).
>>
>> That is pretty hard to believe, since the code we distribute contains
>> neither shared library support nor spinlock support nor any of the
>> other port specific tweaks for OS/2. And someone else is currently
>> working on porting to OS/2. What's up with that?
> Someone posted an 8.0rc4 binary and it is mentioned in the current FAQ,
> so yea, I think it works.
Posting a binary doesn't mean that they didn't have to hack the source.
As of 8.0 the spinlock support should be driven by CPU type not OS type,
so as long as you build using a compiler that defines standard CPU
symbols it should work. In particular I'd expect a gcc build on OS/2 to
not have any issues about spinlocks.
Shared libraries are another story --- there is no way we'd build shared
libraries successfully without a Makefile.os2 and some additions to
Makefile.shlib. Given the importance of plpgsql I don't think we could
consider a build with --disable-shared to be a "supported platform".
regards, tom lane
Tom Lane wrote: > Bruce Momjian <pgman@candle.pha.pa.us> writes: > > Peter Eisentraut wrote: > >> Eric Caloone wrote: > >>> It runs on OS/2 Warp 4 too (Postgres V8 RC3). > >> > >> That is pretty hard to believe, since the code we distribute contains > >> neither shared library support nor spinlock support nor any of the > >> other port specific tweaks for OS/2. And someone else is currently > >> working on porting to OS/2. What's up with that? > > > Someone posted an 8.0rc4 binary and it is mentioned in the current FAQ, > > so yea, I think it works. > > Posting a binary doesn't mean that they didn't have to hack the source. > > As of 8.0 the spinlock support should be driven by CPU type not OS type, > so as long as you build using a compiler that defines standard CPU > symbols it should work. In particular I'd expect a gcc build on OS/2 to > not have any issues about spinlocks. > > Shared libraries are another story --- there is no way we'd build shared > libraries successfully without a Makefile.os2 and some additions to > Makefile.shlib. Given the importance of plpgsql I don't think we could > consider a build with --disable-shared to be a "supported platform". I assume OS/2 is like our Novell port in that it doesn't compile using our CVS source but only with modified source. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce Momjian <pgman@candle.pha.pa.us> writes:
> Tom Lane wrote:
>> Posting a binary doesn't mean that they didn't have to hack the source.
> I assume OS/2 is like our Novell port in that it doesn't compile using
> our CVS source but only with modified source.
We don't have a Novell port --- it is not, and should not be, listed as
a "supported platform" if it requires hacking the source.
There may be a Novell port out there, but I don't take any
responsibility for it ;-)
regards, tom lane
Tom Lane wrote: > Bruce Momjian <pgman@candle.pha.pa.us> writes: > > Tom Lane wrote: > >> Posting a binary doesn't mean that they didn't have to hack the source. > > > I assume OS/2 is like our Novell port in that it doesn't compile using > > our CVS source but only with modified source. > > We don't have a Novell port --- it is not, and should not be, listed as > a "supported platform" if it requires hacking the source. > > There may be a Novell port out there, but I don't take any > responsibility for it ;-) Right, I assume the same for OS/2. It is mentioned in the FAQ and has a URL. That's as far as we can go until the OS/2 port person submits information on what changes he had to make to get it to work and we incorporate his changes in our code. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Hi That would be me doing the OS/2 port. I was waiting until the GA 8.0 release before I submitted the patches for the makefiles and code for the os/2 port. It does support shared libraries, and the spinlock stuff and signals are provided by the GCC 3.3.5 LIBC. I have one major issue with character code conversion to resolve in order to pass the regression tests, as well as a compiler bug that messes up the printf'ing of long long variables. Lorne Sunley -- ----------------------------------------------------------- lsunley@mb.sympatico.ca -----------------------------------------------------------