Re: C function migration from 9.2 to 9.5

Thanks for the responses.

I currently have Postgres installed through apt-get, so I don't think it is
old libraries/header files. Just to double check, I spun up a "blank" VM
with Ubuntu 14.04 (Trusty), which is what we're currently running in our
dev environment, and apt-get installed postgresql-9.3 and
postgresql-server-dev-9.3 (v 9.3.11). I've tried this with 9.5 (from the
PGDG repos) as well, with the same result.

Copied over the two attached files (test.c and test.sql) and basically ran:

> createdb test

> gcc -I/usr/include/postgresql/9.5/server -fPIC -c test.c

This throws a compiler warning on the cast from heap_form_tuple to
HeapTuple, but IIRC it's always done that so not an error?

> gcc -shared -o test.so test.o

> psql -d test -f test.sql
CREATE TYPE
CREATE FUNCTION
psql:test.sql:9: server closed the connection unexpectedly
This probably means the server terminated abnormally
before or while processing the request.
psql:test.sql:9: connection to server was lost

The logs don't seem all that useful:
2016-03-03 23:03:13 UTC STATEMENT: CREATE TYPE sessionType AS ( a int4, b
int4, c int4);
2016-03-03 23:03:13 UTC LOG: server process (PID 9988) was terminated by
signal 11: Segmentation fault
2016-03-03 23:03:13 UTC DETAIL: Failed process was running: SELECT
sessionData();
2016-03-03 23:03:13 UTC LOG: terminating any other active server processes
2016-03-03 23:03:13 UTC WARNING: terminating connection because of crash
of another server process
2016-03-03 23:03:13 UTC DETAIL: The postmaster has commanded this server
process to roll back the current transaction and exit, because another
server process exited abnormally and possibly corrupted shared memory.
2016-03-03 23:03:13 UTC HINT: In a moment you should be able to reconnect
to the database and repeat your command.
2016-03-03 23:03:13 UTC LOG: all server processes terminated;
reinitializing

I haven't attempted to run the stack backtrace with gdb yet (requires a bit
more setup), but that will be the next step if you guys can't see any
obvious mistakes that I've made.

Regards,

Michael A.

On Thu, Mar 3, 2016 at 5:50 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> Michael Omotayo Akinde <michaeloa(at)met(dot)no> writes:
> > We've been having a Postgresql database with some custom C functionality
> > happily running for many years now. It's been running on 9.2, and we wish
> > to upgrade this to the latest version. However, we're seeing some issues
> > with the database process crashing each time.
>
> Like Pavel, I can't see anything wrong with that code --- it's not quite
> according to PG project style, but it certainly looks like it does what
> it needs to. I think he's right to suspect some inconsistency in your
> coding environment. One concrete idea worth considering is that maybe
> you are compiling with headers that postdate commit 3f8c8e3c6 et al and
> trying to use the code in a database that predates that. That'd result
> in successfully compiling a call to a nonexistent core function, which
> might end up as a crash depending on what your dynamic linker does
> about it.
>
> What exactly does the crash look like --- anything interesting in the
> postmaster log? (If your logging setup fails to capture postmaster
> stderr, now would be a good time to fix that.) Have you tried to
> get a back-trace with gdb?
>
> regards, tom lane
>
> PS: for reference, this is the patch I'm wondering about:
>
> Author: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
> Branch: master Release: REL9_4_BR [3f8c8e3c6] 2014-05-01 15:19:06 -0400
> Branch: REL9_3_STABLE Release: REL9_3_5 [b72e90bc3] 2014-05-01 15:19:10
> -0400
> Branch: REL9_2_STABLE Release: REL9_2_9 [8c43980a1] 2014-05-01 15:19:14
> -0400
> Branch: REL9_1_STABLE Release: REL9_1_14 [db1fdc945] 2014-05-01 15:19:17
> -0400
> Branch: REL9_0_STABLE Release: REL9_0_18 [7a4f114f3] 2014-05-01 15:19:20
> -0400
> Branch: REL8_4_STABLE Release: REL8_4_22 [70debcf09] 2014-05-01 15:19:23
> -0400
>
> Fix failure to detoast fields in composite elements of structured
> types.
>