On Sun, Jul 19, 2015 at 04:32:16PM -0400, Noah Misch wrote:
> On Sun, Jul 19, 2015 at 04:22:47PM -0400, Tom Lane wrote:
> > More: if the compiler does have a bug like that, how much confidence can
> > we have, really, that there are no other miscompiled places and won't be
> > any in the future? I
On Sun, Jul 19, 2015 at 04:22:47PM -0400, Tom Lane wrote:
> Andres Freund writes:
> > On July 19, 2015 9:50:33 PM GMT+02:00, Noah Misch wrote:
> >> I propose to expand the gin_private.h "#ifdef PG_USE_INLINE" test to
> >> exclude xlc 32-bit configurations. The last 32-bit AIX kernel exited
> >>
Andres Freund writes:
> On July 19, 2015 9:50:33 PM GMT+02:00, Noah Misch wrote:
>> I propose to expand the gin_private.h "#ifdef PG_USE_INLINE" test to
>> exclude xlc 32-bit configurations. The last 32-bit AIX kernel exited
>> support on 2012-04-30.
> I vote to simply error out in that case th
On July 19, 2015 9:50:33 PM GMT+02:00, Noah Misch wrote:
>In a 32-bit build, xlc 12.1 for AIX miscompiles three inline expansions
>of
>ginCompareItemPointers(), all in ginget.c (line numbers per commit
>9aa6634):
>
>739Assert(!ItemPointerIsValid(&entry->curItem) ||
>740 ginCompar
