Dear #Postgresql fellows, latest release of 9.6.12 has regressions. It's not only forcing existence of Perl (since exactly this version), but also introduced major build issue:
"configure: error: compiler does not report undeclared identifiers"
-
Show this thread
-
Exactly same issue with 10.7 and most likely 11.2.
@PostgreSQL1 reply 0 retweets 1 likeShow this thread -
I really think that big, and responsible players should test their builds on all major platforms. There's
#Linux but#FreeBSD is tier-1 player as well since at least a decade. Seriously important to at least not ignore other big players@PostgreSQL0 replies 7 retweets 7 likesShow this thread -
This Tweet is unavailable.
-
We do test on FreeBSD! Not sure what happened here, will check.
1 reply 0 retweets 1 like -
Replying to @MengTangmu @PostgreSQL
I know because my definition has explicit --without-perl setting, which crashed on .12 - https://github.com/VerKnowSys/sofin-definitions/blob/stable/definitions/postgresql96.def#L8 … After I provided perl as requirement (and getting rid of that --without-perl option), I got that cryptic compiler failure I quoted.
2 replies 0 retweets 0 likes -
Your problem is nothing to do with perl or freebsd. You provided a dozen random compiler options in CFLAGS, one of which (-w) breaks configure. See docs for COPT if you really want to do that. PS: your quoted comment of mine is years out of date.
2 replies 0 retweets 2 likes -
NOTE: My compiler options aren't random. I'm just applying additional hardening features (provided by my production system which is
#HardenedBSD), on Postgresql builds I provide :)1 reply 0 retweets 0 likes -
Having a bunch of -Wblah and -Wno-blah, followed by a -w (which unconditionally suppresses all warnings) seems pretty random to me. Also many of those options duplicate configure's or PG's own choices. Please also remove the outdated quote re. thread safety.
1 reply 0 retweets 0 likes -
It's likely caused by amount of warnings thrown from build process of Postgresql on BSDs? ;) But jokes aside :) I removed outdated comment, but just to be sure: <RhodiumToad> --enable-thread-safety only affects libpq, yes is this part is still relevant? :)
1 reply 0 retweets 0 likes
I do all of my postgres development on freebsd, with -Werror in my COPT flags. --enable-thread-safety is still only for the client-side code, yes.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.