Re: pgsql: Use stdbool.h if suitable - Mailing list pgsql-committers

From Andres Freund
Subject Re: pgsql: Use stdbool.h if suitable
Date
Msg-id 20180323062232.hj2kokl54il72ool@alap3.anarazel.de
Whole thread Raw
In response to Re: pgsql: Use stdbool.h if suitable  (Andres Freund <andres@anarazel.de>)
List pgsql-committers
On 2018-03-22 18:25:01 -0700, Andres Freund wrote:
> Hi,
> 
> On 2018-03-23 00:42:39 +0000, Peter Eisentraut wrote:
> > Use stdbool.h if suitable
> > 
> > Using the standard bool type provided by C allows some recent compilers
> > and debuggers to give better diagnostics.  Also, some extension code and
> > third-party headers are increasingly pulling in stdbool.h, so it's
> > probably saner if everyone uses the same definition.
> 
> Hah. I was setting up LLVM buildfarm animals that run with LLVM
> assertions enabled, and was getting confused why they're failing after
> I'd run a check-world locally with assertions enabled.
> 
> Turns out this broke it. I'll fix it. Just amazed that we've gone weeks
> without a change breaking my JIT work, and then within hours of it
> getting it got broken.  Turns out LLVM represents a stdbool.h boolean as
> an i1 (int with 1 bit) rather than an i8 (int with 8 bits), which makes
> sense given the desired boolean behaviour (still takes 8 bits of
> storage).

Pushed a fix. There's now an array of llvm with assertion using
buildfarm animals. The one I was trying to setup, shows the before/after
now:
https://buildfarm.postgresql.org/cgi-bin/show_history.pl?nm=xenodermus&br=HEAD

- Andres


pgsql-committers by date:

Previous
From: Andres Freund
Date:
Subject: pgsql: Adapt expression JIT to stdbool.h introduction.
Next
From: Peter Eisentraut
Date:
Subject: pgsql: pg_resetwal: Add simple test suite