Feedback-ID: id4a34324:Fastmail
Date: Wed, 18 Oct 2023 11:11:32 -0700
From: Andres Freund <andres@anarazel.de>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Ronan Dunklau <ronan.dunklau@aiven.io>,
	Devrim =?utf-8?B?R8O8bmTDvHo=?= <devrim@gunduz.org>,
	PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>,
	Fabien COELHO <coelho@cri.ensmp.fr>,
	Dmitry Dolgov <9erthalion6@gmail.com>
Subject: Re: LLVM 16 (opaque pointers)
Message-ID: <20231018181132.vux4gg3guv33lgaf@awork3.anarazel.de>
References: <CA+hUKGKNX_=f+1C4r06WETKTq0G4Z_7q4L4Fxn5WWpMycDj9Fw@mail.gmail.com>
 <5716086.DvuYhMxLoT@aivenlaptop>
 <CA+hUKGLxVtLCo3xD1OEnMSQbq04MKx1G_vQ5SZQNMBSFPR_TpA@mail.gmail.com>
 <2700884.mvXUDI8C0e@aivenlaptop>
 <CA+hUKGKnLnJnWrkr=4mSGhE5FuTK55FY15uULR7=zzc=wX4Nqw@mail.gmail.com>
 <CA+hUKGKtUB=-BQZd=u4BFZ-UJWi47fekwiYMSDXDZMyV+z1G-g@mail.gmail.com>
 <CA+hUKG+hopeRBGVwS4JE3uy1UO3TRnhF08HMiSpMrPODgtiw9Q@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <CA+hUKG+hopeRBGVwS4JE3uy1UO3TRnhF08HMiSpMrPODgtiw9Q@mail.gmail.com>
Archived-At: <https://www.postgresql.org/message-id/20231018181132.vux4gg3guv33lgaf%40awork3.anarazel.de>
Precedence: bulk

Hi,

On 2023-10-19 06:20:26 +1300, Thomas Munro wrote:
> On Thu, Oct 19, 2023 at 1:06 AM Thomas Munro <thomas.munro@gmail.com> wrote:
> > On Thu, Oct 19, 2023 at 12:02 AM Thomas Munro <thomas.munro@gmail.com> wrote:
> > > I pushed the first patch, for LLVM 16, and the build farm told me that
> > > some old LLVM versions don't like it.  The problem seems to be the
> > > function LLVMGlobalGetValueType().  I can see that that was only added
> > > to the C API in 2018, so it looks like I may need to back-port that
> > > (trivial) wrapper into our own llvmjit_wrap.cpp for LLVM < 8.
> >
> > Concretely something like the attached should probably fix it, but
> > it'll take me a little while to confirm that...
> 
> Pushed, after digging up some old LLVM skeletons to test, and those
> "old LLVM" animals are turning green now.  I went ahead and pushed the
> much smaller and simpler patch for LLVM 17.

I enabled a new set of buildfarm animals to test LLVM 16 and 17. I initially
forgot to disable them for 11, which means we'll have those failed build on 11
until they age out :/.

Particularly for the LLVM debug builds it'll take a fair bit to run on all
branches. Each branch takes about 3h.

Greetings,

Andres Freund