Build issue on `x86_64-darwin`

[nepet]

Building Core Lightning (state v23.11) on x86_64-darwin returns the following error.

ccan/ccan/fdpass/fdpass.c:16:8: error: variable length array folded to constant array as an extension [-Werror,-Wgnu-folding-constant] char buf[CMSG_SPACE(sizeof(fd))];

[vincenzopalazzo]

We should fix this for the next version. However, I have no device to test it out so I will leave space for someone to take this otherwise when we are close to the next release I will try to fix it 😄

[cdecker]

If @rustyrussell can help us interpret the error message, and what it is complaining about I can give it a shot.

[vincenzopalazzo]

Maybe someone that have access to the machine can test #6950

I was not able to reproduce it so this is just a blinded patch

[chrisguida]

@ddustin you have an intel mac, right?

[ddustin]

I'm on M2 now

[chrisguida]

@ddustin Darn, was hoping you still had it around somewhere...

[vincenzopalazzo]

I think I was wrong.

I was chatting with some friends who are working on GCC and he pointed me to this libuv/libuv#454

I think this is a good point, could we check before the compiler version that @nepet is using? and then this problem happens with clang? or with a new version of GCC?

[nepet]

Mh I just forwarded the issue from the nixpkg builds for v23.11. Never built it myself on a mac. I got an intel mac though, so I can try to build it if you got a potential fix.

gcc --version gives:

Apple clang version 15.0.0 (clang-1500.1.0.2.5)
Target: x86_64-apple-darwin22.6.0

[cdecker]

I think I was wrong.

I was chatting with some friends who are working on GCC and he pointed me to this libuv/libuv#454

I think this is a good point, could we check before the compiler version that @nepet is using? and then this problem happens with clang? or with a new version of GCC?

Good find 👍