Marshall Dawson has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/32027 )
Change subject: Make common macros double-evaluation safe ......................................................................
Patch Set 11:
My crosstools are up to date (e.g. gcc at 8.2.0)...
Okay, this reads a lot like a compiler problem. I assume... ...We intentionally added the whole crossgcc thing to get rid of those issues.
Oh, I'm sorry for being slow. cbfstool.c gets compiled using $(HOSTCC), not any croosstools, so mentioning that was irrelevant.
Can you check what version is on the 17.04 where it works?
6.3.0 on 17.04. I found some more interesting discussion on compiler versions regarding solving the min/max problem in the kernel. I haven't found whether it was solved a different way vs. deciding on whether to raise the version requirement. But 4.6 might be the right minimum version. Middle of the discussion: https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg1633051.html
Do we need to support Ubuntu 14.04? It's 5 years old, and it looks like even official Ubuntu LTS support for it is running out this month.
That's what I'd said to Martin originally -- that the patch landed on the exact month/year the LTS ends, which is why I wasn't making a big deal of it. It's an inconvenience for me, but not the end of the world.
Can we consider this WontFix?
Probably.
(The other option is to #ifdef this out for host utilities, but I'm not really sure we should be held back by ancient compilers. Or maybe we can find a way to have utils built by crossgcc as well, at least as long as we have one for that architecture?)
I envision both those options being ugly. It seems like a better approach might be to decide that we have a minimum gcc version requirement, and be deliberate about what revision that is.