On Tue, Oct 28, 2025, at 3:09 PM, Joel Sherrill via Gcc wrote:
> I'm prone to want it included. If it becomes a problem, it can always be
> removed.
>
> One thing that bothered me was the idea that --enable-languages=all might
> not really mean "all languages". Perhaps the RHS could be something that
> indicates only primary, primary+secondary, etc. That's independent of Algol.
>
> Have there been any reports of testing on architectures besides x86_64?

I regularly build and run the Algol 68 testsuite on aarch64-linux.

Matthias Klose did the kind job of packaging the frontend on Debian on
top of both GCC 15 and trunk. He has been posting the test results to
gcc-testresults list. I recently got an account on the compier farm and
I intend to track/fix the few issues that pop on non- (x86_|aarch)64.

Sam recently worked on packaging the frontend on Gentoo and he uncovered
some issues with the diagnostics code because he used hardening options
in the build flags. That led to Jose rewriting a lot of that code to
stop using hardcoded buffer sizes and I complemented his patches with
some improvements to the way things were doing string size calculations.

As with any software more people building/using it means we get more
chances to fix things :)
 
> --joel

pietro

Reply via email to