George Colpitts <george.colpi...@gmail.com> writes:

> It seems unlikely that the current tests wouldn't find this bug. Is it the
> case that the tests are never run on aarch64-darwin Macs?
>
The tests are certainly run; see, for instance, the 9.8.1-alpha4 release
pipeline [1].

The problem is that #21570 requires very particular (mis)configuration
of the host toolchain (e.g. Richard had multiple, incompatible
toolchains in PATH). We can usually side-step this sort of
misconfiguration by disabling the `configure` script's ld-override
logic, since we using anything but Apple's linker is generally a bad
idea on Darwin. This measure (!8437) is present in the 9.8 branch so I can only
guess that something else is wrong.

Perhaps you could open a ticket and attach config.log?

Cheers,

- Ben


[1]: https://gitlab.haskell.org/ghc/ghc/-/jobs/1668689

Attachment: signature.asc
Description: PGP signature

_______________________________________________
ghc-devs mailing list
ghc-devs@haskell.org
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs

Reply via email to