Closed alexrp closed 2 months ago
This is actually intentional, see 4b8ef36d94841c7743e961e46e2f6fa62bafb92c.
On Windows, where we don't use symlinks for the various frontends, all the extra frontends take up a little space. And as GCC doesn't provide the same, and as far as I know, nothing ever implicitly looks for <triple>-cc
or any of these others that I removed, I removed them.
Ah, I see.
FWIW, I don't see e.g. aarch64-linux-gnu-c++
on my Ubuntu system where I have cross-compilers installed from APT, so I wonder if the same is the case for c++
?
FWIW, I don't see e.g.
aarch64-linux-gnu-c++
on my Ubuntu system where I have cross-compilers installed from APT, so I wonder if the same is the case forc++
?
That seems to be somewhat inconsistent. On Ubuntu 22.04, I indeed see the same, no aarch64-linux-gnu-c++
, but I do have x86_64-w64-mingw32-c++
(which is redirected via Deb package alternatives to e.g. x86_64-w64-mingw32-c++-posix
). But also with a custom handbuilt GCC, targeting mingw-w64, I'm also seeing <triple>-c++
frontends.
So apparently something is inconsistent wrt <triple>-c++
somewhere, either within GCC or within Ubuntu packaging. But for our purposes, there's at least some predecent for keeping that frontend.
Ok, that makes sense (~ish :slightly_smiling_face:). Thanks for confirming!
I'll close this then.
This isn't really a big deal in practice. I just happened to notice this disparity and figured I'd file an issue for it in case it wasn't intentional.