-
Notifications
You must be signed in to change notification settings - Fork 13.3k
rustbuild: is llvm always built with AVR backend? #76470
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Seems like error in #69478 https://github.com/rust-lang/rust/pull/69478/files#diff-9a46b13a265be2ded97369dddf5037f5R147 probably should have been left empty, otherwise this comment is not valid:
cc @dylanmckay |
Yeah, I was looking into |
…s are enabled in config.toml Fixes rust-lang#76470
This is not an issue with AVR specifically, all experimental targets were always built by default, similarly to how all regular targets are built by default. targets = "X86"
experimental-targets = "" rather than just targets = "X86" to disable targets except for x86. |
That seems counterintuitive to me. |
I agree that it's counterintuitive, but not building AVR by default will silently disable Perhaps the better default would be to drop experimental targets only if the |
What about disabling experimental-targets by default but explicitly enabling AVR on the fastest CI test job? |
Hm, I think the comment is probably misleading, but I don't personally see why this is counterintuitive - everything in config.toml.example is commented out, since it's intended to just provide defaults at a point in time. Since many people copy it in its entirety, if they weren't mostly commented we would find it harder to update defaults (and have them actually take effect). On this case, maybe we can merge AVR into the targets list? Rustbuild could manually separate it out into experimental targets for LLVM, but I feel like from a rustc perspective there's no significant difference between experimental and regular targets. I am personally a bit hesitant to modify local configurations in a way that diverges from CI to prevent painful debugging as much as possible - do we have some idea how much time cutting LLVM targets saves? |
See also #76408 |
AVR is no longer an experimental target since LLVM 11. |
I'm going to close this since it seems like it's no longer causing issues. |
I have this in my
config.toml
targets = "X86"
but while
x.py
runs cmake, I sawMaking me wonder if there is something that always causes AVR to be activated (by mistake?)
repo @ e82584a
The text was updated successfully, but these errors were encountered: