dev-lang/rust failure soft blocking dev-lang

[cwc]

Failure of dev-lang/rust
Not sure what to do:

[blocks B ] dev-lang/rust:stable ("dev-lang/rust:stable" is soft blocking dev-lang/rust-1.81.0-r100)

[eschwartz]

bindgen directly depends on virtual/rust unless you update it as well. A full @world rebuild should fix things too.

[depontius]

I'm seeing the same thing, except that emerging cbindgen is fails for me first. I'm not alone in this : https://forums.gentoo.org/viewtopic-t-1171685-highlight-.html
_________________
.sigs waste space and bandwidth

[sam_]

That is not the same thing then. OP's issue is dev-lang/rust failing to build and then it not being able to resume. We need the full log for OP's build failure.

[grumblebear]

Same for me with rust-1.82.
For me it worked to manually remove virtual/rust.

[cwc]

[sam_]

Please sync, try a full world upgrade, then share the full build.log if anything fails.

[depontius]

My conjecture...

The old rust has been masked, and we're supposed to move to slotted rust. I don't know how to do that. When I attempt to update world it fails at the first package that requires rust, namely cbindgen-0.27.1-r1. I suspect that cbindgen needs the new slotted rust in order to build properly, but I can't get there because cbindgen fails first. And as mentioned, I'm not sure how to install the new slotted rust, anyway. All of the old stuff is masked.

[cwc]

I removed virtual/rust
Still failure

I might be missing something : https://wiki.gentoo.org/wiki/Rust

emerge -pqv '=dev-lang/rust-1.81.0-r100::gentoo'

[sam_]

Please don't manually remove anything. As I said in another thread, we need to focus on cbindgen. Please file a bug for it with the full build.log at bugs.gentoo.org. Thanks.

[depontius]

[cwc]

darn. I blindly followed:

[cwc]

[Hu]

I think using package.unmask is premature at this point. I suggest removing relevant package.unmask, package.accept_keywords, and (if any) package.mask entries related to rust, which will get the system closest to the state that the Gentoo developers expect. Next, follow sam_'s request to focus on cbindgen and, if it still fails, report full details.

On its own, virtual/rust doesn't do anything. Virtuals exist to pull in relevant other packages. Let Portage pull in the virtual if and when it needs to do so.

[depontius]

Here's what I used, but note that it's all commented out.

/etc/portage/package.accept_keywords/tempFix

[grumblebear]

After removing virtual/rust you should of course update your @world and not a single package. Then portage should be able to update all packages still depending on virtual/rust. And you should also follow the advice to remove rust from package.unmask.

[depontius]

Another data point. My problems and postings thus far have been on my computer. I haven't updated my wife's computer in a week or so, so I decided to see what's up there. I know it's not a complete update, but I was going to simply see if cbindgen would build successfully on an otherwise undisturbed machine.

[cwc]

got rust installed with: emerge -a cbindgen

[eschwartz]

[depontius]

[eschwartz]

There isn't *supposed* to be a virtual anymore, as the virtual mainly existed so that ebuilds could depend on virtual/rust if they needed "a rust compiler installed" but now the suggestion is to simply inherit the rust eclass and let it handle that.

You can depclean virtual/rust if nothing depends on it, but keep in mind that you will need --with-bdeps=n if you have any packages installed which *built* with the old rust package. In fact, depcleaning all versions of rust and virtual/rust and then removing leftover stray files in /usr/lib/rust might clean up the current orphaned files issue caused by the interrupted update.

[depontius]

Thank you, it appears to be working now. It has completed cbindgen and is proceeding on maturin.

I removed dev-lang/rust-1.81.0 and virtual/rust-1.81.0-r1. After doing that an "emerge -p --depclean" wanted to rip out all things rust pretty completely. After cbindgen completed emerging "emerge -p --depclean" returns nothing, so even the odd fact that rust-common was slated for removal earlier is gone. It al looks good for the moment.

Thanks again for your help. This was a pretty weird stumble-into and it appears that others have stumbled into it as well. I'm not sure exactly what went wrong, I suspect at least part of it had to do with how recently the system had last been updated and what other packages were installed. It also looks as if virtual/rust was a big component of the problem as well, because it was what held onto the old rust, causing there to be two versions installed and making subsequent rust builds fail.
_________________
.sigs waste space and bandwidth

[cwc]

I found this interesting in reference to virtual/rust
Thanks for all the comments. I repeat {Everyone on this forum is so professional. Espeically with my simple questions.}

My dell laptop running xfce4 with no chromium

[depontius]

I have multiple machines here, and after pathfinding on my own, I came down to "emerge -acv --with-bdeps=n virtual/rust" followed by "emerge -a1tv rust", and this seems to be working well. I managed to get past rust and on to other things. I'm now having problems with webkit-gtk not building, but that's clearly a different issue and not relevant here. The rust problems appear to be gone.

Except on the laptop.

[r7l]

Had the same issues on multiple machines. Every Rust based packages was throwing errors after installing the slotted Rust. Packages like "cbindgen" where just the first ones to throw up. Manually removing the old "dev-lang/rust" and "virtual/rust" fixed the issue and everything went on smoothly.