Fórum The Rust Programming Language Forum

yeah as said above an async fn generates a future that persist past the original invocation of the function and so need it's inputs to have lifetime equal or greater than the future generated so the only solution is to either estract the use of fetcher from the future generated or have the arc moved into the function

yurivict: wrkdir is tmpfs. Then that is the issue. The wrkdir got so big that there is no memory remaining to run rustc. You will need to disable tmpfs usage for the rust port. According to Solved - tmpfs for building ports | The FreeBSD Forums there is a TMPFS_BLACKLIST option you can set together with TMPFS_BLACKLIST_TMPDIR . In fact the example usage for this option shows disabling tmpfs for the rust port. yurivict: Can I turn off rustdoc completely? How? I don't think you can without patching the build system of rust. And even if you did, doing so would break cargo test as it will use rustdoc to run doctests. And finally disabling rustdoc would almost certainly still result in an OOM once it goes building cargo.

Can I turn off rustdoc completely? How?

wrkdir is tmpfs. I can try to set USE_TMPFS=OFF for it to go to disk.

Then I would suggest adding the reddit suggestion in the mute-decision announcement. TBH, reddit is banned in my region and thus it is not used in our daily life. Unless it is suggested, I wouldn't know it is the best place for announcement to the global community

Just to check is /wrkdirs a tmpfs or a regular directory on disk? If it is a tmpfs this error makes sense. Rustc needs more than 12GB of disk space to do a full build. If not I'm surprised it is rustdoc that fails. I would have expected rustc_middle to need more RAM.

I suppose you could create a suggestion issue on their GitHub, but seeing the number of issues, I wouldn't expect that to be done quickly, unless you volunteer.

Last 100 lines of log with SIGKILL in them.

Hello, I'm new to Rust and I'm experimenting with it to do a personal web project. I'm at the stage where I want to implement integration tests. For this I've decided to use the fantoccini crate to drive the browser from my rust test code. This works, but there is the annoying problem that if an assertion fails the browser window is not closed. To close the window I'm using #[rocket::async_test] #[serial] async fn test_firefox() -> Result > { // some setup code to start geckodriver let client: fantoccini::Client ... /* connect to the driver */; // test code is here, it uses an assert which can fail if a tests fails client.close().await?; Ok(()) } So the close method on the client is async and closes the browser, but if there is a panic (because of the assert failure) it is not executed. The standard Drop is not async and the Async Drop is not stable. What can I do to implement proper clean up in this situation? I'd prefer if I continue using stable rust!

yurivict: system with only 12 GB I still find it astonishing that people casually talk of "system with only 12 GB" as if that wasn't frikken huge. Coming from a time I could compile my code on machines with 64K bytes or even less this is mind bending. Anyway I have the same problem I run out of disk on a machine with 10GB space and 2GB swap. What to do?

Evian-Zhang: and I was thinking announcement is one of the purposes in the user forum. TBH I think that something reddit-like would (in the abstract at least) be better for announcements so people could vote up interesting things without needing to comment, and to downvote the slop "I did something in an afternoon" ones.

Yes, I agree that a pinned post can help. That post should include the motivation, decision, and instructions for both people who want to post announcements, and people who want to view announcements.