Andy Wang

> reorder stores to different locations. Ah yes that's where I was very mistaken... I hope my updated test case correctly demonstrates the miscompilation. There's also the question about what...

@alexcrichton One thing I'm slightly annoyed about a path-like prefix (which was proposed in the pre-RFC) is the differences between path separators on Windows vs *nix. That's purely aesthetic and...

Coming back to this after a month of busy uni stuff - GCC and Clang both have three separate path remapping [flags](https://reproducible-builds.org/docs/build-path/): - `-fdebug-prefix-map` affects only debuginfo, - `-fmacro-prefix-map` affects...

Further proposal has been added the RFC is ready for more reviews. Could someone add T-compiler label?

Sorry it took me a while to get back to this. Linkers spitting out paths to separate debuginfo file turns out to be a bigger rabbithole than I thought. ~~So...

I have added another scope `split-debuginfo-path`, which takes effect only when `split-debuginfo` is `packed` or `unpacked`. In these cases the path to the split debug information files are embedded into...

Yes, the default behaviour is still full path removal in release profile. There are quite a few sources of absolute paths hence the complexity, but the default is sane and...

I'm not sure what else there is to add. There are a couple of unresolved points which I have noted in the text, but people seem fairly happy with the...

A summary of recent changes: 1. Instead of letting Cargo decide whether or not to sanitise debuginfo based on `-C split-debuginfo` setting (which necessitated Cargo feeding different arguments to `rustc`...

Haven't had any further comments for a while. @joshtriplett could you tag the compiler team (I'm not in rust-lang org so I can't do it) so someone could have a...