rustbuild: add config.toml option to disable codegen tests
Fixes#36232.
I think it worked? Here's a build log where I tried to bootstrap, it crashed, then I added the setting to config.toml and it continued: https://gist.github.com/durka/cbf97cf04b8e065f1a2cfda4c1b6bf95
r? @alexcrichton
Update nightly docs supported Windows versions to match Getting Started page
https://doc.rust-lang.org/book/getting-started.html#tier-1 shows that Windows 7+ is officially supported (implying, for example Windows 10), but the nightly page only listed 7, 8, and Server 2008 R2).
Updated e0493 to new format (+ bonus).
Part of #35233.
Fixes#35999.
r? @jonathandturner
I'm not satisfied with the bonus part, there has to be an easier way to reach into the `Drop`'s span implementation. I'm all ears. :)
Add --Zsave-analysis-api
This is a save-analysis variation which can be used with libraries distributed without their source (e.g., libstd). It will allow IDEs and other tools to get info about types and create URLs to docs and source, without the unnecessary clutter of internal-only save-analysis info. I'm sure we'll iterate somewhat on the design, but this is a first draft.
Add --Zsave-analysis-api
This is a save-analysis variation which can be used with libraries distributed without their source (e.g., libstd). It will allow IDEs and other tools to get info about types and create URLs to docs and source, without the unnecessary clutter of internal-only save-analysis info. I'm sure we'll iterate somewhat on the design, but this is a first draft.
Remove --{enable|disable}-orbit from configure.
Fixes#35956.
r? @eddyb
There are only two buildbots left, though they are both failing. Is there something to be done there other than wait?
Fix optimization regressions for operations on [x; n]-initialized arrays.
Fixes#35662 by using `!=` instead of `<` as the stop condition for `[x; n]` initialization loops.
Also included is cc2009f02d, a hack to run the GVN pass twice, another time after InstCombine.
This hack results in removal of redundant `memset` and `memcpy` calls (from loops over arrays).
cc @nrc Can we get performance numbers on this? Not sure if it regresses anything else.
Rather than saying "struct or union" or adding logic to determine the
type of the item, just change the message to "field is never used",
dropping the "struct".
Update tests accordingly.
Implement untagged unions (RFC 1444)
cc https://github.com/rust-lang/rust/issues/32836
Notes:
- The RFC doesn't talk about `#[packed]` unions, this implementation supports them, packing changes union's alignment to 1 and removes trailing padding.
- The RFC doesn't talk about dynamically sized unions, this implementation doesn't support them and rejects them during wf-checking (similarly, dynamically sized enums are not supported as well).
- The lint for drop fields in unions can't work precisely before monomorphization, so it works pessimistically - non-`Copy` generic fields are reported, types not implementing `Drop` directly, but having non-trivial drop code are reported.
```
struct S(String); // Doesn't implement `Drop`
union U<T> {
a: S, // Reported
b: T, // Reported
}
```
- https://github.com/rust-lang/rust/pull/35764 was indeed helpful and landed timely, I didn't have to implement internal drop flags for unions.
- Unions are not permitted in constant patterns, because matching on union fields is unsafe, I didn't want unsafety checker to dig into all constants to uncover this possible unsafety.
- The RFC doesn't talk about `#[derive]`, generally trait impls cannot be derived for unions, but some of them can. I implemented only `#[derive(Copy)]` so far. In theory shallow `#[derive(Clone)]` can be derived as well if all union fields are `Copy`, I left it for later though, it requires changing how `Clone` impls are generated.
- Moving union fields is implemented as per https://github.com/rust-lang/rust/issues/32836#issuecomment-242511491.
- Testing strategy: union specific behavior is tested, sometimes very basically (e.g. debuginfo), behavior common for all ADTs (e.g. something like coherence
checks) is not generally tested.
r? @eddyb
When using --enable-local-rebuild configure options, the configure
script will test rustc version. But when running it, it will not use the
libraries in the local-rust-root directory.
So use `LD_LIBRARY_PATH` environment variable to correct it.
