We don't actually probe for javac in all circumstances, so if you have
javac installed, but don't have antlr4 installed, and you're on Mac OS
X, then you'll get a message that javac is missing, even though that's
wrong.
To fix this, let's just be a bit more generic in the message, so that
it's the same no matter what part of the lexer tests you're missing.
cc
https://www.reddit.com/r/rust/comments/3m199d/running_make_check_on_the_source_code_says_javac/
We don't actually probe for javac in all circumstances, so if you have
javac installed, but don't have antlr4 installed, and you're on Mac OS
X, then you'll get a message that javac is missing, even though that's
wrong.
To fix this, let's just be a bit more generic in the message, so that
it's the same no matter what part of the lexer tests you're missing.
cc
https://www.reddit.com/r/rust/comments/3m199d/running_make_check_on_the_source_code_says_javac/
This is part of some cleanup I did while investigating #28129.
This also ensures that `on_panic` is run even if the user has registered too many callbacks.
Part of https://github.com/rust-lang/rust/issues/6993
This patch replaces `Ident`s with `Name`s in data structures of HIR and updates the dependent crates to compile and pass `make check`.
Some HIR structures still use `Ident`s, namely `PathSegment`, `PatIdent`, `ExprWhile`, `ExprLoop`, `ExprBreak` and `ExprAgain`, they need them for resolve (but `PathSegment` is special, see https://github.com/rust-lang/rust/issues/6993#issuecomment-141256292).
r? @nrc
Fixes#16264 / #18241.
As far as I can tell, it should be impossible for a trait to be inaccessible if it's in scope, so this check is unnecessary. Are there any cases where this check is actually needed?
Fixes#28279.
Currently
`common_supertype(*mut for<'a> Fn(&'a usize), *mut for<'a> Fn(&'a usize) + 'static)`
equals `*mut Fn(&usize)` which seems to be caused by `higher_ranked_sub()` allowing region variables to escape the comparison. This prevents inference from working properly with stuff like `Rc<Fn(&T)>`.
r? @nikomatsakis
These changes introduce the ability to cross-compile working binaries for NetBSD/amd64. Previous support added in PR #26682 shared all its code with the OpenBSD implementation, and was therefore never functional (e.g. linking against non-existing symbols and using wrong type definitions). Nonetheless, the previous patches were a great starting point and made my work significantly easier. 😃
Because there are no stage0 snapshots for NetBSD (yet), I used a cross-compiler for NetBSD 7.0 RC3 and only tested some toy programs (threading and channels, stack guards, a small TCP/IP echo server and some other platform dependent bits). If someone could point me to documentation on how to generate a stage0 snapshot from a cross-compiler I'm happy to run the full test suite.
A few other notes regarding Rust on NetBSD/amd64:
- To preserve binary compatibility, NetBSD introduces new symbols for system call wrappers on breaking ABI changes and keeps the old (legacy) symbols around, see [this documentation](https://www.netbsd.org/docs/internals/en/chap-processes.html#syscalls_master) for some details. I went ahead and modified the `libc` and `std` crate to use the current (renamed) symbols instead of the legacy ones where I found them, but I might have missed some. Notably using the `sigaction` symbol (deprecated in 1998) instead of `__sigaction14` even triggers SIGSYS (bad syscall) on my amd64 setup. I also changed the type definitions to use the most recent version.
- NetBSD's gdb doesn't really support position independent executables, so you might want to turn that off for debugging, see [NetBSD Problem Report #48250](https://gnats.netbsd.org/48250).
- For binaries invoked using a relative path, NetBSD supports `$ORIGIN` only for short `rpath`s (~64 chars or so, I'm told). If running an executable fails with `execname not specified in AUX vector: No such file or directory`, consider invoking the binary using its full absolute path.
The `register` function is unstable and it is not used anymore, hence
it can be removed (together with the now-unused `Callback` type and
`static` variables).
This patch basically adds a target option for omitting the `-nodefaultlibs` flag when invoking the linker. I am not sure if this is the correct or only way to approach this problem, so any feedback is welcome.
Motivation: I'm currently working on a Rust target specification for the [rumprun](/rumpkernel/rumprun) unikernel. rumprun is based on rump kernels and uses NetBSDs libc and drivers to provide a POSIXy environment. It provides its own linker wrapper that generates binaries which can be "baked" into a unikernel after configuration. Using `-nodefaultlibs` on the rumprun linker will prevent it from selecting the search paths for the rumprun libraries. My current target implementation for rumprun is here: gandro/rust@295744b2ee
Currently, only a target that `is_like_windows` will omit the `-nodefaultlibs` flag, but since rumprun is not like Windows otherwise, I think a separate flag makes more sense. This might be a breaking change for target specifications that have the `is_like_windows` option set to true. Such targets need to set `no_default_libraries` to false in order to restore the old behavior.
The registration of `panicking::on_panic` as a general-purpose
callback is overcomplicated and can fail.
Instead, invoking it explicitly removes the need for locking and paves
the way for further improvements.
If set to false, `-nodefaultlibs` is not passed to the linker. This
was the default behavior on Windows, but it should be configurable
per target.
This is a [breaking-change] for target specifications that have
the `is_like_windows` option set to true. Such targets need to
set `no_default_libraries` to false in order to restore the old
behavior.
Since 1.3.0 the BufWriter has seen tremendous speedups. So when I use it in the shootout benchmarks, I see some nice speedup (which up to 1.2.0 was nixed by the pessimizations during initialization).
With -O2, LLVM's inliner can remove this code, but this does not happen
with -O1 and lower. As a result, dropping Vec<u8> was linear with length,
resulting in abysmal performance for large buffers.
See issue #24280.