This commit adds a ./configure option called `--disable-elf-tls` which disables
ELF based TLS (that which is communicated to LLVM) on platforms which already
support it. OSX 10.6 does not support this form of TLS, and some users of Rust
need to target 10.6 and are unable to do so due to the usage of TLS. The
standard library will continue to use ELF based TLS on OSX by default (as the
officially supported platform is 10.7+), but this adds an option to compile the
standard library in a way that is compatible with 10.6.
Closes#25342
This commit adds a ./configure option called `--disable-elf-tls` which disables
ELF based TLS (that which is communicated to LLVM) on platforms which already
support it. OSX 10.6 does not support this form of TLS, and some users of Rust
need to target 10.6 and are unable to do so due to the usage of TLS. The
standard library will continue to use ELF based TLS on OSX by default (as the
officially supported platform is 10.7+), but this adds an option to compile the
standard library in a way that is compatible with 10.6.
Currently, there are two conditional blocks that exist to check for "clang or gcc"
On line 866:
```
if [ -z "$CFG_ENABLE_CLANG" -a -z "$CFG_GCC" ]
then
err "either clang or gcc is required"
fi
```
and on line 1019:
```
if [ -z "$CC" -a -z "$CFG_ENABLE_CLANG" -a -z "$CFG_GCC" ]
then
err "either clang or gcc is required"
fi
```
Given the order of the clauses, this results in the "either clang or gcc is required" error from the earlier block, (even) when CC is set.
The expected behaviour is to honour user-flags, in this case CC.
Aside from removing all hand-holdy compiler checks in favour of actual compiler *feature* checks, this change removes the redundant former block in favour of the latter block, which appears designed to allow the expected behaviour.
The `run` function passed its argument to `msg` via `"$@"`, but `msg`
only printed its first argument. I think the intention was for `msg` to
print all its arguments. (If not, `run` should only `msg "$1"`.)
Took me a moment to figure out that the appropriate response to
"need program file" was to install the program named "file", not
to think "If I didn't need the program file, why would I be
compiling things?".
the "-c" option of head isn't a posix option, and it isn't supported
under openbsd.
prefer the use of cut -c 1-8 (which is posix) to extract the first 8
chars.
Special thanks to @retep998 for the [excellent writeup](https://github.com/rust-lang/rfcs/issues/1061) of tasks to be done and @ricky26 for initially blazing the trail here!
# MSVC Support
This goal of this series of commits is to add MSVC support to the Rust compiler
and build system, allowing it more easily interoperate with Visual Studio
installations and native libraries compiled outside of MinGW.
The tl;dr; of this change is that there is a new target of the compiler,
`x86_64-pc-windows-msvc`, which will not interact with the MinGW toolchain at
all and will instead use `link.exe` to assemble output artifacts.
## Why try to use MSVC?
With today's Rust distribution, when you install a compiler on Windows you also
install `gcc.exe` and a number of supporting libraries by default (this can be
opted out of). This allows installations to remain independent of MinGW
installations, but it still generally requires native code to be linked with
MinGW instead of MSVC. Some more background can also be found in #1768 about the
incompatibilities between MinGW and MSVC.
Overall the current installation strategy is quite nice so long as you don't
interact with native code, but once you do the usage of a MinGW-based `gcc.exe`
starts to get quite painful.
Relying on a nonstandard Windows toolchain has also been a long-standing "code
smell" of Rust and has been slated for remedy for quite some time now. Using a
standard toolchain is a great motivational factor for improving the
interoperability of Rust code with the native system.
## What does it mean to use MSVC?
"Using MSVC" can be a bit of a nebulous concept, but this PR defines it as:
* The build system for Rust will build as much code as possible with the MSVC
compiler, `cl.exe`.
* The build system will use native MSVC tools for managing archives.
* The compiler will link all output with `link.exe` instead of `gcc.exe`.
None of these are currently implemented today, but all are required for the
compiler to fluently interoperate with MSVC.
## How does this all work?
At the highest level, this PR adds a new target triple to the Rust compiler:
x86_64-pc-windows-msvc
All logic for using MSVC or not is scoped within this triple and code can
conditionally build for MSVC or MinGW via:
#[cfg(target_env = "msvc")]
It is expected that auto builders will be set up for MSVC-based compiles in
addition to the existing MinGW-based compiles, and we will likely soon start
shipping MSVC nightlies where `x86_64-pc-windows-msvc` is the host target triple
of the compiler.
# Summary of changes
Here I'll explain at a high level what many of the changes made were targeted
at, but many more details can be found in the commits themselves. Many thanks to
@retep998 for the excellent writeup in rust-lang/rfcs#1061 and @rick26 for a lot
of the initial proof-of-concept work!
## Build system changes
As is probably expected, a large chunk of this PR is changes to Rust's build
system to build with MSVC. At a high level **it is an explicit non goal** to
enable building outside of a MinGW shell, instead all Makefile infrastructure we
have today is retrofitted with support to use MSVC instead of the standard MSVC
toolchain. Some of the high-level changes are:
* The configure script now detects when MSVC is being targeted and adds a number
of additional requirements about the build environment:
* The `--msvc-root` option must be specified or `cl.exe` must be in PATH to
discover where MSVC is installed. The compiler in use is also required to
target x86_64.
* Once the MSVC root is known, the INCLUDE/LIB environment variables are
scraped so they can be reexported by the build system.
* CMake is required to build LLVM with MSVC (and LLVM is also configured with
CMake instead of the normal configure script).
* jemalloc is currently unconditionally disabled for MSVC targets as jemalloc
isn't a hard requirement and I don't know how to build it with MSVC.
* Invocations of a C and/or C++ compiler are now abstracted behind macros to
appropriately call the underlying compiler with the correct format of
arguments, for example there is now a macro for "assemble an archive from
objects" instead of hard-coded invocations of `$(AR) crus liboutput.a ...`
* The output filenames for standard libraries such as morestack/compiler-rt are
now "more correct" on windows as they are shipped as `foo.lib` instead of
`libfoo.a`.
* Rust targets can now depend on native tools provided by LLVM, and as you'll
see in the commits the entire MSVC target depends on `llvm-ar.exe`.
* Support for custom arbitrary makefile dependencies of Rust targets has been
added. The MSVC target for `rustc_llvm` currently requires a custom `.DEF`
file to be passed to the linker to get further linkages to complete.
## Compiler changes
The modifications made to the compiler have so far largely been minor tweaks
here and there, mostly just adding a layer of abstraction over whether MSVC or a
GNU-like linker is being used. At a high-level these changes are:
* The section name for metadata storage in dynamic libraries is called `.rustc`
for MSVC-based platorms as section names cannot contain more than 8
characters.
* The implementation of `rustc_back::Archive` was refactored, but the
functionality has remained the same.
* Targets can now specify the default `ar` utility to use, and for MSVC this
defaults to `llvm-ar.exe`
* The building of the linker command in `rustc_trans:🔙:link` has been
abstracted behind a trait for the same code path to be used between GNU and
MSVC linkers.
## Standard library changes
Only a few small changes were required to the stadnard library itself, and only
for minor differences between the C runtime of msvcrt.dll and MinGW's libc.a
* Some function names for floating point functions have leading underscores, and
some are not present at all.
* Linkage to the `advapi32` library for crypto-related functions is now
explicit.
* Some small bits of C code here and there were fixed for compatibility with
MSVC's cl.exe compiler.
# Future Work
This commit is not yet a 100% complete port to using MSVC as there are still
some key components missing as well as some unimplemented optimizations. This PR
is already getting large enough that I wanted to draw the line here, but here's
a list of what is not implemented in this PR, on purpose:
## Unwinding
The revision of our LLVM submodule [does not seem to implement][llvm] does not
support lowering SEH exception handling on the Windows MSVC targets, so
unwinding support is not currently implemented for the standard library (it's
lowered to an abort).
[llvm]: https://github.com/rust-lang/llvm/blob/rust-llvm-2015-02-19/lib/CodeGen/Passes.cpp#L454-L461
It looks like, however, that upstream LLVM has quite a bit more support for SEH
unwinding and landing pads than the current revision we have, so adding support
will likely just involve updating LLVM and then adding some shims of our own
here and there.
## dllimport and dllexport
An interesting part of Windows which MSVC forces our hand on (and apparently
MinGW didn't) is the usage of `dllimport` and `dllexport` attributes in LLVM IR
as well as native dependencies (in C these correspond to
`__declspec(dllimport)`).
Whenever a dynamic library is built by MSVC it must have its public interface
specified by functions tagged with `dllexport` or otherwise they're not
available to be linked against. This poses a few problems for the compiler, some
of which are somewhat fundamental, but this commit alters the compiler to attach
the `dllexport` attribute to all LLVM functions that are reachable (e.g. they're
already tagged with external linkage). This is suboptimal for a few reasons:
* If an object file will never be included in a dynamic library, there's no need
to attach the dllexport attribute. Most object files in Rust are not destined
to become part of a dll as binaries are statically linked by default.
* If the compiler is emitting both an rlib and a dylib, the same source object
file is currently used but with MSVC this may be less feasible. The compiler
may be able to get around this, but it may involve some invasive changes to
deal with this.
The flipside of this situation is that whenever you link to a dll and you import
a function from it, the import should be tagged with `dllimport`. At this time,
however, the compiler does not emit `dllimport` for any declarations other than
constants (where it is required), which is again suboptimal for even more
reasons!
* Calling a function imported from another dll without using `dllimport` causes
the linker/compiler to have extra overhead (one `jmp` instruction on x86) when
calling the function.
* The same object file may be used in different circumstances, so a function may
be imported from a dll if the object is linked into a dll, but it may be
just linked against if linked into an rlib.
* The compiler has no knowledge about whether native functions should be tagged
dllimport or not.
For now the compiler takes the perf hit (I do not have any numbers to this
effect) by marking very little as `dllimport` and praying the linker will take
care of everything. Fixing this problem will likely require adding a few
attributes to Rust itself (feature gated at the start) and then strongly
recommending static linkage on Windows! This may also involve shipping a
statically linked compiler on Windows instead of a dynamically linked compiler,
but these sorts of changes are pretty invasive and aren't part of this PR.
## CI integration
Thankfully we don't need to set up a new snapshot bot for the changes made here as our snapshots are freestanding already, we should be able to use the same snapshot to bootstrap both MinGW and MSVC compilers (once a new snapshot is made from these changes).
I plan on setting up a new suite of auto bots which are testing MSVC configurations for now as well, for now they'll just be bootstrapping and not running tests, but once unwinding is implemented they'll start running all tests as well and we'll eventually start gating on them as well.
---
I'd love as many eyes on this as we've got as this was one of my first interactions with MSVC and Visual Studio, so there may be glaring holes that I'm missing here and there!
cc @retep998, @ricky26, @vadimcn, @klutzy
r? @brson
This commit modifies the makefiles to enable building LLVM with cmake and Visual
Studio to generate an LLVM that targets MSVC. Rust's configure script requires
cmake to be installed when targeting MSVC and will configure LLVM with cmake
instead of the normal `./configure` script LLVM provides. The build will then
run cmake to execute the build instead of the normal `make`.
Currently `make clean-llvm` isn't supported on MSVC as I can't figure out how to
run a "clean" target for the Visual Studio files.
This commit starts to add MSVC support to the ./configure script to enable the
build system to detect and build an MSVC target with the cl.exe compiler and
toolchain. The primary change here is a large sanity check when an MSVC target
is requested (and currently only `x86_64-pc-windows-msvc` is recognized).
When building an MSVC target, the configure script either requires the
`--msvc-root` argument or for `cl.exe` to be in `PATH`. It also requires that if
in the path `cl.exe` is the 64-bit version of the compiler.
Once detected the configure script will run the `vcvarsall.bat` script provided
by Visual Studio to learn about the `INCLUDE` and `LIB` variables needed by the
`cl.exe` compiler to run (the default include/lib paths for the
compiler/linker). These variables are then reexported when running `make` to
ensure that our own compiles are running the same toolchain.
The purpose of this detection and environment variable scraping is to avoid
requiring the build itself to be run inside of a `cmd.exe` shell but rather
allow it to run in the currently expected MinGW/MSYS shell.
The code takes a prefix of the MD5 hash of the version string.
Since the hash command differs across GNU and BSD platforms, we scan for
the right one in the configure script.
Closes#25007
The old code simply scanned for the first digit, then munched anything
after that. This didn't work for md5sum, as it would see the "5" and
treat "5sum" as the version instead.
This patch tweaks the algorithm so that it looks for a second
consecutive digit (or dot) after the first. Since "md5sum" has only one
digit, the new code skips over it as intended.
The code takes a prefix of the MD5 hash of the version string.
Since the hash command differs across GNU and BSD platforms, we scan for
the right one in the configure script.
Closes#25007
Then, decouple the question of whether the compiler/stdlib carry
debuginfo (which is controlled via `--enable-debuginfo` and implied by
`--enable-debug`) from the question of whether the tests carry
debuginfo (which now no longer is implied by `--enable-debug` nor
`--enable-debuginfo`, and is off by default).
These commits build on [some great work on reddit](http://www.reddit.com/r/rust/comments/33boew/weekend_experiment_link_rust_programs_against/) for adding MUSL support to the compiler. This goal of this PR is to enable a `--target x86_64-unknown-linux-musl` argument to the compiler to work A-OK. The outcome here is that there are 0 compile-time dependencies for a MUSL-targeting build *except for a linker*. Currently this also assumes that MUSL is being used for statically linked binaries so there is no support for dynamically linked binaries with MUSL.
MUSL support largely just entailed munging around with the linker and where libs are located, and the major highlights are:
* The entirety of `libc.a` is included in `liblibc.rlib` (statically included as an archive).
* The entirety of `libunwind.a` is included in `libstd.rlib` (like with liblibc).
* The target specification for MUSL passes a number of ... flavorful options! Each option is documented in the relevant commit.
* The entire test suite currently passes with MUSL as a target, except for:
* Dynamic linking tests are all ignored as it's not supported with MUSL
* Stack overflow detection is not working MUSL yet (I'm not sure why)
* There is a language change included in this PR to add a `target_env` `#[cfg]` directive. This is used to conditionally build code for only MUSL (or for linux distros not MUSL). I highly suspect that this will also be used by Windows to target MSVC instead of a MinGW-based toolchain.
To build a compiler targeting MUSL you need to follow these steps:
1. Clone the current MUSL repo from `git://git.musl-libc.org/musl`. Build this as usual and install it.
2. Clone and build LLVM's [libcxxabi](http://libcxxabi.llvm.org/) library. Only the `libunwind.a` artifact is needed. I have tried using upstream libunwind's source repo but I have not gotten unwinding to work with it unfortunately. Move `libunwind.a` adjacent to MUSL's `libc.a`
3. Configure a Rust checkout with `--target=x86_64-unknown-linux-musl --musl-root=$MUSL_ROOT` where `MUSL_ROOT` is where you installed MUSL in step 1.
I hope to improve building a copy of libunwind as it's still a little sketchy and difficult to do today, but other than that everything should "just work"! This PR is not intended to include 100% comprehensive support for MUSL, as future modifications will probably be necessary.
This commit adds support to the makefiles, configuration script, and build
system to understand MUSL. This is broken up into a few parts:
* Any target of the form `*-musl` requires the `--musl-root` option to
`./configure` which will indicate the root of the MUSL installation. It is
also expected that there is a libunwind build inside of that installation
built against that MUSL.
* Objects from MUSL are copied into the build tree for Rust to be statically
linked into the appropriate Rust library.
* Objects for binary startup and shutdown are included in each Rust installation
by default for MUSL. This requires MUSL to only be installed on the machine
compiling rust. Only a linker will be necessary for compiling against MUSL on
a target machine.
Eventually a MUSL and/or libunwind build may be integrated by default into the
build but for now they are just always assumed to exist externally.
In most places in mk/tests.mk, it's positioned after rpass-full and
before cfail-full (because rfail comes before cfail). The order of tests
seems a little inconsistent, but reordering everywhere would obscure this
commit.
This commit series starts out with more official test harness support for rustdoc tests, and then each commit afterwards adds a test (where appropriate). Each commit should also test and finish independently of all others (they're all pretty separable).
I've uploaded a [copy of the documentation](http://people.mozilla.org/~acrichton/doc/std/) generated after all these commits were applied, and a double check on issues being closed would be greatly appreciated! I'll also browse the docs a bit and make sure nothing regressed too horribly.
Add a new test directory called 'rustdoc' where all files inside are documented
and run against the `htmldocck` script to have assertions about the output.
Until I can figure out the correct way to configure jemalloc for Bitrig, this patch will correctly disable it. All other build targets remain unchanged.
