bors [Mon, 9 Oct 2017 14:00:12 +0000 (14:00 +0000)]
Auto merge of #45075 - alexcrichton:inline-less, r=michaelwoerister
rustc: Don't inline in CGUs at -O0
This commit tweaks the behavior of inlining functions into multiple codegen
units when rustc is compiling in debug mode. Today rustc will unconditionally
treat `#[inline]` functions by translating them into all codegen units that
they're needed within, marking the linkage as `internal`. This commit changes
the behavior so that in debug mode (compiling at `-O0`) rustc will instead only
translate `#[inline]` functions into *one* codegen unit, forcing all other
codegen units to reference this one copy.
The goal here is to improve debug compile times by reducing the amount of
translation that happens on behalf of multiple codegen units. It was discovered
in #44941 that increasing the number of codegen units had the adverse side
effect of increasing the overal work done by the compiler, and the suspicion
here was that the compiler was inlining, translating, and codegen'ing more
functions with more codegen units (for example `String` would be basically
inlined into all codegen units if used). The strategy in this commit should
reduce the cost of `#[inline]` functions to being equivalent to one codegen
unit, which is only translating and codegen'ing inline functions once.
Collected [data] shows that this does indeed improve the situation from [before]
as the overall cpu-clock time increases at a much slower rate and when pinned to
one core rustc does not consume significantly more wall clock time than with one
codegen unit.
One caveat of this commit is that the symbol names for inlined functions that
are only translated once needed some slight tweaking. These inline functions
could be translated into multiple crates and we need to make sure the symbols
don't collideA so the crate name/disambiguator is mixed in to the symbol name
hash in these situations.
bors [Mon, 9 Oct 2017 04:59:02 +0000 (04:59 +0000)]
Auto merge of #45041 - est31:master, r=alexcrichton
Remove support for the PNaCl target (le32-unknown-nacl)
This removes support for the `le32-unknown-nacl` target which is currently supported by rustc on tier 3. Despite the "nacl" in the name, the target doesn't output native code (x86, ARM, MIPS), instead it outputs binaries in the PNaCl format.
There are two reasons for the removal:
* Google [has announced](https://blog.chromium.org/2017/05/goodbye-pnacl-hello-webassembly.html) deprecation of the PNaCl format. The suggestion is to migrate to wasm. Happens we already have a wasm backend!
* Our PNaCl LLVM backend is provided by the fastcomp patch set that the LLVM fork used by rustc contains in addition to vanilla LLVM (`src/llvm/lib/Target/JSBackend/NaCl`). Upstream LLVM doesn't have PNaCl support. Removing PNaCl support will enable us to move away from fastcomp (#44006) and have a lighter set of patches on top of upstream LLVM inside our LLVM fork. This will help distribution packagers of Rust.
bors [Sun, 8 Oct 2017 18:12:26 +0000 (18:12 +0000)]
Auto merge of #45016 - pnkfelix:mir-borrowck-gather-and-signal-move-errors, r=nikomatsakis
MIR-borrowck: gather and signal any move errors
When building up the `MoveData` structure for a given MIR, also accumulate any erroneous actions, and then report all of those errors when the construction is complete.
This PR adds a host of move-related error constructor methods to `trait BorrowckErrors`. I think I got the notes right; but we should plan to audit all of the notes before turning MIR-borrowck on by default.
bors [Sun, 8 Oct 2017 10:43:45 +0000 (10:43 +0000)]
Auto merge of #45012 - Gankro:noalias, r=arielb1
Add -Zmutable-noalias flag
We disabled noalias on mutable references a long time ago when it was clear that llvm was incorrectly handling this in relation to unwinding edges.
Since then, a few things have happened:
* llvm has cleaned up a bunch of the issues (I'm told)
* we've added a nounwind codegen option
As such, I would like to add this -Z flag so that we can evaluate if the codegen bugs still exist, and if this significantly affects the codegen of different projects, with an eye towards permanently re-enabling it (or at least making it a stable option).
bors [Sun, 8 Oct 2017 08:21:24 +0000 (08:21 +0000)]
Auto merge of #45001 - bgermann:master, r=alexcrichton
Add builder for Solaris and merge it with Fuchsia's builder
The new Solaris builder can be used to build rust-std.
The dilos illumos distribution was chosen, because illumos is free software
as opposed to Oracle Solaris and dilos is the only illumos distribution that
supports x86_64 and sparcv9 at the same level.
kennytm [Sun, 8 Oct 2017 05:39:06 +0000 (13:39 +0800)]
Rollup merge of #45094 - japaric:strict-align, r=alexcrichton
enable strict alignment (+strict-align) on ARMv6
As discovered in #44538 ARMv6 devices may or may not support unaligned memory accesses. ARMv6
Linux *seems* to have no problem with unaligned accesses but this is because the kernel is stepping
in to fix each unaligned memory access -- this incurs in a performance penalty.
This commit enforces aligned memory accesses on all our in-tree ARM targets that may be used with
ARMv6 devices. This should improve performance of Rust programs on ARMv6 devices. For the record,
clang also applies this attribute when targeting ARMv6 devices that are not running Darwin or
NetBSD.
kennytm [Sun, 8 Oct 2017 05:39:05 +0000 (13:39 +0800)]
Rollup merge of #45090 - rust-lang:petrochenkov-patch-1, r=alexcrichton
Document that `-C ar=PATH` doesn't do anything
Are there any plans to use an external archiver in the future?
IIRC, it was used before, but its use was replaced with LLVM's built-in archive management machinery. I can't found a relevant PR though. EDIT: Found it - https://github.com/rust-lang/rust/pull/26926!
The `-C` option is stable so it still can't be removed right away even if there are no plans to use it (but maybe it can be deprecated?).
Target specifications have a field for archiver as well, which is unused too (these ones are unstable, so I guess it can be removed).
kennytm [Sun, 8 Oct 2017 05:39:04 +0000 (13:39 +0800)]
Rollup merge of #45083 - fhartwig:slice-read-to-end, r=bluss
Add read_to_end implementation to &[u8]'s Read impl
The default impl for read_to_end does a bunch of bookkeeping
that isn't necessary for slices and is about 4 times slower
on my machine.
The following benchmark takes about 30 ns before this change and about 7 ns after:
```
#[bench]
fn bench_read_std(b: &mut Bencher) {
let data = vec![0u8; 100];
let mut v = Vec::with_capacity(200);
b.iter(|| {
let mut s = data.as_slice();
v.clear();
s.read_to_end(&mut v).unwrap();
});
}
```
This solves the easy part of https://github.com/rust-lang/rust/issues/44819 (I think extending this to `Take<&[u8]> `would require specialization)
kennytm [Sun, 8 Oct 2017 05:39:01 +0000 (13:39 +0800)]
Rollup merge of #45058 - hunteke:fix_rustc_private_typo_2017Oct, r=shepmaster
Fix typo, per #45057.
This looks like a simple string -- one character -- fix. Given that I'm currently running low on battery, I have not actually compiled and tested this. But I am fully confident this passes muster. If not, I'll be maintainer-educated, yes? ;-)
kennytm [Sun, 8 Oct 2017 05:38:59 +0000 (13:38 +0800)]
Rollup merge of #45052 - steveklabnik:gh44105, r=dtolnay
Modify Rc/Arc language around mutability
There are a few exceptions to the rule that Arc/Rc are immutable. Rather
than dig into the details, add "generally" to hint at this difference,
as it's kind of a distraction at this point in the docs.
Additionally, Arc's docs were slightly different here generally, so add
in both the existing language and the exception.
bors [Sun, 8 Oct 2017 03:02:35 +0000 (03:02 +0000)]
Auto merge of #44983 - vitiral:dirty_clean_groups, r=michaelwoerister
groundwork for rustc_clean/dirty improvements
This is a WIP PR that needs mentoring from @michaelwoerister.
There are several TODOs but no outstanding questions (except for the main one -- **is this the right approach?**)
This is the plumbing for supporing groups in `rustc_clean(labels="...")`, as well as supporting an `except="..."` which will remove the excepted labels in the "clean" check and then assert that they are dirty (this is still TODO).
See the code TODO's and example comments for a rough design.
I'd like to know if this is the design you would like to do, and then I can go about actually filling out the groups and implementing the remaining logic.
Alex Crichton [Fri, 6 Oct 2017 21:59:33 +0000 (14:59 -0700)]
rustc: Don't inline in CGUs at -O0
This commit tweaks the behavior of inlining functions into multiple codegen
units when rustc is compiling in debug mode. Today rustc will unconditionally
treat `#[inline]` functions by translating them into all codegen units that
they're needed within, marking the linkage as `internal`. This commit changes
the behavior so that in debug mode (compiling at `-O0`) rustc will instead only
translate `#[inline]` functions into *one* codegen unit, forcing all other
codegen units to reference this one copy.
The goal here is to improve debug compile times by reducing the amount of
translation that happens on behalf of multiple codegen units. It was discovered
in #44941 that increasing the number of codegen units had the adverse side
effect of increasing the overal work done by the compiler, and the suspicion
here was that the compiler was inlining, translating, and codegen'ing more
functions with more codegen units (for example `String` would be basically
inlined into all codegen units if used). The strategy in this commit should
reduce the cost of `#[inline]` functions to being equivalent to one codegen
unit, which is only translating and codegen'ing inline functions once.
Collected [data] shows that this does indeed improve the situation from [before]
as the overall cpu-clock time increases at a much slower rate and when pinned to
one core rustc does not consume significantly more wall clock time than with one
codegen unit.
One caveat of this commit is that the symbol names for inlined functions that
are only translated once needed some slight tweaking. These inline functions
could be translated into multiple crates and we need to make sure the symbols
don't collideA so the crate name/disambiguator is mixed in to the symbol name
hash in these situations.
bors [Sun, 8 Oct 2017 00:40:58 +0000 (00:40 +0000)]
Auto merge of #44978 - jamesmunns:armv5te-os-atomics, r=alexcrichton
Allow atomic operations up to 32 bits
The ARMv5te platform does not have instruction-level support for atomics, however the kernel provides [user space helpers] which can be used to perform atomic operations. When linked with `libgcc`, the atomic symbols needed by Rust will be provided, rather than CPU level intrinsics.
[user space helpers]: https://www.kernel.org/doc/Documentation/arm/kernel_user_helpers.txt
32-bit versions of these kernel level helpers were introduced in Linux Kernel 2.6.12, and 64-bit version of these kernel level helpers were introduced in Linux Kernel 3.1. I have selected 32 bit versions as std currently only requires Linux version 2.6.18 and above as far as I am aware.
As this target is specifically linux and gnueabi, it is reasonable to assume the Linux Kernel and libc will be available for the target. There is a large performance penalty, as we are not using CPU level intrinsics, however this penalty is likely preferable to not having the target at all.
I have used this change in a custom target (along with xargo) to build std, as well as a number of higher level crates.
## Additional information
For reference, here is what a a code snippet decompiles to:
```rust
use std::sync::atomic::{AtomicIsize, Ordering};
bors [Sat, 7 Oct 2017 22:18:20 +0000 (22:18 +0000)]
Auto merge of #44841 - alexcrichton:thinlto, r=michaelwoerister
rustc: Implement ThinLTO
This commit is an implementation of LLVM's ThinLTO for consumption in rustc
itself. Currently today LTO works by merging all relevant LLVM modules into one
and then running optimization passes. "Thin" LTO operates differently by having
more sharded work and allowing parallelism opportunities between optimizing
codegen units. Further down the road Thin LTO also allows *incremental* LTO
which should enable even faster release builds without compromising on the
performance we have today.
This commit uses a `-Z thinlto` flag to gate whether ThinLTO is enabled. It then
also implements two forms of ThinLTO:
* In one mode we'll *only* perform ThinLTO over the codegen units produced in a
single compilation. That is, we won't load upstream rlibs, but we'll instead
just perform ThinLTO amongst all codegen units produced by the compiler for
the local crate. This is intended to emulate a desired end point where we have
codegen units turned on by default for all crates and ThinLTO allows us to do
this without performance loss.
* In anther mode, like full LTO today, we'll optimize all upstream dependencies
in "thin" mode. Unlike today, however, this LTO step is fully parallelized so
should finish much more quickly.
There's a good bit of comments about what the implementation is doing and where
it came from, but the tl;dr; is that currently most of the support here is
copied from upstream LLVM. This code duplication is done for a number of
reasons:
* Controlling parallelism means we can use the existing jobserver support to
avoid overloading machines.
* We will likely want a slightly different form of incremental caching which
integrates with our own incremental strategy, but this is yet to be
determined.
* This buys us some flexibility about when/where we run ThinLTO, as well as
having it tailored to fit our needs for the time being.
* Finally this allows us to reuse some artifacts such as our `TargetMachine`
creation, where all our options we used today aren't necessarily supported by
upstream LLVM yet.
My hope is that we can get some experience with this copy/paste in tree and then
eventually upstream some work to LLVM itself to avoid the duplication while
still ensuring our needs are met. Otherwise I fear that maintaining these
bindings may be quite costly over the years with LLVM updates!
Jorge Aparicio [Sat, 7 Oct 2017 18:48:25 +0000 (20:48 +0200)]
enable strict alignment (+strict-align) on ARMv6
As discovered in #44538 ARMv6 devices may or may not support unaligned memory accesses. ARMv6
Linux *seems* to have no problem with unaligned accesses but this is because the kernel is stepping
in to fix each unaligned memory access -- this incurs in a performance penalty.
This commit enforces aligned memory accesses on all our in-tree ARM targets that may be used with
ARMv6 devices. This should improve performance of Rust programs on ARMv6 devices. For the record,
clang also applies this attribute when targeting ARMv6 devices that are not running Darwin or
NetBSD.
Alex Crichton [Sun, 23 Jul 2017 15:14:38 +0000 (08:14 -0700)]
rustc: Implement ThinLTO
This commit is an implementation of LLVM's ThinLTO for consumption in rustc
itself. Currently today LTO works by merging all relevant LLVM modules into one
and then running optimization passes. "Thin" LTO operates differently by having
more sharded work and allowing parallelism opportunities between optimizing
codegen units. Further down the road Thin LTO also allows *incremental* LTO
which should enable even faster release builds without compromising on the
performance we have today.
This commit uses a `-Z thinlto` flag to gate whether ThinLTO is enabled. It then
also implements two forms of ThinLTO:
* In one mode we'll *only* perform ThinLTO over the codegen units produced in a
single compilation. That is, we won't load upstream rlibs, but we'll instead
just perform ThinLTO amongst all codegen units produced by the compiler for
the local crate. This is intended to emulate a desired end point where we have
codegen units turned on by default for all crates and ThinLTO allows us to do
this without performance loss.
* In anther mode, like full LTO today, we'll optimize all upstream dependencies
in "thin" mode. Unlike today, however, this LTO step is fully parallelized so
should finish much more quickly.
There's a good bit of comments about what the implementation is doing and where
it came from, but the tl;dr; is that currently most of the support here is
copied from upstream LLVM. This code duplication is done for a number of
reasons:
* Controlling parallelism means we can use the existing jobserver support to
avoid overloading machines.
* We will likely want a slightly different form of incremental caching which
integrates with our own incremental strategy, but this is yet to be
determined.
* This buys us some flexibility about when/where we run ThinLTO, as well as
having it tailored to fit our needs for the time being.
* Finally this allows us to reuse some artifacts such as our `TargetMachine`
creation, where all our options we used today aren't necessarily supported by
upstream LLVM yet.
My hope is that we can get some experience with this copy/paste in tree and then
eventually upstream some work to LLVM itself to avoid the duplication while
still ensuring our needs are met. Otherwise I fear that maintaining these
bindings may be quite costly over the years with LLVM updates!
bors [Sat, 7 Oct 2017 12:13:28 +0000 (12:13 +0000)]
Auto merge of #44913 - leavehouse:patch-1, r=BurntSushi
Fix TcpStream::local_addr docs example code
The local address's port is not 8080 in this example, that's the remote peer address port. On my machine, the local address is different every time, so I've changed `assert_eq` to only test the IP address
bors [Fri, 6 Oct 2017 15:30:32 +0000 (15:30 +0000)]
Auto merge of #45065 - arielb1:not-correct, r=nikomatsakis
fix logic error in #44269's `prune_cache_value_obligations`
We want to retain obligations that *contain* inference variables, not
obligations that *don't contain* them, in order to fix #43132. Because
of surrounding changes to inference, the ICE doesn't occur in its
original case, but I believe it could still be made to occur on master.
Maybe I should try to write a new test case? Certainly not right now
(I'm mainly trying to get us a beta that we can ship) but maybe before
we land this PR on nightly?
This seems to cause a 10% performance regression in my imprecise
attempt to benchmark item-body checking for #43613, but it's better to
be slow and right than fast and wrong. If we want to recover that, I
think we can change the constrained-type-parameter code to actually
give a list of projections that are important for resolving inference
variables and filter everything else out.
fix logic error in #44269's `prune_cache_value_obligations`
We want to retain obligations that *contain* inference variables, not
obligations that *don't contain* them, in order to fix #43132. Because
of surrounding changes to inference, the ICE doesn't occur in its
original case, but I believe it could still be made to occur on master.
Maybe I should try to write a new test case? Certainly not right now
(I'm mainly trying to get us a beta that we can ship) but maybe before
we land this PR on nightly?
This seems to cause a 10% performance regression in my imprecise
attempt to benchmark item-body checking for #43613, but it's better to
be slow and right than fast and wrong. If we want to recover that, I
think we can change the constrained-type-parameter code to actually
give a list of projections that are important for resolving inference
variables and filter everything else out.
bors [Fri, 6 Oct 2017 12:51:11 +0000 (12:51 +0000)]
Auto merge of #44734 - mchlrhw:wip/hashmap-entry-and-then, r=BurntSushi
Implement `and_modify` on `Entry`
## Motivation
`Entry`s are useful for allowing access to existing values in a map while also allowing default values to be inserted for absent keys. The existing API is similar to that of `Option`, where `or` and `or_with` can be used if the option variant is `None`.
The `Entry` API is, however, missing an equivalent of `Option`'s `and_then` method. If it were present it would be possible to modify an existing entry before calling `or_insert` without resorting to matching on the entry variant.
bors [Fri, 6 Oct 2017 10:20:14 +0000 (10:20 +0000)]
Auto merge of #44965 - oconnor663:res_init_glibc, r=dtolnay
replace libc::res_init with res_init_if_glibc_before_2_26
The previous workaround for gibc's res_init bug is not thread-safe on
other implementations of libc, and it can cause crashes. Use a runtime
check to make sure we only call res_init when we need to, which is also
when it's safe. See https://github.com/rust-lang/rust/issues/43592.
~This PR is returning an InvalidData IO error if the glibc version string fails to parse. We could also have treated that case as "not glibc", and gotten rid of the idea that these functions could return an error. (Though I'm not a huge fan of ignoring error returns from `res_init` in any case.) Do other folks agree with these design choices?~
I'm pretty new to hacking on libstd. Is there an easy way to build a toy rust program against my changes to test this, other than doing an entire `sudo make install` on my system? What's the usual workflow?
bors [Fri, 6 Oct 2017 05:37:43 +0000 (05:37 +0000)]
Auto merge of #44818 - petrochenkov:astymac2, r=jseyfried
Improve resolution of associated types in declarative macros 2.0
Make various identifier comparisons for associated types (and sometimes other associated items) hygienic.
Now declarative macros 2.0 can use `Self::AssocTy`, `TyParam::AssocTy`, `Trait<AssocTy = u8>` where `AssocTy` is an associated type of a trait `Trait` visible from the macro. Also, `Trait` can now be implemented inside the macro and specialization should work properly (fixes https://github.com/rust-lang/rust/pull/40847#issuecomment-310867299).
Jack O'Connor [Sun, 1 Oct 2017 21:42:18 +0000 (17:42 -0400)]
replace libc::res_init with res_init_if_glibc_before_2_26
The previous workaround for gibc's res_init bug is not thread-safe on
other implementations of libc, and it can cause crashes. Use a runtime
check to make sure we only call res_init when we need to, which is also
when it's safe. See https://github.com/rust-lang/rust/issues/43592.
steveklabnik [Thu, 5 Oct 2017 20:54:56 +0000 (16:54 -0400)]
Modify Rc/Arc language around mutability
There are a few exceptions to the rule that Arc/Rc are immutable. Rather
than dig into the details, add "generally" to hint at this difference,
as it's kind of a distraction at this point in the docs.
Additionally, Arc's docs were slightly different here generally, so add
in both the existing language and the exception.
kennytm [Thu, 5 Oct 2017 12:22:30 +0000 (20:22 +0800)]
Rollup merge of #44935 - vitiral:dependencies, r=steveklabnik
Add links to headers in README and CONTRIBUTING
this also adds dependencies to CONTRIBUTING
I'm just getting started building the rust compiler and noticed this information/ability was missing.
It was also missing the gdb dependency for running tests. I pulled the information out of `appveyor.yml` and recommended later than 7.1 because that is what [apt ships](https://packages.ubuntu.com/search?suite=trusty&keywords=gdb). Feel free to tell me something different!
kennytm [Thu, 5 Oct 2017 12:22:29 +0000 (20:22 +0800)]
Rollup merge of #44664 - budziq:contributing_external, r=steveklabnik
Corrected the CONTRIBUTING.md "External Dependencies" section
The "External Dependencies" section is a little outdated.
Please see following comments https://github.com/rust-lang/rust/pull/44567#issuecomment-329797331 https://github.com/rust-lang/rust/pull/44567#issuecomment-329799655 for rationale.
bors [Thu, 5 Oct 2017 10:50:11 +0000 (10:50 +0000)]
Auto merge of #45019 - aidanhs:aphs-no-trans-worker-panic, r=alexcrichton
Don't unwrap work item results as the panic trace is useless
Fixes #43402 now there's no multithreaded panic printouts
Also update a comment
--------
Likely regressed in #43506, where the code was changed to panic in worker threads on error.
Unwrapping gives zero extra information since the stack trace is so short, so we may as well just surface that there was an error and exit the thread properly. Because there are then no multithreaded printouts, I think it should mean the output of the test for #26199 is deterministic and not interleaved (thanks to @philipc https://github.com/rust-lang/rust/issues/43402#issuecomment-333835271 for a hint).
Sadly the output is now:
```
thread '<unnamed>' panicked at 'aborting due to worker thread panic', src/librustc_trans/back/write.rs:1643:20
note: Run with `RUST_BACKTRACE=1` for a backtrace.
error: could not write output to : No such file or directory
error: aborting due to previous error
```
but it's an improvement over the multi-panic situation before.
James Tucker [Wed, 4 Oct 2017 22:29:01 +0000 (15:29 -0700)]
zircon: update some more zx and lp constants
* Use size_t where size_t is used, while it's not critical on our
specifically supported architectures, this is more accurate.
* Update HND_SPECIAL_COUNT to the correct value, and give it the size
that enum is likely to be.
projects / rust.git / log