LLVM isn't able to remove the alloca for the unaligned block in the SIMD tail in some cases, so doing this helps SRoA work in cases where it currently doesn't. Found in the `replace_with` RFC discussion.
Change RangeInclusive to a three-field struct.
Fix#45222.
This PR also reverts #48012 (i.e. removed the `try_fold`/`try_rfold` specialization for `RangeInclusive`) because LLVM no longer has trouble recognizing a RangeInclusive loop.
Upgrade to LLVM's master branch (LLVM 7)
### Current status
~~Blocked on a [performance regression](https://github.com/rust-lang/rust/pull/51966#issuecomment-402320576). The performance regression has an [upstream LLVM issue](https://bugs.llvm.org/show_bug.cgi?id=38047) and has also [been bisected](https://reviews.llvm.org/D44282) to an LLVM revision.~~
Ready to merge!
---
This commit upgrades the main LLVM submodule to LLVM's current master branch.
The LLD submodule is updated in tandem as well as compiler-builtins.
Along the way support was also added for LLVM 7's new features. This primarily
includes the support for custom section concatenation natively in LLD so we now
add wasm custom sections in LLVM IR rather than having custom support in rustc
itself for doing so.
Some other miscellaneous changes are:
* We now pass `--gc-sections` to `wasm-ld`
* The optimization level is now passed to `wasm-ld`
* A `--stack-first` option is passed to LLD to have stack overflow always cause
a trap instead of corrupting static data
* The wasm target for LLVM switched to `wasm32-unknown-unknown`.
* The syntax for aligned pointers has changed in LLVM IR and tests are updated
to reflect this.
* ~~The `thumbv6m-none-eabi` target is disabled due to an [LLVM bug][llbug]~~
Nowadays we've been mostly only upgrading whenever there's a major release of
LLVM but enough changes have been happening on the wasm target that there's been
growing motivation for quite some time now to upgrade out version of LLD. To
upgrade LLD, however, we need to upgrade LLVM to avoid needing to build yet
another version of LLVM on the builders.
The revision of LLVM in use here is arbitrarily chosen. We will likely need to
continue to update it over time if and when we discover bugs. Once LLVM 7 is
fully released we can switch to that channel as well.
[llbug]: https://bugs.llvm.org/show_bug.cgi?id=37382
cc #50543
This commit upgrades the main LLVM submodule to LLVM's current master branch.
The LLD submodule is updated in tandem as well as compiler-builtins.
Along the way support was also added for LLVM 7's new features. This primarily
includes the support for custom section concatenation natively in LLD so we now
add wasm custom sections in LLVM IR rather than having custom support in rustc
itself for doing so.
Some other miscellaneous changes are:
* We now pass `--gc-sections` to `wasm-ld`
* The optimization level is now passed to `wasm-ld`
* A `--stack-first` option is passed to LLD to have stack overflow always cause
a trap instead of corrupting static data
* The wasm target for LLVM switched to `wasm32-unknown-unknown`.
* The syntax for aligned pointers has changed in LLVM IR and tests are updated
to reflect this.
* The `thumbv6m-none-eabi` target is disabled due to an [LLVM bug][llbug]
Nowadays we've been mostly only upgrading whenever there's a major release of
LLVM but enough changes have been happening on the wasm target that there's been
growing motivation for quite some time now to upgrade out version of LLD. To
upgrade LLD, however, we need to upgrade LLVM to avoid needing to build yet
another version of LLVM on the builders.
The revision of LLVM in use here is arbitrarily chosen. We will likely need to
continue to update it over time if and when we discover bugs. Once LLVM 7 is
fully released we can switch to that channel as well.
[llbug]: https://bugs.llvm.org/show_bug.cgi?id=37382
Store scalar pair bools as i8 in memory
We represent `bool` as `i1` in a `ScalarPair`, unlike other aggregates,
to optimize IR for checked operators and the like. With this patch, we
still do so when the pair is an immediate value, but we use the `i8`
memory type when the value is loaded or stored as an LLVM aggregate.
So `(bool, bool)` looks like an `{ i1, i1 }` immediate, but `{ i8, i8 }`
in memory. When a pair is a direct function argument, `PassMode::Pair`,
it is still passed using the immediate `i1` type, but as a return value
it will use the `i8` memory type. Also, `bool`-like` enum tags will now
use scalar pairs when possible, where they were previously excluded due
to optimization issues.
Fixes#51516.
Closes#51566.
r? @eddyb
cc @nox
We represent `bool` as `i1` in a `ScalarPair`, unlike other aggregates,
to optimize IR for checked operators and the like. With this patch, we
still do so when the pair is an immediate value, but we use the `i8`
memory type when the value is loaded or stored as an LLVM aggregate.
So `(bool, bool)` looks like an `{ i1, i1 }` immediate, but `{ i8, i8 }`
in memory. When a pair is a direct function argument, `PassMode::Pair`,
it is still passed using the immediate `i1` type, but as a return value
it will use the `i8` memory type. Also, `bool`-like` enum tags will now
use scalar pairs when possible, where they were previously excluded due
to optimization issues.
Add simd math intrinsics and gather/scatter
This PR adds simd math intrinsics for floating-point vectors (sqrt, sin, cos, pow, exp, log, fma, abs, etc.) and the generic simd gather/scatter intrinsics.
OOM can't unwind today, and historically it's been optimized as if it can't
unwind. This accidentally regressed with recent changes to the OOM handler, so
this commit adds in a codegen test to assert that everything gets optimized away
after the OOM function is approrpiately classified as nounwind
Closes#50925
This used to be disabled due to LLVM bugs in the handling of
noalias information in conjunction with unwinding. However,
according to #31681 all known LLVM bugs have been fixed by
LLVM 6.0, so it's probably time to reenable this optimization.
Noalias annotations will not be emitted by default if either
-C panic=abort (as previously) or LLVM >= 6.0 (new).
-Z mutable-noalias=no is left as an escape-hatch to allow
debugging problems suspected to stem from this change.
This reworks the force-frame-pointer PR to explicitly only consider the
value of the flag if it is provided, and use a target default otherwise.
Something that was tried but not kept was renaming the flag to
`frame-pointer`, because for flag `frame-pointer=no`, there is no
guarante, that LLVM will elide *all* the frame pointers; oposite of what
the literal reading of the flag would suggest.
We apparently used to generate bad/incomplete debug info causing
debuggers not to find symbols of stack allocated variables. This was
somehow worked around by having frame pointers.
With the current codegen, this seems no longer necessary, so we can
remove the code that force-enables frame pointers whenever debug info
is requested.
Since certain situations, like profiling code profit from having frame
pointers, we add a -Cforce-frame-pointers flag to always enable frame
pointers.
Fixes#11906
Previously the features specified to LLVM via `-C target-feature` were only
reflected in the `TargetMachine` but this change *also* reflects these and the
base features inside each function itself. This change matches clang and...
Closesrust-lang-nursery/stdsimd#427
Long ago (#40549) we enabled the `uwtable` attribute on Windows by default
(even with `-C panic=abort`) to allow unwinding binaries for [stack unwinding
information][winstack]. It looks like this same issue is [plaguing][arm1]
Gecko's Android platforms [as well][arm2]. This commit applies the same fix
as #40549 except that this time it's applied for all Android targets.
Generating a `-C panic=abort` binary for `armv7-linux-androideabi` before this
commit generated a number of `cantunwind` functions (detected with `readelf -u`)
but after this commit they all list appropriate unwind information.
Closes#49867
[winstack]: https://bugzilla.mozilla.org/show_bug.cgi?id=1302078
[arm1]: https://bugzilla.mozilla.org/show_bug.cgi?id=1453220
[arm2]: https://bugzilla.mozilla.org/show_bug.cgi?id=1451741
Better document the implementors of Clone and Copy
There are two parts to this change. The first part is a change to the compiler and to the standard library (specifically, libcore) to allow implementations of `Clone` and `Copy` to be written for a subset of builtin types. By adding these implementations to libcore, they now show up in the documentation. This is a [breaking-change] for users of `#![no_core]`, because they will now have to supply their own copy of the implementations of `Clone` and `Copy` that were added in libcore.
The second part is purely a documentation change to document the other implementors of `Clone` and `Copy` that cannot be described in Rust code (yet) and are thus provided by the compiler.
Fixes#25893
implement minmax intrinsics
This adds the `simd_{fmin,fmax}` intrinsics, which do a vertical (lane-wise) `min`/`max` for floating point vectors that's equivalent to Rust's `min`/`max` for `f32`/`f64`.
It might make sense to make `{f32,f64}::{min,max}` use the `minnum` and `minmax` intrinsics as well.
---
~~HELP: I need some help with these. Either I should go to sleep or there must be something that I must be missing. AFAICT I am calling the `maxnum` builder correctly, yet rustc/LLVM seem to insert a call to `llvm.minnum` there instead...~~ EDIT: Rust's LLVM version is too old :/
