Clean up various LLVM FFI things in codegen_llvm
cc ```@ZuseZ4``` I touched some autodiff parts
The major change of this PR is [bfd88ce](bfd88cead0) which makes `CodegenCx` generic just like `GenericBuilder`
The other commits mostly took advantage of the new feature of making extern functions safe, but also just used some wrappers that were already there and shrunk unsafe blocks.
best reviewed commit-by-commit
remove `simd_fpow` and `simd_fpowi`
Discussed in https://github.com/rust-lang/rust/issues/137555
These functions are not exposed from `std::intrinsics::simd`, and not used anywhere outside of the compiler. They also don't lower to particularly good code at least on the major ISAs (I checked x86_64, aarch64, s390x, powerpc), where the vector is just spilled to the stack and scalar functions are used for the actual logic.
r? `@RalfJung`
intrinsics: unify rint, roundeven, nearbyint in a single round_ties_even intrinsic
LLVM has three intrinsics here that all do the same thing (when used in the default FP environment). There's no reason Rust needs to copy that historically-grown mess -- let's just have one intrinsic and leave it up to the LLVM backend to decide how to lower that.
Suggested by `@hanna-kruppe` in https://github.com/rust-lang/rust/issues/136459; Cc `@tgross35`
try-job: test-various
This improves the codegen for vector `select`, `gather`, `scatter` and
boolean reduction intrinsics and fixesrust-lang/portable-simd#316.
The current behavior of most mask operations during llvm codegen is to
truncate the mask vector to <N x i1>, telling llvm to use the least
significat bit. The exception is the `simd_bitmask` intrinsics, which
already used the most signifiant bit.
Since sse/avx instructions are defined to use the most significant bit,
truncating means that llvm has to insert a left shift to move the bit
into the most significant position, before the mask can actually be
used.
Similarly on aarch64, mask operations like blend work bit by bit,
repeating the least significant bit across the whole lane involves
shifting it into the sign position and then comparing against zero.
By shifting before truncating to <N x i1>, we tell llvm that we only
consider the most significant bit, removing the need for additional
shift instructions in the assembly.
the behavior of the type system not only depends on the current
assumptions, but also the currentnphase of the compiler. This is
mostly necessary as we need to decide whether and how to reveal
opaque types. We track this via the `TypingMode`.
The initial naming of "Abi" was an awful mistake, conveying wrong ideas
about how psABIs worked and even more about what the enum meant.
It was only meant to represent the way the value would be described to
a codegen backend as it was lowered to that intermediate representation.
It was never meant to mean anything about the actual psABI handling!
The conflation is because LLVM typically will associate a certain form
with a certain ABI, but even that does not hold when the special cases
that actually exist arise, plus the IR annotations that modify the ABI.
Reframe `rustc_abi::Abi` as the `BackendRepr` of the type, and rename
`BackendRepr::Aggregate` as `BackendRepr::Memory`. Unfortunately, due to
the persistent misunderstandings, this too is now incorrect:
- Scattered ABI-relevant code is entangled with BackendRepr
- We do not always pre-compute a correct BackendRepr that reflects how
we "actually" want this value to be handled, so we leave the backend
interface to also inject various special-cases here
- In some cases `BackendRepr::Memory` is a "real" aggregate, but in
others it is in fact using memory, and in some cases it is a scalar!
Our rustc-to-backend lowering code handles this sort of thing right now.
That will eventually be addressed by lifting duplicated lowering code
to either rustc_codegen_ssa or rustc_target as appropriate.
Continue to get rid of `ty::Const::{try_}eval*`
This PR mostly does:
* Removes all of the `try_eval_*` and `eval_*` helpers from `ty::Const`, and replace their usages with `try_to_*`.
* Remove `ty::Const::eval`.
* Rename `ty::Const::normalize` to `ty::Const::normalize_internal`. This function is still used in the normalization code itself.
* Fix some weirdness around the `TransmuteFrom` goal.
I'm happy to split it out further; for example, I could probably land the first part which removes the helpers, or the changes to codegen which are more obvious than the changes to tools.
r? BoxyUwU
Part of https://github.com/rust-lang/rust/issues/130704
Migrate `llvm::set_comdat` and `llvm::SetUniqueComdat` to LLVM-C FFI.
Note, now we can call `llvm::set_comdat` only when the target actually
supports adding comdat. As this has no convenient LLVM-C API, we
implement this as `TargetOptions::supports_comdat`.
Co-authored-by: Stuart Cook <Zalathar@users.noreply.github.com>
Add intrinsics `fmuladd{f16,f32,f64,f128}`. This computes `(a * b) +
c`, to be fused if the code generator determines that (i) the target
instruction set has support for a fused operation, and (ii) that the
fused operation is more efficient than the equivalent, separate pair
of `mul` and `add` instructions.
https://llvm.org/docs/LangRef.html#llvm-fmuladd-intrinsic
MIRI support is included for f32 and f64.
The codegen_cranelift uses the `fma` function from libc, which is a
correct implementation, but without the desired performance semantic. I
think this requires an update to cranelift to expose a suitable
instruction in its IR.
I have not tested with codegen_gcc, but it should behave the same
way (using `fma` from libc).
llvm: replace some deprecated functions
`LLVMMDStringInContext` and `LLVMMDNodeInContext` are deprecated, replace them with `LLVMMDStringInContext2` and `LLVMMDNodeInContext2`.
Also replace `Value` with `Metadata` in some function signatures for better consistency.
Use -0.0 in `intrinsics::simd::reduce_add_unordered`
-0.0 is the actual neutral additive float, not +0.0, and this matters to codegen.
try-job: aarch64-gnu
simd_shuffle intrinsic: allow argument to be passed as vector
See https://github.com/rust-lang/rust/issues/128738 for context.
I'd like to get rid of [this hack](6c0b89dfac/compiler/rustc_codegen_ssa/src/mir/block.rs (L922-L935)). https://github.com/rust-lang/rust/pull/128537 almost lets us do that since constant SIMD vectors will then be passed as immediate arguments. However, simd_shuffle for some reason actually takes an *array* as argument, not a vector, so the hack is still required to ensure that the array becomes an immediate (which then later stages of codegen convert into a vector, as that's what LLVM needs).
This PR prepares simd_shuffle to also support a vector as the `idx` argument. Once this lands, stdarch can hopefully be updated to pass `idx` as a vector, and then support for arrays can be removed, which finally lets us get rid of that hack.
Since LLVM <https://reviews.llvm.org/D99439> (4c7f820b2b20, "Update
@llvm.powi to handle different int sizes for the exponent"), the size of
the integer can be specified for the `powi` intrinsic. Make use of this
so it is more obvious that integer size is consistent across all float
types.
This feature is available since LLVM 13 (October 2021). Based on
bootstrap we currently support >= 17.0, so there should be no support
problems.
Add `select_unpredictable` to force LLVM to use CMOV
Since https://reviews.llvm.org/D118118, LLVM will no longer turn CMOVs into branches if it comes from a `select` marked with an `unpredictable` metadata attribute.
This PR introduces `core::intrinsics::select_unpredictable` which emits such a `select` and uses it in the implementation of `binary_search_by`.
Since https://reviews.llvm.org/D118118, LLVM will no longer turn CMOVs
into branches if it comes from a `select` marked with an `unpredictable`
metadata attribute.
This PR introduces `core::intrinsics::select_unpredictable` which emits
such a `select` and uses it in the implementation of `binary_search_by`.
