Skip to content

Add ADX-related intrinsics #631

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jan 7, 2019
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
46 changes: 46 additions & 0 deletions coresimd/x86/adx.rs
Original file line number Diff line number Diff line change
@@ -0,0 +1,46 @@
#[cfg(test)]
use stdsimd_test::assert_instr;

#[allow(improper_ctypes)]
extern "unadjusted" {
#[link_name = "llvm.x86.addcarry.u32"]
fn llvm_addcarry_u32(a: u8, b: u32, c: u32) -> (u8, u32);
#[link_name = "llvm.x86.subborrow.u32"]
fn llvm_subborrow_u32(a: u8, b: u32, c: u32) -> (u8, u32);
}

/// Add unsigned 32-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry flag), and store the unsigned 32-bit result in out, and the carry-out
/// is returned (carry or overflow flag).
#[inline]
#[cfg_attr(test, assert_instr(adc))]
#[stable(feature = "simd_x86_adx", since = "1.33.0")]
pub unsafe fn _addcarry_u32(c_in: u8, a: u32, b: u32, out: &mut u32) -> u8 {
let (a, b) = llvm_addcarry_u32(c_in, a, b);
*out = b;
a
}

/// Add unsigned 32-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry or overflow flag), and store the unsigned 32-bit result in out, and
/// the carry-out is returned (carry or overflow flag).
#[inline]
#[target_feature(enable = "adx")]
#[cfg_attr(test, assert_instr(adc))]
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think this is correct, it should say adcx

Copy link
Contributor

@gnzlbg gnzlbg Jan 31, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@jethrogb Could you submit a PR with that fix?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think some work is involved in actually making it work that way, see #663

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've opened an issue so that we don't forget about this #666

#[stable(feature = "simd_x86_adx", since = "1.33.0")]
#[cfg(not(stage0))]
pub unsafe fn _addcarryx_u32(c_in: u8, a: u32, b: u32, out: &mut u32) -> u8 {
_addcarry_u32(c_in, a, b, out)
}

/// Add unsigned 32-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry or overflow flag), and store the unsigned 32-bit result in out, and
/// the carry-out is returned (carry or overflow flag).
#[inline]
#[cfg_attr(test, assert_instr(sbb))]
#[stable(feature = "simd_x86_adx", since = "1.33.0")]
pub unsafe fn _subborrow_u32(c_in: u8, a: u32, b: u32, out: &mut u32) -> u8 {
let (a, b) = llvm_subborrow_u32(c_in, a, b);
*out = b;
a
}
3 changes: 3 additions & 0 deletions coresimd/x86/mod.rs
Original file line number Diff line number Diff line change
Expand Up @@ -593,6 +593,9 @@ pub use self::rdrand::*;
mod sha;
pub use self::sha::*;

mod adx;
pub use self::adx::*;

#[cfg(test)]
use stdsimd_test::assert_instr;

Expand Down
46 changes: 46 additions & 0 deletions coresimd/x86_64/adx.rs
Original file line number Diff line number Diff line change
@@ -0,0 +1,46 @@
#[cfg(test)]
use stdsimd_test::assert_instr;

#[allow(improper_ctypes)]
extern "unadjusted" {
#[link_name = "llvm.x86.addcarry.u64"]
fn llvm_addcarry_u64(a: u8, b: u64, c: u64) -> (u8, u64);
#[link_name = "llvm.x86.subborrow.u64"]
fn llvm_subborrow_u64(a: u8, b: u64, c: u64) -> (u8, u64);
}

/// Add unsigned 64-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry flag), and store the unsigned 64-bit result in out, and the carry-out
/// is returned (carry or overflow flag).
#[inline]
#[cfg_attr(test, assert_instr(adc))]
#[stable(feature = "simd_x86_adx", since = "1.33.0")]
pub unsafe fn _addcarry_u64(c_in: u8, a: u64, b: u64, out: &mut u64) -> u8 {
let (a, b) = llvm_addcarry_u64(c_in, a, b);
*out = b;
a
}

/// Add unsigned 64-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry or overflow flag), and store the unsigned 64-bit result in out, and
/// the carry-out is returned (carry or overflow flag).
#[inline]
#[target_feature(enable = "adx")]
#[cfg_attr(test, assert_instr(adc))]
#[stable(feature = "simd_x86_adx", since = "1.33.0")]
#[cfg(not(stage0))]
pub unsafe fn _addcarryx_u64(c_in: u8, a: u64, b: u64, out: &mut u64) -> u8 {
_addcarry_u64(c_in, a, b, out)
}

/// Add unsigned 64-bit integers a and b with unsigned 8-bit carry-in c_in
/// (carry or overflow flag), and store the unsigned 64-bit result in out, and
/// the carry-out is returned (carry or overflow flag).
#[inline]
#[cfg_attr(test, assert_instr(sbb))]
#[stable(feature = "simd_x86_adx", since = "1.33.0")]
pub unsafe fn _subborrow_u64(c_in: u8, a: u64, b: u64, out: &mut u64) -> u8 {
let (a, b) = llvm_subborrow_u64(c_in, a, b);
*out = b;
a
}
3 changes: 3 additions & 0 deletions coresimd/x86_64/mod.rs
Original file line number Diff line number Diff line change
Expand Up @@ -41,3 +41,6 @@ pub use self::rdrand::*;

mod cmpxchg16b;
pub use self::cmpxchg16b::*;

mod adx;
pub use self::adx::*;
3 changes: 2 additions & 1 deletion crates/coresimd/src/lib.rs
Original file line number Diff line number Diff line change
Expand Up @@ -37,7 +37,8 @@
mips_target_feature,
powerpc_target_feature,
wasm_target_feature,
abi_unadjusted
abi_unadjusted,
adx_target_feature
)]
// NB: When running nvptx/nvptx64 cross tests, enabling "integer_atomics" yields
// a compile-time error: 'unknown feature `integer_atomics`'. This ought to be
Expand Down
8 changes: 6 additions & 2 deletions crates/stdsimd-verify/tests/x86-intel.rs
Original file line number Diff line number Diff line change
Expand Up @@ -240,8 +240,12 @@ fn matches(rust: &Function, intel: &Intrinsic) -> Result<(), String> {
// ensuring that we've actually enabled the right instruction
// set for this intrinsic.
match rust.name {
"_bswap" => {}
"_bswap64" => {}
"_bswap" | "_bswap64" => {}

// These don't actually have a target feature unlike their brethren with
// the `x` inside the name which requires adx
"_addcarry_u32" | "_addcarry_u64" | "_subborrow_u32" | "_subborrow_u64" => {}

_ => {
if intel.cpuid.is_empty() {
bail!("missing cpuid for {}", rust.name);
Expand Down
6 changes: 6 additions & 0 deletions stdsimd/arch/detect/arch/x86.rs
Original file line number Diff line number Diff line change
Expand Up @@ -230,6 +230,10 @@ macro_rules! is_x86_feature_detected {
cfg!(target_feature = "cmpxchg16b") || $crate::arch::detect::check_for(
$crate::arch::detect::Feature::cmpxchg16b)
};
("adx") => {
cfg!(target_feature = "adx") || $crate::arch::detect::check_for(
$crate::arch::detect::Feature::adx)
};
($t:tt) => {
compile_error!(concat!("unknown target feature: ", $t))
};
Expand Down Expand Up @@ -322,4 +326,6 @@ pub enum Feature {
xsavec,
/// CMPXCH16B, a 16-byte compare-and-swap instruction
cmpxchg16b,
/// ADX, Intel ADX (Multi-Precision Add-Carry Instruction Extensions)
adx,
}
6 changes: 6 additions & 0 deletions stdsimd/arch/detect/os/x86.rs
Original file line number Diff line number Diff line change
Expand Up @@ -124,6 +124,7 @@ fn detect_features() -> cache::Initializer {
enable(proc_info_ecx, 1, Feature::pclmulqdq);
enable(proc_info_ecx, 30, Feature::rdrand);
enable(extended_features_ebx, 18, Feature::rdseed);
enable(extended_features_ebx, 19, Feature::adx);
enable(proc_info_edx, 4, Feature::tsc);
enable(proc_info_edx, 23, Feature::mmx);
enable(proc_info_edx, 24, Feature::fxsr);
Expand Down Expand Up @@ -290,6 +291,7 @@ mod tests {
println!("xsaves: {:?}", is_x86_feature_detected!("xsaves"));
println!("xsavec: {:?}", is_x86_feature_detected!("xsavec"));
println!("cmpxchg16b: {:?}", is_x86_feature_detected!("cmpxchg16b"));
println!("adx: {:?}", is_x86_feature_detected!("adx"));
}

#[test]
Expand Down Expand Up @@ -350,5 +352,9 @@ mod tests {
is_x86_feature_detected!("cmpxchg16b"),
information.cmpxchg16b(),
);
assert_eq!(
is_x86_feature_detected!("adx"),
information.adx(),
);
}
}