Auto merge of rust-lang#111999 - scottmcm:codegen-less-memcpy, r=compiler-errors

Use `load`+`store` instead of `memcpy` for small integer arrays

I was inspired by rust-lang#98892 to see whether, rather than making `mem::swap` do something smart in the library, we could update MIR assignments like `*_1 = *_2` to do something smarter than `memcpy` for sufficiently-small types that doing it inline is going to be better than a `memcpy` call in assembly anyway.  After all, special code may help `mem::swap`, but if the "obvious" MIR can just result in the correct thing that helps everything -- other code like `mem::replace`, people doing it manually, and just passing around by value in general -- as well as makes MIR inlining happier since it doesn't need to deal with all the complicated library code if it just sees a couple assignments.

LLVM will turn the short, known-length `memcpy`s into direct instructions in the backend, but that's too late for it to be able to remove `alloca`s.  In general, replacing `memcpy`s with typed instructions is hard in the middle-end -- even for `memcpy.inline` where it knows it won't be a function call -- is hard [due to poison propagation issues](https://rust-lang.zulipchat.com/#narrow/stream/187780-t-compiler.2Fwg-llvm/topic/memcpy.20vs.20load-store.20for.20MIR.20assignments/near/360376712).  So because we know more about the type invariants -- these are typed copies -- rustc can emit something more specific, allowing LLVM to `mem2reg` away the `alloca`s in some situations.

rust-lang#52051 previously did something like this in the library for `mem::swap`, but it ended up regressing during enabling mir inlining (rust-lang@cbbf06b), so this has been suboptimal on stable for ≈5 releases now.

The code in this PR is narrowly targeted at just integer arrays in LLVM, but works via a new method on the [`LayoutTypeMethods`](https://doc.rust-lang.org/nightly/nightly-rustc/rustc_codegen_ssa/traits/trait.LayoutTypeMethods.html) trait, so specific backends based on cg_ssa can enable this for more situations over time, as we find them.  I don't want to try to bite off too much in this PR, though.  (Transparent newtypes and simple things like the 3×usize `String` would be obvious candidates for a follow-up.)

Codegen demonstrations: <https://llvm.godbolt.org/z/fK8hT9aqv>

Before:
```llvm
define void `@swap_rgb48_old(ptr` noalias nocapture noundef align 2 dereferenceable(6) %x, ptr noalias nocapture noundef align 2 dereferenceable(6) %y) unnamed_addr #1 {
  %a.i = alloca [3 x i16], align 2
  call void `@llvm.lifetime.start.p0(i64` 6, ptr nonnull %a.i)
  call void `@llvm.memcpy.p0.p0.i64(ptr` noundef nonnull align 2 dereferenceable(6) %a.i, ptr noundef nonnull align 2 dereferenceable(6) %x, i64 6, i1 false)
  tail call void `@llvm.memcpy.p0.p0.i64(ptr` noundef nonnull align 2 dereferenceable(6) %x, ptr noundef nonnull align 2 dereferenceable(6) %y, i64 6, i1 false)
  call void `@llvm.memcpy.p0.p0.i64(ptr` noundef nonnull align 2 dereferenceable(6) %y, ptr noundef nonnull align 2 dereferenceable(6) %a.i, i64 6, i1 false)
  call void `@llvm.lifetime.end.p0(i64` 6, ptr nonnull %a.i)
  ret void
}
```
Note it going to stack:
```nasm
swap_rgb48_old:                         # `@swap_rgb48_old`
        movzx   eax, word ptr [rdi + 4]
        mov     word ptr [rsp - 4], ax
        mov     eax, dword ptr [rdi]
        mov     dword ptr [rsp - 8], eax
        movzx   eax, word ptr [rsi + 4]
        mov     word ptr [rdi + 4], ax
        mov     eax, dword ptr [rsi]
        mov     dword ptr [rdi], eax
        movzx   eax, word ptr [rsp - 4]
        mov     word ptr [rsi + 4], ax
        mov     eax, dword ptr [rsp - 8]
        mov     dword ptr [rsi], eax
        ret
```

Now:
```llvm
define void `@swap_rgb48(ptr` noalias nocapture noundef align 2 dereferenceable(6) %x, ptr noalias nocapture noundef align 2 dereferenceable(6) %y) unnamed_addr #0 {
start:
  %0 = load <3 x i16>, ptr %x, align 2
  %1 = load <3 x i16>, ptr %y, align 2
  store <3 x i16> %1, ptr %x, align 2
  store <3 x i16> %0, ptr %y, align 2
  ret void
}
```
still lowers to `dword`+`word` operations, but has no stack traffic:
```nasm
swap_rgb48:                             # `@swap_rgb48`
        mov     eax, dword ptr [rdi]
        movzx   ecx, word ptr [rdi + 4]
        movzx   edx, word ptr [rsi + 4]
        mov     r8d, dword ptr [rsi]
        mov     dword ptr [rdi], r8d
        mov     word ptr [rdi + 4], dx
        mov     word ptr [rsi + 4], cx
        mov     dword ptr [rsi], eax
        ret
```

And as a demonstration that this isn't just `mem::swap`, a `mem::replace` on a small array (since replace doesn't use swap since rust-lang#83022), which used to be `memcpy`s in LLVM changes in IR
```llvm
define void `@replace_short_array(ptr` noalias nocapture noundef sret([3 x i32]) dereferenceable(12) %0, ptr noalias noundef align 4 dereferenceable(12) %r, ptr noalias nocapture noundef readonly dereferenceable(12) %v) unnamed_addr #0 {
start:
  %1 = load <3 x i32>, ptr %r, align 4
  store <3 x i32> %1, ptr %0, align 4
  %2 = load <3 x i32>, ptr %v, align 4
  store <3 x i32> %2, ptr %r, align 4
  ret void
}
```
but that lowers to reasonable `dword`+`qword` instructions still
```nasm
replace_short_array:                    # `@replace_short_array`
        mov     rax, rdi
        mov     rcx, qword ptr [rsi]
        mov     edi, dword ptr [rsi + 8]
        mov     dword ptr [rax + 8], edi
        mov     qword ptr [rax], rcx
        mov     rcx, qword ptr [rdx]
        mov     edx, dword ptr [rdx + 8]
        mov     dword ptr [rsi + 8], edx
        mov     qword ptr [rsi], rcx
        ret
```

Author: bors

compiler/rustc_codegen_llvm/src/type_.rs

@@ -288,6 +288,9 @@ impl<'ll, 'tcx> LayoutTypeMethods<'tcx> for CodegenCx<'ll, 'tcx> {
     fn reg_backend_type(&self, ty: &Reg) -> &'ll Type {
         ty.llvm_type(self)
     }
+    fn scalar_copy_backend_type(&self, layout: TyAndLayout<'tcx>) -> Option<Self::Type> {
+        layout.scalar_copy_llvm_type(self)
+    }
 }
 
 impl<'ll, 'tcx> TypeMembershipMethods<'tcx> for CodegenCx<'ll, 'tcx> {

compiler/rustc_codegen_llvm/src/type_of.rs

@@ -6,6 +6,7 @@ use rustc_middle::bug;
 use rustc_middle::ty::layout::{FnAbiOf, LayoutOf, TyAndLayout};
 use rustc_middle::ty::print::{with_no_trimmed_paths, with_no_visible_paths};
 use rustc_middle::ty::{self, Ty, TypeVisitableExt};
+use rustc_target::abi::HasDataLayout;
 use rustc_target::abi::{Abi, Align, FieldsShape};
 use rustc_target::abi::{Int, Pointer, F32, F64};
 use rustc_target::abi::{PointeeInfo, Scalar, Size, TyAbiInterface, Variants};
@@ -192,6 +193,7 @@ pub trait LayoutLlvmExt<'tcx> {
     ) -> &'a Type;
     fn llvm_field_index<'a>(&self, cx: &CodegenCx<'a, 'tcx>, index: usize) -> u64;
     fn pointee_info_at<'a>(&self, cx: &CodegenCx<'a, 'tcx>, offset: Size) -> Option<PointeeInfo>;
+    fn scalar_copy_llvm_type<'a>(&self, cx: &CodegenCx<'a, 'tcx>) -> Option<&'a Type>;
 }
 
 impl<'tcx> LayoutLlvmExt<'tcx> for TyAndLayout<'tcx> {
@@ -414,4 +416,35 @@ impl<'tcx> LayoutLlvmExt<'tcx> for TyAndLayout<'tcx> {
         cx.pointee_infos.borrow_mut().insert((self.ty, offset), result);
         result
     }
+
+    fn scalar_copy_llvm_type<'a>(&self, cx: &CodegenCx<'a, 'tcx>) -> Option<&'a Type> {
+        debug_assert!(self.is_sized());
+
+        // FIXME: this is a fairly arbitrary choice, but 128 bits on WASM
+        // (matching the 128-bit SIMD types proposal) and 256 bits on x64
+        // (like AVX2 registers) seems at least like a tolerable starting point.
+        let threshold = cx.data_layout().pointer_size * 4;
+        if self.layout.size() > threshold {
+            return None;
+        }
+
+        // Vectors, even for non-power-of-two sizes, have the same layout as
+        // arrays but don't count as aggregate types
+        if let FieldsShape::Array { count, .. } = self.layout.fields()
+            && let element = self.field(cx, 0)
+            && element.ty.is_integral()
+        {
+            // `cx.type_ix(bits)` is tempting here, but while that works great
+            // for things that *stay* as memory-to-memory copies, it also ends
+            // up suppressing vectorization as it introduces shifts when it
+            // extracts all the individual values.
+
+            let ety = element.llvm_type(cx);
+            return Some(cx.type_vector(ety, *count));
+        }
+
+        // FIXME: The above only handled integer arrays; surely more things
+        // would also be possible. Be careful about provenance, though!
+        None
+    }
 }

compiler/rustc_codegen_ssa/src/base.rs

@@ -380,7 +380,19 @@ pub fn memcpy_ty<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>(
         return;
     }
 
-    bx.memcpy(dst, dst_align, src, src_align, bx.cx().const_usize(size), flags);
+    if flags == MemFlags::empty()
+        && let Some(bty) = bx.cx().scalar_copy_backend_type(layout)
+    {
+        // I look forward to only supporting opaque pointers
+        let pty = bx.type_ptr_to(bty);
+        let src = bx.pointercast(src, pty);
+        let dst = bx.pointercast(dst, pty);
+
+        let temp = bx.load(bty, src, src_align);
+        bx.store(temp, dst, dst_align);
+    } else {
+        bx.memcpy(dst, dst_align, src, src_align, bx.cx().const_usize(size), flags);
+    }
 }
 
 pub fn codegen_instance<'a, 'tcx: 'a, Bx: BuilderMethods<'a, 'tcx>>(

compiler/rustc_codegen_ssa/src/traits/type_.rs

@@ -126,6 +126,28 @@ pub trait LayoutTypeMethods<'tcx>: Backend<'tcx> {
         index: usize,
         immediate: bool,
     ) -> Self::Type;
+
+    /// A type that can be used in a [`super::BuilderMethods::load`] +
+    /// [`super::BuilderMethods::store`] pair to implement a *typed* copy,
+    /// such as a MIR `*_0 = *_1`.
+    ///
+    /// It's always legal to return `None` here, as the provided impl does,
+    /// in which case callers should use [`super::BuilderMethods::memcpy`]
+    /// instead of the `load`+`store` pair.
+    ///
+    /// This can be helpful for things like arrays, where the LLVM backend type
+    /// `[3 x i16]` optimizes to three separate loads and stores, but it can
+    /// instead be copied via an `i48` that stays as the single `load`+`store`.
+    /// (As of 2023-05 LLVM cannot necessarily optimize away a `memcpy` in these
+    /// cases, due to `poison` handling, but in codegen we have more information
+    /// about the type invariants, so can emit something better instead.)
+    ///
+    /// This *should* return `None` for particularly-large types, where leaving
+    /// the `memcpy` may well be important to avoid code size explosion.
+    fn scalar_copy_backend_type(&self, layout: TyAndLayout<'tcx>) -> Option<Self::Type> {
+        let _ = layout;
+        None
+    }
 }
 
 // For backends that support CFI using type membership (i.e., testing whether a given pointer is

tests/codegen/array-codegen.rs

@@ -0,0 +1,35 @@
+// compile-flags: -O -C no-prepopulate-passes
+// min-llvm-version: 15.0 (for opaque pointers)
+
+#![crate_type = "lib"]
+
+// CHECK-LABEL: @array_load
+#[no_mangle]
+pub fn array_load(a: &[u8; 4]) -> [u8; 4] {
+    // CHECK: %0 = alloca [4 x i8], align 1
+    // CHECK: %[[TEMP1:.+]] = load <4 x i8>, ptr %a, align 1
+    // CHECK: store <4 x i8> %[[TEMP1]], ptr %0, align 1
+    // CHECK: %[[TEMP2:.+]] = load i32, ptr %0, align 1
+    // CHECK: ret i32 %[[TEMP2]]
+    *a
+}
+
+// CHECK-LABEL: @array_store
+#[no_mangle]
+pub fn array_store(a: [u8; 4], p: &mut [u8; 4]) {
+    // CHECK: %a = alloca [4 x i8]
+    // CHECK: %[[TEMP:.+]] = load <4 x i8>, ptr %a, align 1
+    // CHECK-NEXT: store <4 x i8> %[[TEMP]], ptr %p, align 1
+    *p = a;
+}
+
+// CHECK-LABEL: @array_copy
+#[no_mangle]
+pub fn array_copy(a: &[u8; 4], p: &mut [u8; 4]) {
+    // CHECK: %[[LOCAL:.+]] = alloca [4 x i8], align 1
+    // CHECK: %[[TEMP1:.+]] = load <4 x i8>, ptr %a, align 1
+    // CHECK: store <4 x i8> %[[TEMP1]], ptr %[[LOCAL]], align 1
+    // CHECK: %[[TEMP2:.+]] = load <4 x i8>, ptr %[[LOCAL]], align 1
+    // CHECK: store <4 x i8> %[[TEMP2]], ptr %p, align 1
+    *p = *a;
+}

tests/codegen/mem-replace-simple-type.rs

@@ -32,3 +32,14 @@ pub fn replace_ref_str<'a>(r: &mut &'a str, v: &'a str) -> &'a str {
     // CHECK: ret { ptr, i64 } %[[P2]]
     std::mem::replace(r, v)
 }
+
+#[no_mangle]
+// CHECK-LABEL: @replace_short_array(
+pub fn replace_short_array(r: &mut [u32; 3], v: [u32; 3]) -> [u32; 3] {
+    // CHECK-NOT: alloca
+    // CHECK: %[[R:.+]] = load <3 x i32>, ptr %r, align 4
+    // CHECK: store <3 x i32> %[[R]], ptr %0
+    // CHECK: %[[V:.+]] = load <3 x i32>, ptr %v, align 4
+    // CHECK: store <3 x i32> %[[V]], ptr %r
+    std::mem::replace(r, v)
+}

tests/codegen/swap-simd-types.rs

@@ -30,3 +30,12 @@ pub fn swap_m256_slice(x: &mut [__m256], y: &mut [__m256]) {
         x.swap_with_slice(y);
     }
 }
+
+// CHECK-LABEL: @swap_bytes32
+#[no_mangle]
+pub fn swap_bytes32(x: &mut [u8; 32], y: &mut [u8; 32]) {
+// CHECK-NOT: alloca
+// CHECK: load <32 x i8>{{.+}}align 1
+// CHECK: store <32 x i8>{{.+}}align 1
+    swap(x, y)
+}

tests/codegen/swap-small-types.rs

@@ -1,4 +1,4 @@
-// compile-flags: -O
+// compile-flags: -O -Z merge-functions=disabled
 // only-x86_64
 // ignore-debug: the debug assertions get in the way
 
@@ -8,13 +8,28 @@ use std::mem::swap;
 
 type RGB48 = [u16; 3];
 
+// CHECK-LABEL: @swap_rgb48_manually(
+#[no_mangle]
+pub fn swap_rgb48_manually(x: &mut RGB48, y: &mut RGB48) {
+    // CHECK-NOT: alloca
+    // CHECK: %[[TEMP0:.+]] = load <3 x i16>, ptr %x, align 2
+    // CHECK: %[[TEMP1:.+]] = load <3 x i16>, ptr %y, align 2
+    // CHECK: store <3 x i16> %[[TEMP1]], ptr %x, align 2
+    // CHECK: store <3 x i16> %[[TEMP0]], ptr %y, align 2
+
+    let temp = *x;
+    *x = *y;
+    *y = temp;
+}
+
 // CHECK-LABEL: @swap_rgb48
 #[no_mangle]
 pub fn swap_rgb48(x: &mut RGB48, y: &mut RGB48) {
-    // FIXME MIR inlining messes up LLVM optimizations.
-// WOULD-CHECK-NOT: alloca
-// WOULD-CHECK: load i48
-// WOULD-CHECK: store i48
+    // CHECK-NOT: alloca
+    // CHECK: load <3 x i16>
+    // CHECK: load <3 x i16>
+    // CHECK: store <3 x i16>
+    // CHECK: store <3 x i16>
     swap(x, y)
 }