codegen: use new {re,de,}allocator annotations in llvm

This obviates the patch that teaches LLVM internals about
_rust_{re,de}alloc functions by putting annotations directly in the IR
for the optimizer.

The sole test change is required to anchor FileCheck to the body of the
`box_uninitialized` method, so it doesn't see the `allocalign` on
`__rust_alloc` and get mad about the string `alloca` showing up. Since I
was there anyway, I added some checks on the attributes to prove the
right attributes got set.

While we're here, we also emit allocator attributes on
__rust_alloc_zeroed. This should allow LLVM to perform more
optimizations for zeroed blocks, and probably fixes #90032. [This
comment](https://github.com/rust-lang/rust/issues/24194#issuecomment-308791157)
mentions "weird UB-like behaviour with bitvec iterators in
rustc_data_structures" so we may need to back this change out if things
go wrong.

The new test cases require LLVM 15, so we copy them into LLVM
14-supporting versions, which we can delete when we drop LLVM 14.

1 files changed, 7 insertions, 0 deletions

diff --git a/src/test/codegen/box-maybe-uninit.rs b/src/test/codegen/box-maybe-uninit.rs
index 5004f787cde..e105e26f16a 100644
--- a/src/test/codegen/box-maybe-uninit.rs
+++ b/src/test/codegen/box-maybe-uninit.rs
@@ -1,4 +1,5 @@
 // compile-flags: -O
+// min-llvm-version: 15.0
 #![crate_type="lib"]
 
 use std::mem::MaybeUninit;
@@ -16,3 +17,9 @@ pub fn box_uninitialized() -> Box<MaybeUninit<usize>> {
 
 // FIXME: add a test for a bigger box. Currently broken, see
 // https://github.com/rust-lang/rust/issues/58201.
+
+// Hide the `allocalign` attribute in the declaration of __rust_alloc
+// from the CHECK-NOT above, and also verify the attributes got set reasonably.
+// CHECK: declare noalias ptr @__rust_alloc(i{{[0-9]+}}, i{{[0-9]+}} allocalign) unnamed_addr [[RUST_ALLOC_ATTRS:#[0-9]+]]
+
+// CHECK-DAG: attributes [[RUST_ALLOC_ATTRS]] = { {{.*}} allockind("alloc,uninitialized,aligned") allocsize(0) uwtable "alloc-family"="__rust_alloc" {{.*}} }