================
@@ -653,13 +631,13 @@ def CIR_AllocaOp : CIR_Op<"alloca", [
}];
let assemblyFormat = [{
- $allocaType `,` qualified(type($addr)) `,`
+ $name
+ `align` `(` $alignment `)`
+ oilist( `init` $init
+ | `const` $constant
+ | `cleanup_dest_slot` $cleanup_dest_slot)
($dynAllocSize^ `:` type($dynAllocSize) `,`)?
- `[` $name
- (`,` `init` $init^)?
- (`,` `const` $constant^)?
- (`,` `cleanup_dest_slot` $cleanup_dest_slot^)?
- `]`
+ $allocaType `->` qualified(type($addr))
----------------
Andres-Salamanca wrote:
I personally think it's good to keep it explicit in LLVM IR it works the other
way around, the allocated type is shown but returning a pointer is implicit. i
think the same applies here, printing the allocated type improves readability
since you can immediately see what's being allocated without having to inspect
the pointer type, especially with nested types like:
```mlir
%0 = cir.alloca "f" align(4) init const !cir.ptr<!cir.func<() -> !s32i>> ->
!cir.ptr<!cir.ptr<!cir.func<() -> !s32i>>>
```
but happy to go either way if you and the others feel strongly about dropping
it.
@erichkeane @andykaylor
https://github.com/llvm/llvm-project/pull/198962
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
