[PATCH] D127762: [Clang][AArch64] Add ACLE attributes for SME.

Tue Jun 28 08:01:56 PDT 2022

sdesmalen added a comment.

Thanks for your patience reviewing this patch @aaron.ballman!

================
Comment at: clang/include/clang/AST/Type.h:4064
     bool HasTrailingReturn : 1;
+    unsigned AArch64SMEAttributes : 8;
     Qualifiers TypeQuals;
----------------
aaron.ballman wrote:
> If we're taking up space in the extra bitfields, why do we need to also take up space in the function prototype itself?
This field is added to ExtProtoInfo struct, which is used to populate the ExtraBitfields in the `FunctionPrototype`'s AST node.

Sorry if this wasn't clear because the diff had no context. `arc diff` wasn't working for me, so manually updated the patch, but then forgot to add full context.

================
Comment at: clang/include/clang/Basic/Attr.td:2333

+def ArmStreamingCompatible : TypeAttr, TargetSpecificAttr<TargetAArch64SME> {
+  let Spellings = [GNU<"arm_streaming_compatible">];
----------------
aaron.ballman wrote:
> You are handling these as declaration attributes in SemaDeclAttr.cpp but declaring them to be type attributes here in Attr.td; that's not okay. If these are type attributes, there should not be code for them in SemaDeclAttr.cpp. (This is why I really think the design needs to be rethought; I think they should be type attributes only, but I think you're trying to match what your specification says.)
Okay, I've removed the handling in SemaDeclAttr, so that it is only handled in SemaType.cpp.

But it still allows the attribute to be applied to any position in the function declaration, i.e.

  __attribute__((arm_streaming)) float foo1(int);
  float __attribute__((arm_streaming)) foo2(int);
  float foo3(int) __attribute__((arm_streaming));

Is that expected?

================
Comment at: clang/include/clang/Basic/Attr.td:2345
+
+def ArmLocallyStreaming : DeclOrStmtAttr, TargetSpecificAttr<TargetAArch64SME> {
+  let Spellings = [GNU<"arm_locally_streaming">];
----------------
aaron.ballman wrote:
> Why is this a *statement* attribute?
You're right that it shouldn't be. I want it to just be a decl attribute and don't want it to be either a Type or a Stmt attribute, but that doesn't seem to exist.
I find their meaning quite confusing, because both allow me to limit their applicability to Functions. What is the right class to derive from here?

================
Comment at: clang/include/clang/Basic/Attr.td:2322

+def ArmStreamingCompatible : DeclOrTypeAttr, TargetSpecificAttr<TargetAArch64> {
+  let Spellings = [Clang<"arm_streaming_compatible">];
----------------
aaron.ballman wrote:
> sdesmalen wrote:
> > aaron.ballman wrote:
> > > sdesmalen wrote:
> > > > aaron.ballman wrote:
> > > > > `DeclOrTypeAttr` is only intended to be used for attributes which were historically a declaration attribute but are semantically type attributes (like calling conventions). It seems to me that each of these is purely a type attribute and we shouldn't allow them to be written in the declaration position. WDYT?
> > > > In this case, I believe the attribute is valid on both the type and the declaration, because the SME ACLE explicitly specifies that the attributes can be applied to both declarations and types.
> > > > 
> > > > The [[https://github.com/ARM-software/acle/pull/188/commits/ce878cf6c43fd824ffc634526e5dfec1ddc1839e#diff-516526d4a18101dc85300bc2033d0f86dc46c505b7510a7694baabea851aedfaR8283-R8294|specification]] says:
> > > > ```Some of the attributes described in this section apply to function types.
> > > > Their semantics are as follows:
> > > > 
> > > > *   If the attribute is attached to a function declaration or definition,
> > > >     it applies to the type of the function.
> > > > 
> > > > *   If the attribute is attached to a function type, it applies to that type.
> > > > 
> > > > *   If the attribute is attached to a pointer to a function type, it applies
> > > >     to that function type.
> > > > 
> > > > *   Otherwise, the attribute is [ill-formed](#ill-formed).```
> > > > In this case, I believe the attribute is valid on both the type and the declaration, because the SME ACLE explicitly specifies that the attributes can be applied to both declarations and types.
> > > 
> > > What are the chances that can change? Because there are problems here:
> > > 
> > > > If the attribute is attached to a function declaration or definition, it applies to the type of the function.
> > > 
> > > This is egregiously opposed to the design of `[[]]` attributes in both C and C++. We came up with `DeclOrTypeAttr` for attributes that previously existed, but that is different than introducing new attributes. It's par for the course for `__attribute__` style attributes, so I don't worry so much there.
> > > 
> > > What's the rationale for this confusing of a design? (e.g., is there some reason you basically have to do that, like historically accepting the attributes in both positions?)
> > The attribute must always apply to the type of the function, because we can't have the streaming-property (or the properties on ZA) being lost between function overrides or function pointer assignments. It's perhaps similar to a calling convention, because the caller may have to set up streaming- or ZA state before the call, and restore state after the call.
> > 
> > I'm not too familiar with the different spellings/syntaxes and their implications, so I've now limited the attribute to `GNU` style attributes (`__attribute__((..))`) and to being a `TypeAttr`, with the exception of the `arm_locally_streaming` attribute, because that one can only be applied to function definitions and is not part of the type.
> > 
> > I've also added some new tests to ensure the properties are correctly propagated (using `decltype()`) and tests to ensure virtual function overloads require the same attributes.
> > 
> > Is this a step in the right direction? :)
> Switching to a GNU-only spelling sort of helps, but I still think this design is confused.
> ```
> *  If the attribute is attached to a function declaration or definition,
>     it applies to the type of the function.
> ```
> This seems like a misdesigned feature and I think it should be fixed in the specification. If it's an attribute that applies to the type of the function, you should not be allowed to specify it on the declaration; what is the benefit to allowing it in an unexpected position?
> If it's an attribute that applies to the type of the function, you should not be allowed to specify it on the declaration
Could you maybe give me an example here of what it looks like to apply the attribute to the function *type* when it's not allowed on the *declaration*? Does this all have to do with the //position// of the attribute in the declaration statement? When writing a function declaration it defines both the name and the type of that function, so there's no way to write a declaration without also defining its type.

Maybe it helps if you have any pointers that explain the difference in syntax between function declaration/function type attributes (https://clang.llvm.org/docs/AttributeReference.html doesn't really address this)

================
Comment at: clang/include/clang/Basic/DiagnosticSemaKinds.td:2000
   "overridden virtual function is here">;
+def err_conflicting_overriding_attributes : Error<
+  "virtual function %0 has different attributes "
----------------
aaron.ballman wrote:
> This error and the one below make me wonder whether an attribute spelling is the appropriate way to surface this functionality as opposed to a keyword. Attributes don't typically don't cause errors in these situations, but because these attributes are strongly coupled to their type system effects I can see why you want these to be errors.
> This error and the one below make me wonder whether an attribute spelling is the appropriate way to surface this functionality as opposed to a keyword. 
I'm not sure I understand what you mean, do you have an example?

================
Comment at: clang/test/Sema/aarch64-sme-func-attrs.c:213
+template short templated<short>(short);
+#endif
+
----------------
aaron.ballman wrote:
> This is an interesting start, but here are some more cases:
> ```
> template <typename Ty>
> struct S {
>   static constexpr int value = 0;
> };
> 
> template <>
> struct S<void (*)()> {
>   static constexpr int value = 1;
> };
> 
> template <>
> struct S<void (* __attribute__((arm_streaming)))()> {
>   static constexpr int value = 2;
> };
> 
> void func() __attribute__((arm_streaming)) {}
> void other_func() {}
> 
> static_assert(C<decltype(+func)>::value == 2, "why are we picking the wrong specialization?");
> static_assert(C<decltype(+other_func)>::value == 1, "why are we picking the wrong specialization?");
> ```
> another interesting test would be to ensure that implicit instantiations do not lose the attribute from the primary template for codegen:
> ```
> template <typename Ty>
> void func(Ty) __attribute__((arm_streaming)) {}
> 
> int main() {
>   func(12); // Does this properly insert the correct IR instructions?
> }
> ```
Great suggestions, thanks! They both seem to be handled correctly.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D127762/new/

https://reviews.llvm.org/D127762