…tbuf arg after the `this` parameter, not before.

…hiscall on required platforms. Still need to reorder the emit so the arguments are emitted correctly.

… JIT can handle it.

…primitive-wrapper struct returning thiscall working.

…eems to optimize them correctly now. Also this enables the JIT to correctly handle native instance method calls that return single primitive field structs.

… since the arguments have already been reversed) on x86.

…a test in interop to validate.

…n for thiscall Reverse P/Invoke stubs, update the stub linker to handle the case where a return value would be an enregistered return on stdcall but is returned via a return buffer for thiscall.

…f-move-to-jit

…s a value type.

Use a type handle instead of a method table.

Don't check size for enums.

…f-move-to-jit

…ove-to-jit

…ct return type based on calling convention. Update runtime side to not assume that the JIT enregisters 8-byte struct returns on Windows x86 (since it doesn't).

… P/Invoke stub linker.

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…T_TYPE_I when jitting on all platforms.

When I removed the struct normalization code from x86, I inadvertantly broke treating these types as ELEMENT_TYPE_I. On all platforms other than x86, this was handled at an earlier point in the code. Remove the if-def and make this happen on all platforms.

…d similarly to RuntimeMethodHandleInternal across the FCall boundary

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…return buffer but the managed calling convention doesn't.

…f-move-to-jit

…on is emitted.

Make better usage of canReturnInRegister to avoid a struct that has been marked as byref being marked as ByValueAsHfa.

Instance call return buffers on Windows ARM64 don't use the standard return buffer register, so don't use use it in that case.

Treat the return buffer as a normal arg and don't use the special retbuf register. Implicitly return the return buffer argument from the function to satisfy the C++ compiler's dependence on the implicit behavior (similar to Windows x64).

Co-authored-by: Jan Kotas <jkotas@microsoft.com>

…b.com:jkoritzinsky/runtime into retbuf-move-to-jit

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…e the case where native expects the returned struct in EAX:EDX but the managed call expects to return the result in a return buffer.

Now that interop isn't bashing the return value to long and we aren't changing the x86 managed calling convention to support returning arbitrary 8-byte structs in EAX:EDX, the managed calling convention expects to return an arbitrary 8-byte struct in a return buffer.
Since in this case, native code expects the return to be enregistered, we need to introduce a return buffer into the stack while executing the stub so the managed calling convention can point to the stub correctly.

…tch the order that a return buffer uses, we can remove creation of our custom return buffer and instead point the return buffer directly into the EAX/EDX slots so our EAX/EDX restore path automatically handles the "native enregistered return, managed retrun buffer return" case.

Not fixing the test since we don't officially support Windows ARM32 anyway and this scenario is already a corner case.

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…nd in the JIT.

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…o-jit

…ll on x86.

… generates a direct jmp instruction).

…o-jit

…malization model. The new model normalizes within the JIT, so the interop IL stubs don't need to account for it.

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Co-authored-by: Carol Eidt <carol.eidt@microsoft.com>

…nt the managed calling convention.

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Co-authored-by: Sergey Andreenko <seandree@microsoft.com>

…ng convention since native calls cannot be tail calls.

Co-authored-by: Andy Ayers <andya@microsoft.com>

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

Signed-off-by: Jeremy Koritzinsky <jekoritz@microsoft.com>

…gedCallConv

…o-jit