llvm dev - Mar 2020 - Memory intrinsics len argument type

Hi,

during instruction selection of the LLVM memory intrinsics 
(llvm.mem{cpy,set,move}) to the corresponding C library functions, 
SelectionDAG ignores the actual type of the "len" argument and asks
the
target for the preferred type (via DataLayout's getIntPtrType()).
This introduces implicit truncations of the len argument in case the i64 
variant is used on a 32-bit target.

In contrast, GlobalISel simply forwards the arguments of the intrinsic 
call, including their types, to the call to the C library function. This 
results in wrong code generation of the i64 versions of these intrinsics 
on 32-bit architectures.

The IRBuilder hard-codes the type of the len argument to i64 for some of 
its functions. In https://reviews.llvm.org/D76283 I propose to change 
those functions to ask the target for its preferred type. However, the 
problem still exists for hand-written IR.

I see two options: either 1) extend GlobalISel to select to the 
appropriate type, potentially introducing truncations, or 2) declare 
this as illegal LLVM IR in the first place.
During the review, additional concerns were raised regarding different 
address spaces having varying bit-widths.

Is there a preferred way to fix this?

Thanks,

Konstantin

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20200318/102a5260/attachment.html>