[PATCH] D47370: AMDGPU: Round up kernel argument allocation size

[Tony Tye via Phabricator via llvm-commits]

t-tye added a comment.

In https://reviews.llvm.org/D47370#1112733, @arsenm wrote:

> In https://reviews.llvm.org/D47370#1112729, @rampitec wrote:
>
> > Are we sure that is what RT(s) do?
>
>
> It doesn't really matter if it does or not, since we're now requesting the larger allocation


Note that technically the HSA spec requires the kernarg segment size to match the size and alignment of the kernel arguments. So it ought not to be larger than deduced from the arguments. In this context the implicit arguments are just extra arguments.


https://reviews.llvm.org/D47370