AMDGPU and 16B stack alignment

Arvind Sankar nivedita at alum.mit.edu
Tue Oct 15 20:26:38 UTC 2019


On Tue, Oct 15, 2019 at 11:05:56AM -0700, Nick Desaulniers wrote:
> Hmmm...I would have liked to remove it outright, as it is an ABI
> mismatch that is likely to result in instability and non-fun-to-debug
> runtime issues in the future.  I suspect my patch does work for GCC
> 7.1+.  The question is: Do we want to either:
> 1. mark AMDGPU broken for GCC < 7.1, or
> 2. continue supporting it via stack alignment mismatch?
> 
> 2 is brittle, and may break at any point in the future, but if it's
> working for someone it does make me feel bad to outright disable it.
> What I'd image 2 looks like is (psuedo code in a Makefile):
> 
> if CC_IS_GCC && GCC_VERSION < 7.1:
>   set stack alignment to 16B and hope for the best
> 
> So my diff would be amended to keep the stack alignment flags, but
> only to support GCC < 7.1.  And that assumes my change compiles with
> GCC 7.1+. (Looks like it does for me locally with GCC 8.3, but I would
> feel even more confident if someone with hardware to test on and GCC
> 7.1+ could boot test).
> -- 
> Thanks,
> ~Nick Desaulniers

If we do keep it, would adding -mstackrealign make it more robust?
That's simple and will only add the alignment to functions that require
16-byte alignment (at least on gcc).

Alternative is to use
__attribute__((force_align_arg_pointer)) on functions that might be
called from 8-byte-aligned code.

It looks like -mstackrealign should work from gcc 5.3 onwards.


More information about the amd-gfx mailing list