[PATCH] ARM big endian function argument passing

[Christian Pirker]

Hi James,

the ABI for 32 bits is different with respect to the 64 ABI, as it 
requires any integral fundamental data types to be expanded to a size of 
4 bytes.

Thanks,
Christian

On 05/08/14 16:56, James Molloy wrote:
> Hi Christian,
>
> This patch does not seem to address sub-word stack arguments. If a 
> stack argument is < 32 bits, surely it should have an offset applied 
> to it in ARMISelLowering::lowerFormalArguments() and friends, similar 
> to what ARM64 does?
>
> Is the ABI for 32-bit different in this respect to the 64-bit ABI?
>
> Cheers,
>
> James
>
>
> On 8 May 2014 15:22, Renato Golin <renato.golin at linaro.org 
> <mailto:renato.golin at linaro.org>> wrote:
>
>     On 8 May 2014 15:20, Christian Pirker <cpirker at a-bix.com
>     <mailto:cpirker at a-bix.com>> wrote:
>     > Test suite have been run with GNU libraries in little and big
>     endian.
>     > I committed this patch as rL208316.
>
>     Thanks!
>
>     --renato
>     _______________________________________________
>     llvm-commits mailing list
>     llvm-commits at cs.uiuc.edu <mailto:llvm-commits at cs.uiuc.edu>
>     http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20140509/7effae26/attachment.html>