Hi Sandeep, > Attached is a patch to llvm-gcc to implement passing of homogenous > aggregates via FP registers per the ARM AAPCS VFP variant (a.k.a. hard > float). please indent using spaces and not tabs. Thanks, Duncan.