Kewen:
On 7/25/24 1:22 AM, Kewen.Lin wrote:
-----------------------------------------------------------------------------------------------------------------
rs6000, Remove __builtin_vec_set_v1ti, __builtin_vec_set_v2df,
__builtin_vec_set_v2di
Remove the built-ins, use the default gimple generation instead.
OK for trunk with better commit log like the above paragraph, thanks!
// Assuming testing on BE goes well too. 🙂
Good point, I hadn't double checked things on BE. I tested the patches
today on BE. The patches do not generate any additional regression testing.
I also investigated the assembly code generation with and without the
patches for -O0 and -O3 using the same scripts as I used previously on
LE. I see the same results. With -O0 the assembly code generations
take one extra instruction for the built-in. With -O3, the code
generated for the vsx set 2df and 2di cases is identical. The code for
the vsx set 1di case requires fewer assembly instructions for the C code
versus the built-in.
                                                        Carl