Alan Lawrence wrote:
Hi,
Comparing 64x1 vector types (defined by hand or from arm_neon.h) using GCC
vector extensions currently generates very poor assembly code, for example
"uint64x1_t foo (uint64x1_t a, uint64x1_t b) { return a >= b; }" generates (at -O3):
fmov x0, d0 // 22 movdi_aarch64/
Hi,
Comparing 64x1 vector types (defined by hand or from arm_neon.h) using GCC
vector extensions currently generates very poor assembly code, for example
"uint64x1_t foo (uint64x1_t a, uint64x1_t b) { return a >= b; }" generates (at -O3):
fmov x0, d0 // 22 movdi_aarch64/12 [length = 4]
fmov x
