Christophe Lyon wrote: > After this patch, I've noticed a regression: > FAIL: gcc.target/aarch64/ldp_vec_64_1.c scan-assembler ldp\td[0-9]+, d[0-9] > You probably need to adjust the scan pattern.
The original code was better and what we should generate. IVOpt chooses to use indexing eventhough it is more expensive, and that is a latent bug. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65068 which shows the same cost calculation issue. Wilco
