https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70308
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jakub at gcc dot gnu.org
--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Much faster where, on which CPU? GCC has lots of different memset/memcpy
strategies, depending on known or expected sizes, what CPU the function is
tuned for etc.
See e.g.
-mno-align-stringops
-minline-all-stringops
-minline-stringops-dynamically
-mstringop-strategy=ALG
-mmemcpy-strategy=STRATEGY
-mmemset-strategy=STRATEGY
options for fine tuning. But, the most important is which CPU you are tuning
for, if it is e.g. -mtune=generic, then it is some compromise between most
commonly used CPUs.
