On 2/7/24 06:48, Alexander Monakov wrote:
- /* Otherwise, use the unaligned memory access functions to
- handle the beginning and end of the buffer, with a couple
+ /* Use unaligned memory access functions to handle
+ the beginning and end of the buffer, with a couple
of loops handling the middle aligned section. */
- uint64_t t = ldq_he_p(buf);
- const uint64_t *p = (uint64_t *)(((uintptr_t)buf + 8) & -8);
- const uint64_t *e = (uint64_t *)(((uintptr_t)buf + len) & -8);
+ uint64_t t = ldq_he_p(buf) | ldq_he_p(buf + len - 8);
+ typedef uint64_t uint64_a __attribute__((may_alias));
+ const uint64_a *p = (void *)(((uintptr_t)buf + 8) & -8);
+ const uint64_a *e = (void *)(((uintptr_t)buf + len - 1) & -8);
You appear to be optimizing this routine for x86, which is not the primary
consumer.
This is going to perform very poorly on hosts that do not support unaligned accesses (e.g.
Sparc and some RISC-V).