void
bcopy(const void *src0, void *dst0, size_t length)
#endif
#endif
{
char *dst = dst0;
const char *src = src0;
size_t t;
if (length == 0 || dst == src) /* nothing to do */
goto done;
/*
* Macros: loop-t-times; and loop-t-times, t>0
*/
#define TLOOP(s) if (t) TLOOP1(s)
#define TLOOP1(s) do { s; } while (--t)
if ((unsigned long)dst < (unsigned long)src) {
/*
* Copy forward.
*/
t = (long)src; /* only need low bits */
if ((t | (long)dst) & wmask) {
/*
* Try to align operands. This cannot be done
* unless the low bits match.
*/
if (length < wsize) {
t = length;
TLOOP1(*dst++ = *src++);
goto done;
}
/*
* Src and dst not at same byte offset
*/
if ((t ^ (long)dst) & wmask) {
t = length;
#if !defined (MEMCOPY)
#if defined(__ARM_NEON__)
memcpy(dst0, src0, length);
goto done;
#else /* BCOPY without NEON */
memcpy(dst0, src0, length);
goto done;
#endif /* __ARM_NEON__ */
#endif /* !MEMCOPY */
} else
t = wsize - (t & wmask);
length -= t;
TLOOP1(*dst++ = *src++);
}
/*
* Copy whole words, then mop up any trailing bytes.
*/
t = length / wsize;
TLOOP(*(word *)dst = *(word *)src; src += wsize; dst += wsize);
t = length & wmask;
TLOOP(*dst++ = *src++);
} else {
/*
* Copy a block of memory, handling overlap.
* This is the routine that actually implements
* (the portable versions of) bcopy, memcpy, and memmove.
*/
void *
memcpy(void *dst0, const void *src0, size_t length)
{
char *dst;
const char *src;
size_t t;
dst = dst0;
src = src0;
if (length == 0 || dst == src) { /* nothing to do */
goto done;
}
/*
* Macros: loop-t-times; and loop-t-times, t>0
*/
#define TLOOP(s) if (t) TLOOP1(s)
#define TLOOP1(s) do { s; } while (--t)
if ((unsigned long)dst < (unsigned long)src) {
/*
* Copy forward.
*/
t = (size_t)src; /* only need low bits */
if ((t | (uintptr_t)dst) & wmask) {
/*
* Try to align operands. This cannot be done
* unless the low bits match.
*/
if ((t ^ (uintptr_t)dst) & wmask || length < wsize) {
t = length;
} else {
t = wsize - (t & wmask);
}
length -= t;
TLOOP1(*dst++ = *src++);
}
/*
* Copy whole words, then mop up any trailing bytes.
*/
t = length / wsize;
TLOOP(*(word *)dst = *(const word *)src; src += wsize;
dst += wsize);
t = length & wmask;
TLOOP(*dst++ = *src++);
} else {
void
bcopy(const void *src0, void *dst0, size_t length)
#endif
{
char *dst = dst0;
const char *src = src0;
size_t t;
unsigned long u;
#if !defined(_KERNEL)
_DIAGASSERT(dst0 != 0);
_DIAGASSERT(src0 != 0);
#endif
if (length == 0 || dst == src) /* nothing to do */
goto done;
/*
* Macros: loop-t-times; and loop-t-times, t>0
*/
#define TLOOP(s) if (t) TLOOP1(s)
#define TLOOP1(s) do { s; } while (--t)
if ((unsigned long)dst < (unsigned long)src) {
/*
* Copy forward.
*/
u = (unsigned long)src; /* only need low bits */
if ((u | (unsigned long)dst) & wmask) {
/*
* Try to align operands. This cannot be done
* unless the low bits match.
*/
if ((u ^ (unsigned long)dst) & wmask || length < wsize)
t = length;
else
t = wsize - (size_t)(u & wmask);
length -= t;
TLOOP1(*dst++ = *src++);
}
/*
* Copy whole words, then mop up any trailing bytes.
*/
t = length / wsize;
TLOOP(*(word *)(void *)dst = *(const word *)(const void *)src; src += wsize; dst += wsize);
t = length & wmask;
TLOOP(*dst++ = *src++);
} else {
void
bcopy(const char *src0, char *dst0, size_t length)
#endif
#endif
{
register char *dst = dst0;
register const char *src = src0;
register size_t t;
if (length == 0 || dst == src) /* nothing to do */
goto retval;
/*
* Macros: loop-t-times; and loop-t-times, t>0
*/
#define TLOOP(s) if (t) TLOOP1(s)
#define TLOOP1(s) do { s; } while (--t)
if ((char *)dst < (char *)src) {
/*
* Copy forward.
*/
t = (int)src; /* only need low bits */
if ((t | (int)dst) & wmask) {
/*
* Try to align operands. This cannot be done
* unless the low bits match.
*/
if ((t ^ (int)dst) & wmask || length < wsize)
t = length;
else
t = wsize - (t & wmask);
length -= t;
TLOOP1(*dst++ = *src++);
}
/*
* Copy whole words, then mop up any trailing bytes.
*/
t = length / wsize;
TLOOP(*(word *)dst = *(word *)src; src += wsize; dst += wsize);
t = length & wmask;
TLOOP(*dst++ = *src++);
} else {
