mirror of
https://github.com/opencv/opencv.git
synced 2024-12-27 03:14:05 +08:00
0de26fd78e
Zlib-ng is zlib replacement with optimizations for "next generation" systems. Its optimization may benifits image library decode and encode speed such as libpng. In our tests, if using zlib-ng and libpng combination on a x86_64 machine with AVX2, the time of `imdecode` amd `imencode` will drop 20% approximately. This patch enables zlib-ng's optimization if `CV_DISABLE_OPTIMIZATION` is OFF. Since Zlib-ng can dispatch intrinsics on the fly, port work is much easier. Related discussion: https://github.com/opencv/opencv/issues/22573
79 lines
1.9 KiB
C
79 lines
1.9 KiB
C
/* crc32_acle.c -- compute the CRC-32 of a data stream
|
|
* Copyright (C) 1995-2006, 2010, 2011, 2012 Mark Adler
|
|
* Copyright (C) 2016 Yang Zhang
|
|
* For conditions of distribution and use, see copyright notice in zlib.h
|
|
*
|
|
*/
|
|
|
|
#ifdef ARM_ACLE
|
|
#include "acle_intrins.h"
|
|
#include "../../zbuild.h"
|
|
|
|
Z_INTERNAL Z_TARGET_CRC uint32_t crc32_acle(uint32_t crc, const uint8_t *buf, size_t len) {
|
|
Z_REGISTER uint32_t c;
|
|
Z_REGISTER const uint16_t *buf2;
|
|
Z_REGISTER const uint32_t *buf4;
|
|
Z_REGISTER const uint64_t *buf8;
|
|
|
|
c = ~crc;
|
|
|
|
if (UNLIKELY(len == 1)) {
|
|
c = __crc32b(c, *buf);
|
|
c = ~c;
|
|
return c;
|
|
}
|
|
|
|
if ((ptrdiff_t)buf & (sizeof(uint64_t) - 1)) {
|
|
if (len && ((ptrdiff_t)buf & 1)) {
|
|
c = __crc32b(c, *buf++);
|
|
len--;
|
|
}
|
|
|
|
if ((len >= sizeof(uint16_t)) && ((ptrdiff_t)buf & sizeof(uint16_t))) {
|
|
buf2 = (const uint16_t *) buf;
|
|
c = __crc32h(c, *buf2++);
|
|
len -= sizeof(uint16_t);
|
|
buf4 = (const uint32_t *) buf2;
|
|
} else {
|
|
buf4 = (const uint32_t *) buf;
|
|
}
|
|
|
|
if ((len >= sizeof(uint32_t)) && ((ptrdiff_t)buf & sizeof(uint32_t))) {
|
|
c = __crc32w(c, *buf4++);
|
|
len -= sizeof(uint32_t);
|
|
}
|
|
|
|
buf8 = (const uint64_t *) buf4;
|
|
} else {
|
|
buf8 = (const uint64_t *) buf;
|
|
}
|
|
|
|
while (len >= sizeof(uint64_t)) {
|
|
c = __crc32d(c, *buf8++);
|
|
len -= sizeof(uint64_t);
|
|
}
|
|
|
|
if (len >= sizeof(uint32_t)) {
|
|
buf4 = (const uint32_t *) buf8;
|
|
c = __crc32w(c, *buf4++);
|
|
len -= sizeof(uint32_t);
|
|
buf2 = (const uint16_t *) buf4;
|
|
} else {
|
|
buf2 = (const uint16_t *) buf8;
|
|
}
|
|
|
|
if (len >= sizeof(uint16_t)) {
|
|
c = __crc32h(c, *buf2++);
|
|
len -= sizeof(uint16_t);
|
|
}
|
|
|
|
buf = (const unsigned char *) buf2;
|
|
if (len) {
|
|
c = __crc32b(c, *buf);
|
|
}
|
|
|
|
c = ~c;
|
|
return c;
|
|
}
|
|
#endif
|