opencv/3rdparty/zlib-ng/arch/x86/slide_hash_sse2.c
FantasqueX 85923c8f30
Merge pull request #26113 from FantasqueX:zlib-ng-2-2-1
Update zlib-ng to 2.2.1 #26113

Release: https://github.com/zlib-ng/zlib-ng/releases/tag/2.2.1
ARM diagnostics patch: https://github.com/zlib-ng/zlib-ng/pull/1774

### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
2024-09-12 16:05:24 +03:00

63 lines
1.7 KiB
C

/*
* SSE optimized hash slide
*
* Copyright (C) 2017 Intel Corporation
* Authors:
* Arjan van de Ven <arjan@linux.intel.com>
* Jim Kukunas <james.t.kukunas@linux.intel.com>
*
* For conditions of distribution and use, see copyright notice in zlib.h
*/
#include "zbuild.h"
#include "deflate.h"
#include <immintrin.h>
#include <assert.h>
static inline void slide_hash_chain(Pos *table0, Pos *table1, uint32_t entries0,
uint32_t entries1, const __m128i wsize) {
uint32_t entries;
Pos *table;
__m128i value0, value1, result0, result1;
int on_chain = 0;
next_chain:
table = (on_chain) ? table1 : table0;
entries = (on_chain) ? entries1 : entries0;
table += entries;
table -= 16;
/* ZALLOC allocates this pointer unless the user chose a custom allocator.
* Our alloc function is aligned to 64 byte boundaries */
do {
value0 = _mm_load_si128((__m128i *)table);
value1 = _mm_load_si128((__m128i *)(table + 8));
result0 = _mm_subs_epu16(value0, wsize);
result1 = _mm_subs_epu16(value1, wsize);
_mm_store_si128((__m128i *)table, result0);
_mm_store_si128((__m128i *)(table + 8), result1);
table -= 16;
entries -= 16;
} while (entries > 0);
++on_chain;
if (on_chain > 1) {
return;
} else {
goto next_chain;
}
}
Z_INTERNAL void slide_hash_sse2(deflate_state *s) {
uint16_t wsize = (uint16_t)s->w_size;
const __m128i xmm_wsize = _mm_set1_epi16((short)wsize);
assert(((uintptr_t)s->head & 15) == 0);
assert(((uintptr_t)s->prev & 15) == 0);
slide_hash_chain(s->head, s->prev, HASH_SIZE, wsize, xmm_wsize);
}