GerbilSoft/zlib-ng
1
Fork 0
mirror of https://github.com/GerbilSoft/zlib-ng.git synced 2025-06-19 03:55:39 -04:00
Code Issues Actions Packages Projects Releases Wiki Activity

cleanup arm/adler32_neon.c code

Browse Source
This commit is contained in:
Sebastian Pop 2019-04-01 15:05:53 -05:00 committed by Hans Kristian Rosbach
parent 3ac4f5de06
commit 098f73a45e
3 changed files with 7 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
adler32.c
View File

@ -13,9 +13,6 @@
uint32_t adler32_c(uint32_t adler, const unsigned char *buf, size_t len);
static uint32_t adler32_combine_(uint32_t adler1, uint32_t adler2, z_off64_t len2);
#define NMAX 5552
/* NMAX is the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1 */
#define DO1(buf, i) {adler += (buf)[i]; sum2 += adler;}
#define DO2(buf, i) DO1(buf, i); DO1(buf, i+1);
#define DO4(buf, i) DO2(buf, i); DO2(buf, i+2);

2
adler32_p.h
View File

@ -9,6 +9,8 @@
#define ADLER32_P_H
#define BASE 65521U /* largest prime smaller than 65536 */
#define NMAX 5552
/* NMAX is the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1 */
/* use NO_DIVIDE if your processor does not do division in hardware --
try it both ways to see which is faster */

17
arch/arm/adler32_neon.c
View File

@ -72,7 +72,6 @@ static void NEON_accum32(uint32_t *s, const unsigned char *buf, size_t len) {
}
static void NEON_handle_tail(uint32_t *pair, const unsigned char *buf, size_t len) {
/* Oldie K&R code integration. */
unsigned int i;
for (i = 0; i < len; ++i) {
pair[0] += buf[i];
@ -97,15 +96,9 @@ uint32_t adler32_neon(uint32_t adler, const unsigned char *buf, size_t len) {
if (len < 16)
return adler32_len_16(adler, buf, len, sum2);
/* The largest prime smaller than 65536. */
const uint32_t M_BASE = 65521;
/* This is the threshold where doing accumulation may overflow. */
const int M_NMAX = 5552;
uint32_t pair[2];
int n = M_NMAX;
int n = NMAX;
unsigned int done = 0;
/* Oldie K&R code integration. */
unsigned int i;
/* Split Adler-32 into component sums, it can be supplied by
@ -122,8 +115,8 @@ uint32_t adler32_neon(uint32_t adler, const unsigned char *buf, size_t len) {
break;
NEON_accum32(pair, buf + i, n / 16);
pair[0] %= M_BASE;
pair[1] %= M_BASE;
pair[0] %= BASE;
pair[1] %= BASE;
done += (n / 16) * 16;
}
@ -131,8 +124,8 @@ uint32_t adler32_neon(uint32_t adler, const unsigned char *buf, size_t len) {
/* Handle the tail elements. */
if (done < len) {
NEON_handle_tail(pair, (buf + done), len - done);
pair[0] %= M_BASE;
pair[1] %= M_BASE;
pair[0] %= BASE;
pair[1] %= BASE;
}
/* D = B * 65536 + A, see: https://en.wikipedia.org/wiki/Adler-32. */