ddidderr/zstd-rs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Move bitwise builtins into bits.h

Browse Source
This commit is contained in:
Elliot Gorokhovsky
2022-02-14 11:16:03 -05:00
parent 970460f67d
commit db2f4a6532
16 changed files with 250 additions and 400 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/compress/zstd_lazy.c
+3 -40
View File
@@ -10,6 +10,7 @@
#include "zstd_compress_internal.h"
#include "zstd_lazy.h"
#include "../common/bits.h" /* ZSTD_countTrailingZeros64 */
/*-*************************************
@@ -765,44 +766,6 @@ size_t ZSTD_HcFindBestMatch(
typedef U64 ZSTD_VecMask; /* Clarifies when we are interacting with a U64 representing a mask of matches */
/* ZSTD_VecMask_next():
* Starting from the LSB, returns the idx of the next non-zero bit.
* Basically counting the nb of trailing zeroes.
*/
static U32 ZSTD_VecMask_next(ZSTD_VecMask val) {
assert(val != 0);
# if defined(_MSC_VER) && defined(_WIN64)
if (val != 0) {
unsigned long r;
_BitScanForward64(&r, val);
return (U32)(r);
} else {
/* Should not reach this code path */
__assume(0);
}
# elif (defined(__GNUC__) && ((__GNUC__ > 3) || ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))
if (sizeof(size_t) == 4) {
U32 mostSignificantWord = (U32)(val >> 32);
U32 leastSignificantWord = (U32)val;
if (leastSignificantWord == 0) {
return 32 + (U32)__builtin_ctz(mostSignificantWord);
} else {
return (U32)__builtin_ctz(leastSignificantWord);
}
} else {
return (U32)__builtin_ctzll(val);
}
# else
/* Software ctz version: http://aggregate.org/MAGIC/#Trailing%20Zero%20Count
* and: https://stackoverflow.com/questions/2709430/count-number-of-bits-in-a-64-bit-long-big-integer
*/
val = ~val & (val - 1ULL); /* Lowest set bit mask */
val = val - ((val >> 1) & 0x5555555555555555);
val = (val & 0x3333333333333333ULL) + ((val >> 2) & 0x3333333333333333ULL);
return (U32)((((val + (val >> 4)) & 0xF0F0F0F0F0F0F0FULL) * 0x101010101010101ULL) >> 56);
# endif
}
/* ZSTD_rotateRight_*():
* Rotates a bitfield to the right by "count" bits.
* https://en.wikipedia.org/w/index.php?title=Circular_shift&oldid=991635599#Implementing_circular_shifts
@@ -1202,7 +1165,7 @@ size_t ZSTD_RowFindBestMatch(
/* Cycle through the matches and prefetch */
for (; (matches > 0) && (nbAttempts > 0); --nbAttempts, matches &= (matches - 1)) {
U32 const matchPos = (head + ZSTD_VecMask_next(matches)) & rowMask;
U32 const matchPos = (head + ZSTD_countTrailingZeros64(matches)) & rowMask;
U32 const matchIndex = row[matchPos];
assert(numMatches < rowEntries);
if (matchIndex < lowLimit)
@@ -1270,7 +1233,7 @@ size_t ZSTD_RowFindBestMatch(
ZSTD_VecMask matches = ZSTD_row_getMatchMask(dmsTagRow, (BYTE)dmsTag, head, rowEntries);
for (; (matches > 0) && (nbAttempts > 0); --nbAttempts, matches &= (matches - 1)) {
U32 const matchPos = (head + ZSTD_VecMask_next(matches)) & rowMask;
U32 const matchPos = (head + ZSTD_countTrailingZeros64(matches)) & rowMask;
U32 const matchIndex = dmsRow[matchPos];
if (matchIndex < dmsLowestIndex)
break;