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

Improvements in zstd decode performance

Browse Source 
Summary: The idea behind wildcopy is that it can be cheaper to copy more bytes (say 8) than it is to copy less (say, 3).  This change takes that further by exploiting some properties:
1. it's almost always OK to copy 16 bytes instead of 8, which means fewer copy instructions, and fewer branches
2. A 16 byte chunk size means that ~90% of wildcopy invocations will have a trip count of 1, so branch prediction will be improved.

Speedup on Xeon E5-2680v4 is in the range of 3-5%.

Measured wildcopy length distributions on silesia.tar:

level	<=8	<=16	<=24	>24
1	78.05%	11.49%	3.52%	6.94%
3	82.14%	8.99%	2.44%	6.43%
6	85.81%	6.51%	2.92%	4.76%
8	83.02%	7.31%	3.64%	6.03%
10	84.13%	6.67%	3.29%	5.91%
15	77.58%	7.55%	5.21%	9.66%
16	80.07%	7.20%	3.98%	8.75%

Test Plan: benchmark silesia, make check
This commit is contained in:
mgrice
2019-08-29 12:25:56 -07:00
parent d944197e79
commit b830599582
3 changed files with 40 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/decompress/zstd_decompress_block.c
+6 -6
View File
@@ -641,7 +641,7 @@ size_t ZSTD_execSequence(BYTE* op,
/* copy Literals */
if (sequence.litLength > 8)
ZSTD_wildcopy_16min(op, (*litPtr), sequence.litLength, ZSTD_no_overlap); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
ZSTD_wildcopy_16min(op, (*litPtr), oend, sequence.litLength, ZSTD_no_overlap); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
else
ZSTD_copy8(op, *litPtr);
op = oLitEnd;
@@ -690,13 +690,13 @@ size_t ZSTD_execSequence(BYTE* op,
if (oMatchEnd > oend-(16-MINMATCH)) {
if (op < oend_w) {
ZSTD_wildcopy(op, match, oend_w - op, ZSTD_overlap_src_before_dst);
ZSTD_wildcopy(op, match, oend, oend_w - op, ZSTD_overlap_src_before_dst);
match += oend_w - op;
op = oend_w;
}
while (op < oMatchEnd) *op++ = *match++;
} else {
ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength-8, ZSTD_overlap_src_before_dst); /* works even if matchLength < 8 */
ZSTD_wildcopy(op, match, oend, (ptrdiff_t)sequence.matchLength-8, ZSTD_overlap_src_before_dst); /* works even if matchLength < 8 */
}
return sequenceLength;
}
@@ -722,7 +722,7 @@ size_t ZSTD_execSequenceLong(BYTE* op,
/* copy Literals */
if (sequence.litLength > 8)
ZSTD_wildcopy_16min(op, *litPtr, sequence.litLength, ZSTD_no_overlap); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
ZSTD_wildcopy_16min(op, *litPtr, oend, sequence.litLength, ZSTD_no_overlap); /* note : since oLitEnd <= oend-WILDCOPY_OVERLENGTH, no risk of overwrite beyond oend */
else
ZSTD_copy8(op, *litPtr); /* note : op <= oLitEnd <= oend_w == oend - 8 */
@@ -772,13 +772,13 @@ size_t ZSTD_execSequenceLong(BYTE* op,
if (oMatchEnd > oend-(16-MINMATCH)) {
if (op < oend_w) {
ZSTD_wildcopy(op, match, oend_w - op, ZSTD_overlap_src_before_dst);
ZSTD_wildcopy(op, match, oend, oend_w - op, ZSTD_overlap_src_before_dst);
match += oend_w - op;
op = oend_w;
}
while (op < oMatchEnd) *op++ = *match++;
} else {
ZSTD_wildcopy(op, match, (ptrdiff_t)sequence.matchLength-8, ZSTD_overlap_src_before_dst); /* works even if matchLength < 8 */
ZSTD_wildcopy(op, match, oend, (ptrdiff_t)sequence.matchLength-8, ZSTD_overlap_src_before_dst); /* works even if matchLength < 8 */
}
return sequenceLength;
}