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

[huf] Add generic C versions of the fast decoding loops

Browse Source 
Add generic C versions of the fast decoding loops to serve architectures
that don't have an assembly implementation. Also allow selecting the C
decoding loop over the assembly decoding loop through a zstd
decompression parameter `ZSTD_d_disableHuffmanAssembly`.

I benchmarked on my Intel i9-9900K and my Macbook Air with an M1 processor.
The benchmark command forces zstd to compress without any matches, using
only literals compression, and measures only Huffman decompression speed:

```
zstd -b1e1 --compress-literals --zstd=tlen=131072 silesia.tar
```

The new fast decoding loops outperform the previous implementation uniformly,
but don't beat the x86-64 assembly. Additionally, the fast C decoding loops suffer
from the same stability problems that we've seen in the past, where the assembly
version doesn't. So even though clang gets close to assembly on x86-64, it still
has stability issues.

| Arch    | Function       | Compiler     | Default (MB/s) | Assembly (MB/s) | Fast (MB/s) |
|---------|----------------|--------------|----------------|-----------------|-------------|
| x86-64  | decompress 4X1 | gcc-12.2.0   |         1029.6 |          1308.1 |      1208.1 |
| x86-64  | decompress 4X1 | clang-14.0.6 |         1019.3 |          1305.6 |      1276.3 |
| x86-64  | decompress 4X2 | gcc-12.2.0   |         1348.5 |          1657.0 |      1374.1 |
| x86-64  | decompress 4X2 | clang-14.0.6 |         1027.6 |          1659.9 |      1468.1 |
| aarch64 | decompress 4X1 | clang-12.0.5 |         1081.0 |             N/A |      1234.9 |
| aarch64 | decompress 4X2 | clang-12.0.5 |         1270.0 |             N/A |      1516.6 |
This commit is contained in:
Nick Terrell
2023-01-25 13:47:51 -08:00
committed by Nick Terrell
parent f3255bfeff
commit 8957fef554
13 changed files with 448 additions and 91 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/fuzz/huf_decompress.c
+2 -1
View File
@@ -33,7 +33,8 @@ int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_optimalDepth : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_preferRepeat : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_suspectUncompressible : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableAsm : 0);
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableAsm : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableFast : 0);
/* Select a random cBufSize - it may be too small */
size_t const dBufSize = FUZZ_dataProducer_uint32Range(producer, 0, 8 * size + 500);
size_t const maxTableLog = FUZZ_dataProducer_uint32Range(producer, 1, HUF_TABLELOG_MAX);
tests/fuzz/huf_round_trip.c
+2 -1
View File
@@ -49,7 +49,8 @@ int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_optimalDepth : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_preferRepeat : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_suspectUncompressible : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableAsm : 0);
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableAsm : 0)
| (FUZZ_dataProducer_int32Range(producer, 0, 1) ? HUF_flags_disableFast : 0);
/* Select a random cBufSize - it may be too small */
size_t const cBufSize = FUZZ_dataProducer_uint32Range(producer, 0, 4 * size);
/* Select a random tableLog - we'll adjust it up later */
tests/fuzzer.c
+11
View File
@@ -684,6 +684,17 @@ static int basicUnitTests(U32 const seed, double compressibility)
if (r != CNBuffSize) goto _output_error; }
DISPLAYLEVEL(3, "OK \n");
DISPLAYLEVEL(3, "test%3i : decompress %u bytes with Huffman assembly disabled : ", testNb++, (unsigned)CNBuffSize);
{
ZSTD_DCtx* dctx = ZSTD_createDCtx();
size_t r;
CHECK_Z(ZSTD_DCtx_setParameter(dctx, ZSTD_d_disableHuffmanAssembly, 1));
r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize);
if (r != CNBuffSize || memcmp(decodedBuffer, CNBuffer, CNBuffSize)) goto _output_error;
ZSTD_freeDCtx(dctx);
}
DISPLAYLEVEL(3, "OK \n");
DISPLAYLEVEL(3, "test%3i : check decompressed result : ", testNb++);
{ size_t u;
for (u=0; u<CNBuffSize; u++) {
tests/zstreamtest.c
+14 -1
View File
@@ -2889,6 +2889,9 @@ static int fuzzerTests_newAPI(U32 seed, int nbTests, int startTest,
CHECK(badParameters(zc, savedParams), "CCtx params are wrong");
/* multi - fragments decompression test */
if (FUZ_rand(&lseed) & 1) {
CHECK_Z(ZSTD_DCtx_reset(zd, ZSTD_reset_session_and_parameters));
}
if (!dictSize /* don't reset if dictionary : could be different */ && (FUZ_rand(&lseed) & 1)) {
DISPLAYLEVEL(5, "resetting DCtx (dict:%p) \n", (void const*)dict);
CHECK_Z( ZSTD_resetDStream(zd) );
@@ -2897,6 +2900,9 @@ static int fuzzerTests_newAPI(U32 seed, int nbTests, int startTest,
DISPLAYLEVEL(5, "using dictionary of size %zu \n", dictSize);
CHECK_Z( ZSTD_initDStream_usingDict(zd, dict, dictSize) );
}
if (FUZ_rand(&lseed) & 1) {
CHECK_Z(ZSTD_DCtx_setParameter(zd, ZSTD_d_disableHuffmanAssembly, FUZ_rand(&lseed) & 1));
}
{ size_t decompressionResult = 1;
ZSTD_inBuffer inBuff = { cBuffer, cSize, 0 };
ZSTD_outBuffer outBuff= { dstBuffer, dstBufferSize, 0 };
@@ -2938,7 +2944,14 @@ static int fuzzerTests_newAPI(U32 seed, int nbTests, int startTest,
} }
/* try decompression on noisy data */
CHECK_Z( ZSTD_initDStream(zd_noise) ); /* note : no dictionary */
if (FUZ_rand(&lseed) & 1) {
CHECK_Z(ZSTD_DCtx_reset(zd_noise, ZSTD_reset_session_and_parameters));
} else {
CHECK_Z(ZSTD_DCtx_reset(zd_noise, ZSTD_reset_session_only));
}
if (FUZ_rand(&lseed) & 1) {
CHECK_Z(ZSTD_DCtx_setParameter(zd_noise, ZSTD_d_disableHuffmanAssembly, FUZ_rand(&lseed) & 1));
}
{ ZSTD_inBuffer inBuff = { cBuffer, cSize, 0 };
ZSTD_outBuffer outBuff= { dstBuffer, dstBufferSize, 0 };
while (outBuff.pos < dstBufferSize) {