[Fuzz] Improve data generation #1723

Converting the rest of the tests to use the new data producer.
This commit is contained in:
Dario Pavlovic
2019-09-10 16:14:43 -07:00
parent ea1ad123da
commit 0630d084cb
13 changed files with 159 additions and 117 deletions
+36 -36
View File
@@ -23,47 +23,47 @@ static void set(ZSTD_CCtx *cctx, ZSTD_cParameter param, int value)
}
static void setRand(ZSTD_CCtx *cctx, ZSTD_cParameter param, unsigned min,
unsigned max, uint32_t *state) {
unsigned const value = FUZZ_rand32(state, min, max);
unsigned max, FUZZ_dataProducer_t *producer) {
unsigned const value = FUZZ_dataProducer_uint32Range(producer, min, max);
set(cctx, param, value);
}
ZSTD_compressionParameters FUZZ_randomCParams(size_t srcSize, uint32_t *state)
ZSTD_compressionParameters FUZZ_randomCParams(size_t srcSize, FUZZ_dataProducer_t *producer)
{
/* Select compression parameters */
ZSTD_compressionParameters cParams;
cParams.windowLog = FUZZ_rand32(state, ZSTD_WINDOWLOG_MIN, 15);
cParams.hashLog = FUZZ_rand32(state, ZSTD_HASHLOG_MIN, 15);
cParams.chainLog = FUZZ_rand32(state, ZSTD_CHAINLOG_MIN, 16);
cParams.searchLog = FUZZ_rand32(state, ZSTD_SEARCHLOG_MIN, 9);
cParams.minMatch = FUZZ_rand32(state, ZSTD_MINMATCH_MIN,
cParams.windowLog = FUZZ_dataProducer_uint32Range(producer, ZSTD_WINDOWLOG_MIN, 15);
cParams.hashLog = FUZZ_dataProducer_uint32Range(producer, ZSTD_HASHLOG_MIN, 15);
cParams.chainLog = FUZZ_dataProducer_uint32Range(producer, ZSTD_CHAINLOG_MIN, 16);
cParams.searchLog = FUZZ_dataProducer_uint32Range(producer, ZSTD_SEARCHLOG_MIN, 9);
cParams.minMatch = FUZZ_dataProducer_uint32Range(producer, ZSTD_MINMATCH_MIN,
ZSTD_MINMATCH_MAX);
cParams.targetLength = FUZZ_rand32(state, 0, 512);
cParams.strategy = FUZZ_rand32(state, ZSTD_STRATEGY_MIN, ZSTD_STRATEGY_MAX);
cParams.targetLength = FUZZ_dataProducer_uint32Range(producer, 0, 512);
cParams.strategy = FUZZ_dataProducer_uint32Range(producer, ZSTD_STRATEGY_MIN, ZSTD_STRATEGY_MAX);
return ZSTD_adjustCParams(cParams, srcSize, 0);
}
ZSTD_frameParameters FUZZ_randomFParams(uint32_t *state)
ZSTD_frameParameters FUZZ_randomFParams(FUZZ_dataProducer_t *producer)
{
/* Select frame parameters */
ZSTD_frameParameters fParams;
fParams.contentSizeFlag = FUZZ_rand32(state, 0, 1);
fParams.checksumFlag = FUZZ_rand32(state, 0, 1);
fParams.noDictIDFlag = FUZZ_rand32(state, 0, 1);
fParams.contentSizeFlag = FUZZ_dataProducer_uint32Range(producer, 0, 1);
fParams.checksumFlag = FUZZ_dataProducer_uint32Range(producer, 0, 1);
fParams.noDictIDFlag = FUZZ_dataProducer_uint32Range(producer, 0, 1);
return fParams;
}
ZSTD_parameters FUZZ_randomParams(size_t srcSize, uint32_t *state)
ZSTD_parameters FUZZ_randomParams(size_t srcSize, FUZZ_dataProducer_t *producer)
{
ZSTD_parameters params;
params.cParams = FUZZ_randomCParams(srcSize, state);
params.fParams = FUZZ_randomFParams(state);
params.cParams = FUZZ_randomCParams(srcSize, producer);
params.fParams = FUZZ_randomFParams(producer);
return params;
}
void FUZZ_setRandomParameters(ZSTD_CCtx *cctx, size_t srcSize, uint32_t *state)
void FUZZ_setRandomParameters(ZSTD_CCtx *cctx, size_t srcSize, FUZZ_dataProducer_t *producer)
{
ZSTD_compressionParameters cParams = FUZZ_randomCParams(srcSize, state);
ZSTD_compressionParameters cParams = FUZZ_randomCParams(srcSize, producer);
set(cctx, ZSTD_c_windowLog, cParams.windowLog);
set(cctx, ZSTD_c_hashLog, cParams.hashLog);
set(cctx, ZSTD_c_chainLog, cParams.chainLog);
@@ -72,30 +72,30 @@ void FUZZ_setRandomParameters(ZSTD_CCtx *cctx, size_t srcSize, uint32_t *state)
set(cctx, ZSTD_c_targetLength, cParams.targetLength);
set(cctx, ZSTD_c_strategy, cParams.strategy);
/* Select frame parameters */
setRand(cctx, ZSTD_c_contentSizeFlag, 0, 1, state);
setRand(cctx, ZSTD_c_checksumFlag, 0, 1, state);
setRand(cctx, ZSTD_c_dictIDFlag, 0, 1, state);
setRand(cctx, ZSTD_c_contentSizeFlag, 0, 1, producer);
setRand(cctx, ZSTD_c_checksumFlag, 0, 1, producer);
setRand(cctx, ZSTD_c_dictIDFlag, 0, 1, producer);
/* Select long distance matching parameters */
setRand(cctx, ZSTD_c_enableLongDistanceMatching, 0, 1, state);
setRand(cctx, ZSTD_c_ldmHashLog, ZSTD_HASHLOG_MIN, 16, state);
setRand(cctx, ZSTD_c_enableLongDistanceMatching, 0, 1, producer);
setRand(cctx, ZSTD_c_ldmHashLog, ZSTD_HASHLOG_MIN, 16, producer);
setRand(cctx, ZSTD_c_ldmMinMatch, ZSTD_LDM_MINMATCH_MIN,
ZSTD_LDM_MINMATCH_MAX, state);
ZSTD_LDM_MINMATCH_MAX, producer);
setRand(cctx, ZSTD_c_ldmBucketSizeLog, 0, ZSTD_LDM_BUCKETSIZELOG_MAX,
state);
producer);
setRand(cctx, ZSTD_c_ldmHashRateLog, ZSTD_LDM_HASHRATELOG_MIN,
ZSTD_LDM_HASHRATELOG_MAX, state);
ZSTD_LDM_HASHRATELOG_MAX, producer);
/* Set misc parameters */
setRand(cctx, ZSTD_c_nbWorkers, 0, 2, state);
setRand(cctx, ZSTD_c_rsyncable, 0, 1, state);
setRand(cctx, ZSTD_c_forceMaxWindow, 0, 1, state);
setRand(cctx, ZSTD_c_literalCompressionMode, 0, 2, state);
setRand(cctx, ZSTD_c_forceAttachDict, 0, 2, state);
if (FUZZ_rand32(state, 0, 1) == 0) {
setRand(cctx, ZSTD_c_srcSizeHint, ZSTD_SRCSIZEHINT_MIN, 2 * srcSize, state);
setRand(cctx, ZSTD_c_nbWorkers, 0, 2, producer);
setRand(cctx, ZSTD_c_rsyncable, 0, 1, producer);
setRand(cctx, ZSTD_c_forceMaxWindow, 0, 1, producer);
setRand(cctx, ZSTD_c_literalCompressionMode, 0, 2, producer);
setRand(cctx, ZSTD_c_forceAttachDict, 0, 2, producer);
if (FUZZ_dataProducer_uint32Range(producer, 0, 1) == 0) {
setRand(cctx, ZSTD_c_srcSizeHint, ZSTD_SRCSIZEHINT_MIN, 2 * srcSize, producer);
}
}
FUZZ_dict_t FUZZ_train(void const* src, size_t srcSize, uint32_t *state)
FUZZ_dict_t FUZZ_train(void const* src, size_t srcSize, FUZZ_dataProducer_t *producer)
{
size_t const dictSize = MAX(srcSize / 8, 1024);
size_t const totalSampleSize = dictSize * 11;
@@ -110,7 +110,7 @@ FUZZ_dict_t FUZZ_train(void const* src, size_t srcSize, uint32_t *state)
for (sample = 0; sample < nbSamples; ++sample) {
size_t const remaining = totalSampleSize - pos;
size_t const offset = FUZZ_rand32(state, 0, MAX(srcSize, 1) - 1);
size_t const offset = FUZZ_dataProducer_uint32Range(producer, 0, MAX(srcSize, 1) - 1);
size_t const limit = MIN(srcSize - offset, remaining);
size_t const toCopy = MIN(limit, remaining / (nbSamples - sample));
memcpy(samples + pos, src + offset, toCopy);