diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-27 18:24:20 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-27 18:24:20 +0000 |
commit | 483eb2f56657e8e7f419ab1a4fab8dce9ade8609 (patch) | |
tree | e5d88d25d870d5dedacb6bbdbe2a966086a0a5cf /src/zstd/tests/fuzz | |
parent | Initial commit. (diff) | |
download | ceph-upstream.tar.xz ceph-upstream.zip |
Adding upstream version 14.2.21.upstream/14.2.21upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r-- | src/zstd/tests/fuzz/.gitignore | 5 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/Makefile | 127 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/README.md | 96 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/block_decompress.c | 51 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/block_round_trip.c | 92 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/default.options | 2 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/fuzz.h | 62 | ||||
-rwxr-xr-x | src/zstd/tests/fuzz/fuzz.py | 818 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/fuzz_helpers.h | 92 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/regression_driver.c | 71 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/simple_decompress.c | 49 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/simple_round_trip.c | 95 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/stream_decompress.c | 85 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/stream_round_trip.c | 162 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/zstd_helpers.c | 84 | ||||
-rw-r--r-- | src/zstd/tests/fuzz/zstd_helpers.h | 35 | ||||
-rw-r--r-- | src/zstd/tests/fuzzer.c | 1643 |
17 files changed, 3569 insertions, 0 deletions
diff --git a/src/zstd/tests/fuzz/.gitignore b/src/zstd/tests/fuzz/.gitignore new file mode 100644 index 00000000..4ff28de9 --- /dev/null +++ b/src/zstd/tests/fuzz/.gitignore @@ -0,0 +1,5 @@ +# test artefacts +corpora +block_decompress +block_round_trip +simple_round_trip diff --git a/src/zstd/tests/fuzz/Makefile b/src/zstd/tests/fuzz/Makefile new file mode 100644 index 00000000..d9b00fd2 --- /dev/null +++ b/src/zstd/tests/fuzz/Makefile @@ -0,0 +1,127 @@ +# ################################################################ +# Copyright (c) 2016-present, Facebook, Inc. +# All rights reserved. +# +# This source code is licensed under both the BSD-style license (found in the +# LICENSE file in the root directory of this source tree) and the GPLv2 (found +# in the COPYING file in the root directory of this source tree). +# ################################################################ + +# Optionally user defined flags +CFLAGS ?= -O3 +CXXFLAGS ?= -O3 +CPPFLAGS ?= +LDFLAGS ?= +ARFLAGS ?= +LIB_FUZZING_ENGINE ?= libregression.a +PYTHON ?= python +ifeq ($(shell uname), Darwin) + DOWNLOAD?=curl -L -o +else + DOWNLOAD?=wget -O +endif +CORPORA_URL_PREFIX:=https://github.com/facebook/zstd/releases/download/fuzz-corpora/ + +ZSTDDIR = ../../lib +PRGDIR = ../../programs + +FUZZ_CPPFLAGS := -I$(ZSTDDIR) -I$(ZSTDDIR)/common -I$(ZSTDDIR)/compress \ + -I$(ZSTDDIR)/dictBuilder -I$(ZSTDDIR)/deprecated -I$(PRGDIR) \ + $(CPPFLAGS) +FUZZ_EXTRA_FLAGS := -Wall -Wextra -Wcast-qual -Wcast-align -Wshadow \ + -Wstrict-aliasing=1 -Wswitch-enum -Wdeclaration-after-statement \ + -Wstrict-prototypes -Wundef -Wformat-security \ + -Wvla -Wformat=2 -Winit-self -Wfloat-equal -Wwrite-strings \ + -Wredundant-decls \ + -g -fno-omit-frame-pointer +FUZZ_CFLAGS := $(FUZZ_EXTRA_FLAGS) $(CFLAGS) +FUZZ_CXXFLAGS := $(FUZZ_EXTRA_FLAGS) -std=c++11 $(CXXFLAGS) +FUZZ_LDFLAGS := $(LDFLAGS) +FUZZ_ARFLAGS := $(ARFLAGS) +FUZZ_TARGET_FLAGS = $(FUZZ_CPPFLAGS) $(FUZZ_CXXFLAGS) $(FUZZ_LDFLAGS) + +FUZZ_HEADERS := fuzz_helpers.h fuzz.h zstd_helpers.h +FUZZ_SRC := zstd_helpers.c + +ZSTDCOMMON_SRC := $(ZSTDDIR)/common/*.c +ZSTDCOMP_SRC := $(ZSTDDIR)/compress/*.c +ZSTDDECOMP_SRC := $(ZSTDDIR)/decompress/*.c +FUZZ_SRC := \ + $(FUZZ_SRC) \ + $(ZSTDDECOMP_SRC) \ + $(ZSTDCOMMON_SRC) \ + $(ZSTDCOMP_SRC) + +FUZZ_OBJ := $(patsubst %.c,%.o, $(wildcard $(FUZZ_SRC))) + + +.PHONY: default all clean cleanall + +default: all + +FUZZ_TARGETS := \ + simple_round_trip \ + stream_round_trip \ + block_round_trip \ + simple_decompress \ + stream_decompress \ + block_decompress + +all: $(FUZZ_TARGETS) + +%.o: %.c + $(CC) $(FUZZ_CPPFLAGS) $(FUZZ_CFLAGS) $^ -c -o $@ + +simple_round_trip: $(FUZZ_HEADERS) $(FUZZ_OBJ) simple_round_trip.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) simple_round_trip.o $(LIB_FUZZING_ENGINE) -o $@ + +stream_round_trip: $(FUZZ_HEADERS) $(FUZZ_OBJ) stream_round_trip.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) stream_round_trip.o $(LIB_FUZZING_ENGINE) -o $@ + +block_round_trip: $(FUZZ_HEADERS) $(FUZZ_OBJ) block_round_trip.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) block_round_trip.o $(LIB_FUZZING_ENGINE) -o $@ + +simple_decompress: $(FUZZ_HEADERS) $(FUZZ_OBJ) simple_decompress.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) simple_decompress.o $(LIB_FUZZING_ENGINE) -o $@ + +stream_decompress: $(FUZZ_HEADERS) $(FUZZ_OBJ) stream_decompress.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) stream_decompress.o $(LIB_FUZZING_ENGINE) -o $@ + +block_decompress: $(FUZZ_HEADERS) $(FUZZ_OBJ) block_decompress.o + $(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) block_decompress.o $(LIB_FUZZING_ENGINE) -o $@ + +libregression.a: $(FUZZ_HEADERS) $(PRGDIR)/util.h regression_driver.o + $(AR) $(FUZZ_ARFLAGS) $@ regression_driver.o + +# Install libfuzzer (not usable for MSAN testing) +# Provided for convienence. To use this library run make libFuzzer and +# set LDFLAGS=-L. +.PHONY: libFuzzer +libFuzzer: + @$(RM) -rf Fuzzer + @git clone https://chromium.googlesource.com/chromium/llvm-project/llvm/lib/Fuzzer + @cd Fuzzer && ./build.sh + +corpora/%_seed_corpus.zip: + @mkdir -p corpora + $(DOWNLOAD) $@ $(CORPORA_URL_PREFIX)$*_seed_corpus.zip + +corpora/%: corpora/%_seed_corpus.zip + unzip -q $^ -d $@ + +.PHONY: corpora +corpora: $(patsubst %,corpora/%,$(FUZZ_TARGETS)) + +regressiontest: corpora + CC="$(CC)" CXX="$(CXX)" CFLAGS="$(CFLAGS)" CXXFLAGS="$(CXXFLAGS)" LDFLAGS="$(LDFLAGS)" $(PYTHON) ./fuzz.py build all + $(PYTHON) ./fuzz.py regression all + +clean: + @$(MAKE) -C $(ZSTDDIR) clean + @$(RM) *.a *.o + @$(RM) simple_round_trip stream_round_trip simple_decompress \ + stream_decompress block_decompress block_round_trip + +cleanall: + @$(RM) -r Fuzzer + @$(RM) -r corpora diff --git a/src/zstd/tests/fuzz/README.md b/src/zstd/tests/fuzz/README.md new file mode 100644 index 00000000..f184be64 --- /dev/null +++ b/src/zstd/tests/fuzz/README.md @@ -0,0 +1,96 @@ +# Fuzzing + +Each fuzzing target can be built with multiple engines. +Zstd provides a fuzz corpus for each target that can be downloaded with +the command: + +``` +make corpora +``` + +It will download each corpus into `./corpora/TARGET`. + +## fuzz.py + +`fuzz.py` is a helper script for building and running fuzzers. +Run `./fuzz.py -h` for the commands and run `./fuzz.py COMMAND -h` for +command specific help. + +### Generating Data + +`fuzz.py` provides a utility to generate seed data for each fuzzer. + +``` +make -C ../tests decodecorpus +./fuzz.py gen TARGET +``` + +By default it outputs 100 samples, each at most 8KB into `corpora/TARGET-seed`, +but that can be configured with the `--number`, `--max-size-log` and `--seed` +flags. + +### Build +It respects the usual build environment variables `CC`, `CFLAGS`, etc. +The environment variables can be overridden with the corresponding flags +`--cc`, `--cflags`, etc. +The specific fuzzing engine is selected with `LIB_FUZZING_ENGINE` or +`--lib-fuzzing-engine`, the default is `libregression.a`. +It has flags that can easily set up sanitizers `--enable-{a,ub,m}san`, and +coverage instrumentation `--enable-coverage`. +It sets sane defaults which can be overriden with flags `--debug`, +`--enable-ubsan-pointer-overlow`, etc. +Run `./fuzz.py build -h` for help. + +### Running Fuzzers + +`./fuzz.py` can run `libfuzzer`, `afl`, and `regression` tests. +See the help of the relevant command for options. +Flags not parsed by `fuzz.py` are passed to the fuzzing engine. +The command used to run the fuzzer is printed for debugging. + +## LibFuzzer + +``` +# Build libfuzzer if necessary +make libFuzzer +# Build the fuzz targets +./fuzz.py build all --enable-coverage --enable-asan --enable-ubsan --lib-fuzzing-engine Fuzzer/libFuzzer.a --cc clang --cxx clang++ +# OR equivalently +CC=clang CXX=clang++ LIB_FUZZING_ENGINE=Fuzzer/libFuzzer.a ./fuzz.py build all --enable-coverage --enable-asan --enable-ubsan +# Run the fuzzer +./fuzz.py libfuzzer TARGET -max_len=8192 -jobs=4 +``` + +where `TARGET` could be `simple_decompress`, `stream_round_trip`, etc. + +### MSAN + +Fuzzing with `libFuzzer` and `MSAN` will require building a C++ standard library +and libFuzzer with MSAN. +`fuzz.py` respects the environment variables / flags `MSAN_EXTRA_CPPFLAGS`, +`MSAN_EXTRA_CFLAGS`, `MSAN_EXTRA_CXXFLAGS`, `MSAN_EXTRA_LDFLAGS` to easily pass +the extra parameters only for MSAN. + +## AFL + +The default `LIB_FUZZING_ENGINE` is `libregression.a`, which produces a binary +that AFL can use. + +``` +# Build the fuzz targets +CC=afl-clang CXX=afl-clang++ ./fuzz.py build all --enable-asan --enable-ubsan +# Run the fuzzer without a memory limit because of ASAN +./fuzz.py afl TARGET -m none +``` + +## Regression Testing + +The regression rest supports the `all` target to run all the fuzzers in one +command. + +``` +CC=clang CXX=clang++ ./fuzz.py build all --enable-asan --enable-ubsan +./fuzz.py regression all +CC=clang CXX=clang++ ./fuzz.py build all --enable-msan +./fuzz.py regression all +``` diff --git a/src/zstd/tests/fuzz/block_decompress.c b/src/zstd/tests/fuzz/block_decompress.c new file mode 100644 index 00000000..3cccc32f --- /dev/null +++ b/src/zstd/tests/fuzz/block_decompress.c @@ -0,0 +1,51 @@ +/** + * Copyright (c) 2016-present, Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target attempts to decompress the fuzzed data with the simple + * decompression function to ensure the decompressor never crashes. + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include "fuzz_helpers.h" +#include "zstd.h" + +static ZSTD_DCtx *dctx = NULL; +static void* rBuf = NULL; +static size_t bufSize = 0; + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + size_t const neededBufSize = ZSTD_BLOCKSIZE_MAX; + + FUZZ_seed(&src, &size); + + /* Allocate all buffers and contexts if not already allocated */ + if (neededBufSize > bufSize) { + free(rBuf); + rBuf = malloc(neededBufSize); + bufSize = neededBufSize; + FUZZ_ASSERT(rBuf); + } + if (!dctx) { + dctx = ZSTD_createDCtx(); + FUZZ_ASSERT(dctx); + } + ZSTD_decompressBegin(dctx); + ZSTD_decompressBlock(dctx, rBuf, neededBufSize, src, size); + +#ifndef STATEFUL_FUZZING + ZSTD_freeDCtx(dctx); dctx = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/block_round_trip.c b/src/zstd/tests/fuzz/block_round_trip.c new file mode 100644 index 00000000..64ca5fc4 --- /dev/null +++ b/src/zstd/tests/fuzz/block_round_trip.c @@ -0,0 +1,92 @@ +/** + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target performs a zstd round-trip test (compress & decompress), + * compares the result with the original, and calls abort() on corruption. + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include "fuzz_helpers.h" +#include "zstd.h" + +static const int kMaxClevel = 19; + +static ZSTD_CCtx *cctx = NULL; +static ZSTD_DCtx *dctx = NULL; +static void* cBuf = NULL; +static void* rBuf = NULL; +static size_t bufSize = 0; +static uint32_t seed; + +static size_t roundTripTest(void *result, size_t resultCapacity, + void *compressed, size_t compressedCapacity, + const void *src, size_t srcSize) +{ + int const cLevel = FUZZ_rand(&seed) % kMaxClevel; + ZSTD_parameters const params = ZSTD_getParams(cLevel, srcSize, 0); + size_t ret = ZSTD_compressBegin_advanced(cctx, NULL, 0, params, srcSize); + FUZZ_ZASSERT(ret); + + ret = ZSTD_compressBlock(cctx, compressed, compressedCapacity, src, srcSize); + FUZZ_ZASSERT(ret); + if (ret == 0) { + FUZZ_ASSERT(resultCapacity >= srcSize); + memcpy(result, src, srcSize); + return srcSize; + } + ZSTD_decompressBegin(dctx); + return ZSTD_decompressBlock(dctx, result, resultCapacity, compressed, ret); +} + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + size_t neededBufSize; + + seed = FUZZ_seed(&src, &size); + neededBufSize = size; + if (size > ZSTD_BLOCKSIZE_MAX) + return 0; + + /* Allocate all buffers and contexts if not already allocated */ + if (neededBufSize > bufSize || !cBuf || !rBuf) { + free(cBuf); + free(rBuf); + cBuf = malloc(neededBufSize); + rBuf = malloc(neededBufSize); + bufSize = neededBufSize; + FUZZ_ASSERT(cBuf && rBuf); + } + if (!cctx) { + cctx = ZSTD_createCCtx(); + FUZZ_ASSERT(cctx); + } + if (!dctx) { + dctx = ZSTD_createDCtx(); + FUZZ_ASSERT(dctx); + } + + { + size_t const result = + roundTripTest(rBuf, neededBufSize, cBuf, neededBufSize, src, size); + FUZZ_ZASSERT(result); + FUZZ_ASSERT_MSG(result == size, "Incorrect regenerated size"); + FUZZ_ASSERT_MSG(!memcmp(src, rBuf, size), "Corruption!"); + } +#ifndef STATEFUL_FUZZING + ZSTD_freeCCtx(cctx); cctx = NULL; + ZSTD_freeDCtx(dctx); dctx = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/default.options b/src/zstd/tests/fuzz/default.options new file mode 100644 index 00000000..8ea85883 --- /dev/null +++ b/src/zstd/tests/fuzz/default.options @@ -0,0 +1,2 @@ +[libfuzzer] +max_len = 8192 diff --git a/src/zstd/tests/fuzz/fuzz.h b/src/zstd/tests/fuzz/fuzz.h new file mode 100644 index 00000000..a6484547 --- /dev/null +++ b/src/zstd/tests/fuzz/fuzz.h @@ -0,0 +1,62 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * Fuzz target interface. + * Fuzz targets have some common parameters passed as macros during compilation. + * Check the documentation for each individual fuzzer for more parameters. + * + * @param STATEFUL_FUZZING: + * Define this to reuse state between fuzzer runs. This can be useful to + * test code paths which are only executed when contexts are reused. + * WARNING: Makes reproducing crashes much harder. + * Default: Not defined. + * @param FUZZ_RNG_SEED_SIZE: + * The number of bytes of the source to look at when constructing a seed + * for the deterministic RNG. These bytes are discarded before passing + * the data to zstd functions. Every fuzzer initializes the RNG exactly + * once before doing anything else, even if it is unused. + * Default: 4. + * @param ZSTD_DEBUG: + * This is a parameter for the zstd library. Defining `ZSTD_DEBUG=1` + * enables assert() statements in the zstd library. Higher levels enable + * logging, so aren't recommended. Defining `ZSTD_DEBUG=1` is + * recommended. + * @param MEM_FORCE_MEMORY_ACCESS: + * This flag controls how the zstd library accesses unaligned memory. + * It can be undefined, or 0 through 2. If it is undefined, it selects + * the method to use based on the compiler. If testing with UBSAN set + * MEM_FORCE_MEMORY_ACCESS=0 to use the standard compliant method. + * @param FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION + * This is the canonical flag to enable deterministic builds for fuzzing. + * Changes to zstd for fuzzing are gated behind this define. + * It is recommended to define this when building zstd for fuzzing. + */ + +#ifndef FUZZ_H +#define FUZZ_H + +#ifndef FUZZ_RNG_SEED_SIZE +# define FUZZ_RNG_SEED_SIZE 4 +#endif + +#include <stddef.h> +#include <stdint.h> + +#ifdef __cplusplus +extern "C" { +#endif + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/zstd/tests/fuzz/fuzz.py b/src/zstd/tests/fuzz/fuzz.py new file mode 100755 index 00000000..b591e4f6 --- /dev/null +++ b/src/zstd/tests/fuzz/fuzz.py @@ -0,0 +1,818 @@ +#!/usr/bin/env python + +# ################################################################ +# Copyright (c) 2016-present, Facebook, Inc. +# All rights reserved. +# +# This source code is licensed under both the BSD-style license (found in the +# LICENSE file in the root directory of this source tree) and the GPLv2 (found +# in the COPYING file in the root directory of this source tree). +# ########################################################################## + +import argparse +import contextlib +import os +import re +import shutil +import subprocess +import sys +import tempfile + + +def abs_join(a, *p): + return os.path.abspath(os.path.join(a, *p)) + + +# Constants +FUZZ_DIR = os.path.abspath(os.path.dirname(__file__)) +CORPORA_DIR = abs_join(FUZZ_DIR, 'corpora') +TARGETS = [ + 'simple_round_trip', + 'stream_round_trip', + 'block_round_trip', + 'simple_decompress', + 'stream_decompress', + 'block_decompress', +] +ALL_TARGETS = TARGETS + ['all'] +FUZZ_RNG_SEED_SIZE = 4 + +# Standard environment variables +CC = os.environ.get('CC', 'cc') +CXX = os.environ.get('CXX', 'c++') +CPPFLAGS = os.environ.get('CPPFLAGS', '') +CFLAGS = os.environ.get('CFLAGS', '-O3') +CXXFLAGS = os.environ.get('CXXFLAGS', CFLAGS) +LDFLAGS = os.environ.get('LDFLAGS', '') +MFLAGS = os.environ.get('MFLAGS', '-j') + +# Fuzzing environment variables +LIB_FUZZING_ENGINE = os.environ.get('LIB_FUZZING_ENGINE', 'libregression.a') +AFL_FUZZ = os.environ.get('AFL_FUZZ', 'afl-fuzz') +DECODECORPUS = os.environ.get('DECODECORPUS', + abs_join(FUZZ_DIR, '..', 'decodecorpus')) + +# Sanitizer environment variables +MSAN_EXTRA_CPPFLAGS = os.environ.get('MSAN_EXTRA_CPPFLAGS', '') +MSAN_EXTRA_CFLAGS = os.environ.get('MSAN_EXTRA_CFLAGS', '') +MSAN_EXTRA_CXXFLAGS = os.environ.get('MSAN_EXTRA_CXXFLAGS', '') +MSAN_EXTRA_LDFLAGS = os.environ.get('MSAN_EXTRA_LDFLAGS', '') + + +def create(r): + d = os.path.abspath(r) + if not os.path.isdir(d): + os.mkdir(d) + return d + + +def check(r): + d = os.path.abspath(r) + if not os.path.isdir(d): + return None + return d + + +@contextlib.contextmanager +def tmpdir(): + dirpath = tempfile.mkdtemp() + try: + yield dirpath + finally: + shutil.rmtree(dirpath, ignore_errors=True) + + +def parse_targets(in_targets): + targets = set() + for target in in_targets: + if not target: + continue + if target == 'all': + targets = targets.union(TARGETS) + elif target in TARGETS: + targets.add(target) + else: + raise RuntimeError('{} is not a valid target'.format(target)) + return list(targets) + + +def targets_parser(args, description): + parser = argparse.ArgumentParser(prog=args.pop(0), description=description) + parser.add_argument( + 'TARGET', + nargs='*', + type=str, + help='Fuzz target(s) to build {{{}}}'.format(', '.join(ALL_TARGETS))) + args, extra = parser.parse_known_args(args) + args.extra = extra + + args.TARGET = parse_targets(args.TARGET) + + return args + + +def parse_env_flags(args, flags): + """ + Look for flags set by environment variables. + """ + san_flags = ','.join(re.findall('-fsanitize=((?:[a-z]+,?)+)', flags)) + nosan_flags = ','.join(re.findall('-fno-sanitize=((?:[a-z]+,?)+)', flags)) + + def set_sanitizer(sanitizer, default, san, nosan): + if sanitizer in san and sanitizer in nosan: + raise RuntimeError('-fno-sanitize={s} and -fsanitize={s} passed'. + format(s=sanitizer)) + if sanitizer in san: + return True + if sanitizer in nosan: + return False + return default + + san = set(san_flags.split(',')) + nosan = set(nosan_flags.split(',')) + + args.asan = set_sanitizer('address', args.asan, san, nosan) + args.msan = set_sanitizer('memory', args.msan, san, nosan) + args.ubsan = set_sanitizer('undefined', args.ubsan, san, nosan) + + args.sanitize = args.asan or args.msan or args.ubsan + + return args + + +def compiler_version(cc, cxx): + """ + Determines the compiler and version. + Only works for clang and gcc. + """ + cc_version_bytes = subprocess.check_output([cc, "--version"]) + cxx_version_bytes = subprocess.check_output([cxx, "--version"]) + if cc_version_bytes.startswith(b'clang'): + assert(cxx_version_bytes.startswith(b'clang')) + compiler = 'clang' + if cc_version_bytes.startswith(b'gcc'): + assert(cxx_version_bytes.startswith(b'g++')) + compiler = 'gcc' + version_regex = b'([0-9])+\.([0-9])+\.([0-9])+' + version_match = re.search(version_regex, cc_version_bytes) + version = tuple(int(version_match.group(i)) for i in range(1, 4)) + return compiler, version + + +def overflow_ubsan_flags(cc, cxx): + compiler, version = compiler_version(cc, cxx) + if compiler == 'gcc': + return ['-fno-sanitize=signed-integer-overflow'] + if compiler == 'clang' and version >= (5, 0, 0): + return ['-fno-sanitize=pointer-overflow'] + return [] + + +def build_parser(args): + description = """ + Cleans the repository and builds a fuzz target (or all). + Many flags default to environment variables (default says $X='y'). + Options that aren't enabling features default to the correct values for + zstd. + Enable sanitizers with --enable-*san. + For regression testing just build. + For libFuzzer set LIB_FUZZING_ENGINE and pass --enable-coverage. + For AFL set CC and CXX to AFL's compilers and set + LIB_FUZZING_ENGINE='libregression.a'. + """ + parser = argparse.ArgumentParser(prog=args.pop(0), description=description) + parser.add_argument( + '--lib-fuzzing-engine', + dest='lib_fuzzing_engine', + type=str, + default=LIB_FUZZING_ENGINE, + help=('The fuzzing engine to use e.g. /path/to/libFuzzer.a ' + "(default: $LIB_FUZZING_ENGINE='{})".format(LIB_FUZZING_ENGINE))) + parser.add_argument( + '--enable-coverage', + dest='coverage', + action='store_true', + help='Enable coverage instrumentation (-fsanitize-coverage)') + parser.add_argument( + '--enable-asan', dest='asan', action='store_true', help='Enable UBSAN') + parser.add_argument( + '--enable-ubsan', + dest='ubsan', + action='store_true', + help='Enable UBSAN') + parser.add_argument( + '--enable-ubsan-pointer-overflow', + dest='ubsan_pointer_overflow', + action='store_true', + help='Enable UBSAN pointer overflow check (known failure)') + parser.add_argument( + '--enable-msan', dest='msan', action='store_true', help='Enable MSAN') + parser.add_argument( + '--enable-msan-track-origins', dest='msan_track_origins', + action='store_true', help='Enable MSAN origin tracking') + parser.add_argument( + '--msan-extra-cppflags', + dest='msan_extra_cppflags', + type=str, + default=MSAN_EXTRA_CPPFLAGS, + help="Extra CPPFLAGS for MSAN (default: $MSAN_EXTRA_CPPFLAGS='{}')". + format(MSAN_EXTRA_CPPFLAGS)) + parser.add_argument( + '--msan-extra-cflags', + dest='msan_extra_cflags', + type=str, + default=MSAN_EXTRA_CFLAGS, + help="Extra CFLAGS for MSAN (default: $MSAN_EXTRA_CFLAGS='{}')".format( + MSAN_EXTRA_CFLAGS)) + parser.add_argument( + '--msan-extra-cxxflags', + dest='msan_extra_cxxflags', + type=str, + default=MSAN_EXTRA_CXXFLAGS, + help="Extra CXXFLAGS for MSAN (default: $MSAN_EXTRA_CXXFLAGS='{}')". + format(MSAN_EXTRA_CXXFLAGS)) + parser.add_argument( + '--msan-extra-ldflags', + dest='msan_extra_ldflags', + type=str, + default=MSAN_EXTRA_LDFLAGS, + help="Extra LDFLAGS for MSAN (default: $MSAN_EXTRA_LDFLAGS='{}')". + format(MSAN_EXTRA_LDFLAGS)) + parser.add_argument( + '--enable-sanitize-recover', + dest='sanitize_recover', + action='store_true', + help='Non-fatal sanitizer errors where possible') + parser.add_argument( + '--debug', + dest='debug', + type=int, + default=1, + help='Set ZSTD_DEBUG (default: 1)') + parser.add_argument( + '--force-memory-access', + dest='memory_access', + type=int, + default=0, + help='Set MEM_FORCE_MEMORY_ACCESS (default: 0)') + parser.add_argument( + '--fuzz-rng-seed-size', + dest='fuzz_rng_seed_size', + type=int, + default=4, + help='Set FUZZ_RNG_SEED_SIZE (default: 4)') + parser.add_argument( + '--disable-fuzzing-mode', + dest='fuzzing_mode', + action='store_false', + help='Do not define FUZZING_BUILD_MORE_UNSAFE_FOR_PRODUCTION') + parser.add_argument( + '--enable-stateful-fuzzing', + dest='stateful_fuzzing', + action='store_true', + help='Reuse contexts between runs (makes reproduction impossible)') + parser.add_argument( + '--cc', + dest='cc', + type=str, + default=CC, + help="CC (default: $CC='{}')".format(CC)) + parser.add_argument( + '--cxx', + dest='cxx', + type=str, + default=CXX, + help="CXX (default: $CXX='{}')".format(CXX)) + parser.add_argument( + '--cppflags', + dest='cppflags', + type=str, + default=CPPFLAGS, + help="CPPFLAGS (default: $CPPFLAGS='{}')".format(CPPFLAGS)) + parser.add_argument( + '--cflags', + dest='cflags', + type=str, + default=CFLAGS, + help="CFLAGS (default: $CFLAGS='{}')".format(CFLAGS)) + parser.add_argument( + '--cxxflags', + dest='cxxflags', + type=str, + default=CXXFLAGS, + help="CXXFLAGS (default: $CXXFLAGS='{}')".format(CXXFLAGS)) + parser.add_argument( + '--ldflags', + dest='ldflags', + type=str, + default=LDFLAGS, + help="LDFLAGS (default: $LDFLAGS='{}')".format(LDFLAGS)) + parser.add_argument( + '--mflags', + dest='mflags', + type=str, + default=MFLAGS, + help="Extra Make flags (default: $MFLAGS='{}')".format(MFLAGS)) + parser.add_argument( + 'TARGET', + nargs='*', + type=str, + help='Fuzz target(s) to build {{{}}}'.format(', '.join(ALL_TARGETS)) + ) + args = parser.parse_args(args) + args = parse_env_flags(args, ' '.join( + [args.cppflags, args.cflags, args.cxxflags, args.ldflags])) + + # Check option sanitiy + if args.msan and (args.asan or args.ubsan): + raise RuntimeError('MSAN may not be used with any other sanitizers') + if args.msan_track_origins and not args.msan: + raise RuntimeError('--enable-msan-track-origins requires MSAN') + if args.ubsan_pointer_overflow and not args.ubsan: + raise RuntimeError('--enable-ubsan-pointer-overlow requires UBSAN') + if args.sanitize_recover and not args.sanitize: + raise RuntimeError('--enable-sanitize-recover but no sanitizers used') + + return args + + +def build(args): + try: + args = build_parser(args) + except Exception as e: + print(e) + return 1 + # The compilation flags we are setting + targets = args.TARGET + cc = args.cc + cxx = args.cxx + cppflags = [args.cppflags] + cflags = [args.cflags] + ldflags = [args.ldflags] + cxxflags = [args.cxxflags] + mflags = [args.mflags] if args.mflags else [] + # Flags to be added to both cflags and cxxflags + common_flags = [] + + cppflags += [ + '-DZSTD_DEBUG={}'.format(args.debug), + '-DMEM_FORCE_MEMORY_ACCESS={}'.format(args.memory_access), + '-DFUZZ_RNG_SEED_SIZE={}'.format(args.fuzz_rng_seed_size), + ] + + mflags += ['LIB_FUZZING_ENGINE={}'.format(args.lib_fuzzing_engine)] + + # Set flags for options + if args.coverage: + common_flags += [ + '-fsanitize-coverage=trace-pc-guard,indirect-calls,trace-cmp' + ] + + if args.sanitize_recover: + recover_flags = ['-fsanitize-recover=all'] + else: + recover_flags = ['-fno-sanitize-recover=all'] + if args.sanitize: + common_flags += recover_flags + + if args.msan: + msan_flags = ['-fsanitize=memory'] + if args.msan_track_origins: + msan_flags += ['-fsanitize-memory-track-origins'] + common_flags += msan_flags + # Append extra MSAN flags (it might require special setup) + cppflags += [args.msan_extra_cppflags] + cflags += [args.msan_extra_cflags] + cxxflags += [args.msan_extra_cxxflags] + ldflags += [args.msan_extra_ldflags] + + if args.asan: + common_flags += ['-fsanitize=address'] + + if args.ubsan: + ubsan_flags = ['-fsanitize=undefined'] + if not args.ubsan_pointer_overflow: + ubsan_flags += overflow_ubsan_flags(cc, cxx) + common_flags += ubsan_flags + + if args.stateful_fuzzing: + cppflags += ['-DSTATEFUL_FUZZING'] + + if args.fuzzing_mode: + cppflags += ['-DFUZZING_BUILD_MORE_UNSAFE_FOR_PRODUCTION'] + + if args.lib_fuzzing_engine == 'libregression.a': + targets = ['libregression.a'] + targets + + # Append the common flags + cflags += common_flags + cxxflags += common_flags + + # Prepare the flags for Make + cc_str = "CC={}".format(cc) + cxx_str = "CXX={}".format(cxx) + cppflags_str = "CPPFLAGS={}".format(' '.join(cppflags)) + cflags_str = "CFLAGS={}".format(' '.join(cflags)) + cxxflags_str = "CXXFLAGS={}".format(' '.join(cxxflags)) + ldflags_str = "LDFLAGS={}".format(' '.join(ldflags)) + + # Print the flags + print('MFLAGS={}'.format(' '.join(mflags))) + print(cc_str) + print(cxx_str) + print(cppflags_str) + print(cflags_str) + print(cxxflags_str) + print(ldflags_str) + + # Clean and build + clean_cmd = ['make', 'clean'] + mflags + print(' '.join(clean_cmd)) + subprocess.check_call(clean_cmd) + build_cmd = [ + 'make', + cc_str, + cxx_str, + cppflags_str, + cflags_str, + cxxflags_str, + ldflags_str, + ] + mflags + targets + print(' '.join(build_cmd)) + subprocess.check_call(build_cmd) + return 0 + + +def libfuzzer_parser(args): + description = """ + Runs a libfuzzer binary. + Passes all extra arguments to libfuzzer. + The fuzzer should have been build with LIB_FUZZING_ENGINE pointing to + libFuzzer.a. + Generates output in the CORPORA directory, puts crashes in the ARTIFACT + directory, and takes extra input from the SEED directory. + To merge AFL's output pass the SEED as AFL's output directory and pass + '-merge=1'. + """ + parser = argparse.ArgumentParser(prog=args.pop(0), description=description) + parser.add_argument( + '--corpora', + type=str, + help='Override the default corpora dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET'))) + parser.add_argument( + '--artifact', + type=str, + help='Override the default artifact dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET-crash'))) + parser.add_argument( + '--seed', + type=str, + help='Override the default seed dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET-seed'))) + parser.add_argument( + 'TARGET', + type=str, + help='Fuzz target(s) to build {{{}}}'.format(', '.join(TARGETS))) + args, extra = parser.parse_known_args(args) + args.extra = extra + + if args.TARGET and args.TARGET not in TARGETS: + raise RuntimeError('{} is not a valid target'.format(args.TARGET)) + + return args + + +def libfuzzer(target, corpora=None, artifact=None, seed=None, extra_args=None): + if corpora is None: + corpora = abs_join(CORPORA_DIR, target) + if artifact is None: + artifact = abs_join(CORPORA_DIR, '{}-crash'.format(target)) + if seed is None: + seed = abs_join(CORPORA_DIR, '{}-seed'.format(target)) + if extra_args is None: + extra_args = [] + + target = abs_join(FUZZ_DIR, target) + + corpora = [create(corpora)] + artifact = create(artifact) + seed = check(seed) + + corpora += [artifact] + if seed is not None: + corpora += [seed] + + cmd = [target, '-artifact_prefix={}/'.format(artifact)] + cmd += corpora + extra_args + print(' '.join(cmd)) + subprocess.check_call(cmd) + + +def libfuzzer_cmd(args): + try: + args = libfuzzer_parser(args) + except Exception as e: + print(e) + return 1 + libfuzzer(args.TARGET, args.corpora, args.artifact, args.seed, args.extra) + return 0 + + +def afl_parser(args): + description = """ + Runs an afl-fuzz job. + Passes all extra arguments to afl-fuzz. + The fuzzer should have been built with CC/CXX set to the AFL compilers, + and with LIB_FUZZING_ENGINE='libregression.a'. + Takes input from CORPORA and writes output to OUTPUT. + Uses AFL_FUZZ as the binary (set from flag or environment variable). + """ + parser = argparse.ArgumentParser(prog=args.pop(0), description=description) + parser.add_argument( + '--corpora', + type=str, + help='Override the default corpora dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET'))) + parser.add_argument( + '--output', + type=str, + help='Override the default AFL output dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET-afl'))) + parser.add_argument( + '--afl-fuzz', + type=str, + default=AFL_FUZZ, + help='AFL_FUZZ (default: $AFL_FUZZ={})'.format(AFL_FUZZ)) + parser.add_argument( + 'TARGET', + type=str, + help='Fuzz target(s) to build {{{}}}'.format(', '.join(TARGETS))) + args, extra = parser.parse_known_args(args) + args.extra = extra + + if args.TARGET and args.TARGET not in TARGETS: + raise RuntimeError('{} is not a valid target'.format(args.TARGET)) + + if not args.corpora: + args.corpora = abs_join(CORPORA_DIR, args.TARGET) + if not args.output: + args.output = abs_join(CORPORA_DIR, '{}-afl'.format(args.TARGET)) + + return args + + +def afl(args): + try: + args = afl_parser(args) + except Exception as e: + print(e) + return 1 + target = abs_join(FUZZ_DIR, args.TARGET) + + corpora = create(args.corpora) + output = create(args.output) + + cmd = [args.afl_fuzz, '-i', corpora, '-o', output] + args.extra + cmd += [target, '@@'] + print(' '.join(cmd)) + subprocess.call(cmd) + return 0 + + +def regression(args): + try: + description = """ + Runs one or more regression tests. + The fuzzer should have been built with with + LIB_FUZZING_ENGINE='libregression.a'. + Takes input from CORPORA. + """ + args = targets_parser(args, description) + except Exception as e: + print(e) + return 1 + for target in args.TARGET: + corpora = create(abs_join(CORPORA_DIR, target)) + target = abs_join(FUZZ_DIR, target) + cmd = [target, corpora] + print(' '.join(cmd)) + subprocess.check_call(cmd) + return 0 + + +def gen_parser(args): + description = """ + Generate a seed corpus appropiate for TARGET with data generated with + decodecorpus. + The fuzz inputs are prepended with a seed before the zstd data, so the + output of decodecorpus shouldn't be used directly. + Generates NUMBER samples prepended with FUZZ_RNG_SEED_SIZE random bytes and + puts the output in SEED. + DECODECORPUS is the decodecorpus binary, and must already be built. + """ + parser = argparse.ArgumentParser(prog=args.pop(0), description=description) + parser.add_argument( + '--number', + '-n', + type=int, + default=100, + help='Number of samples to generate') + parser.add_argument( + '--max-size-log', + type=int, + default=13, + help='Maximum sample size to generate') + parser.add_argument( + '--seed', + type=str, + help='Override the default seed dir (default: {})'.format( + abs_join(CORPORA_DIR, 'TARGET-seed'))) + parser.add_argument( + '--decodecorpus', + type=str, + default=DECODECORPUS, + help="decodecorpus binary (default: $DECODECORPUS='{}')".format( + DECODECORPUS)) + parser.add_argument( + '--fuzz-rng-seed-size', + type=int, + default=4, + help="FUZZ_RNG_SEED_SIZE used for generate the samples (must match)" + ) + parser.add_argument( + 'TARGET', + type=str, + help='Fuzz target(s) to build {{{}}}'.format(', '.join(TARGETS))) + args, extra = parser.parse_known_args(args) + args.extra = extra + + if args.TARGET and args.TARGET not in TARGETS: + raise RuntimeError('{} is not a valid target'.format(args.TARGET)) + + if not args.seed: + args.seed = abs_join(CORPORA_DIR, '{}-seed'.format(args.TARGET)) + + if not os.path.isfile(args.decodecorpus): + raise RuntimeError("{} is not a file run 'make -C {} decodecorpus'". + format(args.decodecorpus, abs_join(FUZZ_DIR, '..'))) + + return args + + +def gen(args): + try: + args = gen_parser(args) + except Exception as e: + print(e) + return 1 + + seed = create(args.seed) + with tmpdir() as compressed: + with tmpdir() as decompressed: + cmd = [ + args.decodecorpus, + '-n{}'.format(args.number), + '-p{}/'.format(compressed), + '-o{}'.format(decompressed), + ] + + if 'block_' in args.TARGET: + cmd += [ + '--gen-blocks', + '--max-block-size-log={}'.format(args.max_size_log) + ] + else: + cmd += ['--max-content-size-log={}'.format(args.max_size_log)] + + print(' '.join(cmd)) + subprocess.check_call(cmd) + + if '_round_trip' in args.TARGET: + print('using decompressed data in {}'.format(decompressed)) + samples = decompressed + elif '_decompress' in args.TARGET: + print('using compressed data in {}'.format(compressed)) + samples = compressed + + # Copy the samples over and prepend the RNG seeds + for name in os.listdir(samples): + samplename = abs_join(samples, name) + outname = abs_join(seed, name) + rng_seed = os.urandom(args.fuzz_rng_seed_size) + with open(samplename, 'rb') as sample: + with open(outname, 'wb') as out: + out.write(rng_seed) + CHUNK_SIZE = 131072 + chunk = sample.read(CHUNK_SIZE) + while len(chunk) > 0: + out.write(chunk) + chunk = sample.read(CHUNK_SIZE) + return 0 + + +def minimize(args): + try: + description = """ + Runs a libfuzzer fuzzer with -merge=1 to build a minimal corpus in + TARGET_seed_corpus. All extra args are passed to libfuzzer. + """ + args = targets_parser(args, description) + except Exception as e: + print(e) + return 1 + + for target in args.TARGET: + # Merge the corpus + anything else into the seed_corpus + corpus = abs_join(CORPORA_DIR, target) + seed_corpus = abs_join(CORPORA_DIR, "{}_seed_corpus".format(target)) + extra_args = [corpus, "-merge=1"] + args.extra + libfuzzer(target, corpora=seed_corpus, extra_args=extra_args) + seeds = set(os.listdir(seed_corpus)) + # Copy all crashes directly into the seed_corpus if not already present + crashes = abs_join(CORPORA_DIR, '{}-crash'.format(target)) + for crash in os.listdir(crashes): + if crash not in seeds: + shutil.copy(abs_join(crashes, crash), seed_corpus) + seeds.add(crash) + + +def zip_cmd(args): + try: + description = """ + Zips up the seed corpus. + """ + args = targets_parser(args, description) + except Exception as e: + print(e) + return 1 + + for target in args.TARGET: + # Zip the seed_corpus + seed_corpus = abs_join(CORPORA_DIR, "{}_seed_corpus".format(target)) + seeds = [abs_join(seed_corpus, f) for f in os.listdir(seed_corpus)] + zip_file = "{}.zip".format(seed_corpus) + cmd = ["zip", "-q", "-j", "-9", zip_file] + print(' '.join(cmd + [abs_join(seed_corpus, '*')])) + subprocess.check_call(cmd + seeds) + + +def list_cmd(args): + print("\n".join(TARGETS)) + + +def short_help(args): + name = args[0] + print("Usage: {} [OPTIONS] COMMAND [ARGS]...\n".format(name)) + + +def help(args): + short_help(args) + print("\tfuzzing helpers (select a command and pass -h for help)\n") + print("Options:") + print("\t-h, --help\tPrint this message") + print("") + print("Commands:") + print("\tbuild\t\tBuild a fuzzer") + print("\tlibfuzzer\tRun a libFuzzer fuzzer") + print("\tafl\t\tRun an AFL fuzzer") + print("\tregression\tRun a regression test") + print("\tgen\t\tGenerate a seed corpus for a fuzzer") + print("\tminimize\tMinimize the test corpora") + print("\tzip\t\tZip the minimized corpora up") + print("\tlist\t\tList the available targets") + + +def main(): + args = sys.argv + if len(args) < 2: + help(args) + return 1 + if args[1] == '-h' or args[1] == '--help' or args[1] == '-H': + help(args) + return 1 + command = args.pop(1) + args[0] = "{} {}".format(args[0], command) + if command == "build": + return build(args) + if command == "libfuzzer": + return libfuzzer_cmd(args) + if command == "regression": + return regression(args) + if command == "afl": + return afl(args) + if command == "gen": + return gen(args) + if command == "minimize": + return minimize(args) + if command == "zip": + return zip_cmd(args) + if command == "list": + return list_cmd(args) + short_help(args) + print("Error: No such command {} (pass -h for help)".format(command)) + return 1 + + +if __name__ == "__main__": + sys.exit(main()) diff --git a/src/zstd/tests/fuzz/fuzz_helpers.h b/src/zstd/tests/fuzz/fuzz_helpers.h new file mode 100644 index 00000000..468c39fb --- /dev/null +++ b/src/zstd/tests/fuzz/fuzz_helpers.h @@ -0,0 +1,92 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * Helper functions for fuzzing. + */ + +#ifndef FUZZ_HELPERS_H +#define FUZZ_HELPERS_H + +#include "fuzz.h" +#include "xxhash.h" +#include "zstd.h" +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> + +#ifdef __cplusplus +extern "C" { +#endif + +#define MIN(a, b) ((a) < (b) ? (a) : (b)) +#define MAX(a, b) ((a) > (b) ? (a) : (b)) + +#define FUZZ_QUOTE_IMPL(str) #str +#define FUZZ_QUOTE(str) FUZZ_QUOTE_IMPL(str) + +/** + * Asserts for fuzzing that are always enabled. + */ +#define FUZZ_ASSERT_MSG(cond, msg) \ + ((cond) ? (void)0 \ + : (fprintf(stderr, "%s: %u: Assertion: `%s' failed. %s\n", __FILE__, \ + __LINE__, FUZZ_QUOTE(cond), (msg)), \ + abort())) +#define FUZZ_ASSERT(cond) FUZZ_ASSERT_MSG((cond), ""); +#define FUZZ_ZASSERT(code) \ + FUZZ_ASSERT_MSG(!ZSTD_isError(code), ZSTD_getErrorName(code)) + +#if defined(__GNUC__) +#define FUZZ_STATIC static __inline __attribute__((unused)) +#elif defined(__cplusplus) || \ + (defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) +#define FUZZ_STATIC static inline +#elif defined(_MSC_VER) +#define FUZZ_STATIC static __inline +#else +#define FUZZ_STATIC static +#endif + +/** + * Determininistically constructs a seed based on the fuzz input. + * Consumes up to the first FUZZ_RNG_SEED_SIZE bytes of the input. + */ +FUZZ_STATIC uint32_t FUZZ_seed(uint8_t const **src, size_t* size) { + uint8_t const *data = *src; + size_t const toHash = MIN(FUZZ_RNG_SEED_SIZE, *size); + *size -= toHash; + *src += toHash; + return XXH32(data, toHash, 0); +} + +#define FUZZ_rotl32(x, r) (((x) << (r)) | ((x) >> (32 - (r)))) + +FUZZ_STATIC uint32_t FUZZ_rand(uint32_t *state) { + static const uint32_t prime1 = 2654435761U; + static const uint32_t prime2 = 2246822519U; + uint32_t rand32 = *state; + rand32 *= prime1; + rand32 += prime2; + rand32 = FUZZ_rotl32(rand32, 13); + *state = rand32; + return rand32 >> 5; +} + +/* Returns a random numer in the range [min, max]. */ +FUZZ_STATIC uint32_t FUZZ_rand32(uint32_t *state, uint32_t min, uint32_t max) { + uint32_t random = FUZZ_rand(state); + return min + (random % (max - min + 1)); +} + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/zstd/tests/fuzz/regression_driver.c b/src/zstd/tests/fuzz/regression_driver.c new file mode 100644 index 00000000..2b714d29 --- /dev/null +++ b/src/zstd/tests/fuzz/regression_driver.c @@ -0,0 +1,71 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +#include "fuzz.h" +#include "fuzz_helpers.h" +#include "util.h" +#include <stddef.h> +#include <stdint.h> +#include <stdio.h> +#include <stdlib.h> + +int main(int argc, char const **argv) { + size_t const kMaxFileSize = (size_t)1 << 20; + int const kFollowLinks = 1; + char *fileNamesBuf = NULL; + char const **files = argv + 1; + unsigned numFiles = argc - 1; + uint8_t *buffer = NULL; + size_t bufferSize = 0; + unsigned i; + int ret; + +#ifdef UTIL_HAS_CREATEFILELIST + files = UTIL_createFileList(files, numFiles, &fileNamesBuf, &numFiles, + kFollowLinks); + if (!files) + numFiles = 0; +#endif + if (numFiles == 0) + fprintf(stderr, "WARNING: No files passed to %s\n", argv[0]); + for (i = 0; i < numFiles; ++i) { + char const *fileName = files[i]; + size_t const fileSize = UTIL_getFileSize(fileName); + size_t readSize; + FILE *file; + + /* Check that it is a regular file, and that the fileSize is valid */ + FUZZ_ASSERT_MSG(UTIL_isRegularFile(fileName), fileName); + FUZZ_ASSERT_MSG(fileSize <= kMaxFileSize, fileName); + /* Ensure we have a large enough buffer allocated */ + if (fileSize > bufferSize) { + free(buffer); + buffer = (uint8_t *)malloc(fileSize); + FUZZ_ASSERT_MSG(buffer, fileName); + bufferSize = fileSize; + } + /* Open the file */ + file = fopen(fileName, "rb"); + FUZZ_ASSERT_MSG(file, fileName); + /* Read the file */ + readSize = fread(buffer, 1, fileSize, file); + FUZZ_ASSERT_MSG(readSize == fileSize, fileName); + /* Close the file */ + fclose(file); + /* Run the fuzz target */ + LLVMFuzzerTestOneInput(buffer, fileSize); + } + + ret = 0; + free(buffer); +#ifdef UTIL_HAS_CREATEFILELIST + UTIL_freeFileList(files, fileNamesBuf); +#endif + return ret; +} diff --git a/src/zstd/tests/fuzz/simple_decompress.c b/src/zstd/tests/fuzz/simple_decompress.c new file mode 100644 index 00000000..bba272c6 --- /dev/null +++ b/src/zstd/tests/fuzz/simple_decompress.c @@ -0,0 +1,49 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target attempts to decompress the fuzzed data with the simple + * decompression function to ensure the decompressor never crashes. + */ + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include "fuzz_helpers.h" +#include "zstd.h" + +static ZSTD_DCtx *dctx = NULL; +static void* rBuf = NULL; +static size_t bufSize = 0; + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + size_t neededBufSize; + + FUZZ_seed(&src, &size); + neededBufSize = MAX(20 * size, (size_t)256 << 10); + + /* Allocate all buffers and contexts if not already allocated */ + if (neededBufSize > bufSize) { + free(rBuf); + rBuf = malloc(neededBufSize); + bufSize = neededBufSize; + FUZZ_ASSERT(rBuf); + } + if (!dctx) { + dctx = ZSTD_createDCtx(); + FUZZ_ASSERT(dctx); + } + ZSTD_decompressDCtx(dctx, rBuf, neededBufSize, src, size); + +#ifndef STATEFUL_FUZZING + ZSTD_freeDCtx(dctx); dctx = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/simple_round_trip.c b/src/zstd/tests/fuzz/simple_round_trip.c new file mode 100644 index 00000000..0921106d --- /dev/null +++ b/src/zstd/tests/fuzz/simple_round_trip.c @@ -0,0 +1,95 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target performs a zstd round-trip test (compress & decompress), + * compares the result with the original, and calls abort() on corruption. + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include "fuzz_helpers.h" +#include "zstd_helpers.h" + +static const int kMaxClevel = 19; + +static ZSTD_CCtx *cctx = NULL; +static ZSTD_DCtx *dctx = NULL; +static void* cBuf = NULL; +static void* rBuf = NULL; +static size_t bufSize = 0; +static uint32_t seed; + +static size_t roundTripTest(void *result, size_t resultCapacity, + void *compressed, size_t compressedCapacity, + const void *src, size_t srcSize) +{ + size_t cSize; + if (FUZZ_rand(&seed) & 1) { + ZSTD_inBuffer in = {src, srcSize, 0}; + ZSTD_outBuffer out = {compressed, compressedCapacity, 0}; + size_t err; + + ZSTD_CCtx_reset(cctx); + FUZZ_setRandomParameters(cctx, srcSize, &seed); + err = ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end); + FUZZ_ZASSERT(err); + FUZZ_ASSERT(err == 0); + cSize = out.pos; + } else { + int const cLevel = FUZZ_rand(&seed) % kMaxClevel; + cSize = ZSTD_compressCCtx( + cctx, compressed, compressedCapacity, src, srcSize, cLevel); + } + FUZZ_ZASSERT(cSize); + return ZSTD_decompressDCtx(dctx, result, resultCapacity, compressed, cSize); +} + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + size_t neededBufSize; + + seed = FUZZ_seed(&src, &size); + neededBufSize = ZSTD_compressBound(size); + + /* Allocate all buffers and contexts if not already allocated */ + if (neededBufSize > bufSize) { + free(cBuf); + free(rBuf); + cBuf = malloc(neededBufSize); + rBuf = malloc(neededBufSize); + bufSize = neededBufSize; + FUZZ_ASSERT(cBuf && rBuf); + } + if (!cctx) { + cctx = ZSTD_createCCtx(); + FUZZ_ASSERT(cctx); + } + if (!dctx) { + dctx = ZSTD_createDCtx(); + FUZZ_ASSERT(dctx); + } + + { + size_t const result = + roundTripTest(rBuf, neededBufSize, cBuf, neededBufSize, src, size); + FUZZ_ZASSERT(result); + FUZZ_ASSERT_MSG(result == size, "Incorrect regenerated size"); + FUZZ_ASSERT_MSG(!memcmp(src, rBuf, size), "Corruption!"); + } +#ifndef STATEFUL_FUZZING + ZSTD_freeCCtx(cctx); cctx = NULL; + ZSTD_freeDCtx(dctx); dctx = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/stream_decompress.c b/src/zstd/tests/fuzz/stream_decompress.c new file mode 100644 index 00000000..7ad57122 --- /dev/null +++ b/src/zstd/tests/fuzz/stream_decompress.c @@ -0,0 +1,85 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target attempts to decompress the fuzzed data with the simple + * decompression function to ensure the decompressor never crashes. + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include "fuzz_helpers.h" +#include "zstd.h" + +static size_t const kBufSize = ZSTD_BLOCKSIZE_MAX; + +static ZSTD_DStream *dstream = NULL; +static void* buf = NULL; +uint32_t seed; + +static ZSTD_outBuffer makeOutBuffer(void) +{ + ZSTD_outBuffer buffer = { buf, 0, 0 }; + + buffer.size = (FUZZ_rand(&seed) % kBufSize) + 1; + FUZZ_ASSERT(buffer.size <= kBufSize); + + return buffer; +} + +static ZSTD_inBuffer makeInBuffer(const uint8_t **src, size_t *size) +{ + ZSTD_inBuffer buffer = { *src, 0, 0 }; + + FUZZ_ASSERT(*size > 0); + buffer.size = (FUZZ_rand(&seed) % *size) + 1; + FUZZ_ASSERT(buffer.size <= *size); + *src += buffer.size; + *size -= buffer.size; + + return buffer; +} + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + seed = FUZZ_seed(&src, &size); + + /* Allocate all buffers and contexts if not already allocated */ + if (!buf) { + buf = malloc(kBufSize); + FUZZ_ASSERT(buf); + } + + if (!dstream) { + dstream = ZSTD_createDStream(); + FUZZ_ASSERT(dstream); + FUZZ_ASSERT(!ZSTD_isError(ZSTD_initDStream(dstream))); + } else { + FUZZ_ASSERT(!ZSTD_isError(ZSTD_resetDStream(dstream))); + } + + while (size > 0) { + ZSTD_inBuffer in = makeInBuffer(&src, &size); + while (in.pos != in.size) { + ZSTD_outBuffer out = makeOutBuffer(); + size_t const rc = ZSTD_decompressStream(dstream, &out, &in); + if (ZSTD_isError(rc)) goto error; + if (rc == 0) FUZZ_ASSERT(!ZSTD_isError(ZSTD_resetDStream(dstream))); + } + } + +error: +#ifndef STATEFUL_FUZZING + ZSTD_freeDStream(dstream); dstream = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/stream_round_trip.c b/src/zstd/tests/fuzz/stream_round_trip.c new file mode 100644 index 00000000..72d70495 --- /dev/null +++ b/src/zstd/tests/fuzz/stream_round_trip.c @@ -0,0 +1,162 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * This fuzz target performs a zstd round-trip test (compress & decompress), + * compares the result with the original, and calls abort() on corruption. + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include <stddef.h> +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include "fuzz_helpers.h" +#include "zstd_helpers.h" + +ZSTD_CCtx *cctx = NULL; +static ZSTD_DCtx *dctx = NULL; +static uint8_t* cBuf = NULL; +static uint8_t* rBuf = NULL; +static size_t bufSize = 0; +static uint32_t seed; + +static ZSTD_outBuffer makeOutBuffer(uint8_t *dst, size_t capacity) +{ + ZSTD_outBuffer buffer = { dst, 0, 0 }; + + FUZZ_ASSERT(capacity > 0); + buffer.size = (FUZZ_rand(&seed) % capacity) + 1; + FUZZ_ASSERT(buffer.size <= capacity); + + return buffer; +} + +static ZSTD_inBuffer makeInBuffer(const uint8_t **src, size_t *size) +{ + ZSTD_inBuffer buffer = { *src, 0, 0 }; + + FUZZ_ASSERT(*size > 0); + buffer.size = (FUZZ_rand(&seed) % *size) + 1; + FUZZ_ASSERT(buffer.size <= *size); + *src += buffer.size; + *size -= buffer.size; + + return buffer; +} + +static size_t compress(uint8_t *dst, size_t capacity, + const uint8_t *src, size_t srcSize) +{ + size_t dstSize = 0; + ZSTD_CCtx_reset(cctx); + FUZZ_setRandomParameters(cctx, srcSize, &seed); + + while (srcSize > 0) { + ZSTD_inBuffer in = makeInBuffer(&src, &srcSize); + /* Mode controls the action. If mode == -1 we pick a new mode */ + int mode = -1; + while (in.pos < in.size) { + ZSTD_outBuffer out = makeOutBuffer(dst, capacity); + /* Previous action finished, pick a new mode. */ + if (mode == -1) mode = FUZZ_rand(&seed) % 10; + switch (mode) { + case 0: /* fall-though */ + case 1: /* fall-though */ + case 2: { + size_t const ret = + ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_flush); + FUZZ_ZASSERT(ret); + if (ret == 0) + mode = -1; + break; + } + case 3: { + size_t ret = + ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end); + FUZZ_ZASSERT(ret); + /* Reset the compressor when the frame is finished */ + if (ret == 0) { + ZSTD_CCtx_reset(cctx); + if ((FUZZ_rand(&seed) & 7) == 0) { + size_t const remaining = in.size - in.pos; + FUZZ_setRandomParameters(cctx, remaining, &seed); + } + mode = -1; + } + break; + } + default: { + size_t const ret = + ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_continue); + FUZZ_ZASSERT(ret); + mode = -1; + } + } + dst += out.pos; + dstSize += out.pos; + capacity -= out.pos; + } + } + for (;;) { + ZSTD_inBuffer in = {NULL, 0, 0}; + ZSTD_outBuffer out = makeOutBuffer(dst, capacity); + size_t const ret = ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end); + FUZZ_ZASSERT(ret); + + dst += out.pos; + dstSize += out.pos; + capacity -= out.pos; + if (ret == 0) + break; + } + return dstSize; +} + +int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size) +{ + size_t neededBufSize; + + seed = FUZZ_seed(&src, &size); + neededBufSize = ZSTD_compressBound(size) * 2; + + /* Allocate all buffers and contexts if not already allocated */ + if (neededBufSize > bufSize) { + free(cBuf); + free(rBuf); + cBuf = (uint8_t*)malloc(neededBufSize); + rBuf = (uint8_t*)malloc(neededBufSize); + bufSize = neededBufSize; + FUZZ_ASSERT(cBuf && rBuf); + } + if (!cctx) { + cctx = ZSTD_createCCtx(); + FUZZ_ASSERT(cctx); + } + if (!dctx) { + dctx = ZSTD_createDCtx(); + FUZZ_ASSERT(dctx); + } + + { + size_t const cSize = compress(cBuf, neededBufSize, src, size); + size_t const rSize = + ZSTD_decompressDCtx(dctx, rBuf, neededBufSize, cBuf, cSize); + FUZZ_ZASSERT(rSize); + FUZZ_ASSERT_MSG(rSize == size, "Incorrect regenerated size"); + FUZZ_ASSERT_MSG(!memcmp(src, rBuf, size), "Corruption!"); + } + +#ifndef STATEFUL_FUZZING + ZSTD_freeCCtx(cctx); cctx = NULL; + ZSTD_freeDCtx(dctx); dctx = NULL; +#endif + return 0; +} diff --git a/src/zstd/tests/fuzz/zstd_helpers.c b/src/zstd/tests/fuzz/zstd_helpers.c new file mode 100644 index 00000000..60289847 --- /dev/null +++ b/src/zstd/tests/fuzz/zstd_helpers.c @@ -0,0 +1,84 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +#define ZSTD_STATIC_LINKING_ONLY + +#include "zstd_helpers.h" +#include "fuzz_helpers.h" +#include "zstd.h" + +static void set(ZSTD_CCtx *cctx, ZSTD_cParameter param, unsigned value) +{ + FUZZ_ZASSERT(ZSTD_CCtx_setParameter(cctx, param, 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); + set(cctx, param, value); +} + +ZSTD_compressionParameters FUZZ_randomCParams(size_t srcSize, uint32_t *state) +{ + /* 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.searchLength = FUZZ_rand32(state, ZSTD_SEARCHLENGTH_MIN, + ZSTD_SEARCHLENGTH_MAX); + cParams.targetLength = FUZZ_rand32(state, ZSTD_TARGETLENGTH_MIN, + ZSTD_TARGETLENGTH_MAX); + cParams.strategy = FUZZ_rand32(state, ZSTD_fast, ZSTD_btultra); + return ZSTD_adjustCParams(cParams, srcSize, 0); +} + +ZSTD_frameParameters FUZZ_randomFParams(uint32_t *state) +{ + /* 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); + return fParams; +} + +ZSTD_parameters FUZZ_randomParams(size_t srcSize, uint32_t *state) +{ + ZSTD_parameters params; + params.cParams = FUZZ_randomCParams(srcSize, state); + params.fParams = FUZZ_randomFParams(state); + return params; +} + +void FUZZ_setRandomParameters(ZSTD_CCtx *cctx, size_t srcSize, uint32_t *state) +{ + ZSTD_compressionParameters cParams = FUZZ_randomCParams(srcSize, state); + set(cctx, ZSTD_p_windowLog, cParams.windowLog); + set(cctx, ZSTD_p_hashLog, cParams.hashLog); + set(cctx, ZSTD_p_chainLog, cParams.chainLog); + set(cctx, ZSTD_p_searchLog, cParams.searchLog); + set(cctx, ZSTD_p_minMatch, cParams.searchLength); + set(cctx, ZSTD_p_targetLength, cParams.targetLength); + set(cctx, ZSTD_p_compressionStrategy, cParams.strategy); + /* Select frame parameters */ + setRand(cctx, ZSTD_p_contentSizeFlag, 0, 1, state); + setRand(cctx, ZSTD_p_checksumFlag, 0, 1, state); + setRand(cctx, ZSTD_p_dictIDFlag, 0, 1, state); + /* Select long distance matchig parameters */ + setRand(cctx, ZSTD_p_enableLongDistanceMatching, 0, 1, state); + setRand(cctx, ZSTD_p_ldmHashLog, ZSTD_HASHLOG_MIN, 16, state); + setRand(cctx, ZSTD_p_ldmMinMatch, ZSTD_LDM_MINMATCH_MIN, + ZSTD_LDM_MINMATCH_MAX, state); + setRand(cctx, ZSTD_p_ldmBucketSizeLog, 0, ZSTD_LDM_BUCKETSIZELOG_MAX, + state); + setRand(cctx, ZSTD_p_ldmHashEveryLog, 0, + ZSTD_WINDOWLOG_MAX - ZSTD_HASHLOG_MIN, state); +} diff --git a/src/zstd/tests/fuzz/zstd_helpers.h b/src/zstd/tests/fuzz/zstd_helpers.h new file mode 100644 index 00000000..3856bebe --- /dev/null +++ b/src/zstd/tests/fuzz/zstd_helpers.h @@ -0,0 +1,35 @@ +/* + * Copyright (c) 2016-present, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + */ + +/** + * Helper functions for fuzzing. + */ + +#ifndef ZSTD_HELPERS_H +#define ZSTD_HELPERS_H + +#include "zstd.h" +#include <stdint.h> + +#ifdef __cplusplus +extern "C" { +#endif + +void FUZZ_setRandomParameters(ZSTD_CCtx *cctx, size_t srcSize, uint32_t *state); + +ZSTD_compressionParameters FUZZ_randomCParams(size_t srcSize, uint32_t *state); +ZSTD_frameParameters FUZZ_randomFParams(uint32_t *state); +ZSTD_parameters FUZZ_randomParams(size_t srcSize, uint32_t *state); + + +#ifdef __cplusplus +} +#endif + +#endif /* ZSTD_HELPERS_H */ diff --git a/src/zstd/tests/fuzzer.c b/src/zstd/tests/fuzzer.c new file mode 100644 index 00000000..76df77af --- /dev/null +++ b/src/zstd/tests/fuzzer.c @@ -0,0 +1,1643 @@ +/* + * Copyright (c) 2015-present, Yann Collet, Facebook, Inc. + * All rights reserved. + * + * This source code is licensed under both the BSD-style license (found in the + * LICENSE file in the root directory of this source tree) and the GPLv2 (found + * in the COPYING file in the root directory of this source tree). + * You may select, at your option, one of the above-listed licenses. + */ + + +/*-************************************ +* Compiler specific +**************************************/ +#ifdef _MSC_VER /* Visual Studio */ +# define _CRT_SECURE_NO_WARNINGS /* fgets */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# pragma warning(disable : 4204) /* disable: C4204: non-constant aggregate initializer */ +#endif + + +/*-************************************ +* Includes +**************************************/ +#include <stdlib.h> /* free */ +#include <stdio.h> /* fgets, sscanf */ +#include <string.h> /* strcmp */ +#include <time.h> /* clock_t */ +#define ZSTD_STATIC_LINKING_ONLY /* ZSTD_compressContinue, ZSTD_compressBlock */ +#include "zstd.h" /* ZSTD_VERSION_STRING */ +#include "zstd_errors.h" /* ZSTD_getErrorCode */ +#include "zstdmt_compress.h" +#define ZDICT_STATIC_LINKING_ONLY +#include "zdict.h" /* ZDICT_trainFromBuffer */ +#include "datagen.h" /* RDG_genBuffer */ +#include "mem.h" +#define XXH_STATIC_LINKING_ONLY +#include "xxhash.h" /* XXH64 */ + + +/*-************************************ +* Constants +**************************************/ +#define KB *(1U<<10) +#define MB *(1U<<20) +#define GB *(1U<<30) + +static const U32 FUZ_compressibility_default = 50; +static const U32 nbTestsDefault = 30000; + + +/*-************************************ +* Display Macros +**************************************/ +#define DISPLAY(...) fprintf(stdout, __VA_ARGS__) +#define DISPLAYLEVEL(l, ...) if (g_displayLevel>=l) { DISPLAY(__VA_ARGS__); } +static U32 g_displayLevel = 2; + +#define DISPLAYUPDATE(l, ...) if (g_displayLevel>=l) { \ + if ((FUZ_clockSpan(g_displayClock) > g_refreshRate) || (g_displayLevel>=4)) \ + { g_displayClock = clock(); DISPLAY(__VA_ARGS__); \ + if (g_displayLevel>=4) fflush(stdout); } } +static const clock_t g_refreshRate = CLOCKS_PER_SEC / 6; +static clock_t g_displayClock = 0; + + +/*-******************************************************* +* Fuzzer functions +*********************************************************/ +#define MIN(a,b) ((a)<(b)?(a):(b)) +#define MAX(a,b) ((a)>(b)?(a):(b)) + +static clock_t FUZ_clockSpan(clock_t cStart) +{ + return clock() - cStart; /* works even when overflow; max span ~ 30mn */ +} + +#define FUZ_rotl32(x,r) ((x << r) | (x >> (32 - r))) +static unsigned FUZ_rand(unsigned* src) +{ + static const U32 prime1 = 2654435761U; + static const U32 prime2 = 2246822519U; + U32 rand32 = *src; + rand32 *= prime1; + rand32 += prime2; + rand32 = FUZ_rotl32(rand32, 13); + *src = rand32; + return rand32 >> 5; +} + +static unsigned FUZ_highbit32(U32 v32) +{ + unsigned nbBits = 0; + if (v32==0) return 0; + while (v32) v32 >>= 1, nbBits++; + return nbBits; +} + + +/*============================================= +* Memory Tests +=============================================*/ +#if defined(__APPLE__) && defined(__MACH__) + +#include <malloc/malloc.h> /* malloc_size */ + +typedef struct { + unsigned long long totalMalloc; + size_t currentMalloc; + size_t peakMalloc; + unsigned nbMalloc; + unsigned nbFree; +} mallocCounter_t; + +static const mallocCounter_t INIT_MALLOC_COUNTER = { 0, 0, 0, 0, 0 }; + +static void* FUZ_mallocDebug(void* counter, size_t size) +{ + mallocCounter_t* const mcPtr = (mallocCounter_t*)counter; + void* const ptr = malloc(size); + if (ptr==NULL) return NULL; + DISPLAYLEVEL(4, "allocating %u KB => effectively %u KB \n", + (U32)(size >> 10), (U32)(malloc_size(ptr) >> 10)); /* OS-X specific */ + mcPtr->totalMalloc += size; + mcPtr->currentMalloc += size; + if (mcPtr->currentMalloc > mcPtr->peakMalloc) + mcPtr->peakMalloc = mcPtr->currentMalloc; + mcPtr->nbMalloc += 1; + return ptr; +} + +static void FUZ_freeDebug(void* counter, void* address) +{ + mallocCounter_t* const mcPtr = (mallocCounter_t*)counter; + DISPLAYLEVEL(4, "freeing %u KB \n", (U32)(malloc_size(address) >> 10)); + mcPtr->nbFree += 1; + mcPtr->currentMalloc -= malloc_size(address); /* OS-X specific */ + free(address); +} + +static void FUZ_displayMallocStats(mallocCounter_t count) +{ + DISPLAYLEVEL(3, "peak:%6u KB, nbMallocs:%2u, total:%6u KB \n", + (U32)(count.peakMalloc >> 10), + count.nbMalloc, + (U32)(count.totalMalloc >> 10)); +} + +#define CHECK_Z(f) { \ + size_t const err = f; \ + if (ZSTD_isError(err)) { \ + DISPLAY("Error => %s : %s ", \ + #f, ZSTD_getErrorName(err)); \ + exit(1); \ +} } + +static int FUZ_mallocTests(unsigned seed, double compressibility, unsigned part) +{ + size_t const inSize = 64 MB + 16 MB + 4 MB + 1 MB + 256 KB + 64 KB; /* 85.3 MB */ + size_t const outSize = ZSTD_compressBound(inSize); + void* const inBuffer = malloc(inSize); + void* const outBuffer = malloc(outSize); + + /* test only played in verbose mode, as they are long */ + if (g_displayLevel<3) return 0; + + /* Create compressible noise */ + if (!inBuffer || !outBuffer) { + DISPLAY("Not enough memory, aborting\n"); + exit(1); + } + RDG_genBuffer(inBuffer, inSize, compressibility, 0. /*auto*/, seed); + + /* simple compression tests */ + if (part <= 1) + { int compressionLevel; + for (compressionLevel=1; compressionLevel<=6; compressionLevel++) { + mallocCounter_t malcount = INIT_MALLOC_COUNTER; + ZSTD_customMem const cMem = { FUZ_mallocDebug, FUZ_freeDebug, &malcount }; + ZSTD_CCtx* const cctx = ZSTD_createCCtx_advanced(cMem); + CHECK_Z( ZSTD_compressCCtx(cctx, outBuffer, outSize, inBuffer, inSize, compressionLevel) ); + ZSTD_freeCCtx(cctx); + DISPLAYLEVEL(3, "compressCCtx level %i : ", compressionLevel); + FUZ_displayMallocStats(malcount); + } } + + /* streaming compression tests */ + if (part <= 2) + { int compressionLevel; + for (compressionLevel=1; compressionLevel<=6; compressionLevel++) { + mallocCounter_t malcount = INIT_MALLOC_COUNTER; + ZSTD_customMem const cMem = { FUZ_mallocDebug, FUZ_freeDebug, &malcount }; + ZSTD_CCtx* const cstream = ZSTD_createCStream_advanced(cMem); + ZSTD_outBuffer out = { outBuffer, outSize, 0 }; + ZSTD_inBuffer in = { inBuffer, inSize, 0 }; + CHECK_Z( ZSTD_initCStream(cstream, compressionLevel) ); + CHECK_Z( ZSTD_compressStream(cstream, &out, &in) ); + CHECK_Z( ZSTD_endStream(cstream, &out) ); + ZSTD_freeCStream(cstream); + DISPLAYLEVEL(3, "compressStream level %i : ", compressionLevel); + FUZ_displayMallocStats(malcount); + } } + + /* advanced MT API test */ + if (part <= 3) + { U32 nbThreads; + for (nbThreads=1; nbThreads<=4; nbThreads++) { + int compressionLevel; + for (compressionLevel=1; compressionLevel<=6; compressionLevel++) { + mallocCounter_t malcount = INIT_MALLOC_COUNTER; + ZSTD_customMem const cMem = { FUZ_mallocDebug, FUZ_freeDebug, &malcount }; + ZSTD_CCtx* const cctx = ZSTD_createCCtx_advanced(cMem); + ZSTD_outBuffer out = { outBuffer, outSize, 0 }; + ZSTD_inBuffer in = { inBuffer, inSize, 0 }; + CHECK_Z( ZSTD_CCtx_setParameter(cctx, ZSTD_p_compressionLevel, (U32)compressionLevel) ); + CHECK_Z( ZSTD_CCtx_setParameter(cctx, ZSTD_p_nbThreads, nbThreads) ); + while ( ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end) ) {} + ZSTD_freeCCtx(cctx); + DISPLAYLEVEL(3, "compress_generic,-T%u,end level %i : ", + nbThreads, compressionLevel); + FUZ_displayMallocStats(malcount); + } } } + + /* advanced MT streaming API test */ + if (part <= 4) + { U32 nbThreads; + for (nbThreads=1; nbThreads<=4; nbThreads++) { + int compressionLevel; + for (compressionLevel=1; compressionLevel<=6; compressionLevel++) { + mallocCounter_t malcount = INIT_MALLOC_COUNTER; + ZSTD_customMem const cMem = { FUZ_mallocDebug, FUZ_freeDebug, &malcount }; + ZSTD_CCtx* const cctx = ZSTD_createCCtx_advanced(cMem); + ZSTD_outBuffer out = { outBuffer, outSize, 0 }; + ZSTD_inBuffer in = { inBuffer, inSize, 0 }; + CHECK_Z( ZSTD_CCtx_setParameter(cctx, ZSTD_p_compressionLevel, (U32)compressionLevel) ); + CHECK_Z( ZSTD_CCtx_setParameter(cctx, ZSTD_p_nbThreads, nbThreads) ); + CHECK_Z( ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_continue) ); + while ( ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end) ) {} + ZSTD_freeCCtx(cctx); + DISPLAYLEVEL(3, "compress_generic,-T%u,continue level %i : ", + nbThreads, compressionLevel); + FUZ_displayMallocStats(malcount); + } } } + + return 0; +} + +#else + +static int FUZ_mallocTests(unsigned seed, double compressibility, unsigned part) +{ + (void)seed; (void)compressibility; (void)part; + return 0; +} + +#endif + +/*============================================= +* Unit tests +=============================================*/ + +#define CHECK_V(var, fn) size_t const var = fn; if (ZSTD_isError(var)) goto _output_error +#define CHECK(fn) { CHECK_V(err, fn); } +#define CHECKPLUS(var, fn, more) { CHECK_V(var, fn); more; } + +static int basicUnitTests(U32 seed, double compressibility) +{ + size_t const CNBuffSize = 5 MB; + void* const CNBuffer = malloc(CNBuffSize); + size_t const compressedBufferSize = ZSTD_compressBound(CNBuffSize); + void* const compressedBuffer = malloc(compressedBufferSize); + void* const decodedBuffer = malloc(CNBuffSize); + ZSTD_DCtx* dctx = ZSTD_createDCtx(); + int testResult = 0; + U32 testNb=0; + size_t cSize; + + /* Create compressible noise */ + if (!CNBuffer || !compressedBuffer || !decodedBuffer) { + DISPLAY("Not enough memory, aborting\n"); + testResult = 1; + goto _end; + } + RDG_genBuffer(CNBuffer, CNBuffSize, compressibility, 0., seed); + + /* Basic tests */ + DISPLAYLEVEL(4, "test%3i : ZSTD_getErrorName : ", testNb++); + { const char* errorString = ZSTD_getErrorName(0); + DISPLAYLEVEL(4, "OK : %s \n", errorString); + } + + DISPLAYLEVEL(4, "test%3i : ZSTD_getErrorName with wrong value : ", testNb++); + { const char* errorString = ZSTD_getErrorName(499); + DISPLAYLEVEL(4, "OK : %s \n", errorString); + } + + + DISPLAYLEVEL(4, "test%3i : compress %u bytes : ", testNb++, (U32)CNBuffSize); + { ZSTD_CCtx* cctx = ZSTD_createCCtx(); + if (cctx==NULL) goto _output_error; + CHECKPLUS(r, ZSTD_compressCCtx(cctx, + compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, 1), + cSize=r ); + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : size of cctx for level 1 : ", testNb++); + { size_t const cctxSize = ZSTD_sizeof_CCtx(cctx); + DISPLAYLEVEL(4, "%u bytes \n", (U32)cctxSize); + } + ZSTD_freeCCtx(cctx); + } + + + DISPLAYLEVEL(4, "test%3i : ZSTD_getFrameContentSize test : ", testNb++); + { unsigned long long const rSize = ZSTD_getFrameContentSize(compressedBuffer, cSize); + if (rSize != CNBuffSize) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : ZSTD_findDecompressedSize test : ", testNb++); + { unsigned long long const rSize = ZSTD_findDecompressedSize(compressedBuffer, cSize); + if (rSize != CNBuffSize) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress %u bytes : ", testNb++, (U32)CNBuffSize); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : check decompressed result : ", testNb++); + { size_t u; + for (u=0; u<CNBuffSize; u++) { + if (((BYTE*)decodedBuffer)[u] != ((BYTE*)CNBuffer)[u]) goto _output_error;; + } } + DISPLAYLEVEL(4, "OK \n"); + + + DISPLAYLEVEL(4, "test%3i : decompress with null dict : ", testNb++); + { size_t const r = ZSTD_decompress_usingDict(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize, NULL, 0); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress with null DDict : ", testNb++); + { size_t const r = ZSTD_decompress_usingDDict(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize, NULL); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress with 1 missing byte : ", testNb++); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize-1); + if (!ZSTD_isError(r)) goto _output_error; + if (ZSTD_getErrorCode((size_t)r) != ZSTD_error_srcSize_wrong) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress with 1 too much byte : ", testNb++); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize+1); + if (!ZSTD_isError(r)) goto _output_error; + if (ZSTD_getErrorCode(r) != ZSTD_error_srcSize_wrong) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + + /* Static CCtx tests */ +#define STATIC_CCTX_LEVEL 3 + DISPLAYLEVEL(4, "test%3i : create static CCtx for level %u :", testNb++, STATIC_CCTX_LEVEL); + { size_t const staticCCtxSize = ZSTD_estimateCStreamSize(STATIC_CCTX_LEVEL); + void* const staticCCtxBuffer = malloc(staticCCtxSize); + size_t const staticDCtxSize = ZSTD_estimateDCtxSize(); + void* const staticDCtxBuffer = malloc(staticDCtxSize); + if (staticCCtxBuffer==NULL || staticDCtxBuffer==NULL) { + free(staticCCtxBuffer); + free(staticDCtxBuffer); + DISPLAY("Not enough memory, aborting\n"); + testResult = 1; + goto _end; + } + { ZSTD_CCtx* staticCCtx = ZSTD_initStaticCCtx(staticCCtxBuffer, staticCCtxSize); + ZSTD_DCtx* staticDCtx = ZSTD_initStaticDCtx(staticDCtxBuffer, staticDCtxSize); + if ((staticCCtx==NULL) || (staticDCtx==NULL)) goto _output_error; + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init CCtx for level %u : ", testNb++, STATIC_CCTX_LEVEL); + { size_t const r = ZSTD_compressBegin(staticCCtx, STATIC_CCTX_LEVEL); + if (ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : simple compression test with static CCtx : ", testNb++); + CHECKPLUS(r, ZSTD_compressCCtx(staticCCtx, + compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, STATIC_CCTX_LEVEL), + cSize=r ); + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", + (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : simple decompression test with static DCtx : ", testNb++); + { size_t const r = ZSTD_decompressDCtx(staticDCtx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : check decompressed result : ", testNb++); + { size_t u; + for (u=0; u<CNBuffSize; u++) { + if (((BYTE*)decodedBuffer)[u] != ((BYTE*)CNBuffer)[u]) + goto _output_error;; + } } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init CCtx for too large level (must fail) : ", testNb++); + { size_t const r = ZSTD_compressBegin(staticCCtx, ZSTD_maxCLevel()); + if (!ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init CCtx for small level %u (should work again) : ", testNb++, 1); + { size_t const r = ZSTD_compressBegin(staticCCtx, 1); + if (ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init CStream for small level %u : ", testNb++, 1); + { size_t const r = ZSTD_initCStream(staticCCtx, 1); + if (ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init CStream with dictionary (should fail) : ", testNb++); + { size_t const r = ZSTD_initCStream_usingDict(staticCCtx, CNBuffer, 64 KB, 1); + if (!ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : init DStream (should fail) : ", testNb++); + { size_t const r = ZSTD_initDStream(staticDCtx); + if (ZSTD_isError(r)) goto _output_error; } + { ZSTD_outBuffer output = { decodedBuffer, CNBuffSize, 0 }; + ZSTD_inBuffer input = { compressedBuffer, ZSTD_FRAMEHEADERSIZE_MAX+1, 0 }; + size_t const r = ZSTD_decompressStream(staticDCtx, &output, &input); + if (!ZSTD_isError(r)) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + } + free(staticCCtxBuffer); + free(staticDCtxBuffer); + } + + + + /* ZSTDMT simple MT compression test */ + DISPLAYLEVEL(4, "test%3i : create ZSTDMT CCtx : ", testNb++); + { ZSTDMT_CCtx* mtctx = ZSTDMT_createCCtx(2); + if (mtctx==NULL) { + DISPLAY("mtctx : mot enough memory, aborting \n"); + testResult = 1; + goto _end; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : compress %u bytes with 2 threads : ", testNb++, (U32)CNBuffSize); + CHECKPLUS(r, ZSTDMT_compressCCtx(mtctx, + compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, + 1), + cSize=r ); + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : decompressed size test : ", testNb++); + { unsigned long long const rSize = ZSTD_getFrameContentSize(compressedBuffer, cSize); + if (rSize != CNBuffSize) { + DISPLAY("ZSTD_getFrameContentSize incorrect : %u != %u \n", (U32)rSize, (U32)CNBuffSize); + goto _output_error; + } } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress %u bytes : ", testNb++, (U32)CNBuffSize); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : check decompressed result : ", testNb++); + { size_t u; + for (u=0; u<CNBuffSize; u++) { + if (((BYTE*)decodedBuffer)[u] != ((BYTE*)CNBuffer)[u]) goto _output_error;; + } } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : compress -T2 with checksum : ", testNb++); + { ZSTD_parameters params = ZSTD_getParams(1, CNBuffSize, 0); + params.fParams.checksumFlag = 1; + params.fParams.contentSizeFlag = 1; + CHECKPLUS(r, ZSTDMT_compress_advanced(mtctx, + compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, + NULL, params, 3 /*overlapRLog*/), + cSize=r ); + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : decompress %u bytes : ", testNb++, (U32)CNBuffSize); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize); + if (r != CNBuffSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + ZSTDMT_freeCCtx(mtctx); + } + + + /* Simple API multiframe test */ + DISPLAYLEVEL(4, "test%3i : compress multiple frames : ", testNb++); + { size_t off = 0; + int i; + int const segs = 4; + /* only use the first half so we don't push against size limit of compressedBuffer */ + size_t const segSize = (CNBuffSize / 2) / segs; + for (i = 0; i < segs; i++) { + CHECK_V(r, ZSTD_compress( + (BYTE *)compressedBuffer + off, CNBuffSize - off, + (BYTE *)CNBuffer + segSize * i, + segSize, 5)); + off += r; + if (i == segs/2) { + /* insert skippable frame */ + const U32 skipLen = 128 KB; + MEM_writeLE32((BYTE*)compressedBuffer + off, ZSTD_MAGIC_SKIPPABLE_START); + MEM_writeLE32((BYTE*)compressedBuffer + off + 4, skipLen); + off += skipLen + ZSTD_skippableHeaderSize; + } + } + cSize = off; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : get decompressed size of multiple frames : ", testNb++); + { unsigned long long const r = ZSTD_findDecompressedSize(compressedBuffer, cSize); + if (r != CNBuffSize / 2) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress multiple frames : ", testNb++); + { CHECK_V(r, ZSTD_decompress(decodedBuffer, CNBuffSize, compressedBuffer, cSize)); + if (r != CNBuffSize / 2) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : check decompressed result : ", testNb++); + if (memcmp(decodedBuffer, CNBuffer, CNBuffSize / 2) != 0) goto _output_error; + DISPLAYLEVEL(4, "OK \n"); + + /* Dictionary and CCtx Duplication tests */ + { ZSTD_CCtx* const ctxOrig = ZSTD_createCCtx(); + ZSTD_CCtx* const ctxDuplicated = ZSTD_createCCtx(); + static const size_t dictSize = 551; + + DISPLAYLEVEL(4, "test%3i : copy context too soon : ", testNb++); + { size_t const copyResult = ZSTD_copyCCtx(ctxDuplicated, ctxOrig, 0); + if (!ZSTD_isError(copyResult)) goto _output_error; } /* error must be detected */ + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : load dictionary into context : ", testNb++); + CHECK( ZSTD_compressBegin_usingDict(ctxOrig, CNBuffer, dictSize, 2) ); + CHECK( ZSTD_copyCCtx(ctxDuplicated, ctxOrig, 0) ); /* Begin_usingDict implies unknown srcSize, so match that */ + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : compress with flat dictionary : ", testNb++); + cSize = 0; + CHECKPLUS(r, ZSTD_compressEnd(ctxOrig, compressedBuffer, compressedBufferSize, + (const char*)CNBuffer + dictSize, CNBuffSize - dictSize), + cSize += r); + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : frame built with flat dictionary should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + CNBuffer, dictSize), + if (r != CNBuffSize - dictSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : compress with duplicated context : ", testNb++); + { size_t const cSizeOrig = cSize; + cSize = 0; + CHECKPLUS(r, ZSTD_compressEnd(ctxDuplicated, compressedBuffer, compressedBufferSize, + (const char*)CNBuffer + dictSize, CNBuffSize - dictSize), + cSize += r); + if (cSize != cSizeOrig) goto _output_error; /* should be identical ==> same size */ + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : frame built with duplicated context should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + CNBuffer, dictSize), + if (r != CNBuffSize - dictSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : decompress with DDict : ", testNb++); + { ZSTD_DDict* const ddict = ZSTD_createDDict(CNBuffer, dictSize); + size_t const r = ZSTD_decompress_usingDDict(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize, ddict); + if (r != CNBuffSize - dictSize) goto _output_error; + DISPLAYLEVEL(4, "OK (size of DDict : %u) \n", (U32)ZSTD_sizeof_DDict(ddict)); + ZSTD_freeDDict(ddict); + } + + DISPLAYLEVEL(4, "test%3i : decompress with static DDict : ", testNb++); + { size_t const ddictBufferSize = ZSTD_estimateDDictSize(dictSize, ZSTD_dlm_byCopy); + void* ddictBuffer = malloc(ddictBufferSize); + if (ddictBuffer == NULL) goto _output_error; + { ZSTD_DDict* const ddict = ZSTD_initStaticDDict(ddictBuffer, ddictBufferSize, CNBuffer, dictSize, ZSTD_dlm_byCopy); + size_t const r = ZSTD_decompress_usingDDict(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize, ddict); + if (r != CNBuffSize - dictSize) goto _output_error; + } + free(ddictBuffer); + DISPLAYLEVEL(4, "OK (size of static DDict : %u) \n", (U32)ddictBufferSize); + } + + DISPLAYLEVEL(4, "test%3i : check content size on duplicated context : ", testNb++); + { size_t const testSize = CNBuffSize / 3; + { ZSTD_parameters p = ZSTD_getParams(2, testSize, dictSize); + p.fParams.contentSizeFlag = 1; + CHECK( ZSTD_compressBegin_advanced(ctxOrig, CNBuffer, dictSize, p, testSize-1) ); + } + CHECK( ZSTD_copyCCtx(ctxDuplicated, ctxOrig, testSize) ); + + CHECKPLUS(r, ZSTD_compressEnd(ctxDuplicated, compressedBuffer, ZSTD_compressBound(testSize), + (const char*)CNBuffer + dictSize, testSize), + cSize = r); + { ZSTD_frameHeader zfh; + if (ZSTD_getFrameHeader(&zfh, compressedBuffer, cSize)) goto _output_error; + if ((zfh.frameContentSize != testSize) && (zfh.frameContentSize != 0)) goto _output_error; + } } + DISPLAYLEVEL(4, "OK \n"); + + ZSTD_freeCCtx(ctxOrig); + ZSTD_freeCCtx(ctxDuplicated); + } + + /* Dictionary and dictBuilder tests */ + { ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + size_t dictSize = 16 KB; + void* dictBuffer = malloc(dictSize); + size_t const totalSampleSize = 1 MB; + size_t const sampleUnitSize = 8 KB; + U32 const nbSamples = (U32)(totalSampleSize / sampleUnitSize); + size_t* const samplesSizes = (size_t*) malloc(nbSamples * sizeof(size_t)); + U32 dictID; + + if (dictBuffer==NULL || samplesSizes==NULL) { + free(dictBuffer); + free(samplesSizes); + goto _output_error; + } + + DISPLAYLEVEL(4, "test%3i : dictBuilder : ", testNb++); + { U32 u; for (u=0; u<nbSamples; u++) samplesSizes[u] = sampleUnitSize; } + dictSize = ZDICT_trainFromBuffer(dictBuffer, dictSize, + CNBuffer, samplesSizes, nbSamples); + if (ZDICT_isError(dictSize)) goto _output_error; + DISPLAYLEVEL(4, "OK, created dictionary of size %u \n", (U32)dictSize); + + DISPLAYLEVEL(4, "test%3i : check dictID : ", testNb++); + dictID = ZDICT_getDictID(dictBuffer, dictSize); + if (dictID==0) goto _output_error; + DISPLAYLEVEL(4, "OK : %u \n", dictID); + + DISPLAYLEVEL(4, "test%3i : compress with dictionary : ", testNb++); + cSize = ZSTD_compress_usingDict(cctx, compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, + dictBuffer, dictSize, 4); + if (ZSTD_isError(cSize)) goto _output_error; + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : retrieve dictID from dictionary : ", testNb++); + { U32 const did = ZSTD_getDictID_fromDict(dictBuffer, dictSize); + if (did != dictID) goto _output_error; /* non-conformant (content-only) dictionary */ + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : retrieve dictID from frame : ", testNb++); + { U32 const did = ZSTD_getDictID_fromFrame(compressedBuffer, cSize); + if (did != dictID) goto _output_error; /* non-conformant (content-only) dictionary */ + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : frame built with dictionary should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + dictBuffer, dictSize), + if (r != CNBuffSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : estimate CDict size : ", testNb++); + { ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + size_t const estimatedSize = ZSTD_estimateCDictSize_advanced(dictSize, cParams, ZSTD_dlm_byRef); + DISPLAYLEVEL(4, "OK : %u \n", (U32)estimatedSize); + } + + DISPLAYLEVEL(4, "test%3i : compress with CDict ", testNb++); + { ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced(dictBuffer, dictSize, + ZSTD_dlm_byRef, ZSTD_dm_auto, + cParams, ZSTD_defaultCMem); + DISPLAYLEVEL(4, "(size : %u) : ", (U32)ZSTD_sizeof_CDict(cdict)); + cSize = ZSTD_compress_usingCDict(cctx, compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, cdict); + ZSTD_freeCDict(cdict); + if (ZSTD_isError(cSize)) goto _output_error; + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : retrieve dictID from frame : ", testNb++); + { U32 const did = ZSTD_getDictID_fromFrame(compressedBuffer, cSize); + if (did != dictID) goto _output_error; /* non-conformant (content-only) dictionary */ + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : frame built with dictionary should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + dictBuffer, dictSize), + if (r != CNBuffSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : compress with static CDict : ", testNb++); + { ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + size_t const cdictSize = ZSTD_estimateCDictSize_advanced(dictSize, cParams, ZSTD_dlm_byCopy); + void* const cdictBuffer = malloc(cdictSize); + if (cdictBuffer==NULL) goto _output_error; + { ZSTD_CDict* const cdict = ZSTD_initStaticCDict(cdictBuffer, cdictSize, + dictBuffer, dictSize, + ZSTD_dlm_byCopy, ZSTD_dm_auto, + cParams); + if (cdict == NULL) { + DISPLAY("ZSTD_initStaticCDict failed "); + goto _output_error; + } + cSize = ZSTD_compress_usingCDict(cctx, + compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, cdict); + if (ZSTD_isError(cSize)) { + DISPLAY("ZSTD_compress_usingCDict failed "); + goto _output_error; + } } + free(cdictBuffer); + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : ZSTD_compress_usingCDict_advanced, no contentSize, no dictID : ", testNb++); + { ZSTD_frameParameters const fParams = { 0 /* frameSize */, 1 /* checksum */, 1 /* noDictID*/ }; + ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dm_auto, cParams, ZSTD_defaultCMem); + cSize = ZSTD_compress_usingCDict_advanced(cctx, compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, cdict, fParams); + ZSTD_freeCDict(cdict); + if (ZSTD_isError(cSize)) goto _output_error; + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : try retrieving contentSize from frame : ", testNb++); + { U64 const contentSize = ZSTD_getFrameContentSize(compressedBuffer, cSize); + if (contentSize != ZSTD_CONTENTSIZE_UNKNOWN) goto _output_error; + } + DISPLAYLEVEL(4, "OK (unknown)\n"); + + DISPLAYLEVEL(4, "test%3i : frame built without dictID should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + dictBuffer, dictSize), + if (r != CNBuffSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : ZSTD_compress_advanced, no dictID : ", testNb++); + { ZSTD_parameters p = ZSTD_getParams(3, CNBuffSize, dictSize); + p.fParams.noDictIDFlag = 1; + cSize = ZSTD_compress_advanced(cctx, compressedBuffer, compressedBufferSize, + CNBuffer, CNBuffSize, + dictBuffer, dictSize, p); + if (ZSTD_isError(cSize)) goto _output_error; + } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/CNBuffSize*100); + + DISPLAYLEVEL(4, "test%3i : frame built without dictID should be decompressible : ", testNb++); + CHECKPLUS(r, ZSTD_decompress_usingDict(dctx, + decodedBuffer, CNBuffSize, + compressedBuffer, cSize, + dictBuffer, dictSize), + if (r != CNBuffSize) goto _output_error); + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : dictionary containing only header should return error : ", testNb++); + { + const size_t ret = ZSTD_decompress_usingDict( + dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize, + "\x37\xa4\x30\xec\x11\x22\x33\x44", 8); + if (ZSTD_getErrorCode(ret) != ZSTD_error_dictionary_corrupted) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Building cdict w/ ZSTD_dm_fullDict on a good dictionary : ", testNb++); + { ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dm_fullDict, cParams, ZSTD_defaultCMem); + if (cdict==NULL) goto _output_error; + ZSTD_freeCDict(cdict); + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Building cdict w/ ZSTD_dm_fullDict on a rawContent (must fail) : ", testNb++); + { ZSTD_compressionParameters const cParams = ZSTD_getCParams(1, CNBuffSize, dictSize); + ZSTD_CDict* const cdict = ZSTD_createCDict_advanced((const char*)dictBuffer+1, dictSize-1, ZSTD_dlm_byRef, ZSTD_dm_fullDict, cParams, ZSTD_defaultCMem); + if (cdict!=NULL) goto _output_error; + ZSTD_freeCDict(cdict); + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Loading rawContent starting with dict header w/ ZSTD_dm_auto should fail : ", testNb++); + { + size_t ret; + MEM_writeLE32((char*)dictBuffer+2, ZSTD_MAGIC_DICTIONARY); + ret = ZSTD_CCtx_loadDictionary_advanced( + cctx, (const char*)dictBuffer+2, dictSize-2, ZSTD_dlm_byRef, ZSTD_dm_auto); + if (!ZSTD_isError(ret)) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Loading rawContent starting with dict header w/ ZSTD_dm_rawContent should pass : ", testNb++); + { + size_t ret; + MEM_writeLE32((char*)dictBuffer+2, ZSTD_MAGIC_DICTIONARY); + ret = ZSTD_CCtx_loadDictionary_advanced( + cctx, (const char*)dictBuffer+2, dictSize-2, ZSTD_dlm_byRef, ZSTD_dm_rawContent); + if (ZSTD_isError(ret)) goto _output_error; + } + DISPLAYLEVEL(4, "OK \n"); + + ZSTD_freeCCtx(cctx); + free(dictBuffer); + free(samplesSizes); + } + + /* COVER dictionary builder tests */ + { ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + size_t dictSize = 16 KB; + size_t optDictSize = dictSize; + void* dictBuffer = malloc(dictSize); + size_t const totalSampleSize = 1 MB; + size_t const sampleUnitSize = 8 KB; + U32 const nbSamples = (U32)(totalSampleSize / sampleUnitSize); + size_t* const samplesSizes = (size_t*) malloc(nbSamples * sizeof(size_t)); + ZDICT_cover_params_t params; + U32 dictID; + + if (dictBuffer==NULL || samplesSizes==NULL) { + free(dictBuffer); + free(samplesSizes); + goto _output_error; + } + + DISPLAYLEVEL(4, "test%3i : ZDICT_trainFromBuffer_cover : ", testNb++); + { U32 u; for (u=0; u<nbSamples; u++) samplesSizes[u] = sampleUnitSize; } + memset(¶ms, 0, sizeof(params)); + params.d = 1 + (FUZ_rand(&seed) % 16); + params.k = params.d + (FUZ_rand(&seed) % 256); + dictSize = ZDICT_trainFromBuffer_cover(dictBuffer, dictSize, + CNBuffer, samplesSizes, nbSamples, + params); + if (ZDICT_isError(dictSize)) goto _output_error; + DISPLAYLEVEL(4, "OK, created dictionary of size %u \n", (U32)dictSize); + + DISPLAYLEVEL(4, "test%3i : check dictID : ", testNb++); + dictID = ZDICT_getDictID(dictBuffer, dictSize); + if (dictID==0) goto _output_error; + DISPLAYLEVEL(4, "OK : %u \n", dictID); + + DISPLAYLEVEL(4, "test%3i : ZDICT_optimizeTrainFromBuffer_cover : ", testNb++); + memset(¶ms, 0, sizeof(params)); + params.steps = 4; + optDictSize = ZDICT_optimizeTrainFromBuffer_cover(dictBuffer, optDictSize, + CNBuffer, samplesSizes, + nbSamples / 4, ¶ms); + if (ZDICT_isError(optDictSize)) goto _output_error; + DISPLAYLEVEL(4, "OK, created dictionary of size %u \n", (U32)optDictSize); + + DISPLAYLEVEL(4, "test%3i : check dictID : ", testNb++); + dictID = ZDICT_getDictID(dictBuffer, optDictSize); + if (dictID==0) goto _output_error; + DISPLAYLEVEL(4, "OK : %u \n", dictID); + + ZSTD_freeCCtx(cctx); + free(dictBuffer); + free(samplesSizes); + } + + /* Decompression defense tests */ + DISPLAYLEVEL(4, "test%3i : Check input length for magic number : ", testNb++); + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, CNBuffer, 3); /* too small input */ + if (!ZSTD_isError(r)) goto _output_error; + if (ZSTD_getErrorCode(r) != ZSTD_error_srcSize_wrong) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Check magic Number : ", testNb++); + ((char*)(CNBuffer))[0] = 1; + { size_t const r = ZSTD_decompress(decodedBuffer, CNBuffSize, CNBuffer, 4); + if (!ZSTD_isError(r)) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + /* content size verification test */ + DISPLAYLEVEL(4, "test%3i : Content size verification : ", testNb++); + { ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + size_t const srcSize = 5000; + size_t const wrongSrcSize = (srcSize + 1000); + ZSTD_parameters params = ZSTD_getParams(1, wrongSrcSize, 0); + params.fParams.contentSizeFlag = 1; + CHECK( ZSTD_compressBegin_advanced(cctx, NULL, 0, params, wrongSrcSize) ); + { size_t const result = ZSTD_compressEnd(cctx, decodedBuffer, CNBuffSize, CNBuffer, srcSize); + if (!ZSTD_isError(result)) goto _output_error; + if (ZSTD_getErrorCode(result) != ZSTD_error_srcSize_wrong) goto _output_error; + DISPLAYLEVEL(4, "OK : %s \n", ZSTD_getErrorName(result)); + } + ZSTD_freeCCtx(cctx); + } + + /* custom formats tests */ + { ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + size_t const inputSize = CNBuffSize / 2; /* won't cause pb with small dict size */ + + /* basic block compression */ + DISPLAYLEVEL(4, "test%3i : magic-less format test : ", testNb++); + CHECK( ZSTD_CCtx_setParameter(cctx, ZSTD_p_format, ZSTD_f_zstd1_magicless) ); + { ZSTD_inBuffer in = { CNBuffer, inputSize, 0 }; + ZSTD_outBuffer out = { compressedBuffer, ZSTD_compressBound(inputSize), 0 }; + size_t const result = ZSTD_compress_generic(cctx, &out, &in, ZSTD_e_end); + if (result != 0) goto _output_error; + if (in.pos != in.size) goto _output_error; + cSize = out.pos; + } + DISPLAYLEVEL(4, "OK (compress : %u -> %u bytes)\n", (U32)inputSize, (U32)cSize); + + DISPLAYLEVEL(4, "test%3i : decompress normally (should fail) : ", testNb++); + { size_t const decodeResult = ZSTD_decompressDCtx(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize); + if (ZSTD_getErrorCode(decodeResult) != ZSTD_error_prefix_unknown) goto _output_error; + DISPLAYLEVEL(4, "OK : %s \n", ZSTD_getErrorName(decodeResult)); + } + + DISPLAYLEVEL(4, "test%3i : decompress with magic-less instruction : ", testNb++); + ZSTD_DCtx_reset(dctx); + CHECK( ZSTD_DCtx_setFormat(dctx, ZSTD_f_zstd1_magicless) ); + { ZSTD_inBuffer in = { compressedBuffer, cSize, 0 }; + ZSTD_outBuffer out = { decodedBuffer, CNBuffSize, 0 }; + size_t const result = ZSTD_decompress_generic(dctx, &out, &in); + if (result != 0) goto _output_error; + if (in.pos != in.size) goto _output_error; + if (out.pos != inputSize) goto _output_error; + DISPLAYLEVEL(4, "OK : regenerated %u bytes \n", (U32)out.pos); + } + + ZSTD_freeCCtx(cctx); + } + + /* block API tests */ + { ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + static const size_t dictSize = 65 KB; + static const size_t blockSize = 100 KB; /* won't cause pb with small dict size */ + size_t cSize2; + + /* basic block compression */ + DISPLAYLEVEL(4, "test%3i : Block compression test : ", testNb++); + CHECK( ZSTD_compressBegin(cctx, 5) ); + CHECK( ZSTD_getBlockSize(cctx) >= blockSize); + cSize = ZSTD_compressBlock(cctx, compressedBuffer, ZSTD_compressBound(blockSize), CNBuffer, blockSize); + if (ZSTD_isError(cSize)) goto _output_error; + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Block decompression test : ", testNb++); + CHECK( ZSTD_decompressBegin(dctx) ); + { CHECK_V(r, ZSTD_decompressBlock(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize) ); + if (r != blockSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + /* dictionary block compression */ + DISPLAYLEVEL(4, "test%3i : Dictionary Block compression test : ", testNb++); + CHECK( ZSTD_compressBegin_usingDict(cctx, CNBuffer, dictSize, 5) ); + cSize = ZSTD_compressBlock(cctx, compressedBuffer, ZSTD_compressBound(blockSize), (char*)CNBuffer+dictSize, blockSize); + if (ZSTD_isError(cSize)) goto _output_error; + cSize2 = ZSTD_compressBlock(cctx, (char*)compressedBuffer+cSize, ZSTD_compressBound(blockSize), (char*)CNBuffer+dictSize+blockSize, blockSize); + if (ZSTD_isError(cSize2)) goto _output_error; + memcpy((char*)compressedBuffer+cSize, (char*)CNBuffer+dictSize+blockSize, blockSize); /* fake non-compressed block */ + cSize2 = ZSTD_compressBlock(cctx, (char*)compressedBuffer+cSize+blockSize, ZSTD_compressBound(blockSize), + (char*)CNBuffer+dictSize+2*blockSize, blockSize); + if (ZSTD_isError(cSize2)) goto _output_error; + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : Dictionary Block decompression test : ", testNb++); + CHECK( ZSTD_decompressBegin_usingDict(dctx, CNBuffer, dictSize) ); + { CHECK_V( r, ZSTD_decompressBlock(dctx, decodedBuffer, CNBuffSize, compressedBuffer, cSize) ); + if (r != blockSize) goto _output_error; } + ZSTD_insertBlock(dctx, (char*)decodedBuffer+blockSize, blockSize); /* insert non-compressed block into dctx history */ + { CHECK_V( r, ZSTD_decompressBlock(dctx, (char*)decodedBuffer+2*blockSize, CNBuffSize, (char*)compressedBuffer+cSize+blockSize, cSize2) ); + if (r != blockSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + ZSTD_freeCCtx(cctx); + } + ZSTD_freeDCtx(dctx); + + /* long rle test */ + { size_t sampleSize = 0; + DISPLAYLEVEL(4, "test%3i : Long RLE test : ", testNb++); + RDG_genBuffer(CNBuffer, sampleSize, compressibility, 0., seed+1); + memset((char*)CNBuffer+sampleSize, 'B', 256 KB - 1); + sampleSize += 256 KB - 1; + RDG_genBuffer((char*)CNBuffer+sampleSize, 96 KB, compressibility, 0., seed+2); + sampleSize += 96 KB; + cSize = ZSTD_compress(compressedBuffer, ZSTD_compressBound(sampleSize), CNBuffer, sampleSize, 1); + if (ZSTD_isError(cSize)) goto _output_error; + { CHECK_V(regenSize, ZSTD_decompress(decodedBuffer, sampleSize, compressedBuffer, cSize)); + if (regenSize!=sampleSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + } + + /* All zeroes test (test bug #137) */ + #define ZEROESLENGTH 100 + DISPLAYLEVEL(4, "test%3i : compress %u zeroes : ", testNb++, ZEROESLENGTH); + memset(CNBuffer, 0, ZEROESLENGTH); + { CHECK_V(r, ZSTD_compress(compressedBuffer, ZSTD_compressBound(ZEROESLENGTH), CNBuffer, ZEROESLENGTH, 1) ); + cSize = r; } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/ZEROESLENGTH*100); + + DISPLAYLEVEL(4, "test%3i : decompress %u zeroes : ", testNb++, ZEROESLENGTH); + { CHECK_V(r, ZSTD_decompress(decodedBuffer, ZEROESLENGTH, compressedBuffer, cSize) ); + if (r != ZEROESLENGTH) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + /* nbSeq limit test */ + #define _3BYTESTESTLENGTH 131000 + #define NB3BYTESSEQLOG 9 + #define NB3BYTESSEQ (1 << NB3BYTESSEQLOG) + #define NB3BYTESSEQMASK (NB3BYTESSEQ-1) + /* creates a buffer full of 3-bytes sequences */ + { BYTE _3BytesSeqs[NB3BYTESSEQ][3]; + U32 rSeed = 1; + + /* create batch of 3-bytes sequences */ + { int i; + for (i=0; i < NB3BYTESSEQ; i++) { + _3BytesSeqs[i][0] = (BYTE)(FUZ_rand(&rSeed) & 255); + _3BytesSeqs[i][1] = (BYTE)(FUZ_rand(&rSeed) & 255); + _3BytesSeqs[i][2] = (BYTE)(FUZ_rand(&rSeed) & 255); + } } + + /* randomly fills CNBuffer with prepared 3-bytes sequences */ + { int i; + for (i=0; i < _3BYTESTESTLENGTH; i += 3) { /* note : CNBuffer size > _3BYTESTESTLENGTH+3 */ + U32 const id = FUZ_rand(&rSeed) & NB3BYTESSEQMASK; + ((BYTE*)CNBuffer)[i+0] = _3BytesSeqs[id][0]; + ((BYTE*)CNBuffer)[i+1] = _3BytesSeqs[id][1]; + ((BYTE*)CNBuffer)[i+2] = _3BytesSeqs[id][2]; + } } } + DISPLAYLEVEL(4, "test%3i : compress lots 3-bytes sequences : ", testNb++); + { CHECK_V(r, ZSTD_compress(compressedBuffer, ZSTD_compressBound(_3BYTESTESTLENGTH), + CNBuffer, _3BYTESTESTLENGTH, 19) ); + cSize = r; } + DISPLAYLEVEL(4, "OK (%u bytes : %.2f%%)\n", (U32)cSize, (double)cSize/_3BYTESTESTLENGTH*100); + + DISPLAYLEVEL(4, "test%3i : decompress lots 3-bytes sequence : ", testNb++); + { CHECK_V(r, ZSTD_decompress(decodedBuffer, _3BYTESTESTLENGTH, compressedBuffer, cSize) ); + if (r != _3BYTESTESTLENGTH) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + DISPLAYLEVEL(4, "test%3i : incompressible data and ill suited dictionary : ", testNb++); + RDG_genBuffer(CNBuffer, CNBuffSize, 0.0, 0.1, seed); + { /* Train a dictionary on low characters */ + size_t dictSize = 16 KB; + void* const dictBuffer = malloc(dictSize); + size_t const totalSampleSize = 1 MB; + size_t const sampleUnitSize = 8 KB; + U32 const nbSamples = (U32)(totalSampleSize / sampleUnitSize); + size_t* const samplesSizes = (size_t*) malloc(nbSamples * sizeof(size_t)); + if (!dictBuffer || !samplesSizes) goto _output_error; + { U32 u; for (u=0; u<nbSamples; u++) samplesSizes[u] = sampleUnitSize; } + dictSize = ZDICT_trainFromBuffer(dictBuffer, dictSize, CNBuffer, samplesSizes, nbSamples); + if (ZDICT_isError(dictSize)) goto _output_error; + /* Reverse the characters to make the dictionary ill suited */ + { U32 u; + for (u = 0; u < CNBuffSize; ++u) { + ((BYTE*)CNBuffer)[u] = 255 - ((BYTE*)CNBuffer)[u]; + } + } + { /* Compress the data */ + size_t const inputSize = 500; + size_t const outputSize = ZSTD_compressBound(inputSize); + void* const outputBuffer = malloc(outputSize); + ZSTD_CCtx* const cctx = ZSTD_createCCtx(); + if (!outputBuffer || !cctx) goto _output_error; + CHECK(ZSTD_compress_usingDict(cctx, outputBuffer, outputSize, CNBuffer, inputSize, dictBuffer, dictSize, 1)); + free(outputBuffer); + ZSTD_freeCCtx(cctx); + } + + free(dictBuffer); + free(samplesSizes); + } + DISPLAYLEVEL(4, "OK \n"); + + + /* findFrameCompressedSize on skippable frames */ + DISPLAYLEVEL(4, "test%3i : frame compressed size of skippable frame : ", testNb++); + { const char* frame = "\x50\x2a\x4d\x18\x05\x0\x0\0abcde"; + size_t const frameSrcSize = 13; + if (ZSTD_findFrameCompressedSize(frame, frameSrcSize) != frameSrcSize) goto _output_error; } + DISPLAYLEVEL(4, "OK \n"); + + /* error string tests */ + DISPLAYLEVEL(4, "test%3i : testing ZSTD error code strings : ", testNb++); + if (strcmp("No error detected", ZSTD_getErrorName((ZSTD_ErrorCode)(0-ZSTD_error_no_error))) != 0) goto _output_error; + if (strcmp("No error detected", ZSTD_getErrorString(ZSTD_error_no_error)) != 0) goto _output_error; + if (strcmp("Unspecified error code", ZSTD_getErrorString((ZSTD_ErrorCode)(0-ZSTD_error_GENERIC))) != 0) goto _output_error; + if (strcmp("Error (generic)", ZSTD_getErrorName((size_t)0-ZSTD_error_GENERIC)) != 0) goto _output_error; + if (strcmp("Error (generic)", ZSTD_getErrorString(ZSTD_error_GENERIC)) != 0) goto _output_error; + if (strcmp("No error detected", ZSTD_getErrorName(ZSTD_error_GENERIC)) != 0) goto _output_error; + DISPLAYLEVEL(4, "OK \n"); + +_end: + free(CNBuffer); + free(compressedBuffer); + free(decodedBuffer); + return testResult; + +_output_error: + testResult = 1; + DISPLAY("Error detected in Unit tests ! \n"); + goto _end; +} + + +static size_t findDiff(const void* buf1, const void* buf2, size_t max) +{ + const BYTE* b1 = (const BYTE*)buf1; + const BYTE* b2 = (const BYTE*)buf2; + size_t u; + for (u=0; u<max; u++) { + if (b1[u] != b2[u]) break; + } + return u; +} + + +static ZSTD_parameters FUZ_makeParams(ZSTD_compressionParameters cParams, ZSTD_frameParameters fParams) +{ + ZSTD_parameters params; + params.cParams = cParams; + params.fParams = fParams; + return params; +} + +static size_t FUZ_rLogLength(U32* seed, U32 logLength) +{ + size_t const lengthMask = ((size_t)1 << logLength) - 1; + return (lengthMask+1) + (FUZ_rand(seed) & lengthMask); +} + +static size_t FUZ_randomLength(U32* seed, U32 maxLog) +{ + U32 const logLength = FUZ_rand(seed) % maxLog; + return FUZ_rLogLength(seed, logLength); +} + +#undef CHECK +#define CHECK(cond, ...) { \ + if (cond) { \ + DISPLAY("Error => "); \ + DISPLAY(__VA_ARGS__); \ + DISPLAY(" (seed %u, test nb %u) \n", seed, testNb); \ + goto _output_error; \ +} } + +#undef CHECK_Z +#define CHECK_Z(f) { \ + size_t const err = f; \ + if (ZSTD_isError(err)) { \ + DISPLAY("Error => %s : %s ", \ + #f, ZSTD_getErrorName(err)); \ + DISPLAY(" (seed %u, test nb %u) \n", seed, testNb); \ + goto _output_error; \ +} } + + +static int fuzzerTests(U32 seed, U32 nbTests, unsigned startTest, U32 const maxDurationS, double compressibility, int bigTests) +{ + static const U32 maxSrcLog = 23; + static const U32 maxSampleLog = 22; + size_t const srcBufferSize = (size_t)1<<maxSrcLog; + size_t const dstBufferSize = (size_t)1<<maxSampleLog; + size_t const cBufferSize = ZSTD_compressBound(dstBufferSize); + BYTE* cNoiseBuffer[5]; + BYTE* srcBuffer; /* jumping pointer */ + BYTE* const cBuffer = (BYTE*) malloc (cBufferSize); + BYTE* const dstBuffer = (BYTE*) malloc (dstBufferSize); + BYTE* const mirrorBuffer = (BYTE*) malloc (dstBufferSize); + ZSTD_CCtx* const refCtx = ZSTD_createCCtx(); + ZSTD_CCtx* const ctx = ZSTD_createCCtx(); + ZSTD_DCtx* const dctx = ZSTD_createDCtx(); + U32 result = 0; + U32 testNb = 0; + U32 coreSeed = seed, lseed = 0; + clock_t const startClock = clock(); + clock_t const maxClockSpan = maxDurationS * CLOCKS_PER_SEC; + int const cLevelLimiter = bigTests ? 3 : 2; + + /* allocation */ + cNoiseBuffer[0] = (BYTE*)malloc (srcBufferSize); + cNoiseBuffer[1] = (BYTE*)malloc (srcBufferSize); + cNoiseBuffer[2] = (BYTE*)malloc (srcBufferSize); + cNoiseBuffer[3] = (BYTE*)malloc (srcBufferSize); + cNoiseBuffer[4] = (BYTE*)malloc (srcBufferSize); + CHECK (!cNoiseBuffer[0] || !cNoiseBuffer[1] || !cNoiseBuffer[2] || !cNoiseBuffer[3] || !cNoiseBuffer[4] + || !dstBuffer || !mirrorBuffer || !cBuffer || !refCtx || !ctx || !dctx, + "Not enough memory, fuzzer tests cancelled"); + + /* Create initial samples */ + RDG_genBuffer(cNoiseBuffer[0], srcBufferSize, 0.00, 0., coreSeed); /* pure noise */ + RDG_genBuffer(cNoiseBuffer[1], srcBufferSize, 0.05, 0., coreSeed); /* barely compressible */ + RDG_genBuffer(cNoiseBuffer[2], srcBufferSize, compressibility, 0., coreSeed); + RDG_genBuffer(cNoiseBuffer[3], srcBufferSize, 0.95, 0., coreSeed); /* highly compressible */ + RDG_genBuffer(cNoiseBuffer[4], srcBufferSize, 1.00, 0., coreSeed); /* sparse content */ + srcBuffer = cNoiseBuffer[2]; + + /* catch up testNb */ + for (testNb=1; testNb < startTest; testNb++) FUZ_rand(&coreSeed); + + /* main test loop */ + for ( ; (testNb <= nbTests) || (FUZ_clockSpan(startClock) < maxClockSpan); testNb++ ) { + size_t sampleSize, maxTestSize, totalTestSize; + size_t cSize, totalCSize, totalGenSize; + U64 crcOrig; + BYTE* sampleBuffer; + const BYTE* dict; + size_t dictSize; + + /* notification */ + if (nbTests >= testNb) { DISPLAYUPDATE(2, "\r%6u/%6u ", testNb, nbTests); } + else { DISPLAYUPDATE(2, "\r%6u ", testNb); } + + FUZ_rand(&coreSeed); + { U32 const prime1 = 2654435761U; lseed = coreSeed ^ prime1; } + + /* srcBuffer selection [0-4] */ + { U32 buffNb = FUZ_rand(&lseed) & 0x7F; + if (buffNb & 7) buffNb=2; /* most common : compressible (P) */ + else { + buffNb >>= 3; + if (buffNb & 7) { + const U32 tnb[2] = { 1, 3 }; /* barely/highly compressible */ + buffNb = tnb[buffNb >> 3]; + } else { + const U32 tnb[2] = { 0, 4 }; /* not compressible / sparse */ + buffNb = tnb[buffNb >> 3]; + } } + srcBuffer = cNoiseBuffer[buffNb]; + } + + /* select src segment */ + sampleSize = FUZ_randomLength(&lseed, maxSampleLog); + + /* create sample buffer (to catch read error with valgrind & sanitizers) */ + sampleBuffer = (BYTE*)malloc(sampleSize); + CHECK(sampleBuffer==NULL, "not enough memory for sample buffer"); + { size_t const sampleStart = FUZ_rand(&lseed) % (srcBufferSize - sampleSize); + memcpy(sampleBuffer, srcBuffer + sampleStart, sampleSize); } + crcOrig = XXH64(sampleBuffer, sampleSize, 0); + + /* compression tests */ + { unsigned const cLevel = + ( FUZ_rand(&lseed) % + (ZSTD_maxCLevel() - (FUZ_highbit32((U32)sampleSize) / cLevelLimiter)) ) + + 1; + cSize = ZSTD_compressCCtx(ctx, cBuffer, cBufferSize, sampleBuffer, sampleSize, cLevel); + CHECK(ZSTD_isError(cSize), "ZSTD_compressCCtx failed : %s", ZSTD_getErrorName(cSize)); + + /* compression failure test : too small dest buffer */ + if (cSize > 3) { + const size_t missing = (FUZ_rand(&lseed) % (cSize-2)) + 1; /* no problem, as cSize > 4 (frameHeaderSizer) */ + const size_t tooSmallSize = cSize - missing; + const U32 endMark = 0x4DC2B1A9; + memcpy(dstBuffer+tooSmallSize, &endMark, 4); + { size_t const errorCode = ZSTD_compressCCtx(ctx, dstBuffer, tooSmallSize, sampleBuffer, sampleSize, cLevel); + CHECK(!ZSTD_isError(errorCode), "ZSTD_compressCCtx should have failed ! (buffer too small : %u < %u)", (U32)tooSmallSize, (U32)cSize); } + { U32 endCheck; memcpy(&endCheck, dstBuffer+tooSmallSize, 4); + CHECK(endCheck != endMark, "ZSTD_compressCCtx : dst buffer overflow"); } + } } + + /* frame header decompression test */ + { ZSTD_frameHeader zfh; + CHECK_Z( ZSTD_getFrameHeader(&zfh, cBuffer, cSize) ); + CHECK(zfh.frameContentSize != sampleSize, "Frame content size incorrect"); + } + + /* Decompressed size test */ + { unsigned long long const rSize = ZSTD_findDecompressedSize(cBuffer, cSize); + CHECK(rSize != sampleSize, "decompressed size incorrect"); + } + + /* successful decompression test */ + { size_t const margin = (FUZ_rand(&lseed) & 1) ? 0 : (FUZ_rand(&lseed) & 31) + 1; + size_t const dSize = ZSTD_decompress(dstBuffer, sampleSize + margin, cBuffer, cSize); + CHECK(dSize != sampleSize, "ZSTD_decompress failed (%s) (srcSize : %u ; cSize : %u)", ZSTD_getErrorName(dSize), (U32)sampleSize, (U32)cSize); + { U64 const crcDest = XXH64(dstBuffer, sampleSize, 0); + CHECK(crcOrig != crcDest, "decompression result corrupted (pos %u / %u)", (U32)findDiff(sampleBuffer, dstBuffer, sampleSize), (U32)sampleSize); + } } + + free(sampleBuffer); /* no longer useful after this point */ + + /* truncated src decompression test */ + { size_t const missing = (FUZ_rand(&lseed) % (cSize-2)) + 1; /* no problem, as cSize > 4 (frameHeaderSizer) */ + size_t const tooSmallSize = cSize - missing; + void* cBufferTooSmall = malloc(tooSmallSize); /* valgrind will catch read overflows */ + CHECK(cBufferTooSmall == NULL, "not enough memory !"); + memcpy(cBufferTooSmall, cBuffer, tooSmallSize); + { size_t const errorCode = ZSTD_decompress(dstBuffer, dstBufferSize, cBufferTooSmall, tooSmallSize); + CHECK(!ZSTD_isError(errorCode), "ZSTD_decompress should have failed ! (truncated src buffer)"); } + free(cBufferTooSmall); + } + + /* too small dst decompression test */ + if (sampleSize > 3) { + size_t const missing = (FUZ_rand(&lseed) % (sampleSize-2)) + 1; /* no problem, as cSize > 4 (frameHeaderSizer) */ + size_t const tooSmallSize = sampleSize - missing; + static const BYTE token = 0xA9; + dstBuffer[tooSmallSize] = token; + { size_t const errorCode = ZSTD_decompress(dstBuffer, tooSmallSize, cBuffer, cSize); + CHECK(!ZSTD_isError(errorCode), "ZSTD_decompress should have failed : %u > %u (dst buffer too small)", (U32)errorCode, (U32)tooSmallSize); } + CHECK(dstBuffer[tooSmallSize] != token, "ZSTD_decompress : dst buffer overflow"); + } + + /* noisy src decompression test */ + if (cSize > 6) { + /* insert noise into src */ + { U32 const maxNbBits = FUZ_highbit32((U32)(cSize-4)); + size_t pos = 4; /* preserve magic number (too easy to detect) */ + for (;;) { + /* keep some original src */ + { U32 const nbBits = FUZ_rand(&lseed) % maxNbBits; + size_t const mask = (1<<nbBits) - 1; + size_t const skipLength = FUZ_rand(&lseed) & mask; + pos += skipLength; + } + if (pos <= cSize) break; + /* add noise */ + { U32 const nbBitsCodes = FUZ_rand(&lseed) % maxNbBits; + U32 const nbBits = nbBitsCodes ? nbBitsCodes-1 : 0; + size_t const mask = (1<<nbBits) - 1; + size_t const rNoiseLength = (FUZ_rand(&lseed) & mask) + 1; + size_t const noiseLength = MIN(rNoiseLength, cSize-pos); + size_t const noiseStart = FUZ_rand(&lseed) % (srcBufferSize - noiseLength); + memcpy(cBuffer + pos, srcBuffer + noiseStart, noiseLength); + pos += noiseLength; + } } } + + /* decompress noisy source */ + { U32 const endMark = 0xA9B1C3D6; + memcpy(dstBuffer+sampleSize, &endMark, 4); + { size_t const decompressResult = ZSTD_decompress(dstBuffer, sampleSize, cBuffer, cSize); + /* result *may* be an unlikely success, but even then, it must strictly respect dst buffer boundaries */ + CHECK((!ZSTD_isError(decompressResult)) && (decompressResult>sampleSize), + "ZSTD_decompress on noisy src : result is too large : %u > %u (dst buffer)", (U32)decompressResult, (U32)sampleSize); + } + { U32 endCheck; memcpy(&endCheck, dstBuffer+sampleSize, 4); + CHECK(endMark!=endCheck, "ZSTD_decompress on noisy src : dst buffer overflow"); + } } } /* noisy src decompression test */ + + /*===== Streaming compression test, scattered segments and dictionary =====*/ + + { U32 const testLog = FUZ_rand(&lseed) % maxSrcLog; + U32 const dictLog = FUZ_rand(&lseed) % maxSrcLog; + int const cLevel = (FUZ_rand(&lseed) % + (ZSTD_maxCLevel() - + (MAX(testLog, dictLog) / cLevelLimiter))) + + 1; + maxTestSize = FUZ_rLogLength(&lseed, testLog); + if (maxTestSize >= dstBufferSize) maxTestSize = dstBufferSize-1; + + dictSize = FUZ_rLogLength(&lseed, dictLog); /* needed also for decompression */ + dict = srcBuffer + (FUZ_rand(&lseed) % (srcBufferSize - dictSize)); + + if (FUZ_rand(&lseed) & 0xF) { + CHECK_Z ( ZSTD_compressBegin_usingDict(refCtx, dict, dictSize, cLevel) ); + } else { + ZSTD_compressionParameters const cPar = ZSTD_getCParams(cLevel, 0, dictSize); + ZSTD_frameParameters const fPar = { FUZ_rand(&lseed)&1 /* contentSizeFlag */, + !(FUZ_rand(&lseed)&3) /* contentChecksumFlag*/, + 0 /*NodictID*/ }; /* note : since dictionary is fake, dictIDflag has no impact */ + ZSTD_parameters const p = FUZ_makeParams(cPar, fPar); + CHECK_Z ( ZSTD_compressBegin_advanced(refCtx, dict, dictSize, p, 0) ); + } + CHECK_Z( ZSTD_copyCCtx(ctx, refCtx, 0) ); + } + + { U32 const nbChunks = (FUZ_rand(&lseed) & 127) + 2; + U32 n; + XXH64_state_t xxhState; + XXH64_reset(&xxhState, 0); + for (totalTestSize=0, cSize=0, n=0 ; n<nbChunks ; n++) { + size_t const segmentSize = FUZ_randomLength(&lseed, maxSampleLog); + size_t const segmentStart = FUZ_rand(&lseed) % (srcBufferSize - segmentSize); + + if (cBufferSize-cSize < ZSTD_compressBound(segmentSize)) break; /* avoid invalid dstBufferTooSmall */ + if (totalTestSize+segmentSize > maxTestSize) break; + + { size_t const compressResult = ZSTD_compressContinue(ctx, cBuffer+cSize, cBufferSize-cSize, srcBuffer+segmentStart, segmentSize); + CHECK (ZSTD_isError(compressResult), "multi-segments compression error : %s", ZSTD_getErrorName(compressResult)); + cSize += compressResult; + } + XXH64_update(&xxhState, srcBuffer+segmentStart, segmentSize); + memcpy(mirrorBuffer + totalTestSize, srcBuffer+segmentStart, segmentSize); + totalTestSize += segmentSize; + } + + { size_t const flushResult = ZSTD_compressEnd(ctx, cBuffer+cSize, cBufferSize-cSize, NULL, 0); + CHECK (ZSTD_isError(flushResult), "multi-segments epilogue error : %s", ZSTD_getErrorName(flushResult)); + cSize += flushResult; + } + crcOrig = XXH64_digest(&xxhState); + } + + /* streaming decompression test */ + /* ensure memory requirement is good enough (should always be true) */ + { ZSTD_frameHeader zfh; + CHECK( ZSTD_getFrameHeader(&zfh, cBuffer, ZSTD_frameHeaderSize_max), + "ZSTD_getFrameHeader(): error retrieving frame information"); + { size_t const roundBuffSize = ZSTD_decodingBufferSize_min(zfh.windowSize, zfh.frameContentSize); + CHECK_Z(roundBuffSize); + CHECK((roundBuffSize > totalTestSize) && (zfh.frameContentSize!=ZSTD_CONTENTSIZE_UNKNOWN), + "ZSTD_decodingBufferSize_min() requires more memory (%u) than necessary (%u)", + (U32)roundBuffSize, (U32)totalTestSize ); + } } + if (dictSize<8) dictSize=0, dict=NULL; /* disable dictionary */ + CHECK_Z( ZSTD_decompressBegin_usingDict(dctx, dict, dictSize) ); + totalCSize = 0; + totalGenSize = 0; + while (totalCSize < cSize) { + size_t const inSize = ZSTD_nextSrcSizeToDecompress(dctx); + size_t const genSize = ZSTD_decompressContinue(dctx, dstBuffer+totalGenSize, dstBufferSize-totalGenSize, cBuffer+totalCSize, inSize); + CHECK (ZSTD_isError(genSize), "ZSTD_decompressContinue error : %s", ZSTD_getErrorName(genSize)); + totalGenSize += genSize; + totalCSize += inSize; + } + CHECK (ZSTD_nextSrcSizeToDecompress(dctx) != 0, "frame not fully decoded"); + CHECK (totalGenSize != totalTestSize, "streaming decompressed data : wrong size") + CHECK (totalCSize != cSize, "compressed data should be fully read") + { U64 const crcDest = XXH64(dstBuffer, totalTestSize, 0); + if (crcDest!=crcOrig) { + size_t const errorPos = findDiff(mirrorBuffer, dstBuffer, totalTestSize); + CHECK (1, "streaming decompressed data corrupted : byte %u / %u (%02X!=%02X)", + (U32)errorPos, (U32)totalTestSize, dstBuffer[errorPos], mirrorBuffer[errorPos]); + } } + } /* for ( ; (testNb <= nbTests) */ + DISPLAY("\r%u fuzzer tests completed \n", testNb-1); + +_cleanup: + ZSTD_freeCCtx(refCtx); + ZSTD_freeCCtx(ctx); + ZSTD_freeDCtx(dctx); + free(cNoiseBuffer[0]); + free(cNoiseBuffer[1]); + free(cNoiseBuffer[2]); + free(cNoiseBuffer[3]); + free(cNoiseBuffer[4]); + free(cBuffer); + free(dstBuffer); + free(mirrorBuffer); + return result; + +_output_error: + result = 1; + goto _cleanup; +} + + +/*_******************************************************* +* Command line +*********************************************************/ +static int FUZ_usage(const char* programName) +{ + DISPLAY( "Usage :\n"); + DISPLAY( " %s [args]\n", programName); + DISPLAY( "\n"); + DISPLAY( "Arguments :\n"); + DISPLAY( " -i# : Nb of tests (default:%u) \n", nbTestsDefault); + DISPLAY( " -s# : Select seed (default:prompt user)\n"); + DISPLAY( " -t# : Select starting test number (default:0)\n"); + DISPLAY( " -P# : Select compressibility in %% (default:%u%%)\n", FUZ_compressibility_default); + DISPLAY( " -v : verbose\n"); + DISPLAY( " -p : pause at the end\n"); + DISPLAY( " -h : display help and exit\n"); + return 0; +} + +/*! readU32FromChar() : + @return : unsigned integer value read from input in `char` format + allows and interprets K, KB, KiB, M, MB and MiB suffix. + Will also modify `*stringPtr`, advancing it to position where it stopped reading. + Note : function result can overflow if digit string > MAX_UINT */ +static unsigned readU32FromChar(const char** stringPtr) +{ + unsigned result = 0; + while ((**stringPtr >='0') && (**stringPtr <='9')) + result *= 10, result += **stringPtr - '0', (*stringPtr)++ ; + if ((**stringPtr=='K') || (**stringPtr=='M')) { + result <<= 10; + if (**stringPtr=='M') result <<= 10; + (*stringPtr)++ ; + if (**stringPtr=='i') (*stringPtr)++; + if (**stringPtr=='B') (*stringPtr)++; + } + return result; +} + +/** longCommandWArg() : + * check if *stringPtr is the same as longCommand. + * If yes, @return 1 and advances *stringPtr to the position which immediately follows longCommand. + * @return 0 and doesn't modify *stringPtr otherwise. + */ +static unsigned longCommandWArg(const char** stringPtr, const char* longCommand) +{ + size_t const comSize = strlen(longCommand); + int const result = !strncmp(*stringPtr, longCommand, comSize); + if (result) *stringPtr += comSize; + return result; +} + +int main(int argc, const char** argv) +{ + U32 seed = 0; + int seedset = 0; + int argNb; + int nbTests = nbTestsDefault; + int testNb = 0; + U32 proba = FUZ_compressibility_default; + int result = 0; + U32 mainPause = 0; + U32 maxDuration = 0; + int bigTests = 1; + U32 memTestsOnly = 0; + const char* const programName = argv[0]; + + /* Check command line */ + for (argNb=1; argNb<argc; argNb++) { + const char* argument = argv[argNb]; + if(!argument) continue; /* Protection if argument empty */ + + /* Handle commands. Aggregated commands are allowed */ + if (argument[0]=='-') { + + if (longCommandWArg(&argument, "--memtest=")) { memTestsOnly = readU32FromChar(&argument); continue; } + + if (!strcmp(argument, "--memtest")) { memTestsOnly=1; continue; } + if (!strcmp(argument, "--no-big-tests")) { bigTests=0; continue; } + + argument++; + while (*argument!=0) { + switch(*argument) + { + case 'h': + return FUZ_usage(programName); + + case 'v': + argument++; + g_displayLevel = 4; + break; + + case 'q': + argument++; + g_displayLevel--; + break; + + case 'p': /* pause at the end */ + argument++; + mainPause = 1; + break; + + case 'i': + argument++; maxDuration = 0; + nbTests = readU32FromChar(&argument); + break; + + case 'T': + argument++; + nbTests = 0; + maxDuration = readU32FromChar(&argument); + if (*argument=='s') argument++; /* seconds */ + if (*argument=='m') maxDuration *= 60, argument++; /* minutes */ + if (*argument=='n') argument++; + break; + + case 's': + argument++; + seedset = 1; + seed = readU32FromChar(&argument); + break; + + case 't': + argument++; + testNb = readU32FromChar(&argument); + break; + + case 'P': /* compressibility % */ + argument++; + proba = readU32FromChar(&argument); + if (proba>100) proba = 100; + break; + + default: + return (FUZ_usage(programName), 1); + } } } } /* for (argNb=1; argNb<argc; argNb++) */ + + /* Get Seed */ + DISPLAY("Starting zstd tester (%i-bits, %s)\n", (int)(sizeof(size_t)*8), ZSTD_VERSION_STRING); + + if (!seedset) { + time_t const t = time(NULL); + U32 const h = XXH32(&t, sizeof(t), 1); + seed = h % 10000; + } + + DISPLAY("Seed = %u\n", seed); + if (proba!=FUZ_compressibility_default) DISPLAY("Compressibility : %u%%\n", proba); + + if (memTestsOnly) { + g_displayLevel = MAX(3, g_displayLevel); + return FUZ_mallocTests(seed, ((double)proba) / 100, memTestsOnly); + } + + if (nbTests < testNb) nbTests = testNb; + + if (testNb==0) + result = basicUnitTests(0, ((double)proba) / 100); /* constant seed for predictability */ + if (!result) + result = fuzzerTests(seed, nbTests, testNb, maxDuration, ((double)proba) / 100, bigTests); + if (mainPause) { + int unused; + DISPLAY("Press Enter \n"); + unused = getchar(); + (void)unused; + } + return result; +} |