Merge pull request #23131 from poettering/shared-compress

move compress.[ch] → src/basic/
This commit is contained in:
Zbigniew Jędrzejewski-Szmek
2022-04-27 10:47:54 +02:00
committed by GitHub
16 changed files with 196 additions and 133 deletions
+10
View File
@@ -172,6 +172,16 @@ tests += [
[files('test-recurse-dir.c')],
[files('test-compress.c'),
[libshared,
libbasic_compress]],
[files('test-compress-benchmark.c'),
[libshared,
libbasic_compress],
[],
[], '', 'timeout=90'],
[files('test-data-fd-util.c')],
[files('test-static-destruct.c')],
+177
View File
@@ -0,0 +1,177 @@
/* SPDX-License-Identifier: LGPL-2.1-or-later */
#include "alloc-util.h"
#include "compress.h"
#include "env-util.h"
#include "macro.h"
#include "memory-util.h"
#include "nulstr-util.h"
#include "parse-util.h"
#include "process-util.h"
#include "random-util.h"
#include "string-util.h"
#include "tests.h"
typedef int (compress_t)(const void *src, uint64_t src_size, void *dst,
size_t dst_alloc_size, size_t *dst_size);
typedef int (decompress_t)(const void *src, uint64_t src_size,
void **dst, size_t* dst_size, size_t dst_max);
#if HAVE_COMPRESSION
static usec_t arg_duration;
static size_t arg_start;
#define MAX_SIZE (1024*1024LU)
#define PRIME 1048571 /* A prime close enough to one megabyte that mod 4 == 3 */
static size_t _permute(size_t x) {
size_t residue;
if (x >= PRIME)
return x;
residue = x*x % PRIME;
if (x <= PRIME / 2)
return residue;
else
return PRIME - residue;
}
static size_t permute(size_t x) {
return _permute((_permute(x) + arg_start) % MAX_SIZE ^ 0xFF345);
}
static char* make_buf(size_t count, const char *type) {
char *buf;
size_t i;
buf = malloc(count);
assert_se(buf);
if (streq(type, "zeros"))
memzero(buf, count);
else if (streq(type, "simple"))
for (i = 0; i < count; i++)
buf[i] = 'a' + i % ('z' - 'a' + 1);
else if (streq(type, "random")) {
size_t step = count / 10;
random_bytes(buf, step);
memzero(buf + 1*step, step);
random_bytes(buf + 2*step, step);
memzero(buf + 3*step, step);
random_bytes(buf + 4*step, step);
memzero(buf + 5*step, step);
random_bytes(buf + 6*step, step);
memzero(buf + 7*step, step);
random_bytes(buf + 8*step, step);
memzero(buf + 9*step, step);
} else
assert_not_reached();
return buf;
}
static void test_compress_decompress(const char* label, const char* type,
compress_t compress, decompress_t decompress) {
usec_t n, n2 = 0;
float dt;
_cleanup_free_ char *text, *buf;
_cleanup_free_ void *buf2 = NULL;
size_t skipped = 0, compressed = 0, total = 0;
text = make_buf(MAX_SIZE, type);
buf = calloc(MAX_SIZE + 1, 1);
assert_se(text && buf);
n = now(CLOCK_MONOTONIC);
for (size_t i = 0; i <= MAX_SIZE; i++) {
size_t j = 0, k = 0, size;
int r;
size = permute(i);
if (size == 0)
continue;
log_debug("%s %zu %zu", type, i, size);
memzero(buf, MIN(size + 1000, MAX_SIZE));
r = compress(text, size, buf, size, &j);
/* assume compression must be successful except for small or random inputs */
assert_se(r > 0 || (size < 2048 && r == -ENOBUFS) || streq(type, "random"));
/* check for overwrites */
assert_se(buf[size] == 0);
if (r < 0) {
skipped += size;
continue;
}
assert_se(j > 0);
if (j >= size)
log_error("%s \"compressed\" %zu -> %zu", label, size, j);
r = decompress(buf, j, &buf2, &k, 0);
assert_se(r == 0);
assert_se(k == size);
assert_se(memcmp(text, buf2, size) == 0);
total += size;
compressed += j;
n2 = now(CLOCK_MONOTONIC);
if (n2 - n > arg_duration)
break;
}
dt = (n2-n) / 1e6;
log_info("%s/%s: compressed & decompressed %zu bytes in %.2fs (%.2fMiB/s), "
"mean compression %.2f%%, skipped %zu bytes",
label, type, total, dt,
total / 1024. / 1024 / dt,
100 - compressed * 100. / total,
skipped);
}
#endif
int main(int argc, char *argv[]) {
#if HAVE_COMPRESSION
test_setup_logging(LOG_INFO);
if (argc >= 2) {
unsigned x;
assert_se(safe_atou(argv[1], &x) >= 0);
arg_duration = x * USEC_PER_SEC;
} else
arg_duration = slow_tests_enabled() ?
2 * USEC_PER_SEC : USEC_PER_SEC / 50;
if (argc == 3)
(void) safe_atozu(argv[2], &arg_start);
else
arg_start = getpid_cached();
const char *i;
NULSTR_FOREACH(i, "zeros\0simple\0random\0") {
#if HAVE_XZ
test_compress_decompress("XZ", i, compress_blob_xz, decompress_blob_xz);
#endif
#if HAVE_LZ4
test_compress_decompress("LZ4", i, compress_blob_lz4, decompress_blob_lz4);
#endif
#if HAVE_ZSTD
test_compress_decompress("ZSTD", i, compress_blob_zstd, decompress_blob_zstd);
#endif
}
return 0;
#else
return log_tests_skipped("No compression feature is enabled");
#endif
}
+382
View File
@@ -0,0 +1,382 @@
/* SPDX-License-Identifier: LGPL-2.1-or-later */
#include <sys/stat.h>
#if HAVE_LZ4
#include <lz4.h>
#endif
#include "alloc-util.h"
#include "compress.h"
#include "fd-util.h"
#include "fs-util.h"
#include "macro.h"
#include "memory-util.h"
#include "path-util.h"
#include "random-util.h"
#include "tests.h"
#include "tmpfile-util.h"
#if HAVE_XZ
# define XZ_OK 0
#else
# define XZ_OK -EPROTONOSUPPORT
#endif
#if HAVE_LZ4
# define LZ4_OK 0
#else
# define LZ4_OK -EPROTONOSUPPORT
#endif
#define HUGE_SIZE (4096*1024)
typedef int (compress_blob_t)(const void *src, uint64_t src_size,
void *dst, size_t dst_alloc_size, size_t *dst_size);
typedef int (decompress_blob_t)(const void *src, uint64_t src_size,
void **dst,
size_t* dst_size, size_t dst_max);
typedef int (decompress_sw_t)(const void *src, uint64_t src_size,
void **buffer,
const void *prefix, size_t prefix_len,
uint8_t extra);
typedef int (compress_stream_t)(int fdf, int fdt, uint64_t max_bytes, uint64_t *uncompressed_size);
typedef int (decompress_stream_t)(int fdf, int fdt, uint64_t max_size);
#if HAVE_COMPRESSION
_unused_ static void test_compress_decompress(
int flag,
const char *compression,
compress_blob_t compress,
decompress_blob_t decompress,
const char *data,
size_t data_len,
bool may_fail) {
char compressed[512];
size_t csize;
_cleanup_free_ char *decompressed = NULL;
int r;
log_info("/* testing %s %s blob compression/decompression */",
compression, data);
r = compress(data, data_len, compressed, sizeof(compressed), &csize);
if (r == -ENOBUFS) {
log_info_errno(r, "compression failed: %m");
assert_se(may_fail);
} else {
assert_se(r == flag);
r = decompress(compressed, csize,
(void **) &decompressed, &csize, 0);
assert_se(r == 0);
assert_se(decompressed);
assert_se(memcmp(decompressed, data, data_len) == 0);
}
r = decompress("garbage", 7,
(void **) &decompressed, &csize, 0);
assert_se(r < 0);
/* make sure to have the minimal lz4 compressed size */
r = decompress("00000000\1g", 9,
(void **) &decompressed, &csize, 0);
assert_se(r < 0);
r = decompress("\100000000g", 9,
(void **) &decompressed, &csize, 0);
assert_se(r < 0);
explicit_bzero_safe(decompressed, MALLOC_SIZEOF_SAFE(decompressed));
}
_unused_ static void test_decompress_startswith(const char *compression,
compress_blob_t compress,
decompress_sw_t decompress_sw,
const char *data,
size_t data_len,
bool may_fail) {
char *compressed;
_cleanup_free_ char *compressed1 = NULL, *compressed2 = NULL, *decompressed = NULL;
size_t csize, len;
int r;
log_info("/* testing decompress_startswith with %s on %.20s text */",
compression, data);
#define BUFSIZE_1 512
#define BUFSIZE_2 20000
compressed = compressed1 = malloc(BUFSIZE_1);
assert_se(compressed1);
r = compress(data, data_len, compressed, BUFSIZE_1, &csize);
if (r == -ENOBUFS) {
log_info_errno(r, "compression failed: %m");
assert_se(may_fail);
compressed = compressed2 = malloc(BUFSIZE_2);
assert_se(compressed2);
r = compress(data, data_len, compressed, BUFSIZE_2, &csize);
}
assert_se(r > 0);
len = strlen(data);
r = decompress_sw(compressed, csize, (void **) &decompressed, data, len, '\0');
assert_se(r > 0);
r = decompress_sw(compressed, csize, (void **) &decompressed, data, len, 'w');
assert_se(r == 0);
r = decompress_sw(compressed, csize, (void **) &decompressed, "barbarbar", 9, ' ');
assert_se(r == 0);
r = decompress_sw(compressed, csize, (void **) &decompressed, data, len - 1, data[len-1]);
assert_se(r > 0);
r = decompress_sw(compressed, csize, (void **) &decompressed, data, len - 1, 'w');
assert_se(r == 0);
r = decompress_sw(compressed, csize, (void **) &decompressed, data, len, '\0');
assert_se(r > 0);
}
_unused_ static void test_decompress_startswith_short(const char *compression,
compress_blob_t compress,
decompress_sw_t decompress_sw) {
#define TEXT "HUGE=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
char buf[1024];
size_t csize;
int r;
log_info("/* %s with %s */", __func__, compression);
r = compress(TEXT, sizeof TEXT, buf, sizeof buf, &csize);
assert_se(r > 0);
for (size_t i = 1; i < strlen(TEXT); i++) {
_cleanup_free_ void *buf2 = NULL;
assert_se(buf2 = malloc(i));
assert_se(decompress_sw(buf, csize, &buf2, TEXT, i, TEXT[i]) == 1);
assert_se(decompress_sw(buf, csize, &buf2, TEXT, i, 'y') == 0);
}
}
_unused_ static void test_compress_stream(int flag,
const char *compression,
const char *cat,
compress_stream_t compress,
decompress_stream_t decompress,
const char *srcfile) {
_cleanup_close_ int src = -1, dst = -1, dst2 = -1;
_cleanup_(unlink_tempfilep) char
pattern[] = "/tmp/systemd-test.compressed.XXXXXX",
pattern2[] = "/tmp/systemd-test.compressed.XXXXXX";
int r;
_cleanup_free_ char *cmd = NULL, *cmd2 = NULL;
struct stat st = {};
uint64_t uncompressed_size;
r = find_executable(cat, NULL);
if (r < 0) {
log_error_errno(r, "Skipping %s, could not find %s binary: %m", __func__, cat);
return;
}
log_debug("/* testing %s compression */", compression);
log_debug("/* create source from %s */", srcfile);
assert_se((src = open(srcfile, O_RDONLY|O_CLOEXEC)) >= 0);
log_debug("/* test compression */");
assert_se((dst = mkostemp_safe(pattern)) >= 0);
assert_se(compress(src, dst, -1, &uncompressed_size) == flag);
if (cat) {
assert_se(asprintf(&cmd, "%s %s | diff %s -", cat, pattern, srcfile) > 0);
assert_se(system(cmd) == 0);
}
log_debug("/* test decompression */");
assert_se((dst2 = mkostemp_safe(pattern2)) >= 0);
assert_se(stat(srcfile, &st) == 0);
assert_se((uint64_t)st.st_size == uncompressed_size);
assert_se(lseek(dst, 0, SEEK_SET) == 0);
r = decompress(dst, dst2, st.st_size);
assert_se(r == 0);
assert_se(asprintf(&cmd2, "diff %s %s", srcfile, pattern2) > 0);
assert_se(system(cmd2) == 0);
log_debug("/* test faulty decompression */");
assert_se(lseek(dst, 1, SEEK_SET) == 1);
r = decompress(dst, dst2, st.st_size);
assert_se(IN_SET(r, 0, -EBADMSG));
assert_se(lseek(dst, 0, SEEK_SET) == 0);
assert_se(lseek(dst2, 0, SEEK_SET) == 0);
r = decompress(dst, dst2, st.st_size - 1);
assert_se(r == -EFBIG);
}
#endif
#if HAVE_LZ4
static void test_lz4_decompress_partial(void) {
char buf[20000], buf2[100];
size_t buf_size = sizeof(buf), compressed;
int r;
_cleanup_free_ char *huge = NULL;
log_debug("/* %s */", __func__);
assert_se(huge = malloc(HUGE_SIZE));
memcpy(huge, "HUGE=", STRLEN("HUGE="));
memset(&huge[STRLEN("HUGE=")], 'x', HUGE_SIZE - STRLEN("HUGE=") - 1);
huge[HUGE_SIZE - 1] = '\0';
r = LZ4_compress_default(huge, buf, HUGE_SIZE, buf_size);
assert_se(r >= 0);
compressed = r;
log_info("Compressed %i → %zu", HUGE_SIZE, compressed);
r = LZ4_decompress_safe(buf, huge, r, HUGE_SIZE);
assert_se(r >= 0);
log_info("Decompressed → %i", r);
r = LZ4_decompress_safe_partial(buf, huge,
compressed,
12, HUGE_SIZE);
assert_se(r >= 0);
log_info("Decompressed partial %i/%i → %i", 12, HUGE_SIZE, r);
for (size_t size = 1; size < sizeof(buf2); size++) {
/* This failed in older lz4s but works in newer ones. */
r = LZ4_decompress_safe_partial(buf, buf2, compressed, size, size);
log_info("Decompressed partial %zu/%zu → %i (%s)", size, size, r,
r < 0 ? "bad" : "good");
if (r >= 0 && LZ4_versionNumber() >= 10803)
/* lz4 <= 1.8.2 should fail that test, let's only check for newer ones */
assert_se(memcmp(buf2, huge, r) == 0);
}
}
#endif
int main(int argc, char *argv[]) {
#if HAVE_COMPRESSION
_unused_ const char text[] =
"text\0foofoofoofoo AAAA aaaaaaaaa ghost busters barbarbar FFF"
"foofoofoofoo AAAA aaaaaaaaa ghost busters barbarbar FFF";
/* The file to test compression on can be specified as the first argument */
const char *srcfile = argc > 1 ? argv[1] : argv[0];
char data[512] = "random\0";
_cleanup_free_ char *huge = NULL;
assert_se(huge = malloc(HUGE_SIZE));
memcpy(huge, "HUGE=", STRLEN("HUGE="));
memset(&huge[STRLEN("HUGE=")], 'x', HUGE_SIZE - STRLEN("HUGE=") - 1);
huge[HUGE_SIZE - 1] = '\0';
test_setup_logging(LOG_DEBUG);
random_bytes(data + 7, sizeof(data) - 7);
#if HAVE_XZ
test_compress_decompress(COMPRESSION_XZ, "XZ",
compress_blob_xz, decompress_blob_xz,
text, sizeof(text), false);
test_compress_decompress(COMPRESSION_XZ, "XZ",
compress_blob_xz, decompress_blob_xz,
data, sizeof(data), true);
test_decompress_startswith("XZ",
compress_blob_xz, decompress_startswith_xz,
text, sizeof(text), false);
test_decompress_startswith("XZ",
compress_blob_xz, decompress_startswith_xz,
data, sizeof(data), true);
test_decompress_startswith("XZ",
compress_blob_xz, decompress_startswith_xz,
huge, HUGE_SIZE, true);
test_compress_stream(COMPRESSION_XZ, "XZ", "xzcat",
compress_stream_xz, decompress_stream_xz, srcfile);
test_decompress_startswith_short("XZ", compress_blob_xz, decompress_startswith_xz);
#else
log_info("/* XZ test skipped */");
#endif
#if HAVE_LZ4
test_compress_decompress(COMPRESSION_LZ4, "LZ4",
compress_blob_lz4, decompress_blob_lz4,
text, sizeof(text), false);
test_compress_decompress(COMPRESSION_LZ4, "LZ4",
compress_blob_lz4, decompress_blob_lz4,
data, sizeof(data), true);
test_decompress_startswith("LZ4",
compress_blob_lz4, decompress_startswith_lz4,
text, sizeof(text), false);
test_decompress_startswith("LZ4",
compress_blob_lz4, decompress_startswith_lz4,
data, sizeof(data), true);
test_decompress_startswith("LZ4",
compress_blob_lz4, decompress_startswith_lz4,
huge, HUGE_SIZE, true);
test_compress_stream(COMPRESSION_LZ4, "LZ4", "lz4cat",
compress_stream_lz4, decompress_stream_lz4, srcfile);
test_lz4_decompress_partial();
test_decompress_startswith_short("LZ4", compress_blob_lz4, decompress_startswith_lz4);
#else
log_info("/* LZ4 test skipped */");
#endif
#if HAVE_ZSTD
test_compress_decompress(COMPRESSION_ZSTD, "ZSTD",
compress_blob_zstd, decompress_blob_zstd,
text, sizeof(text), false);
test_compress_decompress(COMPRESSION_ZSTD, "ZSTD",
compress_blob_zstd, decompress_blob_zstd,
data, sizeof(data), true);
test_decompress_startswith("ZSTD",
compress_blob_zstd, decompress_startswith_zstd,
text, sizeof(text), false);
test_decompress_startswith("ZSTD",
compress_blob_zstd, decompress_startswith_zstd,
data, sizeof(data), true);
test_decompress_startswith("ZSTD",
compress_blob_zstd, decompress_startswith_zstd,
huge, HUGE_SIZE, true);
test_compress_stream(COMPRESSION_ZSTD, "ZSTD", "zstdcat",
compress_stream_zstd, decompress_stream_zstd, srcfile);
test_decompress_startswith_short("ZSTD", compress_blob_zstd, decompress_startswith_zstd);
#else
log_info("/* ZSTD test skipped */");
#endif
return 0;
#else
log_info("/* XZ, LZ4 and ZSTD tests skipped */");
return EXIT_TEST_SKIP;
#endif
}
+1 -2
View File
@@ -122,8 +122,7 @@ int main(int argc, char **argv) {
test_table(unit_load_state, UNIT_LOAD_STATE);
test_table(unit_type, UNIT_TYPE);
test_table(virtualization, VIRTUALIZATION);
test_table_sparse(object_compressed, OBJECT_COMPRESSED);
test_table(compression, COMPRESSION);
assert_cc(sizeof(sd_device_action_t) == sizeof(int64_t));