freebsd-skq/usr.bin/mkuzip/mkuzip.c
Conrad Meyer eefd8f96fb geom_uzip(4), mkuzip(8): Add Zstd image mode
The Zstd format bumps the CLOOP major number to 4 to avoid incompatibility
with older systems.  Support in geom_uzip(4) is conditional on the ZSTDIO
kernel option, which is enabled in amd64 GENERIC, but not all in-tree
configurations.

mkuzip(8) was modified slightly to always initialize the nblocks + 1'th
offset in the CLOOP file format.  Previously, it was only initialized in the
case where the final compressed block happened to be unaligned w.r.t.
DEV_BSIZE.  The "Fake" last+1 block change in r298619 means that the final
compressed block's 'blen' was never correct unless the compressed uzip image
happened to be BSIZE-aligned.  This happened in about 1 out of every 512
cases.  The zlib and lzma decompressors are probably tolerant of extra trash
following the frame they were told to decode, but Zstd complains that the
input size is incorrect.

Correspondingly, geom_uzip(4) was modified slightly to avoid trashing the
nblocks + 1'th offset when it is known to be initialized to a good value.
This corrects the calculated final real cluster compressed length to match
that printed by mkuzip(8).

mkuzip(8) was refactored somewhat to reduce code duplication and increase
ease of adding other compression formats.

  * Input block size validation was pulled out of individual compression
    init routines into main().

  * Init routines now validate a user-provided compression level or select
    an algorithm-specific default, if none was provided.

  * A new interface for calculating the maximal compressed size of an
    incompressible input block was added for each driver.  The generic code
    uses it to validate against MAXPHYS as well as to allocate compression
    result buffers in the generic code.

  * Algorithm selection is now driven by a table lookup, to increase ease of
    adding other formats in the future.

mkuzip(8) gained the ability to explicitly specify a compression level with
'-C'.  The prior defaults -- 9 for zlib and 6 for lzma -- are maintained.
The new zstd default is 9, to match zlib.

Rather than select lzma or zlib with '-L' or its absense, respectively, a
new argument '-A <algorithm>' is provided to select 'zlib', 'lzma', or
'zstd'.  '-L' is considered deprecated, but will probably never be removed.

All of the new features were documented in mkuzip.8; the page was also
cleaned up slightly.

Relnotes:	yes
2019-08-13 23:32:56 +00:00

507 lines
12 KiB
C

/*-
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
*
* Copyright (c) 2004-2016 Maxim Sobolev <sobomax@FreeBSD.org>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/types.h>
#include <sys/endian.h>
#include <sys/param.h>
#include <sys/sysctl.h>
#include <sys/stat.h>
#include <sys/uio.h>
#include <netinet/in.h>
#include <assert.h>
#include <ctype.h>
#include <err.h>
#include <fcntl.h>
#include <pthread.h>
#include <signal.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include "mkuzip.h"
#include "mkuz_cloop.h"
#include "mkuz_blockcache.h"
#include "mkuz_lzma.h"
#include "mkuz_zlib.h"
#include "mkuz_zstd.h"
#include "mkuz_blk.h"
#include "mkuz_cfg.h"
#include "mkuz_conveyor.h"
#include "mkuz_format.h"
#include "mkuz_fqueue.h"
#include "mkuz_time.h"
#include "mkuz_insize.h"
#define DEFAULT_CLSTSIZE 16384
enum UZ_ALGORITHM {
UZ_ZLIB = 0,
UZ_LZMA,
UZ_ZSTD,
UZ_INVALID
};
static const struct mkuz_format uzip_fmts[] = {
[UZ_ZLIB] = {
.option = "zlib",
.magic = CLOOP_MAGIC_ZLIB,
.default_sufx = DEFAULT_SUFX_ZLIB,
.f_compress_bound = mkuz_zlib_cbound,
.f_init = mkuz_zlib_init,
.f_compress = mkuz_zlib_compress,
},
[UZ_LZMA] = {
.option = "lzma",
.magic = CLOOP_MAGIC_LZMA,
.default_sufx = DEFAULT_SUFX_LZMA,
.f_compress_bound = mkuz_lzma_cbound,
.f_init = mkuz_lzma_init,
.f_compress = mkuz_lzma_compress,
},
[UZ_ZSTD] = {
.option = "zstd",
.magic = CLOOP_MAGIC_ZSTD,
.default_sufx = DEFAULT_SUFX_ZSTD,
.f_compress_bound = mkuz_zstd_cbound,
.f_init = mkuz_zstd_init,
.f_compress = mkuz_zstd_compress,
},
};
static struct mkuz_blk *readblock(int, u_int32_t);
static void usage(void);
static void cleanup(void);
static char *cleanfile = NULL;
static int
cmp_blkno(const struct mkuz_blk *bp, void *p)
{
uint32_t *ap;
ap = (uint32_t *)p;
return (bp->info.blkno == *ap);
}
int main(int argc, char **argv)
{
struct mkuz_cfg cfs;
char *oname;
uint64_t *toc;
int i, io, opt, tmp;
struct {
int en;
FILE *f;
} summary;
struct iovec iov[2];
uint64_t offset, last_offset;
struct cloop_header hdr;
struct mkuz_conveyor *cvp;
void *c_ctx;
struct mkuz_blk_info *chit;
size_t ncpusz, ncpu, magiclen;
double st, et;
enum UZ_ALGORITHM comp_alg;
int comp_level;
st = getdtime();
ncpusz = sizeof(size_t);
if (sysctlbyname("hw.ncpu", &ncpu, &ncpusz, NULL, 0) < 0) {
ncpu = 1;
} else if (ncpu > MAX_WORKERS_AUTO) {
ncpu = MAX_WORKERS_AUTO;
}
memset(&hdr, 0, sizeof(hdr));
cfs.blksz = DEFAULT_CLSTSIZE;
oname = NULL;
cfs.verbose = 0;
cfs.no_zcomp = 0;
cfs.en_dedup = 0;
summary.en = 0;
summary.f = stderr;
comp_alg = UZ_ZLIB;
comp_level = USE_DEFAULT_LEVEL;
cfs.nworkers = ncpu;
struct mkuz_blk *iblk, *oblk;
while((opt = getopt(argc, argv, "A:C:o:s:vZdLSj:")) != -1) {
switch(opt) {
case 'A':
for (tmp = UZ_ZLIB; tmp < UZ_INVALID; tmp++) {
if (strcmp(uzip_fmts[tmp].option, optarg) == 0)
break;
}
if (tmp == UZ_INVALID)
errx(1, "invalid algorithm specified: %s",
optarg);
/* Not reached */
comp_alg = tmp;
break;
case 'C':
comp_level = atoi(optarg);
break;
case 'o':
oname = optarg;
break;
case 's':
tmp = atoi(optarg);
if (tmp <= 0) {
errx(1, "invalid cluster size specified: %s",
optarg);
/* Not reached */
}
cfs.blksz = tmp;
break;
case 'v':
cfs.verbose = 1;
break;
case 'Z':
cfs.no_zcomp = 1;
break;
case 'd':
cfs.en_dedup = 1;
break;
case 'L':
comp_alg = UZ_LZMA;
break;
case 'S':
summary.en = 1;
summary.f = stdout;
break;
case 'j':
tmp = atoi(optarg);
if (tmp <= 0) {
errx(1, "invalid number of compression threads"
" specified: %s", optarg);
/* Not reached */
}
cfs.nworkers = tmp;
break;
default:
usage();
/* Not reached */
}
}
argc -= optind;
argv += optind;
if (argc != 1) {
usage();
/* Not reached */
}
cfs.handler = &uzip_fmts[comp_alg];
magiclen = strlcpy(hdr.magic, cfs.handler->magic, sizeof(hdr.magic));
assert(magiclen < sizeof(hdr.magic));
if (cfs.en_dedup != 0) {
/*
* Dedupe requires a version 3 format. Don't downgrade newer
* formats.
*/
if (hdr.magic[CLOOP_OFS_VERSN] == CLOOP_MAJVER_2)
hdr.magic[CLOOP_OFS_VERSN] = CLOOP_MAJVER_3;
hdr.magic[CLOOP_OFS_COMPR] =
tolower(hdr.magic[CLOOP_OFS_COMPR]);
}
if (cfs.blksz % DEV_BSIZE != 0)
errx(1, "cluster size should be multiple of %d", DEV_BSIZE);
cfs.cbound_blksz = cfs.handler->f_compress_bound(cfs.blksz);
if (cfs.cbound_blksz > MAXPHYS)
errx(1, "maximal compressed cluster size %zu greater than MAXPHYS %zu",
cfs.cbound_blksz, (size_t)MAXPHYS);
c_ctx = cfs.handler->f_init(&comp_level);
cfs.comp_level = comp_level;
cfs.iname = argv[0];
if (oname == NULL) {
asprintf(&oname, "%s%s", cfs.iname, cfs.handler->default_sufx);
if (oname == NULL) {
err(1, "can't allocate memory");
/* Not reached */
}
}
signal(SIGHUP, exit);
signal(SIGINT, exit);
signal(SIGTERM, exit);
signal(SIGXCPU, exit);
signal(SIGXFSZ, exit);
atexit(cleanup);
cfs.fdr = open(cfs.iname, O_RDONLY);
if (cfs.fdr < 0) {
err(1, "open(%s)", cfs.iname);
/* Not reached */
}
cfs.isize = mkuz_get_insize(&cfs);
if (cfs.isize < 0) {
errx(1, "can't determine input image size");
/* Not reached */
}
hdr.nblocks = cfs.isize / cfs.blksz;
if ((cfs.isize % cfs.blksz) != 0) {
if (cfs.verbose != 0)
fprintf(stderr, "file size is not multiple "
"of %d, padding data\n", cfs.blksz);
hdr.nblocks++;
}
toc = mkuz_safe_malloc((hdr.nblocks + 1) * sizeof(*toc));
/*
* Initialize last+1 entry with non-heap trash. If final padding is
* added later, it may or may not be overwritten with an offset
* representing the length of the final compressed block. If not,
* initialize to a defined value.
*/
toc[hdr.nblocks] = 0;
cfs.fdw = open(oname, (cfs.en_dedup ? O_RDWR : O_WRONLY) | O_TRUNC | O_CREAT,
S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IXOTH);
if (cfs.fdw < 0) {
err(1, "open(%s)", oname);
/* Not reached */
}
cleanfile = oname;
/* Prepare header that we will write later when we have index ready. */
iov[0].iov_base = (char *)&hdr;
iov[0].iov_len = sizeof(hdr);
iov[1].iov_base = (char *)toc;
iov[1].iov_len = (hdr.nblocks + 1) * sizeof(*toc);
offset = iov[0].iov_len + iov[1].iov_len;
/* Reserve space for header */
lseek(cfs.fdw, offset, SEEK_SET);
if (cfs.verbose != 0) {
fprintf(stderr, "data size %ju bytes, number of clusters "
"%u, index length %zu bytes\n", cfs.isize,
hdr.nblocks, iov[1].iov_len);
}
cvp = mkuz_conveyor_ctor(&cfs);
last_offset = 0;
iblk = oblk = NULL;
for(i = io = 0; iblk != MKUZ_BLK_EOF; i++) {
iblk = readblock(cfs.fdr, cfs.blksz);
mkuz_fqueue_enq(cvp->wrk_queue, iblk);
if (iblk != MKUZ_BLK_EOF &&
(i < (cfs.nworkers * ITEMS_PER_WORKER))) {
continue;
}
drain:
oblk = mkuz_fqueue_deq_when(cvp->results, cmp_blkno, &io);
assert(oblk->info.blkno == (unsigned)io);
oblk->info.offset = offset;
chit = NULL;
if (cfs.en_dedup != 0 && oblk->info.len > 0) {
chit = mkuz_blkcache_regblock(cfs.fdw, oblk);
/*
* There should be at least one non-empty block
* between us and the backref'ed offset, otherwise
* we won't be able to parse that sequence correctly
* as it would be indistinguishible from another
* empty block.
*/
if (chit != NULL && chit->offset == last_offset) {
chit = NULL;
}
}
if (chit != NULL) {
toc[io] = htobe64(chit->offset);
oblk->info.len = 0;
} else {
if (oblk->info.len > 0 && write(cfs.fdw, oblk->data,
oblk->info.len) < 0) {
err(1, "write(%s)", oname);
/* Not reached */
}
toc[io] = htobe64(offset);
last_offset = offset;
offset += oblk->info.len;
}
if (cfs.verbose != 0) {
fprintf(stderr, "cluster #%d, in %u bytes, "
"out len=%lu offset=%lu", io, cfs.blksz,
(u_long)oblk->info.len, (u_long)be64toh(toc[io]));
if (chit != NULL) {
fprintf(stderr, " (backref'ed to #%d)",
chit->blkno);
}
fprintf(stderr, "\n");
}
free(oblk);
io += 1;
if (iblk == MKUZ_BLK_EOF) {
if (io < i)
goto drain;
/* Last block, see if we need to add some padding */
if ((offset % DEV_BSIZE) == 0)
continue;
oblk = mkuz_blk_ctor(DEV_BSIZE - (offset % DEV_BSIZE));
oblk->info.blkno = io;
oblk->info.len = oblk->alen;
if (cfs.verbose != 0) {
fprintf(stderr, "padding data with %lu bytes "
"so that file size is multiple of %d\n",
(u_long)oblk->alen, DEV_BSIZE);
}
mkuz_fqueue_enq(cvp->results, oblk);
goto drain;
}
}
close(cfs.fdr);
if (cfs.verbose != 0 || summary.en != 0) {
et = getdtime();
fprintf(summary.f, "compressed data to %ju bytes, saved %lld "
"bytes, %.2f%% decrease, %.2f bytes/sec.\n", offset,
(long long)(cfs.isize - offset),
100.0 * (long long)(cfs.isize - offset) /
(float)cfs.isize, (float)cfs.isize / (et - st));
}
/* Convert to big endian */
hdr.blksz = htonl(cfs.blksz);
hdr.nblocks = htonl(hdr.nblocks);
/* Write headers into pre-allocated space */
lseek(cfs.fdw, 0, SEEK_SET);
if (writev(cfs.fdw, iov, 2) < 0) {
err(1, "writev(%s)", oname);
/* Not reached */
}
cleanfile = NULL;
close(cfs.fdw);
exit(0);
}
static struct mkuz_blk *
readblock(int fd, u_int32_t clstsize)
{
int numread;
struct mkuz_blk *rval;
static int blockcnt;
off_t cpos;
rval = mkuz_blk_ctor(clstsize);
rval->info.blkno = blockcnt;
blockcnt += 1;
cpos = lseek(fd, 0, SEEK_CUR);
if (cpos < 0) {
err(1, "readblock: lseek() failed");
/* Not reached */
}
rval->info.offset = cpos;
numread = read(fd, rval->data, clstsize);
if (numread < 0) {
err(1, "readblock: read() failed");
/* Not reached */
}
if (numread == 0) {
free(rval);
return MKUZ_BLK_EOF;
}
rval->info.len = numread;
return rval;
}
static void
usage(void)
{
fprintf(stderr, "usage: mkuzip [-vZdLS] [-o outfile] [-s cluster_size] "
"[-j ncompr] infile\n");
exit(1);
}
void *
mkuz_safe_malloc(size_t size)
{
void *retval;
retval = malloc(size);
if (retval == NULL) {
err(1, "can't allocate memory");
/* Not reached */
}
return retval;
}
void *
mkuz_safe_zmalloc(size_t size)
{
void *retval;
retval = mkuz_safe_malloc(size);
bzero(retval, size);
return retval;
}
static void
cleanup(void)
{
if (cleanfile != NULL)
unlink(cleanfile);
}
int
mkuz_memvcmp(const void *memory, unsigned char val, size_t size)
{
const u_char *mm;
mm = (const u_char *)memory;
return (*mm == val) && memcmp(mm, mm + 1, size - 1) == 0;
}