2004-08-01 23:59:26 +02:00
|
|
|
/*
|
|
|
|
* Block driver for the QCOW format
|
2007-09-16 23:08:06 +02:00
|
|
|
*
|
2006-08-01 18:21:11 +02:00
|
|
|
* Copyright (c) 2004-2006 Fabrice Bellard
|
2007-09-16 23:08:06 +02:00
|
|
|
*
|
2004-08-01 23:59:26 +02:00
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
|
|
* of this software and associated documentation files (the "Software"), to deal
|
|
|
|
* in the Software without restriction, including without limitation the rights
|
|
|
|
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
|
|
* copies of the Software, and to permit persons to whom the Software is
|
|
|
|
* furnished to do so, subject to the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice shall be included in
|
|
|
|
* all copies or substantial portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
|
|
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
|
|
* THE SOFTWARE.
|
|
|
|
*/
|
2016-01-18 19:01:42 +01:00
|
|
|
#include "qemu/osdep.h"
|
include/qemu/osdep.h: Don't include qapi/error.h
Commit 57cb38b included qapi/error.h into qemu/osdep.h to get the
Error typedef. Since then, we've moved to include qemu/osdep.h
everywhere. Its file comment explains: "To avoid getting into
possible circular include dependencies, this file should not include
any other QEMU headers, with the exceptions of config-host.h,
compiler.h, os-posix.h and os-win32.h, all of which are doing a
similar job to this file and are under similar constraints."
qapi/error.h doesn't do a similar job, and it doesn't adhere to
similar constraints: it includes qapi-types.h. That's in excess of
100KiB of crap most .c files don't actually need.
Add the typedef to qemu/typedefs.h, and include that instead of
qapi/error.h. Include qapi/error.h in .c files that need it and don't
get it now. Include qapi-types.h in qom/object.h for uint16List.
Update scripts/clean-includes accordingly. Update it further to match
reality: replace config.h by config-target.h, add sysemu/os-posix.h,
sysemu/os-win32.h. Update the list of includes in the qemu/osdep.h
comment quoted above similarly.
This reduces the number of objects depending on qapi/error.h from "all
of them" to less than a third. Unfortunately, the number depending on
qapi-types.h shrinks only a little. More work is needed for that one.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
[Fix compilation without the spice devel packages. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-03-14 09:01:28 +01:00
|
|
|
#include "qapi/error.h"
|
2007-11-11 03:51:17 +01:00
|
|
|
#include "qemu-common.h"
|
2016-03-21 15:11:48 +01:00
|
|
|
#include "qemu/error-report.h"
|
2012-12-17 18:19:44 +01:00
|
|
|
#include "block/block_int.h"
|
2016-03-08 15:57:05 +01:00
|
|
|
#include "sysemu/block-backend.h"
|
2012-12-17 18:20:00 +01:00
|
|
|
#include "qemu/module.h"
|
2016-03-15 17:22:36 +01:00
|
|
|
#include "qemu/bswap.h"
|
2004-09-29 23:30:43 +02:00
|
|
|
#include <zlib.h>
|
2015-03-17 17:22:46 +01:00
|
|
|
#include "qapi/qmp/qerror.h"
|
2015-07-01 19:10:37 +02:00
|
|
|
#include "crypto/cipher.h"
|
2012-12-17 18:19:50 +01:00
|
|
|
#include "migration/migration.h"
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
/**************************************************************/
|
|
|
|
/* QEMU COW block driver with compression and encryption support */
|
|
|
|
|
|
|
|
#define QCOW_MAGIC (('Q' << 24) | ('F' << 16) | ('I' << 8) | 0xfb)
|
|
|
|
#define QCOW_VERSION 1
|
|
|
|
|
|
|
|
#define QCOW_CRYPT_NONE 0
|
|
|
|
#define QCOW_CRYPT_AES 1
|
|
|
|
|
|
|
|
#define QCOW_OFLAG_COMPRESSED (1LL << 63)
|
|
|
|
|
|
|
|
typedef struct QCowHeader {
|
|
|
|
uint32_t magic;
|
|
|
|
uint32_t version;
|
|
|
|
uint64_t backing_file_offset;
|
|
|
|
uint32_t backing_file_size;
|
|
|
|
uint32_t mtime;
|
|
|
|
uint64_t size; /* in bytes */
|
|
|
|
uint8_t cluster_bits;
|
|
|
|
uint8_t l2_bits;
|
2014-05-07 16:56:10 +02:00
|
|
|
uint16_t padding;
|
2004-08-01 23:59:26 +02:00
|
|
|
uint32_t crypt_method;
|
|
|
|
uint64_t l1_table_offset;
|
2014-05-07 16:56:10 +02:00
|
|
|
} QEMU_PACKED QCowHeader;
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
#define L2_CACHE_SIZE 16
|
|
|
|
|
|
|
|
typedef struct BDRVQcowState {
|
|
|
|
int cluster_bits;
|
|
|
|
int cluster_size;
|
|
|
|
int cluster_sectors;
|
|
|
|
int l2_bits;
|
|
|
|
int l2_size;
|
2014-05-08 13:08:20 +02:00
|
|
|
unsigned int l1_size;
|
2004-08-01 23:59:26 +02:00
|
|
|
uint64_t cluster_offset_mask;
|
|
|
|
uint64_t l1_table_offset;
|
|
|
|
uint64_t *l1_table;
|
|
|
|
uint64_t *l2_cache;
|
|
|
|
uint64_t l2_cache_offsets[L2_CACHE_SIZE];
|
|
|
|
uint32_t l2_cache_counts[L2_CACHE_SIZE];
|
|
|
|
uint8_t *cluster_cache;
|
|
|
|
uint8_t *cluster_data;
|
|
|
|
uint64_t cluster_cache_offset;
|
2015-07-01 19:10:37 +02:00
|
|
|
QCryptoCipher *cipher; /* NULL if no key yet */
|
2004-08-01 23:59:26 +02:00
|
|
|
uint32_t crypt_method_header;
|
2011-07-15 16:27:42 +02:00
|
|
|
CoMutex lock;
|
2011-11-22 16:44:45 +01:00
|
|
|
Error *migration_blocker;
|
2004-08-01 23:59:26 +02:00
|
|
|
} BDRVQcowState;
|
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
static int decompress_cluster(BlockDriverState *bs, uint64_t cluster_offset);
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
static int qcow_probe(const uint8_t *buf, int buf_size, const char *filename)
|
|
|
|
{
|
|
|
|
const QCowHeader *cow_header = (const void *)buf;
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2005-04-28 23:09:32 +02:00
|
|
|
if (buf_size >= sizeof(QCowHeader) &&
|
|
|
|
be32_to_cpu(cow_header->magic) == QCOW_MAGIC &&
|
2007-09-16 23:08:06 +02:00
|
|
|
be32_to_cpu(cow_header->version) == QCOW_VERSION)
|
2004-08-01 23:59:26 +02:00
|
|
|
return 100;
|
|
|
|
else
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2013-09-05 14:22:29 +02:00
|
|
|
static int qcow_open(BlockDriverState *bs, QDict *options, int flags,
|
|
|
|
Error **errp)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2014-05-08 13:35:09 +02:00
|
|
|
unsigned int len, i, shift;
|
|
|
|
int ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
QCowHeader header;
|
2006-08-01 18:21:11 +02:00
|
|
|
|
2016-06-20 18:24:02 +02:00
|
|
|
ret = bdrv_pread(bs->file, 0, &header, sizeof(header));
|
2011-12-15 11:14:00 +01:00
|
|
|
if (ret < 0) {
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
be32_to_cpus(&header.magic);
|
|
|
|
be32_to_cpus(&header.version);
|
|
|
|
be64_to_cpus(&header.backing_file_offset);
|
|
|
|
be32_to_cpus(&header.backing_file_size);
|
|
|
|
be32_to_cpus(&header.mtime);
|
|
|
|
be64_to_cpus(&header.size);
|
|
|
|
be32_to_cpus(&header.crypt_method);
|
|
|
|
be64_to_cpus(&header.l1_table_offset);
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2011-12-15 11:14:00 +01:00
|
|
|
if (header.magic != QCOW_MAGIC) {
|
2014-02-17 14:44:06 +01:00
|
|
|
error_setg(errp, "Image not in qcow format");
|
|
|
|
ret = -EINVAL;
|
2011-12-15 11:14:00 +01:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
if (header.version != QCOW_VERSION) {
|
2016-03-16 19:54:33 +01:00
|
|
|
error_setg(errp, "Unsupported qcow version %" PRIu32, header.version);
|
2011-12-15 11:14:00 +01:00
|
|
|
ret = -ENOTSUP;
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
|
|
|
|
2014-05-07 17:30:30 +02:00
|
|
|
if (header.size <= 1) {
|
|
|
|
error_setg(errp, "Image size is too small (must be at least 2 bytes)");
|
2011-12-15 11:14:00 +01:00
|
|
|
ret = -EINVAL;
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2014-05-07 17:30:30 +02:00
|
|
|
if (header.cluster_bits < 9 || header.cluster_bits > 16) {
|
|
|
|
error_setg(errp, "Cluster size must be between 512 and 64k");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2014-05-15 16:10:11 +02:00
|
|
|
/* l2_bits specifies number of entries; storing a uint64_t in each entry,
|
|
|
|
* so bytes = num_entries << 3. */
|
|
|
|
if (header.l2_bits < 9 - 3 || header.l2_bits > 16 - 3) {
|
|
|
|
error_setg(errp, "L2 table size must be between 512 and 64k");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2011-12-15 11:14:00 +01:00
|
|
|
if (header.crypt_method > QCOW_CRYPT_AES) {
|
2014-02-17 14:43:58 +01:00
|
|
|
error_setg(errp, "invalid encryption method in qcow header");
|
2011-12-15 11:14:00 +01:00
|
|
|
ret = -EINVAL;
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2015-07-01 19:10:37 +02:00
|
|
|
if (!qcrypto_cipher_supports(QCRYPTO_CIPHER_ALG_AES_128)) {
|
|
|
|
error_setg(errp, "AES cipher not available");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
s->crypt_method_header = header.crypt_method;
|
2011-12-15 11:14:00 +01:00
|
|
|
if (s->crypt_method_header) {
|
2016-03-21 15:11:48 +01:00
|
|
|
if (bdrv_uses_whitelist() &&
|
|
|
|
s->crypt_method_header == QCOW_CRYPT_AES) {
|
block: drop support for using qcow[2] encryption with system emulators
Back in the 2.3.0 release we declared qcow[2] encryption as
deprecated, warning people that it would be removed in a future
release.
commit a1f688f4152e65260b94f37543521ceff8bfebe4
Author: Markus Armbruster <armbru@redhat.com>
Date: Fri Mar 13 21:09:40 2015 +0100
block: Deprecate QCOW/QCOW2 encryption
The code still exists today, but by a (happy?) accident we entirely
broke the ability to use qcow[2] encryption in the system emulators
in the 2.4.0 release due to
commit 8336aafae1451d54c81dd2b187b45f7c45d2428e
Author: Daniel P. Berrange <berrange@redhat.com>
Date: Tue May 12 17:09:18 2015 +0100
qcow2/qcow: protect against uninitialized encryption key
This commit was designed to prevent future coding bugs which
might cause QEMU to read/write data on an encrypted block
device in plain text mode before a decryption key is set.
It turns out this preventative measure was a little too good,
because we already had a long standing bug where QEMU read
encrypted data in plain text mode during system emulator
startup, in order to guess disk geometry:
Thread 10 (Thread 0x7fffd3fff700 (LWP 30373)):
#0 0x00007fffe90b1a28 in raise () at /lib64/libc.so.6
#1 0x00007fffe90b362a in abort () at /lib64/libc.so.6
#2 0x00007fffe90aa227 in __assert_fail_base () at /lib64/libc.so.6
#3 0x00007fffe90aa2d2 in () at /lib64/libc.so.6
#4 0x000055555587ae19 in qcow2_co_readv (bs=0x5555562accb0, sector_num=0, remaining_sectors=1, qiov=0x7fffffffd260) at block/qcow2.c:1229
#5 0x000055555589b60d in bdrv_aligned_preadv (bs=bs@entry=0x5555562accb0, req=req@entry=0x7fffd3ffea50, offset=offset@entry=0, bytes=bytes@entry=512, align=align@entry=512, qiov=qiov@entry=0x7fffffffd260, flags=0) at block/io.c:908
#6 0x000055555589b8bc in bdrv_co_do_preadv (bs=0x5555562accb0, offset=0, bytes=512, qiov=0x7fffffffd260, flags=<optimized out>) at block/io.c:999
#7 0x000055555589c375 in bdrv_rw_co_entry (opaque=0x7fffffffd210) at block/io.c:544
#8 0x000055555586933b in coroutine_thread (opaque=0x555557876310) at coroutine-gthread.c:134
#9 0x00007ffff64e1835 in g_thread_proxy (data=0x5555562b5590) at gthread.c:778
#10 0x00007ffff6bb760a in start_thread () at /lib64/libpthread.so.0
#11 0x00007fffe917f59d in clone () at /lib64/libc.so.6
Thread 1 (Thread 0x7ffff7ecab40 (LWP 30343)):
#0 0x00007fffe91797a9 in syscall () at /lib64/libc.so.6
#1 0x00007ffff64ff87f in g_cond_wait (cond=cond@entry=0x555555e085f0 <coroutine_cond>, mutex=mutex@entry=0x555555e08600 <coroutine_lock>) at gthread-posix.c:1397
#2 0x00005555558692c3 in qemu_coroutine_switch (co=<optimized out>) at coroutine-gthread.c:117
#3 0x00005555558692c3 in qemu_coroutine_switch (from_=0x5555562b5e30, to_=to_@entry=0x555557876310, action=action@entry=COROUTINE_ENTER) at coroutine-gthread.c:175
#4 0x0000555555868a90 in qemu_coroutine_enter (co=0x555557876310, opaque=0x0) at qemu-coroutine.c:116
#5 0x0000555555859b84 in thread_pool_completion_bh (opaque=0x7fffd40010e0) at thread-pool.c:187
#6 0x0000555555859514 in aio_bh_poll (ctx=ctx@entry=0x5555562953b0) at async.c:85
#7 0x0000555555864d10 in aio_dispatch (ctx=ctx@entry=0x5555562953b0) at aio-posix.c:135
#8 0x0000555555864f75 in aio_poll (ctx=ctx@entry=0x5555562953b0, blocking=blocking@entry=true) at aio-posix.c:291
#9 0x000055555589c40d in bdrv_prwv_co (bs=bs@entry=0x5555562accb0, offset=offset@entry=0, qiov=qiov@entry=0x7fffffffd260, is_write=is_write@entry=false, flags=flags@entry=(unknown: 0)) at block/io.c:591
#10 0x000055555589c503 in bdrv_rw_co (bs=bs@entry=0x5555562accb0, sector_num=sector_num@entry=0, buf=buf@entry=0x7fffffffd2e0 "\321,", nb_sectors=nb_sectors@entry=21845, is_write=is_write@entry=false, flags=flags@entry=(unknown: 0)) at block/io.c:614
#11 0x000055555589c562 in bdrv_read_unthrottled (nb_sectors=21845, buf=0x7fffffffd2e0 "\321,", sector_num=0, bs=0x5555562accb0) at block/io.c:622
#12 0x000055555589c562 in bdrv_read_unthrottled (bs=0x5555562accb0, sector_num=sector_num@entry=0, buf=buf@entry=0x7fffffffd2e0 "\321,", nb_sectors=nb_sectors@entry=21845) at block/io.c:634
nb_sectors@entry=1) at block/block-backend.c:504
#14 0x0000555555752e9f in guess_disk_lchs (blk=blk@entry=0x5555562a5290, pcylinders=pcylinders@entry=0x7fffffffd52c, pheads=pheads@entry=0x7fffffffd530, psectors=psectors@entry=0x7fffffffd534) at hw/block/hd-geometry.c:68
#15 0x0000555555752ff7 in hd_geometry_guess (blk=0x5555562a5290, pcyls=pcyls@entry=0x555557875d1c, pheads=pheads@entry=0x555557875d20, psecs=psecs@entry=0x555557875d24, ptrans=ptrans@entry=0x555557875d28) at hw/block/hd-geometry.c:133
#16 0x0000555555752b87 in blkconf_geometry (conf=conf@entry=0x555557875d00, ptrans=ptrans@entry=0x555557875d28, cyls_max=cyls_max@entry=65536, heads_max=heads_max@entry=16, secs_max=secs_max@entry=255, errp=errp@entry=0x7fffffffd5e0) at hw/block/block.c:71
#17 0x0000555555799bc4 in ide_dev_initfn (dev=0x555557875c80, kind=IDE_HD) at hw/ide/qdev.c:174
#18 0x0000555555768394 in device_realize (dev=0x555557875c80, errp=0x7fffffffd640) at hw/core/qdev.c:247
#19 0x0000555555769a81 in device_set_realized (obj=0x555557875c80, value=<optimized out>, errp=0x7fffffffd730) at hw/core/qdev.c:1058
#20 0x00005555558240ce in property_set_bool (obj=0x555557875c80, v=<optimized out>, opaque=0x555557875de0, name=<optimized out>, errp=0x7fffffffd730)
at qom/object.c:1514
#21 0x0000555555826c87 in object_property_set_qobject (obj=obj@entry=0x555557875c80, value=value@entry=0x55555784bcb0, name=name@entry=0x55555591cb3d "realized", errp=errp@entry=0x7fffffffd730) at qom/qom-qobject.c:24
#22 0x0000555555825760 in object_property_set_bool (obj=obj@entry=0x555557875c80, value=value@entry=true, name=name@entry=0x55555591cb3d "realized", errp=errp@entry=0x7fffffffd730) at qom/object.c:905
#23 0x000055555576897b in qdev_init_nofail (dev=dev@entry=0x555557875c80) at hw/core/qdev.c:380
#24 0x0000555555799ead in ide_create_drive (bus=bus@entry=0x555557629630, unit=unit@entry=0, drive=0x5555562b77e0) at hw/ide/qdev.c:122
#25 0x000055555579a746 in pci_ide_create_devs (dev=dev@entry=0x555557628db0, hd_table=hd_table@entry=0x7fffffffd830) at hw/ide/pci.c:440
#26 0x000055555579b165 in pci_piix3_ide_init (bus=<optimized out>, hd_table=0x7fffffffd830, devfn=<optimized out>) at hw/ide/piix.c:218
#27 0x000055555568ca55 in pc_init1 (machine=0x5555562960a0, pci_enabled=1, kvmclock_enabled=<optimized out>) at /home/berrange/src/virt/qemu/hw/i386/pc_piix.c:256
#28 0x0000555555603ab2 in main (argc=<optimized out>, argv=<optimized out>, envp=<optimized out>) at vl.c:4249
So the safety net is correctly preventing QEMU reading cipher
text as if it were plain text, during startup and aborting QEMU
to avoid bad usage of this data.
For added fun this bug only happens if the encrypted qcow2
file happens to have data written to the first cluster,
otherwise the cluster won't be allocated and so qcow2 would
not try the decryption routines at all, just return all 0's.
That no one even noticed, let alone reported, this bug that
has shipped in 2.4.0, 2.5.0 and 2.6.0 shows that the number
of actual users of encrypted qcow2 is approximately zero.
So rather than fix the crash, and backport it to stable
releases, just go ahead with what we have warned users about
and disable any use of qcow2 encryption in the system
emulators. qemu-img/qemu-io/qemu-nbd are still able to access
qcow2 encrypted images for the sake of data conversion.
In the future, qcow2 will gain support for the alternative
luks format, but when this happens it'll be using the
'-object secret' infrastructure for getting keys, which
avoids this problematic scenario entirely.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2016-06-13 13:30:09 +02:00
|
|
|
error_setg(errp,
|
|
|
|
"Use of AES-CBC encrypted qcow images is no longer "
|
|
|
|
"supported in system emulators");
|
|
|
|
error_append_hint(errp,
|
|
|
|
"You can use 'qemu-img convert' to convert your "
|
|
|
|
"image to an alternative supported format, such "
|
|
|
|
"as unencrypted qcow, or raw with the LUKS "
|
|
|
|
"format instead.\n");
|
|
|
|
ret = -ENOSYS;
|
|
|
|
goto fail;
|
2016-03-21 15:11:48 +01:00
|
|
|
}
|
|
|
|
|
2016-06-24 00:37:26 +02:00
|
|
|
bs->encrypted = true;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
s->cluster_bits = header.cluster_bits;
|
|
|
|
s->cluster_size = 1 << s->cluster_bits;
|
|
|
|
s->cluster_sectors = 1 << (s->cluster_bits - 9);
|
|
|
|
s->l2_bits = header.l2_bits;
|
|
|
|
s->l2_size = 1 << s->l2_bits;
|
|
|
|
bs->total_sectors = header.size / 512;
|
|
|
|
s->cluster_offset_mask = (1LL << (63 - s->cluster_bits)) - 1;
|
|
|
|
|
|
|
|
/* read the level 1 table */
|
|
|
|
shift = s->cluster_bits + s->l2_bits;
|
2014-05-08 13:08:20 +02:00
|
|
|
if (header.size > UINT64_MAX - (1LL << shift)) {
|
|
|
|
error_setg(errp, "Image too large");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
} else {
|
|
|
|
uint64_t l1_size = (header.size + (1LL << shift) - 1) >> shift;
|
|
|
|
if (l1_size > INT_MAX / sizeof(uint64_t)) {
|
|
|
|
error_setg(errp, "Image too large");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
s->l1_size = l1_size;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
s->l1_table_offset = header.l1_table_offset;
|
block: Use g_new() & friends where that makes obvious sense
g_new(T, n) is neater than g_malloc(sizeof(T) * n). It's also safer,
for two reasons. One, it catches multiplication overflowing size_t.
Two, it returns T * rather than void *, which lets the compiler catch
more type errors.
Patch created with Coccinelle, with two manual changes on top:
* Add const to bdrv_iterate_format() to keep the types straight
* Convert the allocation in bdrv_drop_intermediate(), which Coccinelle
inexplicably misses
Coccinelle semantic patch:
@@
type T;
@@
-g_malloc(sizeof(T))
+g_new(T, 1)
@@
type T;
@@
-g_try_malloc(sizeof(T))
+g_try_new(T, 1)
@@
type T;
@@
-g_malloc0(sizeof(T))
+g_new0(T, 1)
@@
type T;
@@
-g_try_malloc0(sizeof(T))
+g_try_new0(T, 1)
@@
type T;
expression n;
@@
-g_malloc(sizeof(T) * (n))
+g_new(T, n)
@@
type T;
expression n;
@@
-g_try_malloc(sizeof(T) * (n))
+g_try_new(T, n)
@@
type T;
expression n;
@@
-g_malloc0(sizeof(T) * (n))
+g_new0(T, n)
@@
type T;
expression n;
@@
-g_try_malloc0(sizeof(T) * (n))
+g_try_new0(T, n)
@@
type T;
expression p, n;
@@
-g_realloc(p, sizeof(T) * (n))
+g_renew(T, p, n)
@@
type T;
expression p, n;
@@
-g_try_realloc(p, sizeof(T) * (n))
+g_try_renew(T, p, n)
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-08-19 10:31:08 +02:00
|
|
|
s->l1_table = g_try_new(uint64_t, s->l1_size);
|
2014-05-20 13:36:05 +02:00
|
|
|
if (s->l1_table == NULL) {
|
|
|
|
error_setg(errp, "Could not allocate memory for L1 table");
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto fail;
|
|
|
|
}
|
2011-12-15 11:14:00 +01:00
|
|
|
|
2016-06-20 18:24:02 +02:00
|
|
|
ret = bdrv_pread(bs->file, s->l1_table_offset, s->l1_table,
|
2011-12-15 11:14:00 +01:00
|
|
|
s->l1_size * sizeof(uint64_t));
|
|
|
|
if (ret < 0) {
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
for(i = 0;i < s->l1_size; i++) {
|
|
|
|
be64_to_cpus(&s->l1_table[i]);
|
|
|
|
}
|
2014-05-20 13:36:05 +02:00
|
|
|
|
|
|
|
/* alloc L2 cache (max. 64k * 16 * 8 = 8 MB) */
|
|
|
|
s->l2_cache =
|
2015-06-16 14:19:22 +02:00
|
|
|
qemu_try_blockalign(bs->file->bs,
|
2014-05-20 13:36:05 +02:00
|
|
|
s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
|
|
|
|
if (s->l2_cache == NULL) {
|
|
|
|
error_setg(errp, "Could not allocate L2 table cache");
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto fail;
|
|
|
|
}
|
2011-08-21 05:09:37 +02:00
|
|
|
s->cluster_cache = g_malloc(s->cluster_size);
|
|
|
|
s->cluster_data = g_malloc(s->cluster_size);
|
2004-08-01 23:59:26 +02:00
|
|
|
s->cluster_cache_offset = -1;
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/* read the backing file name */
|
|
|
|
if (header.backing_file_offset != 0) {
|
|
|
|
len = header.backing_file_size;
|
2015-01-27 14:33:55 +01:00
|
|
|
if (len > 1023 || len >= sizeof(bs->backing_file)) {
|
2014-05-08 13:35:09 +02:00
|
|
|
error_setg(errp, "Backing file name too long");
|
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2016-06-20 18:24:02 +02:00
|
|
|
ret = bdrv_pread(bs->file, header.backing_file_offset,
|
2011-12-15 11:14:00 +01:00
|
|
|
bs->backing_file, len);
|
|
|
|
if (ret < 0) {
|
2004-08-01 23:59:26 +02:00
|
|
|
goto fail;
|
2011-12-15 11:14:00 +01:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
bs->backing_file[len] = '\0';
|
|
|
|
}
|
2011-08-11 23:27:15 +02:00
|
|
|
|
2011-11-22 16:44:45 +01:00
|
|
|
/* Disable migration when qcow images are used */
|
2015-04-08 11:29:19 +02:00
|
|
|
error_setg(&s->migration_blocker, "The qcow format used by node '%s' "
|
|
|
|
"does not support live migration",
|
|
|
|
bdrv_get_device_or_node_name(bs));
|
2011-11-22 16:44:45 +01:00
|
|
|
migrate_add_blocker(s->migration_blocker);
|
|
|
|
|
2011-08-11 23:27:15 +02:00
|
|
|
qemu_co_mutex_init(&s->lock);
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(s->l1_table);
|
2014-05-20 13:36:05 +02:00
|
|
|
qemu_vfree(s->l2_cache);
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(s->cluster_cache);
|
|
|
|
g_free(s->cluster_data);
|
2011-12-15 11:14:00 +01:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2012-09-20 21:13:29 +02:00
|
|
|
|
|
|
|
/* We have nothing to do for QCOW reopen, stubs just return
|
|
|
|
* success */
|
|
|
|
static int qcow_reopen_prepare(BDRVReopenState *state,
|
|
|
|
BlockReopenQueue *queue, Error **errp)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
static int qcow_set_key(BlockDriverState *bs, const char *key)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
uint8_t keybuf[16];
|
|
|
|
int len, i;
|
2015-07-01 19:10:37 +02:00
|
|
|
Error *err;
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
memset(keybuf, 0, 16);
|
|
|
|
len = strlen(key);
|
|
|
|
if (len > 16)
|
|
|
|
len = 16;
|
|
|
|
/* XXX: we could compress the chars to 7 bits to increase
|
|
|
|
entropy */
|
|
|
|
for(i = 0;i < len;i++) {
|
|
|
|
keybuf[i] = key[i];
|
|
|
|
}
|
qcow2/qcow: protect against uninitialized encryption key
When a qcow[2] file is opened, if the header reports an
encryption method, this is used to set the 'crypt_method_header'
field on the BDRVQcow[2]State struct, and the 'encrypted' flag
in the BDRVState struct.
When doing I/O operations, the 'crypt_method' field on the
BDRVQcow[2]State struct is checked to determine if encryption
needs to be applied.
The crypt_method_header value is copied into crypt_method when
the bdrv_set_key() method is called.
The QEMU code which opens a block device is expected to always
do a check
if (bdrv_is_encrypted(bs)) {
bdrv_set_key(bs, ....key...);
}
If code forgets to do this, then 'crypt_method' is never set
and so when I/O is performed, QEMU writes plain text data
into a sector which is expected to contain cipher text, or
when reading, will return cipher text instead of plain
text.
Change the qcow[2] code to consult bs->encrypted when deciding
whether encryption is required, and assert(s->crypt_method)
to protect against cases where the caller forgets to set the
encryption key.
Also put an assert in the set_key methods to protect against
the case where the caller sets an encryption key on a block
device that does not have encryption
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2015-05-12 18:09:18 +02:00
|
|
|
assert(bs->encrypted);
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2015-07-01 19:10:37 +02:00
|
|
|
qcrypto_cipher_free(s->cipher);
|
|
|
|
s->cipher = qcrypto_cipher_new(
|
|
|
|
QCRYPTO_CIPHER_ALG_AES_128,
|
|
|
|
QCRYPTO_CIPHER_MODE_CBC,
|
|
|
|
keybuf, G_N_ELEMENTS(keybuf),
|
|
|
|
&err);
|
|
|
|
|
|
|
|
if (!s->cipher) {
|
|
|
|
/* XXX would be nice if errors in this method could
|
|
|
|
* be properly propagate to the caller. Would need
|
|
|
|
* the bdrv_set_key() API signature to be fixed. */
|
|
|
|
error_free(err);
|
2004-08-01 23:59:26 +02:00
|
|
|
return -1;
|
2015-07-01 19:10:37 +02:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* The crypt function is compatible with the linux cryptoloop
|
|
|
|
algorithm for < 4 GB images. NOTE: out_buf == in_buf is
|
|
|
|
supported */
|
2015-07-01 19:10:37 +02:00
|
|
|
static int encrypt_sectors(BDRVQcowState *s, int64_t sector_num,
|
|
|
|
uint8_t *out_buf, const uint8_t *in_buf,
|
|
|
|
int nb_sectors, bool enc, Error **errp)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
|
|
|
union {
|
|
|
|
uint64_t ll[2];
|
|
|
|
uint8_t b[16];
|
|
|
|
} ivec;
|
|
|
|
int i;
|
2015-07-01 19:10:37 +02:00
|
|
|
int ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
for(i = 0; i < nb_sectors; i++) {
|
|
|
|
ivec.ll[0] = cpu_to_le64(sector_num);
|
|
|
|
ivec.ll[1] = 0;
|
2015-07-01 19:10:37 +02:00
|
|
|
if (qcrypto_cipher_setiv(s->cipher,
|
|
|
|
ivec.b, G_N_ELEMENTS(ivec.b),
|
|
|
|
errp) < 0) {
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
if (enc) {
|
|
|
|
ret = qcrypto_cipher_encrypt(s->cipher,
|
|
|
|
in_buf,
|
|
|
|
out_buf,
|
|
|
|
512,
|
|
|
|
errp);
|
|
|
|
} else {
|
|
|
|
ret = qcrypto_cipher_decrypt(s->cipher,
|
|
|
|
in_buf,
|
|
|
|
out_buf,
|
|
|
|
512,
|
|
|
|
errp);
|
|
|
|
}
|
|
|
|
if (ret < 0) {
|
|
|
|
return -1;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
sector_num++;
|
|
|
|
in_buf += 512;
|
|
|
|
out_buf += 512;
|
|
|
|
}
|
2015-07-01 19:10:37 +02:00
|
|
|
return 0;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* 'allocate' is:
|
|
|
|
*
|
|
|
|
* 0 to not allocate.
|
|
|
|
*
|
|
|
|
* 1 to allocate a normal cluster (for sector indexes 'n_start' to
|
|
|
|
* 'n_end')
|
|
|
|
*
|
|
|
|
* 2 to allocate a compressed cluster of size
|
|
|
|
* 'compressed_size'. 'compressed_size' must be > 0 and <
|
2007-09-16 23:08:06 +02:00
|
|
|
* cluster_size
|
2004-08-01 23:59:26 +02:00
|
|
|
*
|
|
|
|
* return 0 if not allocated.
|
|
|
|
*/
|
|
|
|
static uint64_t get_cluster_offset(BlockDriverState *bs,
|
|
|
|
uint64_t offset, int allocate,
|
|
|
|
int compressed_size,
|
|
|
|
int n_start, int n_end)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int min_index, i, j, l1_index, l2_index;
|
|
|
|
uint64_t l2_offset, *l2_table, cluster_offset, tmp;
|
|
|
|
uint32_t min_count;
|
|
|
|
int new_l2_table;
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
l1_index = offset >> (s->l2_bits + s->cluster_bits);
|
|
|
|
l2_offset = s->l1_table[l1_index];
|
|
|
|
new_l2_table = 0;
|
|
|
|
if (!l2_offset) {
|
|
|
|
if (!allocate)
|
|
|
|
return 0;
|
|
|
|
/* allocate a new l2 entry */
|
2015-06-16 14:19:22 +02:00
|
|
|
l2_offset = bdrv_getlength(bs->file->bs);
|
2004-08-01 23:59:26 +02:00
|
|
|
/* round to cluster size */
|
|
|
|
l2_offset = (l2_offset + s->cluster_size - 1) & ~(s->cluster_size - 1);
|
|
|
|
/* update the L1 entry */
|
|
|
|
s->l1_table[l1_index] = l2_offset;
|
|
|
|
tmp = cpu_to_be64(l2_offset);
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite_sync(bs->file,
|
2010-06-18 16:11:53 +02:00
|
|
|
s->l1_table_offset + l1_index * sizeof(tmp),
|
|
|
|
&tmp, sizeof(tmp)) < 0)
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
|
|
|
new_l2_table = 1;
|
|
|
|
}
|
|
|
|
for(i = 0; i < L2_CACHE_SIZE; i++) {
|
|
|
|
if (l2_offset == s->l2_cache_offsets[i]) {
|
|
|
|
/* increment the hit count */
|
|
|
|
if (++s->l2_cache_counts[i] == 0xffffffff) {
|
|
|
|
for(j = 0; j < L2_CACHE_SIZE; j++) {
|
|
|
|
s->l2_cache_counts[j] >>= 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
l2_table = s->l2_cache + (i << s->l2_bits);
|
|
|
|
goto found;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* not found: load a new entry in the least used one */
|
|
|
|
min_index = 0;
|
|
|
|
min_count = 0xffffffff;
|
|
|
|
for(i = 0; i < L2_CACHE_SIZE; i++) {
|
|
|
|
if (s->l2_cache_counts[i] < min_count) {
|
|
|
|
min_count = s->l2_cache_counts[i];
|
|
|
|
min_index = i;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
l2_table = s->l2_cache + (min_index << s->l2_bits);
|
|
|
|
if (new_l2_table) {
|
|
|
|
memset(l2_table, 0, s->l2_size * sizeof(uint64_t));
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite_sync(bs->file, l2_offset, l2_table,
|
2010-06-18 16:11:53 +02:00
|
|
|
s->l2_size * sizeof(uint64_t)) < 0)
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
|
|
|
} else {
|
2016-06-20 18:24:02 +02:00
|
|
|
if (bdrv_pread(bs->file, l2_offset, l2_table,
|
2015-06-16 14:19:22 +02:00
|
|
|
s->l2_size * sizeof(uint64_t)) !=
|
2004-08-01 23:59:26 +02:00
|
|
|
s->l2_size * sizeof(uint64_t))
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
s->l2_cache_offsets[min_index] = l2_offset;
|
|
|
|
s->l2_cache_counts[min_index] = 1;
|
|
|
|
found:
|
|
|
|
l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
|
|
|
|
cluster_offset = be64_to_cpu(l2_table[l2_index]);
|
2007-09-16 23:08:06 +02:00
|
|
|
if (!cluster_offset ||
|
2004-08-01 23:59:26 +02:00
|
|
|
((cluster_offset & QCOW_OFLAG_COMPRESSED) && allocate == 1)) {
|
|
|
|
if (!allocate)
|
|
|
|
return 0;
|
|
|
|
/* allocate a new cluster */
|
|
|
|
if ((cluster_offset & QCOW_OFLAG_COMPRESSED) &&
|
|
|
|
(n_end - n_start) < s->cluster_sectors) {
|
|
|
|
/* if the cluster is already compressed, we must
|
|
|
|
decompress it in the case it is not completely
|
|
|
|
overwritten */
|
2010-04-14 14:17:38 +02:00
|
|
|
if (decompress_cluster(bs, cluster_offset) < 0)
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
2015-06-16 14:19:22 +02:00
|
|
|
cluster_offset = bdrv_getlength(bs->file->bs);
|
2007-09-16 23:08:06 +02:00
|
|
|
cluster_offset = (cluster_offset + s->cluster_size - 1) &
|
2004-08-01 23:59:26 +02:00
|
|
|
~(s->cluster_size - 1);
|
|
|
|
/* write the cluster content */
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite(bs->file, cluster_offset, s->cluster_cache,
|
2015-06-16 14:19:22 +02:00
|
|
|
s->cluster_size) !=
|
2004-08-01 23:59:26 +02:00
|
|
|
s->cluster_size)
|
|
|
|
return -1;
|
|
|
|
} else {
|
2015-06-16 14:19:22 +02:00
|
|
|
cluster_offset = bdrv_getlength(bs->file->bs);
|
2009-01-08 20:29:03 +01:00
|
|
|
if (allocate == 1) {
|
|
|
|
/* round to cluster size */
|
|
|
|
cluster_offset = (cluster_offset + s->cluster_size - 1) &
|
|
|
|
~(s->cluster_size - 1);
|
2015-06-16 14:19:22 +02:00
|
|
|
bdrv_truncate(bs->file->bs, cluster_offset + s->cluster_size);
|
2009-01-08 20:29:03 +01:00
|
|
|
/* if encrypted, we must initialize the cluster
|
|
|
|
content which won't be written */
|
qcow2/qcow: protect against uninitialized encryption key
When a qcow[2] file is opened, if the header reports an
encryption method, this is used to set the 'crypt_method_header'
field on the BDRVQcow[2]State struct, and the 'encrypted' flag
in the BDRVState struct.
When doing I/O operations, the 'crypt_method' field on the
BDRVQcow[2]State struct is checked to determine if encryption
needs to be applied.
The crypt_method_header value is copied into crypt_method when
the bdrv_set_key() method is called.
The QEMU code which opens a block device is expected to always
do a check
if (bdrv_is_encrypted(bs)) {
bdrv_set_key(bs, ....key...);
}
If code forgets to do this, then 'crypt_method' is never set
and so when I/O is performed, QEMU writes plain text data
into a sector which is expected to contain cipher text, or
when reading, will return cipher text instead of plain
text.
Change the qcow[2] code to consult bs->encrypted when deciding
whether encryption is required, and assert(s->crypt_method)
to protect against cases where the caller forgets to set the
encryption key.
Also put an assert in the set_key methods to protect against
the case where the caller sets an encryption key on a block
device that does not have encryption
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2015-05-12 18:09:18 +02:00
|
|
|
if (bs->encrypted &&
|
2009-01-08 20:29:03 +01:00
|
|
|
(n_end - n_start) < s->cluster_sectors) {
|
|
|
|
uint64_t start_sect;
|
2015-07-01 19:10:37 +02:00
|
|
|
assert(s->cipher);
|
2009-01-08 20:29:03 +01:00
|
|
|
start_sect = (offset & ~(s->cluster_size - 1)) >> 9;
|
|
|
|
memset(s->cluster_data + 512, 0x00, 512);
|
|
|
|
for(i = 0; i < s->cluster_sectors; i++) {
|
|
|
|
if (i < n_start || i >= n_end) {
|
2015-07-01 19:10:37 +02:00
|
|
|
Error *err = NULL;
|
|
|
|
if (encrypt_sectors(s, start_sect + i,
|
|
|
|
s->cluster_data,
|
|
|
|
s->cluster_data + 512, 1,
|
|
|
|
true, &err) < 0) {
|
|
|
|
error_free(err);
|
|
|
|
errno = EIO;
|
|
|
|
return -1;
|
|
|
|
}
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite(bs->file,
|
2015-06-16 14:19:22 +02:00
|
|
|
cluster_offset + i * 512,
|
2009-01-08 20:29:03 +01:00
|
|
|
s->cluster_data, 512) != 512)
|
|
|
|
return -1;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
}
|
2009-01-08 20:29:03 +01:00
|
|
|
} else if (allocate == 2) {
|
|
|
|
cluster_offset |= QCOW_OFLAG_COMPRESSED |
|
|
|
|
(uint64_t)compressed_size << (63 - s->cluster_bits);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
/* update L2 table */
|
|
|
|
tmp = cpu_to_be64(cluster_offset);
|
|
|
|
l2_table[l2_index] = tmp;
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite_sync(bs->file, l2_offset + l2_index * sizeof(tmp),
|
2010-06-18 16:11:53 +02:00
|
|
|
&tmp, sizeof(tmp)) < 0)
|
2004-08-01 23:59:26 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
return cluster_offset;
|
|
|
|
}
|
|
|
|
|
2013-09-04 19:00:28 +02:00
|
|
|
static int64_t coroutine_fn qcow_co_get_block_status(BlockDriverState *bs,
|
2016-01-26 04:58:48 +01:00
|
|
|
int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int index_in_cluster, n;
|
|
|
|
uint64_t cluster_offset;
|
|
|
|
|
2011-11-14 13:44:21 +01:00
|
|
|
qemu_co_mutex_lock(&s->lock);
|
2004-08-01 23:59:26 +02:00
|
|
|
cluster_offset = get_cluster_offset(bs, sector_num << 9, 0, 0, 0, 0);
|
2011-11-14 13:44:21 +01:00
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2004-08-01 23:59:26 +02:00
|
|
|
index_in_cluster = sector_num & (s->cluster_sectors - 1);
|
|
|
|
n = s->cluster_sectors - index_in_cluster;
|
|
|
|
if (n > nb_sectors)
|
|
|
|
n = nb_sectors;
|
|
|
|
*pnum = n;
|
2013-09-04 19:00:30 +02:00
|
|
|
if (!cluster_offset) {
|
|
|
|
return 0;
|
|
|
|
}
|
2015-07-01 19:10:37 +02:00
|
|
|
if ((cluster_offset & QCOW_OFLAG_COMPRESSED) || s->cipher) {
|
2013-09-04 19:00:30 +02:00
|
|
|
return BDRV_BLOCK_DATA;
|
|
|
|
}
|
|
|
|
cluster_offset |= (index_in_cluster << BDRV_SECTOR_BITS);
|
2016-01-26 04:58:49 +01:00
|
|
|
*file = bs->file->bs;
|
2013-09-04 19:00:30 +02:00
|
|
|
return BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | cluster_offset;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
static int decompress_buffer(uint8_t *out_buf, int out_buf_size,
|
|
|
|
const uint8_t *buf, int buf_size)
|
|
|
|
{
|
|
|
|
z_stream strm1, *strm = &strm1;
|
|
|
|
int ret, out_len;
|
|
|
|
|
|
|
|
memset(strm, 0, sizeof(*strm));
|
|
|
|
|
|
|
|
strm->next_in = (uint8_t *)buf;
|
|
|
|
strm->avail_in = buf_size;
|
|
|
|
strm->next_out = out_buf;
|
|
|
|
strm->avail_out = out_buf_size;
|
|
|
|
|
|
|
|
ret = inflateInit2(strm, -12);
|
|
|
|
if (ret != Z_OK)
|
|
|
|
return -1;
|
|
|
|
ret = inflate(strm, Z_FINISH);
|
|
|
|
out_len = strm->next_out - out_buf;
|
|
|
|
if ((ret != Z_STREAM_END && ret != Z_BUF_ERROR) ||
|
|
|
|
out_len != out_buf_size) {
|
|
|
|
inflateEnd(strm);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
inflateEnd(strm);
|
|
|
|
return 0;
|
|
|
|
}
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
static int decompress_cluster(BlockDriverState *bs, uint64_t cluster_offset)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2010-04-14 14:17:38 +02:00
|
|
|
BDRVQcowState *s = bs->opaque;
|
2004-08-01 23:59:26 +02:00
|
|
|
int ret, csize;
|
|
|
|
uint64_t coffset;
|
|
|
|
|
|
|
|
coffset = cluster_offset & s->cluster_offset_mask;
|
|
|
|
if (s->cluster_cache_offset != coffset) {
|
|
|
|
csize = cluster_offset >> (63 - s->cluster_bits);
|
|
|
|
csize &= (s->cluster_size - 1);
|
2016-06-20 18:24:02 +02:00
|
|
|
ret = bdrv_pread(bs->file, coffset, s->cluster_data, csize);
|
2007-09-16 23:08:06 +02:00
|
|
|
if (ret != csize)
|
2004-08-01 23:59:26 +02:00
|
|
|
return -1;
|
|
|
|
if (decompress_buffer(s->cluster_cache, s->cluster_size,
|
|
|
|
s->cluster_data, csize) < 0) {
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
s->cluster_cache_offset = coffset;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2011-11-10 09:23:22 +01:00
|
|
|
static coroutine_fn int qcow_co_readv(BlockDriverState *bs, int64_t sector_num,
|
2011-08-23 15:21:11 +02:00
|
|
|
int nb_sectors, QEMUIOVector *qiov)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int index_in_cluster;
|
2011-08-23 15:21:11 +02:00
|
|
|
int ret = 0, n;
|
2011-08-23 15:21:10 +02:00
|
|
|
uint64_t cluster_offset;
|
2011-08-23 15:21:09 +02:00
|
|
|
struct iovec hd_iov;
|
|
|
|
QEMUIOVector hd_qiov;
|
2011-08-23 15:21:11 +02:00
|
|
|
uint8_t *buf;
|
|
|
|
void *orig_buf;
|
2015-07-01 19:10:37 +02:00
|
|
|
Error *err = NULL;
|
2006-08-01 18:21:11 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
if (qiov->niov > 1) {
|
2014-05-20 13:36:05 +02:00
|
|
|
buf = orig_buf = qemu_try_blockalign(bs, qiov->size);
|
|
|
|
if (buf == NULL) {
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2011-08-23 15:21:11 +02:00
|
|
|
} else {
|
|
|
|
orig_buf = NULL;
|
|
|
|
buf = (uint8_t *)qiov->iov->iov_base;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
|
|
|
|
while (nb_sectors != 0) {
|
|
|
|
/* prepare next request */
|
|
|
|
cluster_offset = get_cluster_offset(bs, sector_num << 9,
|
|
|
|
0, 0, 0, 0);
|
|
|
|
index_in_cluster = sector_num & (s->cluster_sectors - 1);
|
|
|
|
n = s->cluster_sectors - index_in_cluster;
|
|
|
|
if (n > nb_sectors) {
|
|
|
|
n = nb_sectors;
|
|
|
|
}
|
2006-08-07 04:38:06 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
if (!cluster_offset) {
|
2015-06-17 14:55:21 +02:00
|
|
|
if (bs->backing) {
|
2011-08-23 15:21:11 +02:00
|
|
|
/* read from the base image */
|
|
|
|
hd_iov.iov_base = (void *)buf;
|
|
|
|
hd_iov.iov_len = n * 512;
|
|
|
|
qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
|
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2016-05-24 17:21:22 +02:00
|
|
|
ret = bdrv_co_readv(bs->backing, sector_num, n, &hd_qiov);
|
2011-08-23 15:21:11 +02:00
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/* Note: in this case, no need to wait */
|
|
|
|
memset(buf, 0, 512 * n);
|
|
|
|
}
|
|
|
|
} else if (cluster_offset & QCOW_OFLAG_COMPRESSED) {
|
|
|
|
/* add AIO support for compressed blocks ? */
|
|
|
|
if (decompress_cluster(bs, cluster_offset) < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
memcpy(buf,
|
|
|
|
s->cluster_cache + index_in_cluster * 512, 512 * n);
|
|
|
|
} else {
|
|
|
|
if ((cluster_offset & 511) != 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
hd_iov.iov_base = (void *)buf;
|
2011-08-23 15:21:09 +02:00
|
|
|
hd_iov.iov_len = n * 512;
|
|
|
|
qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
|
2011-07-15 16:27:42 +02:00
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2016-05-24 17:21:22 +02:00
|
|
|
ret = bdrv_co_readv(bs->file,
|
2011-08-23 15:21:11 +02:00
|
|
|
(cluster_offset >> 9) + index_in_cluster,
|
2011-08-23 15:21:09 +02:00
|
|
|
n, &hd_qiov);
|
2011-07-15 16:27:42 +02:00
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
if (ret < 0) {
|
2011-08-23 15:21:11 +02:00
|
|
|
break;
|
|
|
|
}
|
qcow2/qcow: protect against uninitialized encryption key
When a qcow[2] file is opened, if the header reports an
encryption method, this is used to set the 'crypt_method_header'
field on the BDRVQcow[2]State struct, and the 'encrypted' flag
in the BDRVState struct.
When doing I/O operations, the 'crypt_method' field on the
BDRVQcow[2]State struct is checked to determine if encryption
needs to be applied.
The crypt_method_header value is copied into crypt_method when
the bdrv_set_key() method is called.
The QEMU code which opens a block device is expected to always
do a check
if (bdrv_is_encrypted(bs)) {
bdrv_set_key(bs, ....key...);
}
If code forgets to do this, then 'crypt_method' is never set
and so when I/O is performed, QEMU writes plain text data
into a sector which is expected to contain cipher text, or
when reading, will return cipher text instead of plain
text.
Change the qcow[2] code to consult bs->encrypted when deciding
whether encryption is required, and assert(s->crypt_method)
to protect against cases where the caller forgets to set the
encryption key.
Also put an assert in the set_key methods to protect against
the case where the caller sets an encryption key on a block
device that does not have encryption
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2015-05-12 18:09:18 +02:00
|
|
|
if (bs->encrypted) {
|
2015-07-01 19:10:37 +02:00
|
|
|
assert(s->cipher);
|
|
|
|
if (encrypt_sectors(s, sector_num, buf, buf,
|
|
|
|
n, false, &err) < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2011-02-19 22:18:12 +01:00
|
|
|
}
|
|
|
|
}
|
2011-08-23 15:21:11 +02:00
|
|
|
ret = 0;
|
2009-04-07 20:43:24 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
nb_sectors -= n;
|
|
|
|
sector_num += n;
|
|
|
|
buf += n * 512;
|
2011-08-23 15:21:10 +02:00
|
|
|
}
|
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
done:
|
2011-07-15 16:27:42 +02:00
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
if (qiov->niov > 1) {
|
allow qemu_iovec_from_buffer() to specify offset from which to start copying
Similar to
qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
int c, size_t bytes);
the new prototype is:
qemu_iovec_from_buf(QEMUIOVector *qiov, size_t offset,
const void *buf, size_t bytes);
The processing starts at offset bytes within qiov.
This way, we may copy a bounce buffer directly to
a middle of qiov.
This is exactly the same function as iov_from_buf() from
iov.c, so use the existing implementation and rename it
to qemu_iovec_from_buf() to be shorter and to match the
utility function.
As with utility implementation, we now assert that the
offset is inside actual iovec. Nothing changed for
current callers, because `offset' parameter is new.
While at it, stop using "bounce-qiov" in block/qcow2.c
and copy decrypted data directly from cluster_data
instead of recreating a temp qiov for doing that.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-06-07 18:17:55 +02:00
|
|
|
qemu_iovec_from_buf(qiov, 0, orig_buf, qiov->size);
|
2011-08-23 15:21:11 +02:00
|
|
|
qemu_vfree(orig_buf);
|
2011-06-07 15:20:44 +02:00
|
|
|
}
|
|
|
|
|
2011-07-15 16:27:42 +02:00
|
|
|
return ret;
|
2011-08-23 15:21:11 +02:00
|
|
|
|
|
|
|
fail:
|
2015-07-01 19:10:37 +02:00
|
|
|
error_free(err);
|
2011-08-23 15:21:11 +02:00
|
|
|
ret = -EIO;
|
|
|
|
goto done;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2011-11-10 09:23:22 +01:00
|
|
|
static coroutine_fn int qcow_co_writev(BlockDriverState *bs, int64_t sector_num,
|
2011-08-23 15:21:11 +02:00
|
|
|
int nb_sectors, QEMUIOVector *qiov)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
int index_in_cluster;
|
|
|
|
uint64_t cluster_offset;
|
|
|
|
const uint8_t *src_buf;
|
2011-08-23 15:21:11 +02:00
|
|
|
int ret = 0, n;
|
2011-08-23 15:21:09 +02:00
|
|
|
uint8_t *cluster_data = NULL;
|
|
|
|
struct iovec hd_iov;
|
|
|
|
QEMUIOVector hd_qiov;
|
2011-08-23 15:21:11 +02:00
|
|
|
uint8_t *buf;
|
|
|
|
void *orig_buf;
|
2006-08-07 04:38:06 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
s->cluster_cache_offset = -1; /* disable compressed cache */
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
if (qiov->niov > 1) {
|
2014-05-20 13:36:05 +02:00
|
|
|
buf = orig_buf = qemu_try_blockalign(bs, qiov->size);
|
|
|
|
if (buf == NULL) {
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2012-06-07 18:21:06 +02:00
|
|
|
qemu_iovec_to_buf(qiov, 0, buf, qiov->size);
|
2006-08-01 18:21:11 +02:00
|
|
|
} else {
|
2011-08-23 15:21:11 +02:00
|
|
|
orig_buf = NULL;
|
|
|
|
buf = (uint8_t *)qiov->iov->iov_base;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
2009-04-07 20:43:20 +02:00
|
|
|
|
2011-07-15 16:27:42 +02:00
|
|
|
qemu_co_mutex_lock(&s->lock);
|
2011-08-23 15:21:10 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
while (nb_sectors != 0) {
|
2006-08-01 18:21:11 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
index_in_cluster = sector_num & (s->cluster_sectors - 1);
|
|
|
|
n = s->cluster_sectors - index_in_cluster;
|
|
|
|
if (n > nb_sectors) {
|
|
|
|
n = nb_sectors;
|
|
|
|
}
|
|
|
|
cluster_offset = get_cluster_offset(bs, sector_num << 9, 1, 0,
|
|
|
|
index_in_cluster,
|
|
|
|
index_in_cluster + n);
|
|
|
|
if (!cluster_offset || (cluster_offset & 511) != 0) {
|
|
|
|
ret = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
qcow2/qcow: protect against uninitialized encryption key
When a qcow[2] file is opened, if the header reports an
encryption method, this is used to set the 'crypt_method_header'
field on the BDRVQcow[2]State struct, and the 'encrypted' flag
in the BDRVState struct.
When doing I/O operations, the 'crypt_method' field on the
BDRVQcow[2]State struct is checked to determine if encryption
needs to be applied.
The crypt_method_header value is copied into crypt_method when
the bdrv_set_key() method is called.
The QEMU code which opens a block device is expected to always
do a check
if (bdrv_is_encrypted(bs)) {
bdrv_set_key(bs, ....key...);
}
If code forgets to do this, then 'crypt_method' is never set
and so when I/O is performed, QEMU writes plain text data
into a sector which is expected to contain cipher text, or
when reading, will return cipher text instead of plain
text.
Change the qcow[2] code to consult bs->encrypted when deciding
whether encryption is required, and assert(s->crypt_method)
to protect against cases where the caller forgets to set the
encryption key.
Also put an assert in the set_key methods to protect against
the case where the caller sets an encryption key on a block
device that does not have encryption
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2015-05-12 18:09:18 +02:00
|
|
|
if (bs->encrypted) {
|
2015-07-01 19:10:37 +02:00
|
|
|
Error *err = NULL;
|
|
|
|
assert(s->cipher);
|
2011-08-23 15:21:11 +02:00
|
|
|
if (!cluster_data) {
|
|
|
|
cluster_data = g_malloc0(s->cluster_size);
|
|
|
|
}
|
2015-07-01 19:10:37 +02:00
|
|
|
if (encrypt_sectors(s, sector_num, cluster_data, buf,
|
|
|
|
n, true, &err) < 0) {
|
|
|
|
error_free(err);
|
|
|
|
ret = -EIO;
|
|
|
|
break;
|
|
|
|
}
|
2011-08-23 15:21:11 +02:00
|
|
|
src_buf = cluster_data;
|
|
|
|
} else {
|
|
|
|
src_buf = buf;
|
|
|
|
}
|
2006-08-01 18:21:11 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
hd_iov.iov_base = (void *)src_buf;
|
|
|
|
hd_iov.iov_len = n * 512;
|
|
|
|
qemu_iovec_init_external(&hd_qiov, &hd_iov, 1);
|
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2016-05-24 17:21:22 +02:00
|
|
|
ret = bdrv_co_writev(bs->file,
|
2011-08-23 15:21:11 +02:00
|
|
|
(cluster_offset >> 9) + index_in_cluster,
|
|
|
|
n, &hd_qiov);
|
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
if (ret < 0) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
ret = 0;
|
2009-05-25 15:45:37 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
nb_sectors -= n;
|
|
|
|
sector_num += n;
|
|
|
|
buf += n * 512;
|
|
|
|
}
|
2011-07-15 16:27:42 +02:00
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2011-08-23 15:21:11 +02:00
|
|
|
if (qiov->niov > 1) {
|
|
|
|
qemu_vfree(orig_buf);
|
2011-06-07 15:20:44 +02:00
|
|
|
}
|
2011-10-07 07:32:47 +02:00
|
|
|
g_free(cluster_data);
|
2011-06-07 15:20:44 +02:00
|
|
|
|
2011-07-15 16:27:42 +02:00
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2004-09-18 21:32:11 +02:00
|
|
|
static void qcow_close(BlockDriverState *bs)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
2011-11-22 16:44:45 +01:00
|
|
|
|
2015-07-01 19:10:37 +02:00
|
|
|
qcrypto_cipher_free(s->cipher);
|
|
|
|
s->cipher = NULL;
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(s->l1_table);
|
2014-05-20 13:36:05 +02:00
|
|
|
qemu_vfree(s->l2_cache);
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(s->cluster_cache);
|
|
|
|
g_free(s->cluster_data);
|
2011-11-22 16:44:45 +01:00
|
|
|
|
|
|
|
migrate_del_blocker(s->migration_blocker);
|
|
|
|
error_free(s->migration_blocker);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2014-06-05 11:20:57 +02:00
|
|
|
static int qcow_create(const char *filename, QemuOpts *opts, Error **errp)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2011-11-21 08:40:39 +01:00
|
|
|
int header_size, backing_filename_len, l1_size, shift, i;
|
2004-08-01 23:59:26 +02:00
|
|
|
QCowHeader header;
|
2011-11-21 08:40:39 +01:00
|
|
|
uint8_t *tmp;
|
2009-05-18 16:42:10 +02:00
|
|
|
int64_t total_size = 0;
|
2014-06-05 11:20:57 +02:00
|
|
|
char *backing_file = NULL;
|
2009-05-18 16:42:10 +02:00
|
|
|
int flags = 0;
|
2013-09-05 14:45:29 +02:00
|
|
|
Error *local_err = NULL;
|
2010-01-20 00:56:12 +01:00
|
|
|
int ret;
|
2016-03-08 15:57:05 +01:00
|
|
|
BlockBackend *qcow_blk;
|
2009-05-18 16:42:10 +02:00
|
|
|
|
|
|
|
/* Read out options */
|
2014-09-10 11:05:46 +02:00
|
|
|
total_size = ROUND_UP(qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0),
|
|
|
|
BDRV_SECTOR_SIZE);
|
2014-06-05 11:20:57 +02:00
|
|
|
backing_file = qemu_opt_get_del(opts, BLOCK_OPT_BACKING_FILE);
|
|
|
|
if (qemu_opt_get_bool_del(opts, BLOCK_OPT_ENCRYPT, false)) {
|
|
|
|
flags |= BLOCK_FLAG_ENCRYPT;
|
2009-05-18 16:42:10 +02:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2014-06-05 11:21:11 +02:00
|
|
|
ret = bdrv_create_file(filename, opts, &local_err);
|
2011-11-21 08:40:39 +01:00
|
|
|
if (ret < 0) {
|
2014-02-17 14:43:58 +01:00
|
|
|
error_propagate(errp, local_err);
|
2014-06-05 11:20:57 +02:00
|
|
|
goto cleanup;
|
2011-11-21 08:40:39 +01:00
|
|
|
}
|
|
|
|
|
2016-03-16 19:54:38 +01:00
|
|
|
qcow_blk = blk_new_open(filename, NULL, NULL,
|
2016-03-15 14:34:37 +01:00
|
|
|
BDRV_O_RDWR | BDRV_O_PROTOCOL, &local_err);
|
2016-03-08 15:57:05 +01:00
|
|
|
if (qcow_blk == NULL) {
|
2014-02-17 14:43:58 +01:00
|
|
|
error_propagate(errp, local_err);
|
2016-03-08 15:57:05 +01:00
|
|
|
ret = -EIO;
|
2014-06-05 11:20:57 +02:00
|
|
|
goto cleanup;
|
2011-11-21 08:40:39 +01:00
|
|
|
}
|
|
|
|
|
2016-03-08 15:57:05 +01:00
|
|
|
blk_set_allow_write_beyond_eof(qcow_blk, true);
|
|
|
|
|
|
|
|
ret = blk_truncate(qcow_blk, 0);
|
2011-11-21 08:40:39 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
memset(&header, 0, sizeof(header));
|
|
|
|
header.magic = cpu_to_be32(QCOW_MAGIC);
|
|
|
|
header.version = cpu_to_be32(QCOW_VERSION);
|
2014-09-10 11:05:46 +02:00
|
|
|
header.size = cpu_to_be64(total_size);
|
2004-08-01 23:59:26 +02:00
|
|
|
header_size = sizeof(header);
|
|
|
|
backing_filename_len = 0;
|
|
|
|
if (backing_file) {
|
2008-03-18 07:52:48 +01:00
|
|
|
if (strcmp(backing_file, "fat:")) {
|
|
|
|
header.backing_file_offset = cpu_to_be64(header_size);
|
|
|
|
backing_filename_len = strlen(backing_file);
|
|
|
|
header.backing_file_size = cpu_to_be32(backing_filename_len);
|
|
|
|
header_size += backing_filename_len;
|
|
|
|
} else {
|
|
|
|
/* special backing file for vvfat */
|
|
|
|
backing_file = NULL;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
header.cluster_bits = 9; /* 512 byte cluster to avoid copying
|
2014-03-24 09:30:17 +01:00
|
|
|
unmodified sectors */
|
2004-08-01 23:59:26 +02:00
|
|
|
header.l2_bits = 12; /* 32 KB L2 tables */
|
|
|
|
} else {
|
|
|
|
header.cluster_bits = 12; /* 4 KB clusters */
|
|
|
|
header.l2_bits = 9; /* 4 KB L2 tables */
|
|
|
|
}
|
|
|
|
header_size = (header_size + 7) & ~7;
|
|
|
|
shift = header.cluster_bits + header.l2_bits;
|
2014-09-10 11:05:46 +02:00
|
|
|
l1_size = (total_size + (1LL << shift) - 1) >> shift;
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
header.l1_table_offset = cpu_to_be64(header_size);
|
2007-09-16 23:59:02 +02:00
|
|
|
if (flags & BLOCK_FLAG_ENCRYPT) {
|
2004-08-01 23:59:26 +02:00
|
|
|
header.crypt_method = cpu_to_be32(QCOW_CRYPT_AES);
|
|
|
|
} else {
|
|
|
|
header.crypt_method = cpu_to_be32(QCOW_CRYPT_NONE);
|
|
|
|
}
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/* write all the data */
|
2016-05-06 18:26:27 +02:00
|
|
|
ret = blk_pwrite(qcow_blk, 0, &header, sizeof(header), 0);
|
2010-01-20 00:56:12 +01:00
|
|
|
if (ret != sizeof(header)) {
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
if (backing_file) {
|
2016-03-08 15:57:05 +01:00
|
|
|
ret = blk_pwrite(qcow_blk, sizeof(header),
|
2016-05-06 18:26:27 +02:00
|
|
|
backing_file, backing_filename_len, 0);
|
2010-01-20 00:56:12 +01:00
|
|
|
if (ret != backing_filename_len) {
|
|
|
|
goto exit;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2011-11-21 08:40:39 +01:00
|
|
|
|
|
|
|
tmp = g_malloc0(BDRV_SECTOR_SIZE);
|
2016-05-31 18:35:52 +02:00
|
|
|
for (i = 0; i < DIV_ROUND_UP(sizeof(uint64_t) * l1_size, BDRV_SECTOR_SIZE);
|
|
|
|
i++) {
|
2016-05-06 18:26:27 +02:00
|
|
|
ret = blk_pwrite(qcow_blk, header_size + BDRV_SECTOR_SIZE * i,
|
|
|
|
tmp, BDRV_SECTOR_SIZE, 0);
|
2011-11-21 08:40:39 +01:00
|
|
|
if (ret != BDRV_SECTOR_SIZE) {
|
|
|
|
g_free(tmp);
|
2010-01-20 00:56:12 +01:00
|
|
|
goto exit;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2010-01-20 00:56:12 +01:00
|
|
|
|
2011-11-21 08:40:39 +01:00
|
|
|
g_free(tmp);
|
2010-01-20 00:56:12 +01:00
|
|
|
ret = 0;
|
|
|
|
exit:
|
2016-03-08 15:57:05 +01:00
|
|
|
blk_unref(qcow_blk);
|
2014-06-05 11:20:57 +02:00
|
|
|
cleanup:
|
|
|
|
g_free(backing_file);
|
2010-01-20 00:56:12 +01:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2006-08-05 23:32:10 +02:00
|
|
|
static int qcow_make_empty(BlockDriverState *bs)
|
2005-12-18 19:28:15 +01:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
uint32_t l1_length = s->l1_size * sizeof(uint64_t);
|
2006-08-01 18:21:11 +02:00
|
|
|
int ret;
|
2005-12-18 19:28:15 +01:00
|
|
|
|
|
|
|
memset(s->l1_table, 0, l1_length);
|
2016-06-20 20:09:15 +02:00
|
|
|
if (bdrv_pwrite_sync(bs->file, s->l1_table_offset, s->l1_table,
|
2010-06-18 16:11:53 +02:00
|
|
|
l1_length) < 0)
|
|
|
|
return -1;
|
2015-06-16 14:19:22 +02:00
|
|
|
ret = bdrv_truncate(bs->file->bs, s->l1_table_offset + l1_length);
|
2006-08-01 18:21:11 +02:00
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
2005-12-18 19:28:15 +01:00
|
|
|
|
|
|
|
memset(s->l2_cache, 0, s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
|
|
|
|
memset(s->l2_cache_offsets, 0, L2_CACHE_SIZE * sizeof(uint64_t));
|
|
|
|
memset(s->l2_cache_counts, 0, L2_CACHE_SIZE * sizeof(uint32_t));
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-05-31 14:42:08 +02:00
|
|
|
typedef struct QcowWriteCo {
|
|
|
|
BlockDriverState *bs;
|
|
|
|
int64_t sector_num;
|
|
|
|
const uint8_t *buf;
|
|
|
|
int nb_sectors;
|
|
|
|
int ret;
|
|
|
|
} QcowWriteCo;
|
|
|
|
|
|
|
|
static void qcow_write_co_entry(void *opaque)
|
|
|
|
{
|
|
|
|
QcowWriteCo *co = opaque;
|
|
|
|
QEMUIOVector qiov;
|
|
|
|
|
|
|
|
struct iovec iov = (struct iovec) {
|
|
|
|
.iov_base = (uint8_t*) co->buf,
|
|
|
|
.iov_len = co->nb_sectors * BDRV_SECTOR_SIZE,
|
|
|
|
};
|
|
|
|
qemu_iovec_init_external(&qiov, &iov, 1);
|
|
|
|
|
|
|
|
co->ret = qcow_co_writev(co->bs, co->sector_num, co->nb_sectors, &qiov);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Wrapper for non-coroutine contexts */
|
|
|
|
static int qcow_write(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
const uint8_t *buf, int nb_sectors)
|
|
|
|
{
|
|
|
|
Coroutine *co;
|
|
|
|
AioContext *aio_context = bdrv_get_aio_context(bs);
|
|
|
|
QcowWriteCo data = {
|
|
|
|
.bs = bs,
|
|
|
|
.sector_num = sector_num,
|
|
|
|
.buf = buf,
|
|
|
|
.nb_sectors = nb_sectors,
|
|
|
|
.ret = -EINPROGRESS,
|
|
|
|
};
|
coroutine: move entry argument to qemu_coroutine_create
In practice the entry argument is always known at creation time, and
it is confusing that sometimes qemu_coroutine_enter is used with a
non-NULL argument to re-enter a coroutine (this happens in
block/sheepdog.c and tests/test-coroutine.c). So pass the opaque value
at creation time, for consistency with e.g. aio_bh_new.
Mostly done with the following semantic patch:
@ entry1 @
expression entry, arg, co;
@@
- co = qemu_coroutine_create(entry);
+ co = qemu_coroutine_create(entry, arg);
...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);
@ entry2 @
expression entry, arg;
identifier co;
@@
- Coroutine *co = qemu_coroutine_create(entry);
+ Coroutine *co = qemu_coroutine_create(entry, arg);
...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);
@ entry3 @
expression entry, arg;
@@
- qemu_coroutine_enter(qemu_coroutine_create(entry), arg);
+ qemu_coroutine_enter(qemu_coroutine_create(entry, arg));
@ reentry @
expression co;
@@
- qemu_coroutine_enter(co, NULL);
+ qemu_coroutine_enter(co);
except for the aforementioned few places where the semantic patch
stumbled (as expected) and for test_co_queue, which would otherwise
produce an uninitialized variable warning.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2016-07-04 19:10:01 +02:00
|
|
|
co = qemu_coroutine_create(qcow_write_co_entry, &data);
|
|
|
|
qemu_coroutine_enter(co);
|
2016-05-31 14:42:08 +02:00
|
|
|
while (data.ret == -EINPROGRESS) {
|
|
|
|
aio_poll(aio_context, true);
|
|
|
|
}
|
|
|
|
return data.ret;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/* XXX: put compressed sectors first, then all the cluster aligned
|
|
|
|
tables to avoid losing bytes in alignment */
|
2007-09-16 23:08:06 +02:00
|
|
|
static int qcow_write_compressed(BlockDriverState *bs, int64_t sector_num,
|
2006-08-05 23:32:10 +02:00
|
|
|
const uint8_t *buf, int nb_sectors)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
z_stream strm;
|
|
|
|
int ret, out_len;
|
|
|
|
uint8_t *out_buf;
|
|
|
|
uint64_t cluster_offset;
|
|
|
|
|
2013-04-15 17:17:32 +02:00
|
|
|
if (nb_sectors != s->cluster_sectors) {
|
|
|
|
ret = -EINVAL;
|
|
|
|
|
|
|
|
/* Zero-pad last write if image size is not cluster aligned */
|
|
|
|
if (sector_num + nb_sectors == bs->total_sectors &&
|
|
|
|
nb_sectors < s->cluster_sectors) {
|
|
|
|
uint8_t *pad_buf = qemu_blockalign(bs, s->cluster_size);
|
|
|
|
memset(pad_buf, 0, s->cluster_size);
|
|
|
|
memcpy(pad_buf, buf, nb_sectors * BDRV_SECTOR_SIZE);
|
|
|
|
ret = qcow_write_compressed(bs, sector_num,
|
|
|
|
pad_buf, s->cluster_sectors);
|
|
|
|
qemu_vfree(pad_buf);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2016-07-14 18:59:25 +02:00
|
|
|
out_buf = g_malloc(s->cluster_size);
|
2004-08-01 23:59:26 +02:00
|
|
|
|
|
|
|
/* best compression, small window, no zlib header */
|
|
|
|
memset(&strm, 0, sizeof(strm));
|
|
|
|
ret = deflateInit2(&strm, Z_DEFAULT_COMPRESSION,
|
2007-09-16 23:08:06 +02:00
|
|
|
Z_DEFLATED, -12,
|
2004-08-01 23:59:26 +02:00
|
|
|
9, Z_DEFAULT_STRATEGY);
|
|
|
|
if (ret != 0) {
|
2011-10-26 11:21:50 +02:00
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
strm.avail_in = s->cluster_size;
|
|
|
|
strm.next_in = (uint8_t *)buf;
|
|
|
|
strm.avail_out = s->cluster_size;
|
|
|
|
strm.next_out = out_buf;
|
|
|
|
|
|
|
|
ret = deflate(&strm, Z_FINISH);
|
|
|
|
if (ret != Z_STREAM_END && ret != Z_OK) {
|
|
|
|
deflateEnd(&strm);
|
2011-10-26 11:21:50 +02:00
|
|
|
ret = -EINVAL;
|
|
|
|
goto fail;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
out_len = strm.next_out - out_buf;
|
|
|
|
|
|
|
|
deflateEnd(&strm);
|
|
|
|
|
|
|
|
if (ret != Z_STREAM_END || out_len >= s->cluster_size) {
|
|
|
|
/* could not compress: write normal cluster */
|
2016-05-31 14:42:08 +02:00
|
|
|
ret = qcow_write(bs, sector_num, buf, s->cluster_sectors);
|
2011-10-26 11:21:50 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
} else {
|
2007-09-16 23:08:06 +02:00
|
|
|
cluster_offset = get_cluster_offset(bs, sector_num << 9, 2,
|
2004-08-01 23:59:26 +02:00
|
|
|
out_len, 0, 0);
|
2011-10-26 11:21:50 +02:00
|
|
|
if (cluster_offset == 0) {
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
cluster_offset &= s->cluster_offset_mask;
|
2016-06-20 20:09:15 +02:00
|
|
|
ret = bdrv_pwrite(bs->file, cluster_offset, out_buf, out_len);
|
2011-10-26 11:21:50 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
}
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2011-10-26 11:21:50 +02:00
|
|
|
ret = 0;
|
|
|
|
fail:
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(out_buf);
|
2011-10-26 11:21:50 +02:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2006-08-05 23:32:10 +02:00
|
|
|
static int qcow_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
|
|
|
|
{
|
|
|
|
BDRVQcowState *s = bs->opaque;
|
|
|
|
bdi->cluster_size = s->cluster_size;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-06-05 11:20:57 +02:00
|
|
|
static QemuOptsList qcow_create_opts = {
|
|
|
|
.name = "qcow-create-opts",
|
|
|
|
.head = QTAILQ_HEAD_INITIALIZER(qcow_create_opts.head),
|
|
|
|
.desc = {
|
|
|
|
{
|
|
|
|
.name = BLOCK_OPT_SIZE,
|
|
|
|
.type = QEMU_OPT_SIZE,
|
|
|
|
.help = "Virtual disk size"
|
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = BLOCK_OPT_BACKING_FILE,
|
|
|
|
.type = QEMU_OPT_STRING,
|
|
|
|
.help = "File name of a base image"
|
|
|
|
},
|
|
|
|
{
|
|
|
|
.name = BLOCK_OPT_ENCRYPT,
|
|
|
|
.type = QEMU_OPT_BOOL,
|
|
|
|
.help = "Encrypt the image",
|
|
|
|
.def_value_str = "off"
|
|
|
|
},
|
|
|
|
{ /* end of list */ }
|
|
|
|
}
|
2009-05-18 16:42:10 +02:00
|
|
|
};
|
|
|
|
|
2009-05-10 00:03:42 +02:00
|
|
|
static BlockDriver bdrv_qcow = {
|
2009-03-07 23:00:29 +01:00
|
|
|
.format_name = "qcow",
|
|
|
|
.instance_size = sizeof(BDRVQcowState),
|
|
|
|
.bdrv_probe = qcow_probe,
|
|
|
|
.bdrv_open = qcow_open,
|
|
|
|
.bdrv_close = qcow_close,
|
2014-06-05 11:20:57 +02:00
|
|
|
.bdrv_reopen_prepare = qcow_reopen_prepare,
|
2014-06-05 11:21:11 +02:00
|
|
|
.bdrv_create = qcow_create,
|
2013-06-28 12:47:42 +02:00
|
|
|
.bdrv_has_zero_init = bdrv_has_zero_init_1,
|
2014-06-04 15:09:35 +02:00
|
|
|
.supports_backing = true,
|
2011-11-10 17:25:44 +01:00
|
|
|
|
|
|
|
.bdrv_co_readv = qcow_co_readv,
|
|
|
|
.bdrv_co_writev = qcow_co_writev,
|
2013-09-04 19:00:28 +02:00
|
|
|
.bdrv_co_get_block_status = qcow_co_get_block_status,
|
2011-11-10 17:25:44 +01:00
|
|
|
|
|
|
|
.bdrv_set_key = qcow_set_key,
|
|
|
|
.bdrv_make_empty = qcow_make_empty,
|
|
|
|
.bdrv_write_compressed = qcow_write_compressed,
|
|
|
|
.bdrv_get_info = qcow_get_info,
|
2009-05-18 16:42:10 +02:00
|
|
|
|
2014-06-05 11:20:57 +02:00
|
|
|
.create_opts = &qcow_create_opts,
|
2004-08-01 23:59:26 +02:00
|
|
|
};
|
2009-05-10 00:03:42 +02:00
|
|
|
|
|
|
|
static void bdrv_qcow_init(void)
|
|
|
|
{
|
|
|
|
bdrv_register(&bdrv_qcow);
|
|
|
|
}
|
|
|
|
|
|
|
|
block_init(bdrv_qcow_init);
|