2009-05-28 18:07:05 +04:00
|
|
|
/*
|
|
|
|
* Block driver for the QCOW version 2 format
|
|
|
|
*
|
|
|
|
* Copyright (c) 2004-2006 Fabrice Bellard
|
|
|
|
*
|
|
|
|
* 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 21:01:42 +03:00
|
|
|
#include "qemu/osdep.h"
|
2009-05-28 18:07:05 +04:00
|
|
|
#include <zlib.h>
|
|
|
|
|
2018-03-06 19:14:08 +03:00
|
|
|
#include "qapi/error.h"
|
2018-05-03 22:50:20 +03:00
|
|
|
#include "qcow2.h"
|
2016-03-15 19:22:36 +03:00
|
|
|
#include "qemu/bswap.h"
|
2012-03-01 21:36:21 +04:00
|
|
|
#include "trace.h"
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2017-09-18 15:42:29 +03:00
|
|
|
int qcow2_shrink_l1_table(BlockDriverState *bs, uint64_t exact_size)
|
|
|
|
{
|
|
|
|
BDRVQcow2State *s = bs->opaque;
|
|
|
|
int new_l1_size, i, ret;
|
|
|
|
|
|
|
|
if (exact_size >= s->l1_size) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
new_l1_size = exact_size;
|
|
|
|
|
|
|
|
#ifdef DEBUG_ALLOC2
|
|
|
|
fprintf(stderr, "shrink l1_table from %d to %d\n", s->l1_size, new_l1_size);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_SHRINK_WRITE_TABLE);
|
|
|
|
ret = bdrv_pwrite_zeroes(bs->file, s->l1_table_offset +
|
|
|
|
new_l1_size * sizeof(uint64_t),
|
|
|
|
(s->l1_size - new_l1_size) * sizeof(uint64_t), 0);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = bdrv_flush(bs->file->bs);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_SHRINK_FREE_L2_CLUSTERS);
|
|
|
|
for (i = s->l1_size - 1; i > new_l1_size - 1; i--) {
|
|
|
|
if ((s->l1_table[i] & L1E_OFFSET_MASK) == 0) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
qcow2_free_clusters(bs, s->l1_table[i] & L1E_OFFSET_MASK,
|
|
|
|
s->cluster_size, QCOW2_DISCARD_ALWAYS);
|
|
|
|
s->l1_table[i] = 0;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
/*
|
|
|
|
* If the write in the l1_table failed the image may contain a partially
|
|
|
|
* overwritten l1_table. In this case it would be better to clear the
|
|
|
|
* l1_table in memory to avoid possible image corruption.
|
|
|
|
*/
|
|
|
|
memset(s->l1_table + new_l1_size, 0,
|
|
|
|
(s->l1_size - new_l1_size) * sizeof(uint64_t));
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2013-05-14 18:14:33 +04:00
|
|
|
int qcow2_grow_l1_table(BlockDriverState *bs, uint64_t min_size,
|
|
|
|
bool exact_size)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-05-14 18:14:33 +04:00
|
|
|
int new_l1_size2, ret, i;
|
2009-05-28 18:07:05 +04:00
|
|
|
uint64_t *new_l1_table;
|
2013-09-30 19:57:21 +04:00
|
|
|
int64_t old_l1_table_offset, old_l1_size;
|
2013-05-14 18:14:33 +04:00
|
|
|
int64_t new_l1_table_offset, new_l1_size;
|
2009-05-28 18:07:05 +04:00
|
|
|
uint8_t data[12];
|
|
|
|
|
2010-10-18 19:53:53 +04:00
|
|
|
if (min_size <= s->l1_size)
|
2009-05-28 18:07:05 +04:00
|
|
|
return 0;
|
2010-10-18 19:53:53 +04:00
|
|
|
|
qcow2: Check min_size in qcow2_grow_l1_table()
First, new_l1_size is an int64_t, whereas min_size is a uint64_t.
Therefore, during the loop which adjusts new_l1_size until it equals or
exceeds min_size, new_l1_size might overflow and become negative. The
comparison in the loop condition however will take it as an unsigned
value (because min_size is unsigned) and therefore recognize it as
exceeding min_size. Therefore, the loop is left with a negative
new_l1_size, which is not correct. This could be fixed by making
new_l1_size uint64_t.
On the other hand, however, by doing this, the while loop may take
forever. If min_size is e.g. UINT64_MAX, it will take new_l1_size
probably multiple overflows to reach the exact same value (if it reaches
it at all). Then, right after the loop, new_l1_size will be recognized
as being too big anyway.
Both problems require a ridiculously high min_size value, which is very
unlikely to occur; but both problems are also simply avoided by checking
whether min_size is sane before calculating new_l1_size (which should
still be checked separately, though).
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-04-29 21:03:14 +04:00
|
|
|
/* Do a sanity check on min_size before trying to calculate new_l1_size
|
|
|
|
* (this prevents overflows during the while loop for the calculation of
|
|
|
|
* new_l1_size) */
|
|
|
|
if (min_size > INT_MAX / sizeof(uint64_t)) {
|
|
|
|
return -EFBIG;
|
|
|
|
}
|
|
|
|
|
2010-10-18 19:53:53 +04:00
|
|
|
if (exact_size) {
|
|
|
|
new_l1_size = min_size;
|
|
|
|
} else {
|
|
|
|
/* Bump size up to reduce the number of times we have to grow */
|
|
|
|
new_l1_size = s->l1_size;
|
|
|
|
if (new_l1_size == 0) {
|
|
|
|
new_l1_size = 1;
|
|
|
|
}
|
|
|
|
while (min_size > new_l1_size) {
|
2017-06-22 14:04:16 +03:00
|
|
|
new_l1_size = DIV_ROUND_UP(new_l1_size * 3, 2);
|
2010-10-18 19:53:53 +04:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
2010-10-18 19:53:53 +04:00
|
|
|
|
2016-06-15 18:36:30 +03:00
|
|
|
QEMU_BUILD_BUG_ON(QCOW_MAX_L1_SIZE > INT_MAX);
|
|
|
|
if (new_l1_size > QCOW_MAX_L1_SIZE / sizeof(uint64_t)) {
|
2013-05-14 18:14:33 +04:00
|
|
|
return -EFBIG;
|
|
|
|
}
|
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
#ifdef DEBUG_ALLOC2
|
2013-05-14 18:14:33 +04:00
|
|
|
fprintf(stderr, "grow l1_table from %d to %" PRId64 "\n",
|
|
|
|
s->l1_size, new_l1_size);
|
2009-05-28 18:07:05 +04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
new_l1_size2 = sizeof(uint64_t) * new_l1_size;
|
2020-01-18 22:09:26 +03:00
|
|
|
new_l1_table = qemu_try_blockalign(bs->file->bs, new_l1_size2);
|
2014-05-20 19:12:47 +04:00
|
|
|
if (new_l1_table == NULL) {
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2020-01-18 22:09:26 +03:00
|
|
|
memset(new_l1_table, 0, new_l1_size2);
|
2014-05-20 19:12:47 +04:00
|
|
|
|
2016-09-13 11:56:27 +03:00
|
|
|
if (s->l1_size) {
|
|
|
|
memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t));
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
/* write new table (align to cluster) */
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_ALLOC_TABLE);
|
2009-05-28 18:07:07 +04:00
|
|
|
new_l1_table_offset = qcow2_alloc_clusters(bs, new_l1_size2);
|
2010-01-20 17:04:01 +03:00
|
|
|
if (new_l1_table_offset < 0) {
|
2014-05-20 19:12:47 +04:00
|
|
|
qemu_vfree(new_l1_table);
|
2010-01-20 17:04:01 +03:00
|
|
|
return new_l1_table_offset;
|
|
|
|
}
|
2011-01-10 19:17:28 +03:00
|
|
|
|
|
|
|
ret = qcow2_cache_flush(bs, s->refcount_block_cache);
|
|
|
|
if (ret < 0) {
|
2011-06-01 12:50:00 +04:00
|
|
|
goto fail;
|
2011-01-10 19:17:28 +03:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2013-08-30 16:34:26 +04:00
|
|
|
/* the L1 position has not yet been updated, so these clusters must
|
|
|
|
* indeed be completely free */
|
2013-10-10 13:09:23 +04:00
|
|
|
ret = qcow2_pre_write_overlap_check(bs, 0, new_l1_table_offset,
|
2019-01-15 22:39:06 +03:00
|
|
|
new_l1_size2, false);
|
2013-08-30 16:34:26 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_WRITE_TABLE);
|
2009-05-28 18:07:05 +04:00
|
|
|
for(i = 0; i < s->l1_size; i++)
|
|
|
|
new_l1_table[i] = cpu_to_be64(new_l1_table[i]);
|
2016-06-20 21:09:15 +03:00
|
|
|
ret = bdrv_pwrite_sync(bs->file, new_l1_table_offset,
|
2015-06-16 15:19:22 +03:00
|
|
|
new_l1_table, new_l1_size2);
|
2010-06-16 19:44:35 +04:00
|
|
|
if (ret < 0)
|
2009-05-28 18:07:05 +04:00
|
|
|
goto fail;
|
|
|
|
for(i = 0; i < s->l1_size; i++)
|
|
|
|
new_l1_table[i] = be64_to_cpu(new_l1_table[i]);
|
|
|
|
|
|
|
|
/* set new table */
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_ACTIVATE_TABLE);
|
2016-06-16 19:06:17 +03:00
|
|
|
stl_be_p(data, new_l1_size);
|
2013-11-05 20:38:36 +04:00
|
|
|
stq_be_p(data + 4, new_l1_table_offset);
|
2016-06-20 21:09:15 +03:00
|
|
|
ret = bdrv_pwrite_sync(bs->file, offsetof(QCowHeader, l1_size),
|
2015-06-16 15:19:22 +03:00
|
|
|
data, sizeof(data));
|
2010-06-16 19:44:35 +04:00
|
|
|
if (ret < 0) {
|
2009-05-28 18:07:05 +04:00
|
|
|
goto fail;
|
2010-01-20 17:02:58 +03:00
|
|
|
}
|
2014-05-20 19:12:47 +04:00
|
|
|
qemu_vfree(s->l1_table);
|
2013-09-30 19:57:21 +04:00
|
|
|
old_l1_table_offset = s->l1_table_offset;
|
2009-05-28 18:07:05 +04:00
|
|
|
s->l1_table_offset = new_l1_table_offset;
|
|
|
|
s->l1_table = new_l1_table;
|
2013-09-30 19:57:21 +04:00
|
|
|
old_l1_size = s->l1_size;
|
2009-05-28 18:07:05 +04:00
|
|
|
s->l1_size = new_l1_size;
|
2013-09-30 19:57:21 +04:00
|
|
|
qcow2_free_clusters(bs, old_l1_table_offset, old_l1_size * sizeof(uint64_t),
|
|
|
|
QCOW2_DISCARD_OTHER);
|
2009-05-28 18:07:05 +04:00
|
|
|
return 0;
|
|
|
|
fail:
|
2014-05-20 19:12:47 +04:00
|
|
|
qemu_vfree(new_l1_table);
|
2013-06-19 15:44:18 +04:00
|
|
|
qcow2_free_clusters(bs, new_l1_table_offset, new_l1_size2,
|
|
|
|
QCOW2_DISCARD_OTHER);
|
2010-06-16 19:44:35 +04:00
|
|
|
return ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* l2_load
|
|
|
|
*
|
2018-02-05 17:33:15 +03:00
|
|
|
* @bs: The BlockDriverState
|
|
|
|
* @offset: A guest offset, used to calculate what slice of the L2
|
|
|
|
* table to load.
|
|
|
|
* @l2_offset: Offset to the L2 table in the image file.
|
|
|
|
* @l2_slice: Location to store the pointer to the L2 slice.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2018-02-05 17:33:15 +03:00
|
|
|
* Loads a L2 slice into memory (L2 slices are the parts of L2 tables
|
|
|
|
* that are loaded by the qcow2 cache). If the slice is in the cache,
|
|
|
|
* the cache is used; otherwise the L2 slice is loaded from the image
|
|
|
|
* file.
|
2009-05-28 18:07:05 +04:00
|
|
|
*/
|
2018-02-05 17:33:15 +03:00
|
|
|
static int l2_load(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t l2_offset, uint64_t **l2_slice)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2018-02-05 17:33:15 +03:00
|
|
|
int start_of_slice = sizeof(uint64_t) *
|
|
|
|
(offset_to_l2_index(s, offset) - offset_to_l2_slice_index(s, offset));
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:15 +03:00
|
|
|
return qcow2_cache_get(bs, s->l2_table_cache, l2_offset + start_of_slice,
|
|
|
|
(void **)l2_slice);
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2009-06-16 13:31:28 +04:00
|
|
|
/*
|
2020-01-18 22:09:28 +03:00
|
|
|
* Writes an L1 entry to disk (note that depending on the alignment
|
|
|
|
* requirements this function may write more that just one entry in
|
|
|
|
* order to prevent bdrv_pwrite from performing a read-modify-write)
|
2009-06-16 13:31:28 +04:00
|
|
|
*/
|
2013-08-30 16:34:28 +04:00
|
|
|
int qcow2_write_l1_entry(BlockDriverState *bs, int l1_index)
|
2009-06-16 13:31:28 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2009-06-16 13:31:28 +04:00
|
|
|
int l1_start_index;
|
2010-03-23 19:28:22 +03:00
|
|
|
int i, ret;
|
2020-01-18 22:09:28 +03:00
|
|
|
int bufsize = MAX(sizeof(uint64_t),
|
|
|
|
MIN(bs->file->bs->bl.request_alignment, s->cluster_size));
|
|
|
|
int nentries = bufsize / sizeof(uint64_t);
|
|
|
|
g_autofree uint64_t *buf = g_try_new0(uint64_t, nentries);
|
2009-06-16 13:31:28 +04:00
|
|
|
|
2020-01-18 22:09:28 +03:00
|
|
|
if (buf == NULL) {
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
l1_start_index = QEMU_ALIGN_DOWN(l1_index, nentries);
|
|
|
|
for (i = 0; i < MIN(nentries, s->l1_size - l1_start_index); i++) {
|
2009-06-16 13:31:28 +04:00
|
|
|
buf[i] = cpu_to_be64(s->l1_table[l1_start_index + i]);
|
|
|
|
}
|
|
|
|
|
2013-10-10 13:09:23 +04:00
|
|
|
ret = qcow2_pre_write_overlap_check(bs, QCOW2_OL_ACTIVE_L1,
|
2020-01-18 22:09:28 +03:00
|
|
|
s->l1_table_offset + 8 * l1_start_index, bufsize, false);
|
2013-08-30 16:34:26 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L1_UPDATE);
|
2016-06-20 21:09:15 +03:00
|
|
|
ret = bdrv_pwrite_sync(bs->file,
|
2015-06-16 15:19:22 +03:00
|
|
|
s->l1_table_offset + 8 * l1_start_index,
|
2020-01-18 22:09:28 +03:00
|
|
|
buf, bufsize);
|
2010-03-23 19:28:22 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2009-06-16 13:31:28 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
/*
|
|
|
|
* l2_allocate
|
|
|
|
*
|
|
|
|
* Allocate a new l2 entry in the file. If l1_index points to an already
|
|
|
|
* used entry in the L2 table (i.e. we are doing a copy on write for the L2
|
|
|
|
* table) copy the contents of the old L2 table into the newly allocated one.
|
|
|
|
* Otherwise the new table is initialized with zeros.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
static int l2_allocate(BlockDriverState *bs, int l1_index)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2009-06-16 13:31:28 +04:00
|
|
|
uint64_t old_l2_offset;
|
2018-02-05 17:33:17 +03:00
|
|
|
uint64_t *l2_slice = NULL;
|
|
|
|
unsigned slice, slice_size2, n_slices;
|
2010-02-02 17:20:57 +03:00
|
|
|
int64_t l2_offset;
|
2010-03-23 19:41:24 +03:00
|
|
|
int ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
old_l2_offset = s->l1_table[l1_index];
|
|
|
|
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_l2_allocate(bs, l1_index);
|
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
/* allocate a new l2 entry */
|
|
|
|
|
2009-05-28 18:07:07 +04:00
|
|
|
l2_offset = qcow2_alloc_clusters(bs, s->l2_size * sizeof(uint64_t));
|
2010-01-20 17:04:01 +03:00
|
|
|
if (l2_offset < 0) {
|
2013-09-25 18:37:20 +04:00
|
|
|
ret = l2_offset;
|
|
|
|
goto fail;
|
2010-01-20 17:04:01 +03:00
|
|
|
}
|
2011-01-10 19:17:28 +03:00
|
|
|
|
2019-02-08 18:44:53 +03:00
|
|
|
/* The offset must fit in the offset field of the L1 table entry */
|
|
|
|
assert((l2_offset & L1E_OFFSET_MASK) == l2_offset);
|
|
|
|
|
2017-11-03 17:18:51 +03:00
|
|
|
/* If we're allocating the table at offset 0 then something is wrong */
|
|
|
|
if (l2_offset == 0) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "Preventing invalid "
|
|
|
|
"allocation of L2 table at offset 0");
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2011-01-10 19:17:28 +03:00
|
|
|
ret = qcow2_cache_flush(bs, s->refcount_block_cache);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
/* allocate a new entry in the l2 cache */
|
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
slice_size2 = s->l2_slice_size * sizeof(uint64_t);
|
|
|
|
n_slices = s->cluster_size / slice_size2;
|
|
|
|
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_l2_allocate_get_empty(bs, l1_index);
|
2018-02-05 17:33:17 +03:00
|
|
|
for (slice = 0; slice < n_slices; slice++) {
|
2018-02-05 17:33:16 +03:00
|
|
|
ret = qcow2_cache_get_empty(bs, s->l2_table_cache,
|
2018-02-05 17:33:17 +03:00
|
|
|
l2_offset + slice * slice_size2,
|
|
|
|
(void **) &l2_slice);
|
2018-02-05 17:33:16 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2011-01-10 19:17:28 +03:00
|
|
|
|
2018-02-05 17:33:16 +03:00
|
|
|
if ((old_l2_offset & L1E_OFFSET_MASK) == 0) {
|
2018-02-05 17:33:17 +03:00
|
|
|
/* if there was no old l2 table, clear the new slice */
|
|
|
|
memset(l2_slice, 0, slice_size2);
|
2018-02-05 17:33:16 +03:00
|
|
|
} else {
|
2018-02-05 17:33:17 +03:00
|
|
|
uint64_t *old_slice;
|
|
|
|
uint64_t old_l2_slice_offset =
|
|
|
|
(old_l2_offset & L1E_OFFSET_MASK) + slice * slice_size2;
|
2011-01-10 19:17:28 +03:00
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
/* if there was an old l2 table, read a slice from the disk */
|
2018-02-05 17:33:16 +03:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_ALLOC_COW_READ);
|
2018-02-05 17:33:17 +03:00
|
|
|
ret = qcow2_cache_get(bs, s->l2_table_cache, old_l2_slice_offset,
|
|
|
|
(void **) &old_slice);
|
2018-02-05 17:33:16 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
memcpy(l2_slice, old_slice, slice_size2);
|
2018-02-05 17:33:16 +03:00
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &old_slice);
|
2011-01-10 19:17:28 +03:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:17 +03:00
|
|
|
/* write the l2 slice to the file */
|
2018-02-05 17:33:16 +03:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_ALLOC_WRITE);
|
2011-01-10 19:17:28 +03:00
|
|
|
|
2018-02-05 17:33:16 +03:00
|
|
|
trace_qcow2_l2_allocate_write_l2(bs, l1_index);
|
2018-02-05 17:33:17 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
2011-01-10 19:17:28 +03:00
|
|
|
|
|
|
|
ret = qcow2_cache_flush(bs, s->l2_table_cache);
|
2010-03-23 19:41:24 +03:00
|
|
|
if (ret < 0) {
|
2010-05-12 18:23:26 +04:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* update the L1 entry */
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_l2_allocate_write_l1(bs, l1_index);
|
2010-05-12 18:23:26 +04:00
|
|
|
s->l1_table[l1_index] = l2_offset | QCOW_OFLAG_COPIED;
|
2013-08-30 16:34:28 +04:00
|
|
|
ret = qcow2_write_l1_entry(bs, l1_index);
|
2010-05-12 18:23:26 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
2010-03-23 19:41:24 +03:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_l2_allocate_done(bs, l1_index, 0);
|
2010-03-23 19:41:24 +03:00
|
|
|
return 0;
|
2010-05-12 18:23:26 +04:00
|
|
|
|
|
|
|
fail:
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_l2_allocate_done(bs, l1_index, ret);
|
2018-02-05 17:33:17 +03:00
|
|
|
if (l2_slice != NULL) {
|
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2013-09-25 18:37:18 +04:00
|
|
|
}
|
2010-06-07 18:43:22 +04:00
|
|
|
s->l1_table[l1_index] = old_l2_offset;
|
2013-09-25 18:37:19 +04:00
|
|
|
if (l2_offset > 0) {
|
|
|
|
qcow2_free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t),
|
|
|
|
QCOW2_DISCARD_ALWAYS);
|
|
|
|
}
|
2010-05-12 18:23:26 +04:00
|
|
|
return ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2012-03-15 19:37:40 +04:00
|
|
|
/*
|
2018-02-05 17:33:33 +03:00
|
|
|
* Checks how many clusters in a given L2 slice are contiguous in the image
|
2012-03-15 19:37:40 +04:00
|
|
|
* file. As soon as one of the flags in the bitmask stop_flags changes compared
|
|
|
|
* to the first cluster, the search is stopped and the cluster is not counted
|
|
|
|
* as contiguous. (This allows it, for example, to stop at the first compressed
|
|
|
|
* cluster which may require a different handling)
|
|
|
|
*/
|
2019-01-17 20:42:40 +03:00
|
|
|
static int count_contiguous_clusters(BlockDriverState *bs, int nb_clusters,
|
|
|
|
int cluster_size, uint64_t *l2_slice, uint64_t stop_flags)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
|
|
|
int i;
|
2017-05-07 03:05:45 +03:00
|
|
|
QCow2ClusterType first_cluster_type;
|
2013-11-12 16:48:07 +04:00
|
|
|
uint64_t mask = stop_flags | L2E_OFFSET_MASK | QCOW_OFLAG_COMPRESSED;
|
2018-02-05 17:33:33 +03:00
|
|
|
uint64_t first_entry = be64_to_cpu(l2_slice[0]);
|
2013-09-27 14:14:15 +04:00
|
|
|
uint64_t offset = first_entry & mask;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2019-01-18 15:44:40 +03:00
|
|
|
first_cluster_type = qcow2_get_cluster_type(bs, first_entry);
|
|
|
|
if (first_cluster_type == QCOW2_CLUSTER_UNALLOCATED) {
|
2009-05-28 18:07:05 +04:00
|
|
|
return 0;
|
2017-05-04 02:11:18 +03:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2017-05-04 02:11:18 +03:00
|
|
|
/* must be allocated */
|
|
|
|
assert(first_cluster_type == QCOW2_CLUSTER_NORMAL ||
|
2017-05-07 03:05:46 +03:00
|
|
|
first_cluster_type == QCOW2_CLUSTER_ZERO_ALLOC);
|
2013-09-27 14:14:15 +04:00
|
|
|
|
2013-09-27 15:36:11 +04:00
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
2018-02-05 17:33:33 +03:00
|
|
|
uint64_t l2_entry = be64_to_cpu(l2_slice[i]) & mask;
|
2012-03-15 19:37:40 +04:00
|
|
|
if (offset + (uint64_t) i * cluster_size != l2_entry) {
|
2009-05-28 18:07:05 +04:00
|
|
|
break;
|
2012-03-15 19:37:40 +04:00
|
|
|
}
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-12-14 01:37:37 +03:00
|
|
|
return i;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2017-05-07 03:05:44 +03:00
|
|
|
/*
|
|
|
|
* Checks how many consecutive unallocated clusters in a given L2
|
2018-02-05 17:33:34 +03:00
|
|
|
* slice have the same cluster type.
|
2017-05-07 03:05:44 +03:00
|
|
|
*/
|
2019-01-17 20:42:40 +03:00
|
|
|
static int count_contiguous_clusters_unallocated(BlockDriverState *bs,
|
|
|
|
int nb_clusters,
|
2018-02-05 17:33:34 +03:00
|
|
|
uint64_t *l2_slice,
|
2017-05-07 03:05:45 +03:00
|
|
|
QCow2ClusterType wanted_type)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2012-03-15 19:37:40 +04:00
|
|
|
int i;
|
|
|
|
|
2017-05-07 03:05:46 +03:00
|
|
|
assert(wanted_type == QCOW2_CLUSTER_ZERO_PLAIN ||
|
2017-05-07 03:05:44 +03:00
|
|
|
wanted_type == QCOW2_CLUSTER_UNALLOCATED);
|
2012-03-15 19:37:40 +04:00
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
2018-02-05 17:33:34 +03:00
|
|
|
uint64_t entry = be64_to_cpu(l2_slice[i]);
|
2019-01-17 20:42:40 +03:00
|
|
|
QCow2ClusterType type = qcow2_get_cluster_type(bs, entry);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2017-05-07 03:05:46 +03:00
|
|
|
if (type != wanted_type) {
|
2012-03-15 19:37:40 +04:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
static int coroutine_fn do_perform_cow_read(BlockDriverState *bs,
|
|
|
|
uint64_t src_cluster_offset,
|
|
|
|
unsigned offset_in_cluster,
|
2017-06-19 16:40:07 +03:00
|
|
|
QEMUIOVector *qiov)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2016-06-01 16:21:05 +03:00
|
|
|
int ret;
|
2011-09-19 13:26:48 +04:00
|
|
|
|
2017-06-19 16:40:07 +03:00
|
|
|
if (qiov->size == 0) {
|
2017-06-19 16:40:04 +03:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_COW_READ);
|
2011-11-30 16:23:41 +04:00
|
|
|
|
2014-03-11 02:44:07 +04:00
|
|
|
if (!bs->drv) {
|
2017-06-19 16:40:05 +03:00
|
|
|
return -ENOMEDIUM;
|
2014-03-11 02:44:07 +04:00
|
|
|
}
|
|
|
|
|
2011-11-30 16:23:41 +04:00
|
|
|
/* Call .bdrv_co_readv() directly instead of using the public block-layer
|
|
|
|
* interface. This avoids double I/O throttling and request tracking,
|
|
|
|
* which can lead to deadlock when block layer copy-on-read is enabled.
|
|
|
|
*/
|
2019-06-04 19:15:13 +03:00
|
|
|
ret = bs->drv->bdrv_co_preadv_part(bs,
|
|
|
|
src_cluster_offset + offset_in_cluster,
|
|
|
|
qiov->size, qiov, 0, 0);
|
2011-09-19 13:26:48 +04:00
|
|
|
if (ret < 0) {
|
2017-06-19 16:40:05 +03:00
|
|
|
return ret;
|
2011-09-19 13:26:48 +04:00
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int coroutine_fn do_perform_cow_write(BlockDriverState *bs,
|
|
|
|
uint64_t cluster_offset,
|
|
|
|
unsigned offset_in_cluster,
|
2017-06-19 16:40:07 +03:00
|
|
|
QEMUIOVector *qiov)
|
2017-06-19 16:40:05 +03:00
|
|
|
{
|
2019-01-15 22:39:06 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2017-06-19 16:40:05 +03:00
|
|
|
int ret;
|
|
|
|
|
2017-06-19 16:40:07 +03:00
|
|
|
if (qiov->size == 0) {
|
2017-06-19 16:40:05 +03:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2013-10-10 13:09:23 +04:00
|
|
|
ret = qcow2_pre_write_overlap_check(bs, 0,
|
2019-01-15 22:39:06 +03:00
|
|
|
cluster_offset + offset_in_cluster, qiov->size, true);
|
2013-08-30 16:34:26 +04:00
|
|
|
if (ret < 0) {
|
2017-06-19 16:40:05 +03:00
|
|
|
return ret;
|
2013-08-30 16:34:26 +04:00
|
|
|
}
|
|
|
|
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_COW_WRITE);
|
2019-01-15 22:39:06 +03:00
|
|
|
ret = bdrv_co_pwritev(s->data_file, cluster_offset + offset_in_cluster,
|
2017-06-19 16:40:07 +03:00
|
|
|
qiov->size, qiov, 0);
|
2011-09-19 13:26:48 +04:00
|
|
|
if (ret < 0) {
|
2017-06-19 16:40:05 +03:00
|
|
|
return ret;
|
2011-09-19 13:26:48 +04:00
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
return 0;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get_cluster_offset
|
|
|
|
*
|
2016-05-31 17:13:07 +03:00
|
|
|
* For a given offset of the virtual disk, find the cluster type and offset in
|
|
|
|
* the qcow2 file. The offset is stored in *cluster_offset.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2016-05-31 17:13:07 +03:00
|
|
|
* On entry, *bytes is the maximum number of contiguous bytes starting at
|
|
|
|
* offset that we are interested in.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2016-05-31 17:13:07 +03:00
|
|
|
* On exit, *bytes is the number of bytes starting at offset that have the same
|
|
|
|
* cluster type and (if applicable) are stored contiguously in the image file.
|
|
|
|
* Compressed clusters are always returned one by one.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2012-03-14 22:15:03 +04:00
|
|
|
* Returns the cluster type (QCOW2_CLUSTER_*) on success, -errno in error
|
|
|
|
* cases.
|
2009-05-28 18:07:05 +04:00
|
|
|
*/
|
2010-05-21 19:59:36 +04:00
|
|
|
int qcow2_get_cluster_offset(BlockDriverState *bs, uint64_t offset,
|
2016-05-31 17:13:07 +03:00
|
|
|
unsigned int *bytes, uint64_t *cluster_offset)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-05-14 18:14:33 +04:00
|
|
|
unsigned int l2_index;
|
2018-02-05 17:33:20 +03:00
|
|
|
uint64_t l1_index, l2_offset, *l2_slice;
|
|
|
|
int c;
|
2016-06-20 17:26:23 +03:00
|
|
|
unsigned int offset_in_cluster;
|
|
|
|
uint64_t bytes_available, bytes_needed, nb_clusters;
|
2017-05-07 03:05:45 +03:00
|
|
|
QCow2ClusterType type;
|
2010-05-21 20:25:20 +04:00
|
|
|
int ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2016-05-31 17:41:09 +03:00
|
|
|
offset_in_cluster = offset_into_cluster(s, offset);
|
2016-05-31 17:13:07 +03:00
|
|
|
bytes_needed = (uint64_t) *bytes + offset_in_cluster;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2016-05-31 17:41:09 +03:00
|
|
|
/* compute how many bytes there are between the start of the cluster
|
2018-02-05 17:33:20 +03:00
|
|
|
* containing offset and the end of the l2 slice that contains
|
|
|
|
* the entry pointing to it */
|
|
|
|
bytes_available =
|
|
|
|
((uint64_t) (s->l2_slice_size - offset_to_l2_slice_index(s, offset)))
|
|
|
|
<< s->cluster_bits;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2016-05-31 17:41:09 +03:00
|
|
|
if (bytes_needed > bytes_available) {
|
|
|
|
bytes_needed = bytes_available;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2010-05-21 19:59:36 +04:00
|
|
|
*cluster_offset = 0;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2015-08-26 14:17:13 +03:00
|
|
|
/* seek to the l2 offset in the l1 table */
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:12 +03:00
|
|
|
l1_index = offset_to_l1_index(s, offset);
|
2012-03-14 22:15:03 +04:00
|
|
|
if (l1_index >= s->l1_size) {
|
2017-05-07 03:05:45 +03:00
|
|
|
type = QCOW2_CLUSTER_UNALLOCATED;
|
2009-05-28 18:07:05 +04:00
|
|
|
goto out;
|
2012-03-14 22:15:03 +04:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2012-03-14 22:15:03 +04:00
|
|
|
l2_offset = s->l1_table[l1_index] & L1E_OFFSET_MASK;
|
|
|
|
if (!l2_offset) {
|
2017-05-07 03:05:45 +03:00
|
|
|
type = QCOW2_CLUSTER_UNALLOCATED;
|
2009-05-28 18:07:05 +04:00
|
|
|
goto out;
|
2012-03-14 22:15:03 +04:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2014-09-05 18:07:18 +04:00
|
|
|
if (offset_into_cluster(s, l2_offset)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "L2 table offset %#" PRIx64
|
|
|
|
" unaligned (L1 index: %#" PRIx64 ")",
|
|
|
|
l2_offset, l1_index);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:20 +03:00
|
|
|
/* load the l2 slice in memory */
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:20 +03:00
|
|
|
ret = l2_load(bs, offset, l2_offset, &l2_slice);
|
2010-05-21 20:25:20 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2010-05-21 19:59:36 +04:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
/* find the cluster offset for the given disk offset */
|
|
|
|
|
2018-02-05 17:33:20 +03:00
|
|
|
l2_index = offset_to_l2_slice_index(s, offset);
|
|
|
|
*cluster_offset = be64_to_cpu(l2_slice[l2_index]);
|
2015-09-14 17:39:47 +03:00
|
|
|
|
2016-05-31 17:41:09 +03:00
|
|
|
nb_clusters = size_to_clusters(s, bytes_needed);
|
2016-06-20 17:26:23 +03:00
|
|
|
/* bytes_needed <= *bytes + offset_in_cluster, both of which are unsigned
|
|
|
|
* integers; the minimum cluster size is 512, so this assertion is always
|
|
|
|
* true */
|
|
|
|
assert(nb_clusters <= INT_MAX);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2019-01-17 20:42:40 +03:00
|
|
|
type = qcow2_get_cluster_type(bs, *cluster_offset);
|
2017-05-07 03:05:46 +03:00
|
|
|
if (s->qcow_version < 3 && (type == QCOW2_CLUSTER_ZERO_PLAIN ||
|
|
|
|
type == QCOW2_CLUSTER_ZERO_ALLOC)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "Zero cluster entry found"
|
|
|
|
" in pre-v3 image (L2 offset: %#" PRIx64
|
|
|
|
", L2 index: %#x)", l2_offset, l2_index);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
2017-05-07 03:05:45 +03:00
|
|
|
switch (type) {
|
2012-03-14 22:15:03 +04:00
|
|
|
case QCOW2_CLUSTER_COMPRESSED:
|
2019-01-15 22:39:06 +03:00
|
|
|
if (has_data_file(bs)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "Compressed cluster "
|
|
|
|
"entry found in image with external data "
|
|
|
|
"file (L2 offset: %#" PRIx64 ", L2 index: "
|
|
|
|
"%#x)", l2_offset, l2_index);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
2012-03-14 22:15:03 +04:00
|
|
|
/* Compressed clusters can only be processed one by one */
|
|
|
|
c = 1;
|
|
|
|
*cluster_offset &= L2E_COMPRESSED_OFFSET_SIZE_MASK;
|
|
|
|
break;
|
2017-05-07 03:05:46 +03:00
|
|
|
case QCOW2_CLUSTER_ZERO_PLAIN:
|
2012-03-14 22:15:03 +04:00
|
|
|
case QCOW2_CLUSTER_UNALLOCATED:
|
2009-05-28 18:07:05 +04:00
|
|
|
/* how many empty clusters ? */
|
2019-01-17 20:42:40 +03:00
|
|
|
c = count_contiguous_clusters_unallocated(bs, nb_clusters,
|
2018-02-05 17:33:20 +03:00
|
|
|
&l2_slice[l2_index], type);
|
2012-03-14 22:15:03 +04:00
|
|
|
*cluster_offset = 0;
|
|
|
|
break;
|
2017-05-07 03:05:46 +03:00
|
|
|
case QCOW2_CLUSTER_ZERO_ALLOC:
|
2012-03-14 22:15:03 +04:00
|
|
|
case QCOW2_CLUSTER_NORMAL:
|
2009-05-28 18:07:05 +04:00
|
|
|
/* how many allocated clusters ? */
|
2019-01-17 20:42:40 +03:00
|
|
|
c = count_contiguous_clusters(bs, nb_clusters, s->cluster_size,
|
2018-02-05 17:33:20 +03:00
|
|
|
&l2_slice[l2_index], QCOW_OFLAG_ZERO);
|
2012-03-14 22:15:03 +04:00
|
|
|
*cluster_offset &= L2E_OFFSET_MASK;
|
2014-09-05 18:07:18 +04:00
|
|
|
if (offset_into_cluster(s, *cluster_offset)) {
|
2017-05-07 03:05:46 +03:00
|
|
|
qcow2_signal_corruption(bs, true, -1, -1,
|
|
|
|
"Cluster allocation offset %#"
|
2014-09-05 18:07:18 +04:00
|
|
|
PRIx64 " unaligned (L2 offset: %#" PRIx64
|
|
|
|
", L2 index: %#x)", *cluster_offset,
|
|
|
|
l2_offset, l2_index);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
2019-01-15 22:39:06 +03:00
|
|
|
if (has_data_file(bs) && *cluster_offset != offset - offset_in_cluster)
|
|
|
|
{
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1,
|
|
|
|
"External data file host cluster offset %#"
|
|
|
|
PRIx64 " does not match guest cluster "
|
|
|
|
"offset: %#" PRIx64
|
|
|
|
", L2 index: %#x)", *cluster_offset,
|
|
|
|
offset - offset_in_cluster, l2_index);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
2012-03-14 22:15:03 +04:00
|
|
|
break;
|
2012-06-15 15:43:18 +04:00
|
|
|
default:
|
|
|
|
abort();
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:20 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2011-01-10 19:17:28 +03:00
|
|
|
|
2016-06-20 17:26:23 +03:00
|
|
|
bytes_available = (int64_t)c * s->cluster_size;
|
2012-03-14 22:15:03 +04:00
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
out:
|
2016-05-31 17:41:09 +03:00
|
|
|
if (bytes_available > bytes_needed) {
|
|
|
|
bytes_available = bytes_needed;
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2016-06-20 17:26:23 +03:00
|
|
|
/* bytes_available <= bytes_needed <= *bytes + offset_in_cluster;
|
|
|
|
* subtracting offset_in_cluster will therefore definitely yield something
|
|
|
|
* not exceeding UINT_MAX */
|
|
|
|
assert(bytes_available - offset_in_cluster <= UINT_MAX);
|
2016-05-31 17:13:07 +03:00
|
|
|
*bytes = bytes_available - offset_in_cluster;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2017-05-07 03:05:45 +03:00
|
|
|
return type;
|
2014-09-05 18:07:18 +04:00
|
|
|
|
|
|
|
fail:
|
2018-02-05 17:33:20 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **)&l2_slice);
|
2014-09-05 18:07:18 +04:00
|
|
|
return ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get_cluster_table
|
|
|
|
*
|
|
|
|
* for a given disk offset, load (and allocate if needed)
|
2018-02-05 17:33:19 +03:00
|
|
|
* the appropriate slice of its l2 table.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2018-02-05 17:33:19 +03:00
|
|
|
* the cluster index in the l2 slice is given to the caller.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2010-01-20 17:03:00 +03:00
|
|
|
* Returns 0 on success, -errno in failure case
|
2009-05-28 18:07:05 +04:00
|
|
|
*/
|
|
|
|
static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
|
2018-02-05 17:33:19 +03:00
|
|
|
uint64_t **new_l2_slice,
|
2009-05-28 18:07:05 +04:00
|
|
|
int *new_l2_index)
|
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-05-14 18:14:33 +04:00
|
|
|
unsigned int l2_index;
|
|
|
|
uint64_t l1_index, l2_offset;
|
2018-02-05 17:33:19 +03:00
|
|
|
uint64_t *l2_slice = NULL;
|
2009-09-15 14:30:43 +04:00
|
|
|
int ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2015-08-26 14:17:13 +03:00
|
|
|
/* seek to the l2 offset in the l1 table */
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:12 +03:00
|
|
|
l1_index = offset_to_l1_index(s, offset);
|
2009-05-28 18:07:05 +04:00
|
|
|
if (l1_index >= s->l1_size) {
|
2010-10-18 19:53:53 +04:00
|
|
|
ret = qcow2_grow_l1_table(bs, l1_index + 1, false);
|
2010-01-20 17:03:00 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
2012-02-23 18:40:55 +04:00
|
|
|
|
2013-05-14 18:14:33 +04:00
|
|
|
assert(l1_index < s->l1_size);
|
2012-02-23 18:40:55 +04:00
|
|
|
l2_offset = s->l1_table[l1_index] & L1E_OFFSET_MASK;
|
2014-09-05 18:07:18 +04:00
|
|
|
if (offset_into_cluster(s, l2_offset)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "L2 table offset %#" PRIx64
|
|
|
|
" unaligned (L1 index: %#" PRIx64 ")",
|
|
|
|
l2_offset, l1_index);
|
|
|
|
return -EIO;
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:18 +03:00
|
|
|
if (!(s->l1_table[l1_index] & QCOW_OFLAG_COPIED)) {
|
2011-02-09 19:36:19 +03:00
|
|
|
/* First allocate a new L2 table (and do COW if needed) */
|
2018-02-05 17:33:17 +03:00
|
|
|
ret = l2_allocate(bs, l1_index);
|
2010-03-23 19:41:24 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2010-01-20 17:03:00 +03:00
|
|
|
}
|
2011-02-09 19:36:19 +03:00
|
|
|
|
|
|
|
/* Then decrease the refcount of the old table */
|
|
|
|
if (l2_offset) {
|
2013-06-19 15:44:18 +04:00
|
|
|
qcow2_free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t),
|
|
|
|
QCOW2_DISCARD_OTHER);
|
2011-02-09 19:36:19 +03:00
|
|
|
}
|
2018-02-05 17:33:17 +03:00
|
|
|
|
|
|
|
/* Get the offset of the newly-allocated l2 table */
|
|
|
|
l2_offset = s->l1_table[l1_index] & L1E_OFFSET_MASK;
|
|
|
|
assert(offset_into_cluster(s, l2_offset) == 0);
|
2018-02-05 17:33:18 +03:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:19 +03:00
|
|
|
/* load the l2 slice in memory */
|
|
|
|
ret = l2_load(bs, offset, l2_offset, &l2_slice);
|
2018-02-05 17:33:18 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* find the cluster offset for the given disk offset */
|
|
|
|
|
2018-02-05 17:33:19 +03:00
|
|
|
l2_index = offset_to_l2_slice_index(s, offset);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:19 +03:00
|
|
|
*new_l2_slice = l2_slice;
|
2009-05-28 18:07:05 +04:00
|
|
|
*new_l2_index = l2_index;
|
|
|
|
|
2010-01-20 17:03:00 +03:00
|
|
|
return 0;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* alloc_compressed_cluster_offset
|
|
|
|
*
|
2019-02-27 12:26:24 +03:00
|
|
|
* For a given offset on the virtual disk, allocate a new compressed cluster
|
|
|
|
* and put the host offset of the cluster into *host_offset. If a cluster is
|
|
|
|
* already allocated at the offset, return an error.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2019-02-27 12:26:24 +03:00
|
|
|
* Return 0 on success and -errno in error cases
|
2009-05-28 18:07:05 +04:00
|
|
|
*/
|
2019-02-27 12:26:24 +03:00
|
|
|
int qcow2_alloc_compressed_cluster_offset(BlockDriverState *bs,
|
|
|
|
uint64_t offset,
|
|
|
|
int compressed_size,
|
|
|
|
uint64_t *host_offset)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2009-05-28 18:07:05 +04:00
|
|
|
int l2_index, ret;
|
2018-02-05 17:33:32 +03:00
|
|
|
uint64_t *l2_slice;
|
2010-02-02 17:20:57 +03:00
|
|
|
int64_t cluster_offset;
|
2009-05-28 18:07:05 +04:00
|
|
|
int nb_csectors;
|
|
|
|
|
2019-01-15 22:39:06 +03:00
|
|
|
if (has_data_file(bs)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:32 +03:00
|
|
|
ret = get_cluster_table(bs, offset, &l2_slice, &l2_index);
|
2010-01-20 17:03:00 +03:00
|
|
|
if (ret < 0) {
|
2019-02-27 12:26:24 +03:00
|
|
|
return ret;
|
2010-01-20 17:03:00 +03:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2012-03-15 20:20:11 +04:00
|
|
|
/* Compression can't overwrite anything. Fail if the cluster was already
|
|
|
|
* allocated. */
|
2018-02-05 17:33:32 +03:00
|
|
|
cluster_offset = be64_to_cpu(l2_slice[l2_index]);
|
2012-03-15 20:20:11 +04:00
|
|
|
if (cluster_offset & L2E_OFFSET_MASK) {
|
2018-02-05 17:33:32 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2019-02-27 12:26:24 +03:00
|
|
|
return -EIO;
|
2011-10-18 19:12:44 +04:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2009-05-28 18:07:07 +04:00
|
|
|
cluster_offset = qcow2_alloc_bytes(bs, compressed_size);
|
2010-01-20 17:04:01 +03:00
|
|
|
if (cluster_offset < 0) {
|
2018-02-05 17:33:32 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2019-02-27 12:26:24 +03:00
|
|
|
return cluster_offset;
|
2010-01-20 17:04:01 +03:00
|
|
|
}
|
|
|
|
|
2019-05-10 19:22:54 +03:00
|
|
|
nb_csectors =
|
|
|
|
(cluster_offset + compressed_size - 1) / QCOW2_COMPRESSED_SECTOR_SIZE -
|
|
|
|
(cluster_offset / QCOW2_COMPRESSED_SECTOR_SIZE);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2020-01-13 19:11:46 +03:00
|
|
|
/* The offset and size must fit in their fields of the L2 table entry */
|
|
|
|
assert((cluster_offset & s->cluster_offset_mask) == cluster_offset);
|
|
|
|
assert((nb_csectors & s->csize_mask) == nb_csectors);
|
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
cluster_offset |= QCOW_OFLAG_COMPRESSED |
|
|
|
|
((uint64_t)nb_csectors << s->csize_shift);
|
|
|
|
|
|
|
|
/* update L2 table */
|
|
|
|
|
|
|
|
/* compressed clusters never have the copied flag */
|
|
|
|
|
2010-04-14 16:17:38 +04:00
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_L2_UPDATE_COMPRESSED);
|
2018-02-05 17:33:32 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
|
|
|
l2_slice[l2_index] = cpu_to_be64(cluster_offset);
|
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2009-06-16 13:31:29 +04:00
|
|
|
|
2019-02-27 12:26:24 +03:00
|
|
|
*host_offset = cluster_offset & s->cluster_offset_mask;
|
|
|
|
return 0;
|
2009-06-16 13:31:29 +04:00
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:04 +03:00
|
|
|
static int perform_cow(BlockDriverState *bs, QCowL2Meta *m)
|
2012-12-07 21:08:43 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2017-06-19 16:40:04 +03:00
|
|
|
Qcow2COWRegion *start = &m->cow_start;
|
|
|
|
Qcow2COWRegion *end = &m->cow_end;
|
2017-06-19 16:40:05 +03:00
|
|
|
unsigned buffer_size;
|
2017-06-19 16:40:06 +03:00
|
|
|
unsigned data_bytes = end->offset - (start->offset + start->nb_bytes);
|
|
|
|
bool merge_reads;
|
2017-06-19 16:40:05 +03:00
|
|
|
uint8_t *start_buffer, *end_buffer;
|
2017-06-19 16:40:07 +03:00
|
|
|
QEMUIOVector qiov;
|
2012-12-07 21:08:43 +04:00
|
|
|
int ret;
|
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
assert(start->nb_bytes <= UINT_MAX - end->nb_bytes);
|
2017-06-19 16:40:06 +03:00
|
|
|
assert(start->nb_bytes + end->nb_bytes <= UINT_MAX - data_bytes);
|
|
|
|
assert(start->offset + start->nb_bytes <= end->offset);
|
2017-06-19 16:40:05 +03:00
|
|
|
|
2019-05-16 17:27:49 +03:00
|
|
|
if ((start->nb_bytes == 0 && end->nb_bytes == 0) || m->skip_cow) {
|
2012-12-07 21:08:43 +04:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:06 +03:00
|
|
|
/* If we have to read both the start and end COW regions and the
|
|
|
|
* middle region is not too large then perform just one read
|
|
|
|
* operation */
|
|
|
|
merge_reads = start->nb_bytes && end->nb_bytes && data_bytes <= 16384;
|
|
|
|
if (merge_reads) {
|
|
|
|
buffer_size = start->nb_bytes + data_bytes + end->nb_bytes;
|
|
|
|
} else {
|
|
|
|
/* If we have to do two reads, add some padding in the middle
|
|
|
|
* if necessary to make sure that the end region is optimally
|
|
|
|
* aligned. */
|
|
|
|
size_t align = bdrv_opt_mem_align(bs);
|
|
|
|
assert(align > 0 && align <= UINT_MAX);
|
|
|
|
assert(QEMU_ALIGN_UP(start->nb_bytes, align) <=
|
|
|
|
UINT_MAX - end->nb_bytes);
|
|
|
|
buffer_size = QEMU_ALIGN_UP(start->nb_bytes, align) + end->nb_bytes;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Reserve a buffer large enough to store all the data that we're
|
|
|
|
* going to read */
|
2017-06-19 16:40:05 +03:00
|
|
|
start_buffer = qemu_try_blockalign(bs, buffer_size);
|
|
|
|
if (start_buffer == NULL) {
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
/* The part of the buffer where the end region is located */
|
|
|
|
end_buffer = start_buffer + buffer_size - end->nb_bytes;
|
|
|
|
|
2019-06-04 19:15:14 +03:00
|
|
|
qemu_iovec_init(&qiov, 2 + (m->data_qiov ?
|
|
|
|
qemu_iovec_subvec_niov(m->data_qiov,
|
|
|
|
m->data_qiov_offset,
|
|
|
|
data_bytes)
|
|
|
|
: 0));
|
2017-06-19 16:40:07 +03:00
|
|
|
|
2012-12-07 21:08:43 +04:00
|
|
|
qemu_co_mutex_unlock(&s->lock);
|
2017-06-19 16:40:06 +03:00
|
|
|
/* First we read the existing data from both COW regions. We
|
|
|
|
* either read the whole region in one go, or the start and end
|
|
|
|
* regions separately. */
|
|
|
|
if (merge_reads) {
|
2017-06-19 16:40:07 +03:00
|
|
|
qemu_iovec_add(&qiov, start_buffer, buffer_size);
|
|
|
|
ret = do_perform_cow_read(bs, m->offset, start->offset, &qiov);
|
2017-06-19 16:40:06 +03:00
|
|
|
} else {
|
2017-06-19 16:40:07 +03:00
|
|
|
qemu_iovec_add(&qiov, start_buffer, start->nb_bytes);
|
|
|
|
ret = do_perform_cow_read(bs, m->offset, start->offset, &qiov);
|
2017-06-19 16:40:06 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2017-06-19 16:40:05 +03:00
|
|
|
|
2017-06-19 16:40:07 +03:00
|
|
|
qemu_iovec_reset(&qiov);
|
|
|
|
qemu_iovec_add(&qiov, end_buffer, end->nb_bytes);
|
|
|
|
ret = do_perform_cow_read(bs, m->offset, end->offset, &qiov);
|
2017-06-19 16:40:06 +03:00
|
|
|
}
|
2012-12-07 21:08:43 +04:00
|
|
|
if (ret < 0) {
|
2017-06-19 16:40:04 +03:00
|
|
|
goto fail;
|
2012-12-07 21:08:43 +04:00
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
/* Encrypt the data if necessary before writing it */
|
|
|
|
if (bs->encrypted) {
|
2019-09-15 23:36:54 +03:00
|
|
|
ret = qcow2_co_encrypt(bs,
|
|
|
|
m->alloc_offset + start->offset,
|
|
|
|
m->offset + start->offset,
|
|
|
|
start_buffer, start->nb_bytes);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = qcow2_co_encrypt(bs,
|
|
|
|
m->alloc_offset + end->offset,
|
|
|
|
m->offset + end->offset,
|
|
|
|
end_buffer, end->nb_bytes);
|
|
|
|
if (ret < 0) {
|
2017-06-19 16:40:05 +03:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-06-19 16:40:08 +03:00
|
|
|
/* And now we can write everything. If we have the guest data we
|
|
|
|
* can write everything in one single operation */
|
|
|
|
if (m->data_qiov) {
|
|
|
|
qemu_iovec_reset(&qiov);
|
|
|
|
if (start->nb_bytes) {
|
|
|
|
qemu_iovec_add(&qiov, start_buffer, start->nb_bytes);
|
|
|
|
}
|
2019-06-04 19:15:14 +03:00
|
|
|
qemu_iovec_concat(&qiov, m->data_qiov, m->data_qiov_offset, data_bytes);
|
2017-06-19 16:40:08 +03:00
|
|
|
if (end->nb_bytes) {
|
|
|
|
qemu_iovec_add(&qiov, end_buffer, end->nb_bytes);
|
|
|
|
}
|
|
|
|
/* NOTE: we have a write_aio blkdebug event here followed by
|
|
|
|
* a cow_write one in do_perform_cow_write(), but there's only
|
|
|
|
* one single I/O operation */
|
|
|
|
BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
|
|
|
|
ret = do_perform_cow_write(bs, m->alloc_offset, start->offset, &qiov);
|
|
|
|
} else {
|
|
|
|
/* If there's no guest data then write both COW regions separately */
|
|
|
|
qemu_iovec_reset(&qiov);
|
|
|
|
qemu_iovec_add(&qiov, start_buffer, start->nb_bytes);
|
|
|
|
ret = do_perform_cow_write(bs, m->alloc_offset, start->offset, &qiov);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
qemu_iovec_reset(&qiov);
|
|
|
|
qemu_iovec_add(&qiov, end_buffer, end->nb_bytes);
|
|
|
|
ret = do_perform_cow_write(bs, m->alloc_offset, end->offset, &qiov);
|
2017-06-19 16:40:05 +03:00
|
|
|
}
|
2017-06-19 16:40:04 +03:00
|
|
|
|
|
|
|
fail:
|
|
|
|
qemu_co_mutex_lock(&s->lock);
|
|
|
|
|
2012-12-07 21:08:43 +04:00
|
|
|
/*
|
|
|
|
* Before we update the L2 table to actually point to the new cluster, we
|
|
|
|
* need to be sure that the refcounts have been increased and COW was
|
|
|
|
* handled.
|
|
|
|
*/
|
2017-06-19 16:40:04 +03:00
|
|
|
if (ret == 0) {
|
|
|
|
qcow2_cache_depends_on_flush(s->l2_table_cache);
|
|
|
|
}
|
2012-12-07 21:08:43 +04:00
|
|
|
|
2017-06-19 16:40:05 +03:00
|
|
|
qemu_vfree(start_buffer);
|
2017-06-19 16:40:07 +03:00
|
|
|
qemu_iovec_destroy(&qiov);
|
2017-06-19 16:40:04 +03:00
|
|
|
return ret;
|
2012-12-07 21:08:43 +04:00
|
|
|
}
|
|
|
|
|
2010-01-20 17:03:01 +03:00
|
|
|
int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2009-05-28 18:07:05 +04:00
|
|
|
int i, j = 0, l2_index, ret;
|
2018-02-05 17:33:21 +03:00
|
|
|
uint64_t *old_cluster, *l2_slice;
|
2012-03-02 17:10:54 +04:00
|
|
|
uint64_t cluster_offset = m->alloc_offset;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2012-03-01 21:36:21 +04:00
|
|
|
trace_qcow2_cluster_link_l2(qemu_coroutine_self(), m->nb_clusters);
|
2012-12-07 21:08:46 +04:00
|
|
|
assert(m->nb_clusters > 0);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
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 12:31:08 +04:00
|
|
|
old_cluster = g_try_new(uint64_t, m->nb_clusters);
|
2014-05-20 19:12:47 +04:00
|
|
|
if (old_cluster == NULL) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto err;
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
/* copy content of unmodified sectors */
|
2017-06-19 16:40:04 +03:00
|
|
|
ret = perform_cow(bs, m);
|
2012-12-07 21:08:43 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto err;
|
2011-01-10 19:17:28 +03:00
|
|
|
}
|
|
|
|
|
2012-12-07 21:08:43 +04:00
|
|
|
/* Update L2 table. */
|
2013-03-15 13:35:08 +04:00
|
|
|
if (s->use_lazy_refcounts) {
|
2012-12-07 21:08:47 +04:00
|
|
|
qcow2_mark_dirty(bs);
|
|
|
|
}
|
2012-07-27 12:05:22 +04:00
|
|
|
if (qcow2_need_accurate_refcounts(s)) {
|
|
|
|
qcow2_cache_set_dependency(bs, s->l2_table_cache,
|
|
|
|
s->refcount_block_cache);
|
|
|
|
}
|
2012-12-07 21:08:47 +04:00
|
|
|
|
2018-02-05 17:33:21 +03:00
|
|
|
ret = get_cluster_table(bs, m->offset, &l2_slice, &l2_index);
|
2010-01-20 17:03:00 +03:00
|
|
|
if (ret < 0) {
|
2009-05-28 18:07:05 +04:00
|
|
|
goto err;
|
2010-01-20 17:03:00 +03:00
|
|
|
}
|
2018-02-05 17:33:21 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2018-02-05 17:33:21 +03:00
|
|
|
assert(l2_index + m->nb_clusters <= s->l2_slice_size);
|
2009-05-28 18:07:05 +04:00
|
|
|
for (i = 0; i < m->nb_clusters; i++) {
|
2020-01-13 19:11:46 +03:00
|
|
|
uint64_t offset = cluster_offset + (i << s->cluster_bits);
|
2009-05-28 18:07:05 +04:00
|
|
|
/* if two concurrent writes happen to the same unallocated cluster
|
2016-06-01 16:21:05 +03:00
|
|
|
* each write allocates separate cluster and writes data concurrently.
|
|
|
|
* The first one to complete updates l2 table with pointer to its
|
|
|
|
* cluster the second one has to do RMW (which is done above by
|
|
|
|
* perform_cow()), update l2 table with its cluster pointer and free
|
|
|
|
* old cluster. This is what this loop does */
|
2018-02-05 17:33:21 +03:00
|
|
|
if (l2_slice[l2_index + i] != 0) {
|
|
|
|
old_cluster[j++] = l2_slice[l2_index + i];
|
2016-06-01 16:21:05 +03:00
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2020-01-13 19:11:46 +03:00
|
|
|
/* The offset must fit in the offset field of the L2 table entry */
|
|
|
|
assert((offset & L2E_OFFSET_MASK) == offset);
|
|
|
|
|
|
|
|
l2_slice[l2_index + i] = cpu_to_be64(offset | QCOW_OFLAG_COPIED);
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2010-09-17 19:02:09 +04:00
|
|
|
|
2018-02-05 17:33:21 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2010-09-01 14:40:52 +04:00
|
|
|
/*
|
|
|
|
* If this was a COW, we need to decrease the refcount of the old cluster.
|
2013-06-19 15:44:18 +04:00
|
|
|
*
|
|
|
|
* Don't discard clusters that reach a refcount of 0 (e.g. compressed
|
|
|
|
* clusters), the next write will reuse them anyway.
|
2010-09-01 14:40:52 +04:00
|
|
|
*/
|
2017-05-04 02:11:18 +03:00
|
|
|
if (!m->keep_old_clusters && j != 0) {
|
2010-09-01 14:40:52 +04:00
|
|
|
for (i = 0; i < j; i++) {
|
2013-06-19 15:44:18 +04:00
|
|
|
qcow2_free_any_clusters(bs, be64_to_cpu(old_cluster[i]), 1,
|
|
|
|
QCOW2_DISCARD_NEVER);
|
2010-09-01 14:40:52 +04:00
|
|
|
}
|
|
|
|
}
|
2009-05-28 18:07:05 +04:00
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
err:
|
2011-08-21 07:09:37 +04:00
|
|
|
g_free(old_cluster);
|
2009-05-28 18:07:05 +04:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2018-06-28 18:05:45 +03:00
|
|
|
/**
|
|
|
|
* Frees the allocated clusters because the request failed and they won't
|
|
|
|
* actually be linked.
|
|
|
|
*/
|
|
|
|
void qcow2_alloc_cluster_abort(BlockDriverState *bs, QCowL2Meta *m)
|
|
|
|
{
|
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2020-02-25 17:31:28 +03:00
|
|
|
if (!has_data_file(bs) && !m->keep_old_clusters) {
|
2020-02-11 12:48:59 +03:00
|
|
|
qcow2_free_clusters(bs, m->alloc_offset,
|
|
|
|
m->nb_clusters << s->cluster_bits,
|
|
|
|
QCOW2_DISCARD_NEVER);
|
|
|
|
}
|
2018-06-28 18:05:45 +03:00
|
|
|
}
|
|
|
|
|
2012-03-02 22:27:53 +04:00
|
|
|
/*
|
|
|
|
* Returns the number of contiguous clusters that can be used for an allocating
|
|
|
|
* write, but require COW to be performed (this includes yet unallocated space,
|
|
|
|
* which must copy from the backing file)
|
|
|
|
*/
|
2019-01-17 20:42:40 +03:00
|
|
|
static int count_cow_clusters(BlockDriverState *bs, int nb_clusters,
|
2018-02-05 17:33:35 +03:00
|
|
|
uint64_t *l2_slice, int l2_index)
|
2012-03-02 22:27:53 +04:00
|
|
|
{
|
2012-03-27 15:17:22 +04:00
|
|
|
int i;
|
2012-03-02 22:27:53 +04:00
|
|
|
|
2012-03-27 15:17:22 +04:00
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
2018-02-05 17:33:35 +03:00
|
|
|
uint64_t l2_entry = be64_to_cpu(l2_slice[l2_index + i]);
|
2019-01-17 20:42:40 +03:00
|
|
|
QCow2ClusterType cluster_type = qcow2_get_cluster_type(bs, l2_entry);
|
2012-03-27 15:17:22 +04:00
|
|
|
|
|
|
|
switch(cluster_type) {
|
|
|
|
case QCOW2_CLUSTER_NORMAL:
|
|
|
|
if (l2_entry & QCOW_OFLAG_COPIED) {
|
|
|
|
goto out;
|
|
|
|
}
|
2012-03-02 22:27:53 +04:00
|
|
|
break;
|
2012-03-27 15:17:22 +04:00
|
|
|
case QCOW2_CLUSTER_UNALLOCATED:
|
|
|
|
case QCOW2_CLUSTER_COMPRESSED:
|
2017-05-07 03:05:46 +03:00
|
|
|
case QCOW2_CLUSTER_ZERO_PLAIN:
|
|
|
|
case QCOW2_CLUSTER_ZERO_ALLOC:
|
2012-03-02 22:27:53 +04:00
|
|
|
break;
|
2012-03-27 15:17:22 +04:00
|
|
|
default:
|
|
|
|
abort();
|
|
|
|
}
|
2012-03-02 22:27:53 +04:00
|
|
|
}
|
|
|
|
|
2012-03-27 15:17:22 +04:00
|
|
|
out:
|
2012-03-02 22:27:53 +04:00
|
|
|
assert(i <= nb_clusters);
|
|
|
|
return i;
|
|
|
|
}
|
|
|
|
|
2012-03-02 17:10:54 +04:00
|
|
|
/*
|
2012-12-07 21:08:49 +04:00
|
|
|
* Check if there already is an AIO write request in flight which allocates
|
|
|
|
* the same cluster. In this case we need to wait until the previous
|
|
|
|
* request has completed and updated the L2 table accordingly.
|
2013-03-26 20:49:58 +04:00
|
|
|
*
|
|
|
|
* Returns:
|
|
|
|
* 0 if there was no dependency. *cur_bytes indicates the number of
|
|
|
|
* bytes from guest_offset that can be read before the next
|
|
|
|
* dependency must be processed (or the request is complete)
|
|
|
|
*
|
|
|
|
* -EAGAIN if we had to wait for another request, previously gathered
|
|
|
|
* information on cluster allocation may be invalid now. The caller
|
|
|
|
* must start over anyway, so consider *cur_bytes undefined.
|
2012-03-02 17:10:54 +04:00
|
|
|
*/
|
2012-12-07 21:08:49 +04:00
|
|
|
static int handle_dependencies(BlockDriverState *bs, uint64_t guest_offset,
|
2013-03-27 14:43:49 +04:00
|
|
|
uint64_t *cur_bytes, QCowL2Meta **m)
|
2012-03-02 17:10:54 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2012-03-02 17:10:54 +04:00
|
|
|
QCowL2Meta *old_alloc;
|
2013-03-26 20:49:58 +04:00
|
|
|
uint64_t bytes = *cur_bytes;
|
2012-03-02 17:10:54 +04:00
|
|
|
|
|
|
|
QLIST_FOREACH(old_alloc, &s->cluster_allocs, next_in_flight) {
|
|
|
|
|
2013-03-26 20:49:58 +04:00
|
|
|
uint64_t start = guest_offset;
|
|
|
|
uint64_t end = start + bytes;
|
|
|
|
uint64_t old_start = l2meta_cow_start(old_alloc);
|
|
|
|
uint64_t old_end = l2meta_cow_end(old_alloc);
|
2012-03-02 17:10:54 +04:00
|
|
|
|
2013-03-26 20:49:57 +04:00
|
|
|
if (end <= old_start || start >= old_end) {
|
2012-03-02 17:10:54 +04:00
|
|
|
/* No intersection */
|
|
|
|
} else {
|
|
|
|
if (start < old_start) {
|
|
|
|
/* Stop at the start of a running allocation */
|
2013-03-26 20:49:58 +04:00
|
|
|
bytes = old_start - start;
|
2012-03-02 17:10:54 +04:00
|
|
|
} else {
|
2013-03-26 20:49:58 +04:00
|
|
|
bytes = 0;
|
2012-03-02 17:10:54 +04:00
|
|
|
}
|
|
|
|
|
2013-03-27 14:43:49 +04:00
|
|
|
/* Stop if already an l2meta exists. After yielding, it wouldn't
|
|
|
|
* be valid any more, so we'd have to clean up the old L2Metas
|
|
|
|
* and deal with requests depending on them before starting to
|
|
|
|
* gather new ones. Not worth the trouble. */
|
|
|
|
if (bytes == 0 && *m) {
|
|
|
|
*cur_bytes = 0;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2013-03-26 20:49:58 +04:00
|
|
|
if (bytes == 0) {
|
2012-03-02 17:10:54 +04:00
|
|
|
/* Wait for the dependency to complete. We need to recheck
|
|
|
|
* the free/allocated clusters when we continue. */
|
2017-02-13 21:12:43 +03:00
|
|
|
qemu_co_queue_wait(&old_alloc->dependent_requests, &s->lock);
|
2012-03-02 17:10:54 +04:00
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-03-26 20:49:58 +04:00
|
|
|
/* Make sure that existing clusters and new allocations are only used up to
|
|
|
|
* the next dependency if we shortened the request above */
|
|
|
|
*cur_bytes = bytes;
|
2012-03-02 17:10:54 +04:00
|
|
|
|
2012-12-07 21:08:49 +04:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2013-03-26 20:50:05 +04:00
|
|
|
/*
|
|
|
|
* Checks how many already allocated clusters that don't require a copy on
|
2019-01-18 16:40:36 +03:00
|
|
|
* write there are at the given guest_offset (up to *bytes). If *host_offset is
|
|
|
|
* not INV_OFFSET, only physically contiguous clusters beginning at this host
|
|
|
|
* offset are counted.
|
2013-03-26 20:50:05 +04:00
|
|
|
*
|
2013-03-26 20:50:09 +04:00
|
|
|
* Note that guest_offset may not be cluster aligned. In this case, the
|
|
|
|
* returned *host_offset points to exact byte referenced by guest_offset and
|
|
|
|
* therefore isn't cluster aligned as well.
|
2013-03-26 20:50:05 +04:00
|
|
|
*
|
|
|
|
* Returns:
|
|
|
|
* 0: if no allocated clusters are available at the given offset.
|
|
|
|
* *bytes is normally unchanged. It is set to 0 if the cluster
|
|
|
|
* is allocated and doesn't need COW, but doesn't have the right
|
|
|
|
* physical offset.
|
|
|
|
*
|
|
|
|
* 1: if allocated clusters that don't require a COW are available at
|
|
|
|
* the requested offset. *bytes may have decreased and describes
|
|
|
|
* the length of the area that can be written to.
|
|
|
|
*
|
|
|
|
* -errno: in error cases
|
|
|
|
*/
|
|
|
|
static int handle_copied(BlockDriverState *bs, uint64_t guest_offset,
|
2013-03-26 20:50:07 +04:00
|
|
|
uint64_t *host_offset, uint64_t *bytes, QCowL2Meta **m)
|
2013-03-26 20:50:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-03-26 20:50:05 +04:00
|
|
|
int l2_index;
|
|
|
|
uint64_t cluster_offset;
|
2018-02-05 17:33:22 +03:00
|
|
|
uint64_t *l2_slice;
|
2015-09-14 17:39:47 +03:00
|
|
|
uint64_t nb_clusters;
|
2013-03-26 20:50:07 +04:00
|
|
|
unsigned int keep_clusters;
|
2015-05-11 15:54:58 +03:00
|
|
|
int ret;
|
2013-03-26 20:50:05 +04:00
|
|
|
|
|
|
|
trace_qcow2_handle_copied(qemu_coroutine_self(), guest_offset, *host_offset,
|
|
|
|
*bytes);
|
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
assert(*host_offset == INV_OFFSET || offset_into_cluster(s, guest_offset)
|
|
|
|
== offset_into_cluster(s, *host_offset));
|
2013-03-26 20:50:09 +04:00
|
|
|
|
2013-03-26 20:50:06 +04:00
|
|
|
/*
|
2018-02-05 17:33:22 +03:00
|
|
|
* Calculate the number of clusters to look for. We stop at L2 slice
|
2013-03-26 20:50:06 +04:00
|
|
|
* boundaries to keep things simple.
|
|
|
|
*/
|
|
|
|
nb_clusters =
|
|
|
|
size_to_clusters(s, offset_into_cluster(s, guest_offset) + *bytes);
|
|
|
|
|
2018-02-05 17:33:22 +03:00
|
|
|
l2_index = offset_to_l2_slice_index(s, guest_offset);
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_slice_size - l2_index);
|
2015-09-14 17:39:47 +03:00
|
|
|
assert(nb_clusters <= INT_MAX);
|
2013-03-26 20:50:06 +04:00
|
|
|
|
2013-03-26 20:50:05 +04:00
|
|
|
/* Find L2 entry for the first involved cluster */
|
2018-02-05 17:33:22 +03:00
|
|
|
ret = get_cluster_table(bs, guest_offset, &l2_slice, &l2_index);
|
2013-03-26 20:50:05 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:22 +03:00
|
|
|
cluster_offset = be64_to_cpu(l2_slice[l2_index]);
|
2013-03-26 20:50:05 +04:00
|
|
|
|
|
|
|
/* Check how many clusters are already allocated and don't need COW */
|
2019-01-17 20:42:40 +03:00
|
|
|
if (qcow2_get_cluster_type(bs, cluster_offset) == QCOW2_CLUSTER_NORMAL
|
2013-03-26 20:50:05 +04:00
|
|
|
&& (cluster_offset & QCOW_OFLAG_COPIED))
|
|
|
|
{
|
2013-03-26 20:50:08 +04:00
|
|
|
/* If a specific host_offset is required, check it */
|
|
|
|
bool offset_matches =
|
|
|
|
(cluster_offset & L2E_OFFSET_MASK) == *host_offset;
|
|
|
|
|
2014-09-05 18:07:18 +04:00
|
|
|
if (offset_into_cluster(s, cluster_offset & L2E_OFFSET_MASK)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "Data cluster offset "
|
|
|
|
"%#llx unaligned (guest offset: %#" PRIx64
|
|
|
|
")", cluster_offset & L2E_OFFSET_MASK,
|
|
|
|
guest_offset);
|
|
|
|
ret = -EIO;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
if (*host_offset != INV_OFFSET && !offset_matches) {
|
2013-03-26 20:50:08 +04:00
|
|
|
*bytes = 0;
|
|
|
|
ret = 0;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2013-03-26 20:50:05 +04:00
|
|
|
/* We keep all QCOW_OFLAG_COPIED clusters */
|
2013-03-26 20:50:07 +04:00
|
|
|
keep_clusters =
|
2019-01-17 20:42:40 +03:00
|
|
|
count_contiguous_clusters(bs, nb_clusters, s->cluster_size,
|
2018-02-05 17:33:22 +03:00
|
|
|
&l2_slice[l2_index],
|
2013-03-26 20:50:05 +04:00
|
|
|
QCOW_OFLAG_COPIED | QCOW_OFLAG_ZERO);
|
2013-03-26 20:50:07 +04:00
|
|
|
assert(keep_clusters <= nb_clusters);
|
|
|
|
|
|
|
|
*bytes = MIN(*bytes,
|
|
|
|
keep_clusters * s->cluster_size
|
|
|
|
- offset_into_cluster(s, guest_offset));
|
2013-03-26 20:50:05 +04:00
|
|
|
|
|
|
|
ret = 1;
|
|
|
|
} else {
|
|
|
|
ret = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Cleanup */
|
2013-03-26 20:50:08 +04:00
|
|
|
out:
|
2018-02-05 17:33:22 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2013-03-26 20:50:05 +04:00
|
|
|
|
2013-03-26 20:50:08 +04:00
|
|
|
/* Only return a host offset if we actually made progress. Otherwise we
|
|
|
|
* would make requirements for handle_alloc() that it can't fulfill */
|
2014-09-05 18:07:18 +04:00
|
|
|
if (ret > 0) {
|
2013-03-26 20:50:09 +04:00
|
|
|
*host_offset = (cluster_offset & L2E_OFFSET_MASK)
|
|
|
|
+ offset_into_cluster(s, guest_offset);
|
2013-03-26 20:50:08 +04:00
|
|
|
}
|
|
|
|
|
2013-03-26 20:50:05 +04:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-12-07 21:08:49 +04:00
|
|
|
/*
|
|
|
|
* Allocates new clusters for the given guest_offset.
|
|
|
|
*
|
|
|
|
* At most *nb_clusters are allocated, and on return *nb_clusters is updated to
|
|
|
|
* contain the number of clusters that have been allocated and are contiguous
|
|
|
|
* in the image file.
|
|
|
|
*
|
2019-01-18 16:40:36 +03:00
|
|
|
* If *host_offset is not INV_OFFSET, it specifies the offset in the image file
|
|
|
|
* at which the new clusters must start. *nb_clusters can be 0 on return in
|
|
|
|
* this case if the cluster at host_offset is already in use. If *host_offset
|
|
|
|
* is INV_OFFSET, the clusters can be allocated anywhere in the image file.
|
2012-12-07 21:08:49 +04:00
|
|
|
*
|
|
|
|
* *host_offset is updated to contain the offset into the image file at which
|
|
|
|
* the first allocated cluster starts.
|
|
|
|
*
|
|
|
|
* Return 0 on success and -errno in error cases. -EAGAIN means that the
|
|
|
|
* function has been waiting for another request and the allocation must be
|
|
|
|
* restarted, but the whole request should not be failed.
|
|
|
|
*/
|
|
|
|
static int do_alloc_cluster_offset(BlockDriverState *bs, uint64_t guest_offset,
|
2015-09-14 17:39:47 +03:00
|
|
|
uint64_t *host_offset, uint64_t *nb_clusters)
|
2012-12-07 21:08:49 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2012-12-07 21:08:49 +04:00
|
|
|
|
|
|
|
trace_qcow2_do_alloc_clusters_offset(qemu_coroutine_self(), guest_offset,
|
|
|
|
*host_offset, *nb_clusters);
|
|
|
|
|
2019-01-15 22:39:06 +03:00
|
|
|
if (has_data_file(bs)) {
|
|
|
|
assert(*host_offset == INV_OFFSET ||
|
|
|
|
*host_offset == start_of_cluster(s, guest_offset));
|
|
|
|
*host_offset = start_of_cluster(s, guest_offset);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-03-02 17:10:54 +04:00
|
|
|
/* Allocate new clusters */
|
|
|
|
trace_qcow2_cluster_alloc_phys(qemu_coroutine_self());
|
2019-01-18 16:40:36 +03:00
|
|
|
if (*host_offset == INV_OFFSET) {
|
2012-05-24 14:56:32 +04:00
|
|
|
int64_t cluster_offset =
|
|
|
|
qcow2_alloc_clusters(bs, *nb_clusters * s->cluster_size);
|
|
|
|
if (cluster_offset < 0) {
|
|
|
|
return cluster_offset;
|
|
|
|
}
|
|
|
|
*host_offset = cluster_offset;
|
|
|
|
return 0;
|
2012-03-02 17:10:54 +04:00
|
|
|
} else {
|
2015-09-14 17:39:47 +03:00
|
|
|
int64_t ret = qcow2_alloc_clusters_at(bs, *host_offset, *nb_clusters);
|
2012-05-24 14:56:32 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
*nb_clusters = ret;
|
|
|
|
return 0;
|
2012-03-02 17:10:54 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-03-26 20:50:00 +04:00
|
|
|
/*
|
|
|
|
* Allocates new clusters for an area that either is yet unallocated or needs a
|
2019-01-18 16:40:36 +03:00
|
|
|
* copy on write. If *host_offset is not INV_OFFSET, clusters are only
|
|
|
|
* allocated if the new allocation can match the specified host offset.
|
2013-03-26 20:50:00 +04:00
|
|
|
*
|
2013-03-26 20:50:09 +04:00
|
|
|
* Note that guest_offset may not be cluster aligned. In this case, the
|
|
|
|
* returned *host_offset points to exact byte referenced by guest_offset and
|
|
|
|
* therefore isn't cluster aligned as well.
|
2013-03-26 20:50:00 +04:00
|
|
|
*
|
|
|
|
* Returns:
|
|
|
|
* 0: if no clusters could be allocated. *bytes is set to 0,
|
|
|
|
* *host_offset is left unchanged.
|
|
|
|
*
|
|
|
|
* 1: if new clusters were allocated. *bytes may be decreased if the
|
|
|
|
* new allocation doesn't cover all of the requested area.
|
|
|
|
* *host_offset is updated to contain the host offset of the first
|
|
|
|
* newly allocated cluster.
|
|
|
|
*
|
|
|
|
* -errno: in error cases
|
|
|
|
*/
|
|
|
|
static int handle_alloc(BlockDriverState *bs, uint64_t guest_offset,
|
2013-03-26 20:50:03 +04:00
|
|
|
uint64_t *host_offset, uint64_t *bytes, QCowL2Meta **m)
|
2013-03-26 20:50:00 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-03-26 20:50:00 +04:00
|
|
|
int l2_index;
|
2018-02-05 17:33:23 +03:00
|
|
|
uint64_t *l2_slice;
|
2013-03-26 20:50:00 +04:00
|
|
|
uint64_t entry;
|
2015-09-14 17:39:47 +03:00
|
|
|
uint64_t nb_clusters;
|
2013-03-26 20:50:00 +04:00
|
|
|
int ret;
|
2017-05-04 02:11:18 +03:00
|
|
|
bool keep_old_clusters = false;
|
2013-03-26 20:50:00 +04:00
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
uint64_t alloc_cluster_offset = INV_OFFSET;
|
2013-03-26 20:50:00 +04:00
|
|
|
|
|
|
|
trace_qcow2_handle_alloc(qemu_coroutine_self(), guest_offset, *host_offset,
|
|
|
|
*bytes);
|
|
|
|
assert(*bytes > 0);
|
|
|
|
|
2013-03-26 20:50:01 +04:00
|
|
|
/*
|
2018-02-05 17:33:23 +03:00
|
|
|
* Calculate the number of clusters to look for. We stop at L2 slice
|
2013-03-26 20:50:01 +04:00
|
|
|
* boundaries to keep things simple.
|
|
|
|
*/
|
2013-03-26 20:50:03 +04:00
|
|
|
nb_clusters =
|
|
|
|
size_to_clusters(s, offset_into_cluster(s, guest_offset) + *bytes);
|
|
|
|
|
2018-02-05 17:33:23 +03:00
|
|
|
l2_index = offset_to_l2_slice_index(s, guest_offset);
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_slice_size - l2_index);
|
2015-09-14 17:39:47 +03:00
|
|
|
assert(nb_clusters <= INT_MAX);
|
2013-03-26 20:50:01 +04:00
|
|
|
|
2019-10-10 13:08:57 +03:00
|
|
|
/* Limit total allocation byte count to INT_MAX */
|
|
|
|
nb_clusters = MIN(nb_clusters, INT_MAX >> s->cluster_bits);
|
|
|
|
|
2013-03-26 20:50:00 +04:00
|
|
|
/* Find L2 entry for the first involved cluster */
|
2018-02-05 17:33:23 +03:00
|
|
|
ret = get_cluster_table(bs, guest_offset, &l2_slice, &l2_index);
|
2013-03-26 20:50:00 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:23 +03:00
|
|
|
entry = be64_to_cpu(l2_slice[l2_index]);
|
2019-09-11 18:16:26 +03:00
|
|
|
nb_clusters = count_cow_clusters(bs, nb_clusters, l2_slice, l2_index);
|
2013-03-26 20:50:00 +04:00
|
|
|
|
2013-03-27 14:43:49 +04:00
|
|
|
/* This function is only called when there were no non-COW clusters, so if
|
|
|
|
* we can't find any unallocated or COW clusters either, something is
|
|
|
|
* wrong with our code. */
|
|
|
|
assert(nb_clusters > 0);
|
|
|
|
|
2019-01-17 20:42:40 +03:00
|
|
|
if (qcow2_get_cluster_type(bs, entry) == QCOW2_CLUSTER_ZERO_ALLOC &&
|
2017-05-07 03:05:46 +03:00
|
|
|
(entry & QCOW_OFLAG_COPIED) &&
|
2019-01-18 16:40:36 +03:00
|
|
|
(*host_offset == INV_OFFSET ||
|
2017-05-04 02:11:18 +03:00
|
|
|
start_of_cluster(s, *host_offset) == (entry & L2E_OFFSET_MASK)))
|
|
|
|
{
|
2017-11-10 23:31:08 +03:00
|
|
|
int preallocated_nb_clusters;
|
|
|
|
|
|
|
|
if (offset_into_cluster(s, entry & L2E_OFFSET_MASK)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "Preallocated zero "
|
|
|
|
"cluster offset %#llx unaligned (guest "
|
|
|
|
"offset: %#" PRIx64 ")",
|
|
|
|
entry & L2E_OFFSET_MASK, guest_offset);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2017-05-04 02:11:18 +03:00
|
|
|
/* Try to reuse preallocated zero clusters; contiguous normal clusters
|
|
|
|
* would be fine, too, but count_cow_clusters() above has limited
|
|
|
|
* nb_clusters already to a range of COW clusters */
|
2017-11-10 23:31:08 +03:00
|
|
|
preallocated_nb_clusters =
|
2019-01-17 20:42:40 +03:00
|
|
|
count_contiguous_clusters(bs, nb_clusters, s->cluster_size,
|
2018-02-05 17:33:23 +03:00
|
|
|
&l2_slice[l2_index], QCOW_OFLAG_COPIED);
|
2017-05-04 02:11:18 +03:00
|
|
|
assert(preallocated_nb_clusters > 0);
|
2013-03-26 20:50:00 +04:00
|
|
|
|
2017-05-04 02:11:18 +03:00
|
|
|
nb_clusters = preallocated_nb_clusters;
|
|
|
|
alloc_cluster_offset = entry & L2E_OFFSET_MASK;
|
2013-03-26 20:50:00 +04:00
|
|
|
|
2017-05-04 02:11:18 +03:00
|
|
|
/* We want to reuse these clusters, so qcow2_alloc_cluster_link_l2()
|
|
|
|
* should not free them. */
|
|
|
|
keep_old_clusters = true;
|
2013-03-26 20:50:00 +04:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:23 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2017-05-04 02:11:18 +03:00
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
if (alloc_cluster_offset == INV_OFFSET) {
|
2017-05-04 02:11:18 +03:00
|
|
|
/* Allocate, if necessary at a given offset in the image file */
|
2019-01-18 16:40:36 +03:00
|
|
|
alloc_cluster_offset = *host_offset == INV_OFFSET ? INV_OFFSET :
|
|
|
|
start_of_cluster(s, *host_offset);
|
2017-05-04 02:11:18 +03:00
|
|
|
ret = do_alloc_cluster_offset(bs, guest_offset, &alloc_cluster_offset,
|
|
|
|
&nb_clusters);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Can't extend contiguous allocation */
|
|
|
|
if (nb_clusters == 0) {
|
|
|
|
*bytes = 0;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
assert(alloc_cluster_offset != INV_OFFSET);
|
2014-08-08 00:47:53 +04:00
|
|
|
}
|
|
|
|
|
2013-03-26 20:50:04 +04:00
|
|
|
/*
|
|
|
|
* Save info needed for meta data update.
|
|
|
|
*
|
2016-06-01 15:08:56 +03:00
|
|
|
* requested_bytes: Number of bytes from the start of the first
|
2013-03-26 20:50:04 +04:00
|
|
|
* newly allocated cluster to the end of the (possibly shortened
|
|
|
|
* before) write request.
|
|
|
|
*
|
2016-06-01 15:08:56 +03:00
|
|
|
* avail_bytes: Number of bytes from the start of the first
|
2013-03-26 20:50:04 +04:00
|
|
|
* newly allocated to the end of the last newly allocated cluster.
|
|
|
|
*
|
2016-06-01 15:08:56 +03:00
|
|
|
* nb_bytes: The number of bytes from the start of the first
|
2013-03-26 20:50:04 +04:00
|
|
|
* newly allocated cluster to the end of the area that the write
|
|
|
|
* request actually writes to (excluding COW at the end)
|
|
|
|
*/
|
2016-06-01 15:08:56 +03:00
|
|
|
uint64_t requested_bytes = *bytes + offset_into_cluster(s, guest_offset);
|
2019-10-10 13:08:57 +03:00
|
|
|
int avail_bytes = nb_clusters << s->cluster_bits;
|
2016-06-01 15:08:56 +03:00
|
|
|
int nb_bytes = MIN(requested_bytes, avail_bytes);
|
2013-03-26 20:50:11 +04:00
|
|
|
QCowL2Meta *old_m = *m;
|
2013-03-26 20:50:04 +04:00
|
|
|
|
|
|
|
*m = g_malloc0(sizeof(**m));
|
|
|
|
|
|
|
|
**m = (QCowL2Meta) {
|
2013-03-26 20:50:11 +04:00
|
|
|
.next = old_m,
|
|
|
|
|
2013-03-26 20:50:09 +04:00
|
|
|
.alloc_offset = alloc_cluster_offset,
|
2013-03-26 20:50:04 +04:00
|
|
|
.offset = start_of_cluster(s, guest_offset),
|
|
|
|
.nb_clusters = nb_clusters,
|
|
|
|
|
2017-05-04 02:11:18 +03:00
|
|
|
.keep_old_clusters = keep_old_clusters,
|
|
|
|
|
2013-03-26 20:50:04 +04:00
|
|
|
.cow_start = {
|
|
|
|
.offset = 0,
|
2016-06-01 15:08:56 +03:00
|
|
|
.nb_bytes = offset_into_cluster(s, guest_offset),
|
2013-03-26 20:50:04 +04:00
|
|
|
},
|
|
|
|
.cow_end = {
|
2016-06-01 15:08:56 +03:00
|
|
|
.offset = nb_bytes,
|
|
|
|
.nb_bytes = avail_bytes - nb_bytes,
|
2013-03-26 20:50:04 +04:00
|
|
|
},
|
|
|
|
};
|
|
|
|
qemu_co_queue_init(&(*m)->dependent_requests);
|
|
|
|
QLIST_INSERT_HEAD(&s->cluster_allocs, *m, next_in_flight);
|
|
|
|
|
2013-03-26 20:50:09 +04:00
|
|
|
*host_offset = alloc_cluster_offset + offset_into_cluster(s, guest_offset);
|
2016-06-01 15:08:56 +03:00
|
|
|
*bytes = MIN(*bytes, nb_bytes - offset_into_cluster(s, guest_offset));
|
2013-03-26 20:50:04 +04:00
|
|
|
assert(*bytes != 0);
|
|
|
|
|
2013-03-26 20:50:00 +04:00
|
|
|
return 1;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
if (*m && (*m)->nb_clusters > 0) {
|
|
|
|
QLIST_REMOVE(*m, next_in_flight);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-05-28 18:07:05 +04:00
|
|
|
/*
|
|
|
|
* alloc_cluster_offset
|
|
|
|
*
|
2012-03-02 17:10:54 +04:00
|
|
|
* For a given offset on the virtual disk, find the cluster offset in qcow2
|
|
|
|
* file. If the offset is not found, allocate a new cluster.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2012-03-02 17:10:54 +04:00
|
|
|
* If the cluster was already allocated, m->nb_clusters is set to 0 and
|
2011-09-07 18:19:17 +04:00
|
|
|
* other fields in m are meaningless.
|
2010-01-20 17:03:01 +03:00
|
|
|
*
|
|
|
|
* If the cluster is newly allocated, m->nb_clusters is set to the number of
|
2011-06-30 19:42:09 +04:00
|
|
|
* contiguous clusters that have been allocated. In this case, the other
|
|
|
|
* fields of m are valid and contain information about the first allocated
|
|
|
|
* cluster.
|
2009-05-28 18:07:05 +04:00
|
|
|
*
|
2011-06-30 19:42:09 +04:00
|
|
|
* If the request conflicts with another write request in flight, the coroutine
|
|
|
|
* is queued and will be reentered when the dependency has completed.
|
2010-01-20 17:03:01 +03:00
|
|
|
*
|
|
|
|
* Return 0 on success and -errno in error cases
|
2009-05-28 18:07:05 +04:00
|
|
|
*/
|
2010-02-02 17:20:57 +03:00
|
|
|
int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
|
2016-06-01 17:55:05 +03:00
|
|
|
unsigned int *bytes, uint64_t *host_offset,
|
|
|
|
QCowL2Meta **m)
|
2009-05-28 18:07:05 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-03-26 20:50:10 +04:00
|
|
|
uint64_t start, remaining;
|
2012-03-02 17:10:54 +04:00
|
|
|
uint64_t cluster_offset;
|
2013-03-26 20:49:58 +04:00
|
|
|
uint64_t cur_bytes;
|
2013-03-26 20:50:10 +04:00
|
|
|
int ret;
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2016-06-01 17:55:05 +03:00
|
|
|
trace_qcow2_alloc_clusters_offset(qemu_coroutine_self(), offset, *bytes);
|
2013-03-26 20:50:10 +04:00
|
|
|
|
2012-04-24 18:10:56 +04:00
|
|
|
again:
|
2014-01-26 07:12:37 +04:00
|
|
|
start = offset;
|
2016-06-01 17:55:05 +03:00
|
|
|
remaining = *bytes;
|
2019-01-18 16:40:36 +03:00
|
|
|
cluster_offset = INV_OFFSET;
|
|
|
|
*host_offset = INV_OFFSET;
|
2013-03-27 14:43:49 +04:00
|
|
|
cur_bytes = 0;
|
|
|
|
*m = NULL;
|
2013-03-26 20:50:05 +04:00
|
|
|
|
2013-03-26 20:50:12 +04:00
|
|
|
while (true) {
|
2013-03-27 14:43:49 +04:00
|
|
|
|
2019-01-18 16:40:36 +03:00
|
|
|
if (*host_offset == INV_OFFSET && cluster_offset != INV_OFFSET) {
|
2013-03-27 14:43:49 +04:00
|
|
|
*host_offset = start_of_cluster(s, cluster_offset);
|
|
|
|
}
|
|
|
|
|
|
|
|
assert(remaining >= cur_bytes);
|
|
|
|
|
|
|
|
start += cur_bytes;
|
|
|
|
remaining -= cur_bytes;
|
2019-01-18 16:40:36 +03:00
|
|
|
|
|
|
|
if (cluster_offset != INV_OFFSET) {
|
|
|
|
cluster_offset += cur_bytes;
|
|
|
|
}
|
2013-03-27 14:43:49 +04:00
|
|
|
|
|
|
|
if (remaining == 0) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
cur_bytes = remaining;
|
|
|
|
|
2013-03-26 20:50:12 +04:00
|
|
|
/*
|
|
|
|
* Now start gathering as many contiguous clusters as possible:
|
|
|
|
*
|
|
|
|
* 1. Check for overlaps with in-flight allocations
|
|
|
|
*
|
|
|
|
* a) Overlap not in the first cluster -> shorten this request and
|
|
|
|
* let the caller handle the rest in its next loop iteration.
|
|
|
|
*
|
|
|
|
* b) Real overlaps of two requests. Yield and restart the search
|
|
|
|
* for contiguous clusters (the situation could have changed
|
|
|
|
* while we were sleeping)
|
|
|
|
*
|
|
|
|
* c) TODO: Request starts in the same cluster as the in-flight
|
|
|
|
* allocation ends. Shorten the COW of the in-fight allocation,
|
|
|
|
* set cluster_offset to write to the same cluster and set up
|
|
|
|
* the right synchronisation between the in-flight request and
|
|
|
|
* the new one.
|
|
|
|
*/
|
2013-03-27 14:43:49 +04:00
|
|
|
ret = handle_dependencies(bs, start, &cur_bytes, m);
|
2013-03-26 20:50:12 +04:00
|
|
|
if (ret == -EAGAIN) {
|
2013-03-27 14:43:49 +04:00
|
|
|
/* Currently handle_dependencies() doesn't yield if we already had
|
|
|
|
* an allocation. If it did, we would have to clean up the L2Meta
|
|
|
|
* structs before starting over. */
|
|
|
|
assert(*m == NULL);
|
2013-03-26 20:50:12 +04:00
|
|
|
goto again;
|
|
|
|
} else if (ret < 0) {
|
|
|
|
return ret;
|
2013-03-27 14:43:49 +04:00
|
|
|
} else if (cur_bytes == 0) {
|
|
|
|
break;
|
2013-03-26 20:50:12 +04:00
|
|
|
} else {
|
|
|
|
/* handle_dependencies() may have decreased cur_bytes (shortened
|
|
|
|
* the allocations below) so that the next dependency is processed
|
|
|
|
* correctly during the next loop iteration. */
|
2013-03-26 20:50:05 +04:00
|
|
|
}
|
2013-03-26 20:50:10 +04:00
|
|
|
|
2013-03-26 20:50:12 +04:00
|
|
|
/*
|
|
|
|
* 2. Count contiguous COPIED clusters.
|
|
|
|
*/
|
|
|
|
ret = handle_copied(bs, start, &cluster_offset, &cur_bytes, m);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
} else if (ret) {
|
2013-03-27 14:43:49 +04:00
|
|
|
continue;
|
2013-03-26 20:50:12 +04:00
|
|
|
} else if (cur_bytes == 0) {
|
|
|
|
break;
|
|
|
|
}
|
2012-12-07 21:08:45 +04:00
|
|
|
|
2013-03-26 20:50:12 +04:00
|
|
|
/*
|
|
|
|
* 3. If the request still hasn't completed, allocate new clusters,
|
|
|
|
* considering any cluster_offset of steps 1c or 2.
|
|
|
|
*/
|
|
|
|
ret = handle_alloc(bs, start, &cluster_offset, &cur_bytes, m);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
} else if (ret) {
|
2013-03-27 14:43:49 +04:00
|
|
|
continue;
|
2013-03-26 20:50:12 +04:00
|
|
|
} else {
|
|
|
|
assert(cur_bytes == 0);
|
|
|
|
break;
|
|
|
|
}
|
2013-03-26 20:50:01 +04:00
|
|
|
}
|
2013-03-26 20:50:00 +04:00
|
|
|
|
2016-06-01 17:55:05 +03:00
|
|
|
*bytes -= remaining;
|
|
|
|
assert(*bytes > 0);
|
2019-01-18 16:40:36 +03:00
|
|
|
assert(*host_offset != INV_OFFSET);
|
2009-05-28 18:07:05 +04:00
|
|
|
|
2010-01-20 17:03:01 +03:00
|
|
|
return 0;
|
2009-05-28 18:07:05 +04:00
|
|
|
}
|
|
|
|
|
2011-01-26 18:56:48 +03:00
|
|
|
/*
|
|
|
|
* This discards as many clusters of nb_clusters as possible at once (i.e.
|
2018-02-05 17:33:24 +03:00
|
|
|
* all clusters in the same L2 slice) and returns the number of discarded
|
2011-01-26 18:56:48 +03:00
|
|
|
* clusters.
|
|
|
|
*/
|
2018-02-05 17:33:24 +03:00
|
|
|
static int discard_in_l2_slice(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t nb_clusters,
|
|
|
|
enum qcow2_discard_type type, bool full_discard)
|
2011-01-26 18:56:48 +03:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2018-02-05 17:33:24 +03:00
|
|
|
uint64_t *l2_slice;
|
2011-01-26 18:56:48 +03:00
|
|
|
int l2_index;
|
|
|
|
int ret;
|
|
|
|
int i;
|
|
|
|
|
2018-02-05 17:33:24 +03:00
|
|
|
ret = get_cluster_table(bs, offset, &l2_slice, &l2_index);
|
2011-01-26 18:56:48 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:24 +03:00
|
|
|
/* Limit nb_clusters to one L2 slice */
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_slice_size - l2_index);
|
2015-09-14 17:39:47 +03:00
|
|
|
assert(nb_clusters <= INT_MAX);
|
2011-01-26 18:56:48 +03:00
|
|
|
|
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
2014-04-29 18:12:30 +04:00
|
|
|
uint64_t old_l2_entry;
|
2011-01-26 18:56:48 +03:00
|
|
|
|
2018-02-05 17:33:24 +03:00
|
|
|
old_l2_entry = be64_to_cpu(l2_slice[l2_index + i]);
|
2014-02-08 17:38:33 +04:00
|
|
|
|
|
|
|
/*
|
2014-10-24 17:57:30 +04:00
|
|
|
* If full_discard is false, make sure that a discarded area reads back
|
|
|
|
* as zeroes for v3 images (we cannot do it for v2 without actually
|
|
|
|
* writing a zero-filled buffer). We can skip the operation if the
|
|
|
|
* cluster is already marked as zero, or if it's unallocated and we
|
|
|
|
* don't have a backing file.
|
2014-02-08 17:38:33 +04:00
|
|
|
*
|
block: Convert bdrv_get_block_status() to bytes
We are gradually moving away from sector-based interfaces, towards
byte-based. In the common case, allocation is unlikely to ever use
values that are not naturally sector-aligned, but it is possible
that byte-based values will let us be more precise about allocation
at the end of an unaligned file that can do byte-based access.
Changing the name of the function from bdrv_get_block_status() to
bdrv_block_status() ensures that the compiler enforces that all
callers are updated. For now, the io.c layer still assert()s that
all callers are sector-aligned, but that can be relaxed when a later
patch implements byte-based block status in the drivers.
There was an inherent limitation in returning the offset via the
return value: we only have room for BDRV_BLOCK_OFFSET_MASK bits, which
means an offset can only be mapped for sector-aligned queries (or,
if we declare that non-aligned input is at the same relative position
modulo 512 of the answer), so the new interface also changes things to
return the offset via output through a parameter by reference rather
than mashed into the return value. We'll have some glue code that
munges between the two styles until we finish converting all uses.
For the most part this patch is just the addition of scaling at the
callers followed by inverse scaling at bdrv_block_status(), coupled
with the tweak in calling convention. But some code, particularly
bdrv_is_allocated(), gets a lot simpler because it no longer has to
mess with sectors.
For ease of review, bdrv_get_block_status_above() will be tackled
separately.
Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2017-10-12 06:47:03 +03:00
|
|
|
* TODO We might want to use bdrv_block_status(bs) here, but we're
|
2014-02-08 17:38:33 +04:00
|
|
|
* holding s->lock, so that doesn't work today.
|
2014-10-24 17:57:30 +04:00
|
|
|
*
|
|
|
|
* If full_discard is true, the sector should not read back as zeroes,
|
|
|
|
* but rather fall through to the backing file.
|
2014-02-08 17:38:33 +04:00
|
|
|
*/
|
2019-01-17 20:42:40 +03:00
|
|
|
switch (qcow2_get_cluster_type(bs, old_l2_entry)) {
|
2017-05-07 03:05:42 +03:00
|
|
|
case QCOW2_CLUSTER_UNALLOCATED:
|
|
|
|
if (full_discard || !bs->backing) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
2017-05-07 03:05:46 +03:00
|
|
|
case QCOW2_CLUSTER_ZERO_PLAIN:
|
|
|
|
if (!full_discard) {
|
2017-05-07 03:05:42 +03:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
2017-05-07 03:05:46 +03:00
|
|
|
case QCOW2_CLUSTER_ZERO_ALLOC:
|
2017-05-07 03:05:42 +03:00
|
|
|
case QCOW2_CLUSTER_NORMAL:
|
|
|
|
case QCOW2_CLUSTER_COMPRESSED:
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
abort();
|
2011-01-26 18:56:48 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
/* First remove L2 entries */
|
2018-02-05 17:33:24 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
2014-10-24 17:57:30 +04:00
|
|
|
if (!full_discard && s->qcow_version >= 3) {
|
2018-02-05 17:33:24 +03:00
|
|
|
l2_slice[l2_index + i] = cpu_to_be64(QCOW_OFLAG_ZERO);
|
2014-02-08 17:38:33 +04:00
|
|
|
} else {
|
2018-02-05 17:33:24 +03:00
|
|
|
l2_slice[l2_index + i] = cpu_to_be64(0);
|
2014-02-08 17:38:33 +04:00
|
|
|
}
|
2011-01-26 18:56:48 +03:00
|
|
|
|
|
|
|
/* Then decrease the refcount */
|
2014-04-29 18:12:30 +04:00
|
|
|
qcow2_free_any_clusters(bs, old_l2_entry, 1, type);
|
2011-01-26 18:56:48 +03:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:24 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2011-01-26 18:56:48 +03:00
|
|
|
|
|
|
|
return nb_clusters;
|
|
|
|
}
|
|
|
|
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
int qcow2_cluster_discard(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t bytes, enum qcow2_discard_type type,
|
|
|
|
bool full_discard)
|
2011-01-26 18:56:48 +03:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
uint64_t end_offset = offset + bytes;
|
2015-09-14 17:39:47 +03:00
|
|
|
uint64_t nb_clusters;
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
int64_t cleared;
|
2011-01-26 18:56:48 +03:00
|
|
|
int ret;
|
|
|
|
|
2017-05-07 03:05:51 +03:00
|
|
|
/* Caller must pass aligned values, except at image end */
|
2017-03-31 21:53:55 +03:00
|
|
|
assert(QEMU_IS_ALIGNED(offset, s->cluster_size));
|
2017-05-07 03:05:51 +03:00
|
|
|
assert(QEMU_IS_ALIGNED(end_offset, s->cluster_size) ||
|
|
|
|
end_offset == bs->total_sectors << BDRV_SECTOR_BITS);
|
2011-01-26 18:56:48 +03:00
|
|
|
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
nb_clusters = size_to_clusters(s, bytes);
|
2011-01-26 18:56:48 +03:00
|
|
|
|
2013-06-19 15:44:20 +04:00
|
|
|
s->cache_discards = true;
|
|
|
|
|
2018-02-05 17:33:24 +03:00
|
|
|
/* Each L2 slice is handled by its own loop iteration */
|
2011-01-26 18:56:48 +03:00
|
|
|
while (nb_clusters > 0) {
|
2018-02-05 17:33:24 +03:00
|
|
|
cleared = discard_in_l2_slice(bs, offset, nb_clusters, type,
|
|
|
|
full_discard);
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
if (cleared < 0) {
|
|
|
|
ret = cleared;
|
2013-06-19 15:44:20 +04:00
|
|
|
goto fail;
|
2011-01-26 18:56:48 +03:00
|
|
|
}
|
|
|
|
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
nb_clusters -= cleared;
|
|
|
|
offset += (cleared * s->cluster_size);
|
2011-01-26 18:56:48 +03:00
|
|
|
}
|
|
|
|
|
2013-06-19 15:44:20 +04:00
|
|
|
ret = 0;
|
|
|
|
fail:
|
|
|
|
s->cache_discards = false;
|
|
|
|
qcow2_process_discards(bs, ret);
|
|
|
|
|
|
|
|
return ret;
|
2011-01-26 18:56:48 +03:00
|
|
|
}
|
2012-03-20 18:12:58 +04:00
|
|
|
|
|
|
|
/*
|
|
|
|
* This zeroes as many clusters of nb_clusters as possible at once (i.e.
|
2018-02-05 17:33:25 +03:00
|
|
|
* all clusters in the same L2 slice) and returns the number of zeroed
|
2012-03-20 18:12:58 +04:00
|
|
|
* clusters.
|
|
|
|
*/
|
2018-02-05 17:33:25 +03:00
|
|
|
static int zero_in_l2_slice(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t nb_clusters, int flags)
|
2012-03-20 18:12:58 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2018-02-05 17:33:25 +03:00
|
|
|
uint64_t *l2_slice;
|
2012-03-20 18:12:58 +04:00
|
|
|
int l2_index;
|
|
|
|
int ret;
|
|
|
|
int i;
|
qcow2: Optimize zero_single_l2() to minimize L2 churn
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:47 +03:00
|
|
|
bool unmap = !!(flags & BDRV_REQ_MAY_UNMAP);
|
2012-03-20 18:12:58 +04:00
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
ret = get_cluster_table(bs, offset, &l2_slice, &l2_index);
|
2012-03-20 18:12:58 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
/* Limit nb_clusters to one L2 slice */
|
|
|
|
nb_clusters = MIN(nb_clusters, s->l2_slice_size - l2_index);
|
2015-09-14 17:39:47 +03:00
|
|
|
assert(nb_clusters <= INT_MAX);
|
2012-03-20 18:12:58 +04:00
|
|
|
|
|
|
|
for (i = 0; i < nb_clusters; i++) {
|
|
|
|
uint64_t old_offset;
|
qcow2: Optimize zero_single_l2() to minimize L2 churn
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:47 +03:00
|
|
|
QCow2ClusterType cluster_type;
|
2012-03-20 18:12:58 +04:00
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
old_offset = be64_to_cpu(l2_slice[l2_index + i]);
|
2012-03-20 18:12:58 +04:00
|
|
|
|
qcow2: Optimize zero_single_l2() to minimize L2 churn
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:47 +03:00
|
|
|
/*
|
|
|
|
* Minimize L2 changes if the cluster already reads back as
|
|
|
|
* zeroes with correct allocation.
|
|
|
|
*/
|
2019-01-17 20:42:40 +03:00
|
|
|
cluster_type = qcow2_get_cluster_type(bs, old_offset);
|
qcow2: Optimize zero_single_l2() to minimize L2 churn
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:47 +03:00
|
|
|
if (cluster_type == QCOW2_CLUSTER_ZERO_PLAIN ||
|
|
|
|
(cluster_type == QCOW2_CLUSTER_ZERO_ALLOC && !unmap)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
qcow2: Optimize zero_single_l2() to minimize L2 churn
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:47 +03:00
|
|
|
if (cluster_type == QCOW2_CLUSTER_COMPRESSED || unmap) {
|
2018-02-05 17:33:25 +03:00
|
|
|
l2_slice[l2_index + i] = cpu_to_be64(QCOW_OFLAG_ZERO);
|
2013-06-19 15:44:18 +04:00
|
|
|
qcow2_free_any_clusters(bs, old_offset, 1, QCOW2_DISCARD_REQUEST);
|
2012-03-20 18:12:58 +04:00
|
|
|
} else {
|
2018-02-05 17:33:25 +03:00
|
|
|
l2_slice[l2_index + i] |= cpu_to_be64(QCOW_OFLAG_ZERO);
|
2012-03-20 18:12:58 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2012-03-20 18:12:58 +04:00
|
|
|
|
|
|
|
return nb_clusters;
|
|
|
|
}
|
|
|
|
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
int qcow2_cluster_zeroize(BlockDriverState *bs, uint64_t offset,
|
|
|
|
uint64_t bytes, int flags)
|
2012-03-20 18:12:58 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
uint64_t end_offset = offset + bytes;
|
2015-09-14 17:39:47 +03:00
|
|
|
uint64_t nb_clusters;
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
int64_t cleared;
|
2012-03-20 18:12:58 +04:00
|
|
|
int ret;
|
|
|
|
|
2019-02-22 16:29:38 +03:00
|
|
|
/* If we have to stay in sync with an external data file, zero out
|
|
|
|
* s->data_file first. */
|
|
|
|
if (data_file_is_raw(bs)) {
|
|
|
|
assert(has_data_file(bs));
|
|
|
|
ret = bdrv_co_pwrite_zeroes(s->data_file, offset, bytes, flags);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-05-07 03:05:51 +03:00
|
|
|
/* Caller must pass aligned values, except at image end */
|
|
|
|
assert(QEMU_IS_ALIGNED(offset, s->cluster_size));
|
|
|
|
assert(QEMU_IS_ALIGNED(end_offset, s->cluster_size) ||
|
2020-04-24 15:54:42 +03:00
|
|
|
end_offset >= bs->total_sectors << BDRV_SECTOR_BITS);
|
2017-05-07 03:05:51 +03:00
|
|
|
|
2012-03-20 18:12:58 +04:00
|
|
|
/* The zero flag is only supported by version 3 and newer */
|
|
|
|
if (s->qcow_version < 3) {
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:25 +03:00
|
|
|
/* Each L2 slice is handled by its own loop iteration */
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
nb_clusters = size_to_clusters(s, bytes);
|
2012-03-20 18:12:58 +04:00
|
|
|
|
2013-06-19 15:44:20 +04:00
|
|
|
s->cache_discards = true;
|
|
|
|
|
2012-03-20 18:12:58 +04:00
|
|
|
while (nb_clusters > 0) {
|
2018-02-05 17:33:25 +03:00
|
|
|
cleared = zero_in_l2_slice(bs, offset, nb_clusters, flags);
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
if (cleared < 0) {
|
|
|
|
ret = cleared;
|
2013-06-19 15:44:20 +04:00
|
|
|
goto fail;
|
2012-03-20 18:12:58 +04:00
|
|
|
}
|
|
|
|
|
qcow2: Discard/zero clusters by byte count
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
2017-05-07 03:05:52 +03:00
|
|
|
nb_clusters -= cleared;
|
|
|
|
offset += (cleared * s->cluster_size);
|
2012-03-20 18:12:58 +04:00
|
|
|
}
|
|
|
|
|
2013-06-19 15:44:20 +04:00
|
|
|
ret = 0;
|
|
|
|
fail:
|
|
|
|
s->cache_discards = false;
|
|
|
|
qcow2_process_discards(bs, ret);
|
|
|
|
|
|
|
|
return ret;
|
2012-03-20 18:12:58 +04:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Expands all zero clusters in a specific L1 table (or deallocates them, for
|
|
|
|
* non-backed non-pre-allocated zero clusters).
|
|
|
|
*
|
2014-10-27 13:12:53 +03:00
|
|
|
* l1_entries and *visited_l1_entries are used to keep track of progress for
|
|
|
|
* status_cb(). l1_entries contains the total number of L1 entries and
|
|
|
|
* *visited_l1_entries counts all visited L1 entries.
|
2013-09-03 12:09:52 +04:00
|
|
|
*/
|
|
|
|
static int expand_zero_clusters_in_l1(BlockDriverState *bs, uint64_t *l1_table,
|
2014-10-27 13:12:55 +03:00
|
|
|
int l1_size, int64_t *visited_l1_entries,
|
2014-10-27 13:12:53 +03:00
|
|
|
int64_t l1_entries,
|
2015-07-27 18:51:32 +03:00
|
|
|
BlockDriverAmendStatusCB *status_cb,
|
|
|
|
void *cb_opaque)
|
2013-09-03 12:09:52 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-09-03 12:09:52 +04:00
|
|
|
bool is_active_l1 = (l1_table == s->l1_table);
|
2018-02-05 17:33:30 +03:00
|
|
|
uint64_t *l2_slice = NULL;
|
|
|
|
unsigned slice, slice_size2, n_slices;
|
2013-09-03 12:09:52 +04:00
|
|
|
int ret;
|
|
|
|
int i, j;
|
|
|
|
|
2018-02-05 17:33:30 +03:00
|
|
|
slice_size2 = s->l2_slice_size * sizeof(uint64_t);
|
|
|
|
n_slices = s->cluster_size / slice_size2;
|
|
|
|
|
2013-09-03 12:09:52 +04:00
|
|
|
if (!is_active_l1) {
|
|
|
|
/* inactive L2 tables require a buffer to be stored in when loading
|
|
|
|
* them from disk */
|
2018-02-05 17:33:30 +03:00
|
|
|
l2_slice = qemu_try_blockalign(bs->file->bs, slice_size2);
|
|
|
|
if (l2_slice == NULL) {
|
2014-05-20 19:12:47 +04:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < l1_size; i++) {
|
|
|
|
uint64_t l2_offset = l1_table[i] & L1E_OFFSET_MASK;
|
2015-02-10 23:28:48 +03:00
|
|
|
uint64_t l2_refcount;
|
2013-09-03 12:09:52 +04:00
|
|
|
|
|
|
|
if (!l2_offset) {
|
|
|
|
/* unallocated */
|
2014-10-27 13:12:53 +03:00
|
|
|
(*visited_l1_entries)++;
|
|
|
|
if (status_cb) {
|
2015-07-27 18:51:32 +03:00
|
|
|
status_cb(bs, *visited_l1_entries, l1_entries, cb_opaque);
|
2014-10-27 13:12:53 +03:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2015-01-19 23:49:03 +03:00
|
|
|
if (offset_into_cluster(s, l2_offset)) {
|
|
|
|
qcow2_signal_corruption(bs, true, -1, -1, "L2 table offset %#"
|
|
|
|
PRIx64 " unaligned (L1 index: %#x)",
|
|
|
|
l2_offset, i);
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:28 +03:00
|
|
|
ret = qcow2_get_refcount(bs, l2_offset >> s->cluster_bits,
|
|
|
|
&l2_refcount);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:30 +03:00
|
|
|
for (slice = 0; slice < n_slices; slice++) {
|
|
|
|
uint64_t slice_offset = l2_offset + slice * slice_size2;
|
|
|
|
bool l2_dirty = false;
|
2018-02-05 17:33:29 +03:00
|
|
|
if (is_active_l1) {
|
|
|
|
/* get active L2 tables from cache */
|
2018-02-05 17:33:30 +03:00
|
|
|
ret = qcow2_cache_get(bs, s->l2_table_cache, slice_offset,
|
|
|
|
(void **)&l2_slice);
|
2018-02-05 17:33:29 +03:00
|
|
|
} else {
|
|
|
|
/* load inactive L2 tables from disk */
|
2018-02-05 17:33:30 +03:00
|
|
|
ret = bdrv_pread(bs->file, slice_offset, l2_slice, slice_size2);
|
2018-02-05 17:33:29 +03:00
|
|
|
}
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:30 +03:00
|
|
|
for (j = 0; j < s->l2_slice_size; j++) {
|
|
|
|
uint64_t l2_entry = be64_to_cpu(l2_slice[j]);
|
2018-02-05 17:33:29 +03:00
|
|
|
int64_t offset = l2_entry & L2E_OFFSET_MASK;
|
|
|
|
QCow2ClusterType cluster_type =
|
2019-01-17 20:42:40 +03:00
|
|
|
qcow2_get_cluster_type(bs, l2_entry);
|
2018-02-05 17:33:29 +03:00
|
|
|
|
|
|
|
if (cluster_type != QCOW2_CLUSTER_ZERO_PLAIN &&
|
|
|
|
cluster_type != QCOW2_CLUSTER_ZERO_ALLOC) {
|
2013-09-03 12:09:52 +04:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
if (cluster_type == QCOW2_CLUSTER_ZERO_PLAIN) {
|
|
|
|
if (!bs->backing) {
|
|
|
|
/* not backed; therefore we can simply deallocate the
|
|
|
|
* cluster */
|
2018-02-05 17:33:30 +03:00
|
|
|
l2_slice[j] = 0;
|
2018-02-05 17:33:29 +03:00
|
|
|
l2_dirty = true;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
offset = qcow2_alloc_clusters(bs, s->cluster_size);
|
|
|
|
if (offset < 0) {
|
|
|
|
ret = offset;
|
|
|
|
goto fail;
|
|
|
|
}
|
2014-10-27 13:12:55 +03:00
|
|
|
|
2020-01-13 19:11:46 +03:00
|
|
|
/* The offset must fit in the offset field */
|
|
|
|
assert((offset & L2E_OFFSET_MASK) == offset);
|
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
if (l2_refcount > 1) {
|
|
|
|
/* For shared L2 tables, set the refcount accordingly
|
|
|
|
* (it is already 1 and needs to be l2_refcount) */
|
|
|
|
ret = qcow2_update_cluster_refcount(
|
|
|
|
bs, offset >> s->cluster_bits,
|
2015-02-10 23:28:47 +03:00
|
|
|
refcount_diff(1, l2_refcount), false,
|
2014-10-27 13:12:55 +03:00
|
|
|
QCOW2_DISCARD_OTHER);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
qcow2_free_clusters(bs, offset, s->cluster_size,
|
|
|
|
QCOW2_DISCARD_OTHER);
|
|
|
|
goto fail;
|
|
|
|
}
|
2014-10-27 13:12:55 +03:00
|
|
|
}
|
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
if (offset_into_cluster(s, offset)) {
|
2018-02-05 17:33:30 +03:00
|
|
|
int l2_index = slice * s->l2_slice_size + j;
|
2018-02-05 17:33:29 +03:00
|
|
|
qcow2_signal_corruption(
|
|
|
|
bs, true, -1, -1,
|
|
|
|
"Cluster allocation offset "
|
|
|
|
"%#" PRIx64 " unaligned (L2 offset: %#"
|
|
|
|
PRIx64 ", L2 index: %#x)", offset,
|
2018-02-05 17:33:30 +03:00
|
|
|
l2_offset, l2_index);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (cluster_type == QCOW2_CLUSTER_ZERO_PLAIN) {
|
|
|
|
qcow2_free_clusters(bs, offset, s->cluster_size,
|
|
|
|
QCOW2_DISCARD_ALWAYS);
|
|
|
|
}
|
|
|
|
ret = -EIO;
|
|
|
|
goto fail;
|
2015-01-19 23:49:03 +03:00
|
|
|
}
|
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
ret = qcow2_pre_write_overlap_check(bs, 0, offset,
|
2019-01-15 22:39:06 +03:00
|
|
|
s->cluster_size, true);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
if (cluster_type == QCOW2_CLUSTER_ZERO_PLAIN) {
|
|
|
|
qcow2_free_clusters(bs, offset, s->cluster_size,
|
|
|
|
QCOW2_DISCARD_ALWAYS);
|
|
|
|
}
|
|
|
|
goto fail;
|
2013-09-27 12:21:48 +04:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2019-01-15 22:39:06 +03:00
|
|
|
ret = bdrv_pwrite_zeroes(s->data_file, offset,
|
|
|
|
s->cluster_size, 0);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
if (cluster_type == QCOW2_CLUSTER_ZERO_PLAIN) {
|
|
|
|
qcow2_free_clusters(bs, offset, s->cluster_size,
|
|
|
|
QCOW2_DISCARD_ALWAYS);
|
|
|
|
}
|
|
|
|
goto fail;
|
2013-09-27 12:21:48 +04:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
if (l2_refcount == 1) {
|
2018-02-05 17:33:30 +03:00
|
|
|
l2_slice[j] = cpu_to_be64(offset | QCOW_OFLAG_COPIED);
|
2018-02-05 17:33:29 +03:00
|
|
|
} else {
|
2018-02-05 17:33:30 +03:00
|
|
|
l2_slice[j] = cpu_to_be64(offset);
|
2018-02-05 17:33:29 +03:00
|
|
|
}
|
|
|
|
l2_dirty = true;
|
2013-09-25 14:07:22 +04:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-02-05 17:33:29 +03:00
|
|
|
if (is_active_l1) {
|
|
|
|
if (l2_dirty) {
|
2018-02-05 17:33:30 +03:00
|
|
|
qcow2_cache_entry_mark_dirty(s->l2_table_cache, l2_slice);
|
2018-02-05 17:33:29 +03:00
|
|
|
qcow2_cache_depends_on_flush(s->l2_table_cache);
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
2018-02-05 17:33:30 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2018-02-05 17:33:29 +03:00
|
|
|
} else {
|
|
|
|
if (l2_dirty) {
|
|
|
|
ret = qcow2_pre_write_overlap_check(
|
|
|
|
bs, QCOW2_OL_INACTIVE_L2 | QCOW2_OL_ACTIVE_L2,
|
2019-01-15 22:39:06 +03:00
|
|
|
slice_offset, slice_size2, false);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-02-05 17:33:30 +03:00
|
|
|
ret = bdrv_pwrite(bs->file, slice_offset,
|
|
|
|
l2_slice, slice_size2);
|
2018-02-05 17:33:29 +03:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2014-10-27 13:12:53 +03:00
|
|
|
|
|
|
|
(*visited_l1_entries)++;
|
|
|
|
if (status_cb) {
|
2015-07-27 18:51:32 +03:00
|
|
|
status_cb(bs, *visited_l1_entries, l1_entries, cb_opaque);
|
2014-10-27 13:12:53 +03:00
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
fail:
|
2018-02-05 17:33:30 +03:00
|
|
|
if (l2_slice) {
|
2013-09-03 12:09:52 +04:00
|
|
|
if (!is_active_l1) {
|
2018-02-05 17:33:30 +03:00
|
|
|
qemu_vfree(l2_slice);
|
2013-09-03 12:09:52 +04:00
|
|
|
} else {
|
2018-02-05 17:33:30 +03:00
|
|
|
qcow2_cache_put(s->l2_table_cache, (void **) &l2_slice);
|
2013-09-03 12:09:52 +04:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* For backed images, expands all zero clusters on the image. For non-backed
|
|
|
|
* images, deallocates all non-pre-allocated zero clusters (and claims the
|
|
|
|
* allocation for pre-allocated ones). This is important for downgrading to a
|
|
|
|
* qcow2 version which doesn't yet support metadata zero clusters.
|
|
|
|
*/
|
2014-10-27 13:12:53 +03:00
|
|
|
int qcow2_expand_zero_clusters(BlockDriverState *bs,
|
2015-07-27 18:51:32 +03:00
|
|
|
BlockDriverAmendStatusCB *status_cb,
|
|
|
|
void *cb_opaque)
|
2013-09-03 12:09:52 +04:00
|
|
|
{
|
2015-09-07 18:12:56 +03:00
|
|
|
BDRVQcow2State *s = bs->opaque;
|
2013-09-03 12:09:52 +04:00
|
|
|
uint64_t *l1_table = NULL;
|
2014-10-27 13:12:53 +03:00
|
|
|
int64_t l1_entries = 0, visited_l1_entries = 0;
|
2013-09-03 12:09:52 +04:00
|
|
|
int ret;
|
|
|
|
int i, j;
|
|
|
|
|
2014-10-27 13:12:53 +03:00
|
|
|
if (status_cb) {
|
|
|
|
l1_entries = s->l1_size;
|
|
|
|
for (i = 0; i < s->nb_snapshots; i++) {
|
|
|
|
l1_entries += s->snapshots[i].l1_size;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-09-03 12:09:52 +04:00
|
|
|
ret = expand_zero_clusters_in_l1(bs, s->l1_table, s->l1_size,
|
2014-10-27 13:12:53 +03:00
|
|
|
&visited_l1_entries, l1_entries,
|
2015-07-27 18:51:32 +03:00
|
|
|
status_cb, cb_opaque);
|
2013-09-03 12:09:52 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Inactive L1 tables may point to active L2 tables - therefore it is
|
|
|
|
* necessary to flush the L2 table cache before trying to access the L2
|
|
|
|
* tables pointed to by inactive L1 entries (else we might try to expand
|
|
|
|
* zero clusters that have already been expanded); furthermore, it is also
|
|
|
|
* necessary to empty the L2 table cache, since it may contain tables which
|
|
|
|
* are now going to be modified directly on disk, bypassing the cache.
|
|
|
|
* qcow2_cache_empty() does both for us. */
|
|
|
|
ret = qcow2_cache_empty(bs, s->l2_table_cache);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (i = 0; i < s->nb_snapshots; i++) {
|
2018-03-06 19:14:08 +03:00
|
|
|
int l1_size2;
|
|
|
|
uint64_t *new_l1_table;
|
|
|
|
Error *local_err = NULL;
|
|
|
|
|
|
|
|
ret = qcow2_validate_table(bs, s->snapshots[i].l1_table_offset,
|
|
|
|
s->snapshots[i].l1_size, sizeof(uint64_t),
|
|
|
|
QCOW_MAX_L1_SIZE, "Snapshot L1 table",
|
|
|
|
&local_err);
|
|
|
|
if (ret < 0) {
|
|
|
|
error_report_err(local_err);
|
|
|
|
goto fail;
|
|
|
|
}
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-03-06 19:14:08 +03:00
|
|
|
l1_size2 = s->snapshots[i].l1_size * sizeof(uint64_t);
|
|
|
|
new_l1_table = g_try_realloc(l1_table, l1_size2);
|
2018-02-09 17:42:22 +03:00
|
|
|
|
|
|
|
if (!new_l1_table) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
l1_table = new_l1_table;
|
2013-09-03 12:09:52 +04:00
|
|
|
|
2018-03-06 19:14:08 +03:00
|
|
|
ret = bdrv_pread(bs->file, s->snapshots[i].l1_table_offset,
|
|
|
|
l1_table, l1_size2);
|
2013-09-03 12:09:52 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (j = 0; j < s->snapshots[i].l1_size; j++) {
|
|
|
|
be64_to_cpus(&l1_table[j]);
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = expand_zero_clusters_in_l1(bs, l1_table, s->snapshots[i].l1_size,
|
2014-10-27 13:12:53 +03:00
|
|
|
&visited_l1_entries, l1_entries,
|
2015-07-27 18:51:32 +03:00
|
|
|
status_cb, cb_opaque);
|
2013-09-03 12:09:52 +04:00
|
|
|
if (ret < 0) {
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
g_free(l1_table);
|
|
|
|
return ret;
|
|
|
|
}
|