qemu/block.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

8433 lines
247 KiB
C
Raw Normal View History

/*
* QEMU System Emulator block driver
*
* Copyright (c) 2003 Fabrice Bellard
* Copyright (c) 2020 Virtuozzo International GmbH.
*
* 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.
*/
#include "qemu/osdep.h"
#include "block/trace.h"
#include "block/block_int.h"
#include "block/blockjob.h"
#include "block/dirty-bitmap.h"
#include "block/fuse.h"
#include "block/nbd.h"
#include "block/qdict.h"
#include "qemu/error-report.h"
#include "block/module_block.h"
#include "qemu/main-loop.h"
#include "qemu/module.h"
#include "qapi/error.h"
#include "qapi/qmp/qdict.h"
#include "qapi/qmp/qjson.h"
#include "qapi/qmp/qnull.h"
#include "qapi/qmp/qstring.h"
#include "qapi/qobject-output-visitor.h"
#include "qapi/qapi-visit-block-core.h"
#include "sysemu/block-backend.h"
#include "qemu/notify.h"
#include "qemu/option.h"
#include "qemu/coroutine.h"
#include "block/qapi.h"
#include "qemu/timer.h"
#include "qemu/cutils.h"
#include "qemu/id.h"
block: block-status cache for data regions As we have attempted before (https://lists.gnu.org/archive/html/qemu-devel/2019-01/msg06451.html, "file-posix: Cache lseek result for data regions"; https://lists.nongnu.org/archive/html/qemu-block/2021-02/msg00934.html, "file-posix: Cache next hole"), this patch seeks to reduce the number of SEEK_DATA/HOLE operations the file-posix driver has to perform. The main difference is that this time it is implemented as part of the general block layer code. The problem we face is that on some filesystems or in some circumstances, SEEK_DATA/HOLE is unreasonably slow. Given the implementation is outside of qemu, there is little we can do about its performance. We have already introduced the want_zero parameter to bdrv_co_block_status() to reduce the number of SEEK_DATA/HOLE calls unless we really want zero information; but sometimes we do want that information, because for files that consist largely of zero areas, special-casing those areas can give large performance boosts. So the real problem is with files that consist largely of data, so that inquiring the block status does not gain us much performance, but where such an inquiry itself takes a lot of time. To address this, we want to cache data regions. Most of the time, when bad performance is reported, it is in places where the image is iterated over from start to end (qemu-img convert or the mirror job), so a simple yet effective solution is to cache only the current data region. (Note that only caching data regions but not zero regions means that returning false information from the cache is not catastrophic: Treating zeroes as data is fine. While we try to invalidate the cache on zero writes and discards, such incongruences may still occur when there are other processes writing to the image.) We only use the cache for nodes without children (i.e. protocol nodes), because that is where the problem is: Drivers that rely on block-status implementations outside of qemu (e.g. SEEK_DATA/HOLE). Resolves: https://gitlab.com/qemu-project/qemu/-/issues/307 Signed-off-by: Hanna Reitz <hreitz@redhat.com> Message-Id: <20210812084148.14458-3-hreitz@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> [hreitz: Added `local_file == bs` assertion, as suggested by Vladimir] Signed-off-by: Hanna Reitz <hreitz@redhat.com>
2021-08-12 11:41:44 +03:00
#include "qemu/range.h"
#include "qemu/rcu.h"
#include "block/coroutines.h"
#ifdef CONFIG_BSD
#include <sys/ioctl.h>
#include <sys/queue.h>
#if defined(HAVE_SYS_DISK_H)
#include <sys/disk.h>
#endif
#endif
#ifdef _WIN32
#include <windows.h>
#endif
#define NOT_DONE 0x7fffffff /* used while emulated sync operation in progress */
/* Protected by BQL */
static QTAILQ_HEAD(, BlockDriverState) graph_bdrv_states =
QTAILQ_HEAD_INITIALIZER(graph_bdrv_states);
/* Protected by BQL */
static QTAILQ_HEAD(, BlockDriverState) all_bdrv_states =
QTAILQ_HEAD_INITIALIZER(all_bdrv_states);
/* Protected by BQL */
static QLIST_HEAD(, BlockDriver) bdrv_drivers =
QLIST_HEAD_INITIALIZER(bdrv_drivers);
static BlockDriverState *bdrv_open_inherit(const char *filename,
const char *reference,
QDict *options, int flags,
BlockDriverState *parent,
const BdrvChildClass *child_class,
BdrvChildRole child_role,
block: Parse filenames only when explicitly requested When handling image filenames from legacy options such as -drive or from tools, these filenames are parsed for protocol prefixes, including for the json:{} pseudo-protocol. This behaviour is intended for filenames that come directly from the command line and for backing files, which may come from the image file itself. Higher level management tools generally take care to verify that untrusted images don't contain a bad (or any) backing file reference; 'qemu-img info' is a suitable tool for this. However, for other files that can be referenced in images, such as qcow2 data files or VMDK extents, the string from the image file is usually not verified by management tools - and 'qemu-img info' wouldn't be suitable because in contrast to backing files, it already opens these other referenced files. So here the string should be interpreted as a literal local filename. More complex configurations need to be specified explicitly on the command line or in QMP. This patch changes bdrv_open_inherit() so that it only parses filenames if a new parameter parse_filename is true. It is set for the top level in bdrv_open(), for the file child and for the backing file child. All other callers pass false and disable filename parsing this way. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
2024-04-25 15:56:02 +03:00
bool parse_filename,
Error **errp);
static bool bdrv_recurse_has_child(BlockDriverState *bs,
BlockDriverState *child);
static void GRAPH_WRLOCK
bdrv_replace_child_noperm(BdrvChild *child, BlockDriverState *new_bs);
static void GRAPH_WRLOCK
bdrv_remove_child(BdrvChild *child, Transaction *tran);
static int bdrv_reopen_prepare(BDRVReopenState *reopen_state,
BlockReopenQueue *queue,
Transaction *change_child_tran, Error **errp);
static void bdrv_reopen_commit(BDRVReopenState *reopen_state);
static void bdrv_reopen_abort(BDRVReopenState *reopen_state);
static bool bdrv_backing_overridden(BlockDriverState *bs);
block: use transactions as a replacement of ->{can_}set_aio_context() Simplify the way the aiocontext can be changed in a BDS graph. There are currently two problems in bdrv_try_set_aio_context: - There is a confusion of AioContext locks taken and released, because we assume that old aiocontext is always taken and new one is taken inside. - It doesn't look very safe to call bdrv_drained_begin while some nodes have already switched to the new aiocontext and others haven't. This could be especially dangerous because bdrv_drained_begin polls, so something else could be executed while graph is in an inconsistent state. Additional minor nitpick: can_set and set_ callbacks both traverse the graph, both using the ignored list of visited nodes in a different way. Therefore, get rid of all of this and introduce a new callback, change_aio_context, that uses transactions to efficiently, cleanly and most importantly safely change the aiocontext of a graph. This new callback is a "merge" of the two previous ones: - Just like can_set_aio_context, recursively traverses the graph. Marks all nodes that are visited using a GList, and checks if they *could* change the aio_context. - For each node that passes the above check, drain it and add a new transaction that implements a callback that effectively changes the aiocontext. - Once done, the recursive function returns if *all* nodes can change the AioContext. If so, commit the above transactions. Regardless of the outcome, call transaction.clean() to undo all drains done in the recursion. - The transaction list is scanned only after all nodes are being drained, so we are sure that they all are in the same context, and then we switch their AioContext, concluding the drain only after all nodes switched to the new AioContext. In this way we make sure that bdrv_drained_begin() is always called under the old AioContext, and bdrv_drained_end() under the new one. - Because of the above, we don't need to release and re-acquire the old AioContext every time, as everything is done once (and not per-node drain and aiocontext change). Note that the "change" API is not yet invoked anywhere. Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com> Message-Id: <20221025084952.2139888-3-eesposit@redhat.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2022-10-25 11:49:44 +03:00
static bool bdrv_change_aio_context(BlockDriverState *bs, AioContext *ctx,
GHashTable *visited, Transaction *tran,
block: use transactions as a replacement of ->{can_}set_aio_context() Simplify the way the aiocontext can be changed in a BDS graph. There are currently two problems in bdrv_try_set_aio_context: - There is a confusion of AioContext locks taken and released, because we assume that old aiocontext is always taken and new one is taken inside. - It doesn't look very safe to call bdrv_drained_begin while some nodes have already switched to the new aiocontext and others haven't. This could be especially dangerous because bdrv_drained_begin polls, so something else could be executed while graph is in an inconsistent state. Additional minor nitpick: can_set and set_ callbacks both traverse the graph, both using the ignored list of visited nodes in a different way. Therefore, get rid of all of this and introduce a new callback, change_aio_context, that uses transactions to efficiently, cleanly and most importantly safely change the aiocontext of a graph. This new callback is a "merge" of the two previous ones: - Just like can_set_aio_context, recursively traverses the graph. Marks all nodes that are visited using a GList, and checks if they *could* change the aio_context. - For each node that passes the above check, drain it and add a new transaction that implements a callback that effectively changes the aiocontext. - Once done, the recursive function returns if *all* nodes can change the AioContext. If so, commit the above transactions. Regardless of the outcome, call transaction.clean() to undo all drains done in the recursion. - The transaction list is scanned only after all nodes are being drained, so we are sure that they all are in the same context, and then we switch their AioContext, concluding the drain only after all nodes switched to the new AioContext. In this way we make sure that bdrv_drained_begin() is always called under the old AioContext, and bdrv_drained_end() under the new one. - Because of the above, we don't need to release and re-acquire the old AioContext every time, as everything is done once (and not per-node drain and aiocontext change). Note that the "change" API is not yet invoked anywhere. Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com> Message-Id: <20221025084952.2139888-3-eesposit@redhat.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2022-10-25 11:49:44 +03:00
Error **errp);
/* If non-zero, use only whitelisted block drivers */
static int use_bdrv_whitelist;
#ifdef _WIN32
static int is_windows_drive_prefix(const char *filename)
{
return (((filename[0] >= 'a' && filename[0] <= 'z') ||
(filename[0] >= 'A' && filename[0] <= 'Z')) &&
filename[1] == ':');
}
int is_windows_drive(const char *filename)
{
if (is_windows_drive_prefix(filename) &&
filename[2] == '\0')
return 1;
if (strstart(filename, "\\\\.\\", NULL) ||
strstart(filename, "//./", NULL))
return 1;
return 0;
}
#endif
size_t bdrv_opt_mem_align(BlockDriverState *bs)
{
if (!bs || !bs->drv) {
block: align bounce buffers to page The following sequence int fd = open(argv[1], O_RDWR | O_CREAT | O_DIRECT, 0644); for (i = 0; i < 100000; i++) write(fd, buf, 4096); performs 5% better if buf is aligned to 4096 bytes. The difference is quite reliable. On the other hand we do not want at the moment to enforce bounce buffering if guest request is aligned to 512 bytes. The patch changes default bounce buffer optimal alignment to MAX(page size, 4k). 4k is chosen as maximal known sector size on real HDD. The justification of the performance improve is quite interesting. From the kernel point of view each request to the disk was split by two. This could be seen by blktrace like this: 9,0 11 1 0.000000000 11151 Q WS 312737792 + 1023 [qemu-img] 9,0 11 2 0.000007938 11151 Q WS 312738815 + 8 [qemu-img] 9,0 11 3 0.000030735 11151 Q WS 312738823 + 1016 [qemu-img] 9,0 11 4 0.000032482 11151 Q WS 312739839 + 8 [qemu-img] 9,0 11 5 0.000041379 11151 Q WS 312739847 + 1016 [qemu-img] 9,0 11 6 0.000042818 11151 Q WS 312740863 + 8 [qemu-img] 9,0 11 7 0.000051236 11151 Q WS 312740871 + 1017 [qemu-img] 9,0 5 1 0.169071519 11151 Q WS 312741888 + 1023 [qemu-img] After the patch the pattern becomes normal: 9,0 6 1 0.000000000 12422 Q WS 314834944 + 1024 [qemu-img] 9,0 6 2 0.000038527 12422 Q WS 314835968 + 1024 [qemu-img] 9,0 6 3 0.000072849 12422 Q WS 314836992 + 1024 [qemu-img] 9,0 6 4 0.000106276 12422 Q WS 314838016 + 1024 [qemu-img] and the amount of requests sent to disk (could be calculated counting number of lines in the output of blktrace) is reduced about 2 times. Both qemu-img and qemu-io are affected while qemu-kvm is not. The guest does his job well and real requests comes properly aligned (to page). Signed-off-by: Denis V. Lunev <den@openvz.org> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-id: 1431441056-26198-3-git-send-email-den@openvz.org CC: Paolo Bonzini <pbonzini@redhat.com> CC: Kevin Wolf <kwolf@redhat.com> CC: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2015-05-12 17:30:56 +03:00
/* page size or 4k (hdd sector size) should be on the safe side */
return MAX(4096, qemu_real_host_page_size());
}
IO_CODE();
return bs->bl.opt_mem_alignment;
}
size_t bdrv_min_mem_align(BlockDriverState *bs)
{
if (!bs || !bs->drv) {
block: align bounce buffers to page The following sequence int fd = open(argv[1], O_RDWR | O_CREAT | O_DIRECT, 0644); for (i = 0; i < 100000; i++) write(fd, buf, 4096); performs 5% better if buf is aligned to 4096 bytes. The difference is quite reliable. On the other hand we do not want at the moment to enforce bounce buffering if guest request is aligned to 512 bytes. The patch changes default bounce buffer optimal alignment to MAX(page size, 4k). 4k is chosen as maximal known sector size on real HDD. The justification of the performance improve is quite interesting. From the kernel point of view each request to the disk was split by two. This could be seen by blktrace like this: 9,0 11 1 0.000000000 11151 Q WS 312737792 + 1023 [qemu-img] 9,0 11 2 0.000007938 11151 Q WS 312738815 + 8 [qemu-img] 9,0 11 3 0.000030735 11151 Q WS 312738823 + 1016 [qemu-img] 9,0 11 4 0.000032482 11151 Q WS 312739839 + 8 [qemu-img] 9,0 11 5 0.000041379 11151 Q WS 312739847 + 1016 [qemu-img] 9,0 11 6 0.000042818 11151 Q WS 312740863 + 8 [qemu-img] 9,0 11 7 0.000051236 11151 Q WS 312740871 + 1017 [qemu-img] 9,0 5 1 0.169071519 11151 Q WS 312741888 + 1023 [qemu-img] After the patch the pattern becomes normal: 9,0 6 1 0.000000000 12422 Q WS 314834944 + 1024 [qemu-img] 9,0 6 2 0.000038527 12422 Q WS 314835968 + 1024 [qemu-img] 9,0 6 3 0.000072849 12422 Q WS 314836992 + 1024 [qemu-img] 9,0 6 4 0.000106276 12422 Q WS 314838016 + 1024 [qemu-img] and the amount of requests sent to disk (could be calculated counting number of lines in the output of blktrace) is reduced about 2 times. Both qemu-img and qemu-io are affected while qemu-kvm is not. The guest does his job well and real requests comes properly aligned (to page). Signed-off-by: Denis V. Lunev <den@openvz.org> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-id: 1431441056-26198-3-git-send-email-den@openvz.org CC: Paolo Bonzini <pbonzini@redhat.com> CC: Kevin Wolf <kwolf@redhat.com> CC: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2015-05-12 17:30:56 +03:00
/* page size or 4k (hdd sector size) should be on the safe side */
return MAX(4096, qemu_real_host_page_size());
}
IO_CODE();
return bs->bl.min_mem_alignment;
}
/* check if the path starts with "<protocol>:" */
int path_has_protocol(const char *path)
{
const char *p;
#ifdef _WIN32
if (is_windows_drive(path) ||
is_windows_drive_prefix(path)) {
return 0;
}
p = path + strcspn(path, ":/\\");
#else
p = path + strcspn(path, ":/");
#endif
return *p == ':';
}
int path_is_absolute(const char *path)
{
#ifdef _WIN32
/* specific case for names like: "\\.\d:" */
if (is_windows_drive(path) || is_windows_drive_prefix(path)) {
return 1;
}
return (*path == '/' || *path == '\\');
#else
return (*path == '/');
#endif
}
/* if filename is absolute, just return its duplicate. Otherwise, build a
path to it by considering it is relative to base_path. URL are
supported. */
char *path_combine(const char *base_path, const char *filename)
{
const char *protocol_stripped = NULL;
const char *p, *p1;
char *result;
int len;
if (path_is_absolute(filename)) {
return g_strdup(filename);
}
if (path_has_protocol(base_path)) {
protocol_stripped = strchr(base_path, ':');
if (protocol_stripped) {
protocol_stripped++;
}
}
p = protocol_stripped ?: base_path;
p1 = strrchr(base_path, '/');
#ifdef _WIN32
{
const char *p2;
p2 = strrchr(base_path, '\\');
if (!p1 || p2 > p1) {
p1 = p2;
}
}
#endif
if (p1) {
p1++;
} else {
p1 = base_path;
}
if (p1 > p) {
p = p1;
}
len = p - base_path;
result = g_malloc(len + strlen(filename) + 1);
memcpy(result, base_path, len);
strcpy(result + len, filename);
return result;
}
2017-05-22 22:52:16 +03:00
/*
* Helper function for bdrv_parse_filename() implementations to remove optional
* protocol prefixes (especially "file:") from a filename and for putting the
* stripped filename into the options QDict if there is such a prefix.
*/
void bdrv_parse_filename_strip_prefix(const char *filename, const char *prefix,
QDict *options)
{
if (strstart(filename, prefix, &filename)) {
/* Stripping the explicit protocol prefix may result in a protocol
* prefix being (wrongly) detected (if the filename contains a colon) */
if (path_has_protocol(filename)) {
GString *fat_filename;
2017-05-22 22:52:16 +03:00
/* This means there is some colon before the first slash; therefore,
* this cannot be an absolute path */
assert(!path_is_absolute(filename));
/* And we can thus fix the protocol detection issue by prefixing it
* by "./" */
fat_filename = g_string_new("./");
g_string_append(fat_filename, filename);
2017-05-22 22:52:16 +03:00
assert(!path_has_protocol(fat_filename->str));
2017-05-22 22:52:16 +03:00
qdict_put(options, "filename",
qstring_from_gstring(fat_filename));
2017-05-22 22:52:16 +03:00
} else {
/* If no protocol prefix was detected, we can use the shortened
* filename as-is */
qdict_put_str(options, "filename", filename);
}
}
}
/* Returns whether the image file is opened as read-only. Note that this can
* return false and writing to the image file is still not possible because the
* image is inactivated. */
bool bdrv_is_read_only(BlockDriverState *bs)
{
IO_CODE();
return !(bs->open_flags & BDRV_O_RDWR);
}
static int GRAPH_RDLOCK
bdrv_can_set_read_only(BlockDriverState *bs, bool read_only,
bool ignore_allow_rdw, Error **errp)
{
IO_CODE();
/* Do not set read_only if copy_on_read is enabled */
if (bs->copy_on_read && read_only) {
error_setg(errp, "Can't set node '%s' to r/o with copy-on-read enabled",
bdrv_get_device_or_node_name(bs));
return -EINVAL;
}
/* Do not clear read_only if it is prohibited */
if (!read_only && !(bs->open_flags & BDRV_O_ALLOW_RDWR) &&
!ignore_allow_rdw)
{
error_setg(errp, "Node '%s' is read only",
bdrv_get_device_or_node_name(bs));
return -EPERM;
}
return 0;
}
/*
* Called by a driver that can only provide a read-only image.
*
* Returns 0 if the node is already read-only or it could switch the node to
* read-only because BDRV_O_AUTO_RDONLY is set.
*
* Returns -EACCES if the node is read-write and BDRV_O_AUTO_RDONLY is not set
* or bdrv_can_set_read_only() forbids making the node read-only. If @errmsg
* is not NULL, it is used as the error message for the Error object.
*/
int bdrv_apply_auto_read_only(BlockDriverState *bs, const char *errmsg,
Error **errp)
{
int ret = 0;
IO_CODE();
if (!(bs->open_flags & BDRV_O_RDWR)) {
return 0;
}
if (!(bs->open_flags & BDRV_O_AUTO_RDONLY)) {
goto fail;
}
ret = bdrv_can_set_read_only(bs, true, false, NULL);
if (ret < 0) {
goto fail;
}
bs->open_flags &= ~BDRV_O_RDWR;
return 0;
fail:
error_setg(errp, "%s", errmsg ?: "Image is read-only");
return -EACCES;
}
/*
* If @backing is empty, this function returns NULL without setting
* @errp. In all other cases, NULL will only be returned with @errp
* set.
*
* Therefore, a return value of NULL without @errp set means that
* there is no backing file; if @errp is set, there is one but its
* absolute filename cannot be generated.
*/
char *bdrv_get_full_backing_filename_from_filename(const char *backed,
const char *backing,
Error **errp)
{
if (backing[0] == '\0') {
return NULL;
} else if (path_has_protocol(backing) || path_is_absolute(backing)) {
return g_strdup(backing);
} else if (backed[0] == '\0' || strstart(backed, "json:", NULL)) {
error_setg(errp, "Cannot use relative backing file names for '%s'",
backed);
return NULL;
} else {
return path_combine(backed, backing);
}
}
/*
* If @filename is empty or NULL, this function returns NULL without
* setting @errp. In all other cases, NULL will only be returned with
* @errp set.
*/
static char * GRAPH_RDLOCK
bdrv_make_absolute_filename(BlockDriverState *relative_to,
const char *filename, Error **errp)
{
char *dir, *full_name;
if (!filename || filename[0] == '\0') {
return NULL;
} else if (path_has_protocol(filename) || path_is_absolute(filename)) {
return g_strdup(filename);
}
dir = bdrv_dirname(relative_to, errp);
if (!dir) {
return NULL;
}
full_name = g_strconcat(dir, filename, NULL);
g_free(dir);
return full_name;
}
char *bdrv_get_full_backing_filename(BlockDriverState *bs, Error **errp)
{
GLOBAL_STATE_CODE();
return bdrv_make_absolute_filename(bs, bs->backing_file, errp);
}
void bdrv_register(BlockDriver *bdrv)
{
assert(bdrv->format_name);
GLOBAL_STATE_CODE();
QLIST_INSERT_HEAD(&bdrv_drivers, bdrv, list);
}
BlockDriverState *bdrv_new(void)
{
BlockDriverState *bs;
int i;
GLOBAL_STATE_CODE();
bs = g_new0(BlockDriverState, 1);
QLIST_INIT(&bs->dirty_bitmaps);
for (i = 0; i < BLOCK_OP_TYPE_MAX; i++) {
QLIST_INIT(&bs->op_blockers[i]);
}
qemu_mutex_init(&bs->reqs_lock);
qemu_mutex_init(&bs->dirty_bitmap_mutex);
bs->refcnt = 1;
bs->aio_context = qemu_get_aio_context();
qemu_co_queue_init(&bs->flush_queue);
block: block-status cache for data regions As we have attempted before (https://lists.gnu.org/archive/html/qemu-devel/2019-01/msg06451.html, "file-posix: Cache lseek result for data regions"; https://lists.nongnu.org/archive/html/qemu-block/2021-02/msg00934.html, "file-posix: Cache next hole"), this patch seeks to reduce the number of SEEK_DATA/HOLE operations the file-posix driver has to perform. The main difference is that this time it is implemented as part of the general block layer code. The problem we face is that on some filesystems or in some circumstances, SEEK_DATA/HOLE is unreasonably slow. Given the implementation is outside of qemu, there is little we can do about its performance. We have already introduced the want_zero parameter to bdrv_co_block_status() to reduce the number of SEEK_DATA/HOLE calls unless we really want zero information; but sometimes we do want that information, because for files that consist largely of zero areas, special-casing those areas can give large performance boosts. So the real problem is with files that consist largely of data, so that inquiring the block status does not gain us much performance, but where such an inquiry itself takes a lot of time. To address this, we want to cache data regions. Most of the time, when bad performance is reported, it is in places where the image is iterated over from start to end (qemu-img convert or the mirror job), so a simple yet effective solution is to cache only the current data region. (Note that only caching data regions but not zero regions means that returning false information from the cache is not catastrophic: Treating zeroes as data is fine. While we try to invalidate the cache on zero writes and discards, such incongruences may still occur when there are other processes writing to the image.) We only use the cache for nodes without children (i.e. protocol nodes), because that is where the problem is: Drivers that rely on block-status implementations outside of qemu (e.g. SEEK_DATA/HOLE). Resolves: https://gitlab.com/qemu-project/qemu/-/issues/307 Signed-off-by: Hanna Reitz <hreitz@redhat.com> Message-Id: <20210812084148.14458-3-hreitz@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> [hreitz: Added `local_file == bs` assertion, as suggested by Vladimir] Signed-off-by: Hanna Reitz <hreitz@redhat.com>
2021-08-12 11:41:44 +03:00
qemu_co_mutex_init(&bs->bsc_modify_lock);
bs->block_status_cache = g_new0(BdrvBlockStatusCache, 1);
for (i = 0; i < bdrv_drain_all_count; i++) {
bdrv_drained_begin(bs);
}
QTAILQ_INSERT_TAIL(&all_bdrv_states, bs, bs_list);
return bs;
}
2016-08-12 16:27:03 +03:00
static BlockDriver *bdrv_do_find_format(const char *format_name)
{
BlockDriver *drv1;
GLOBAL_STATE_CODE();
2016-08-12 16:27:03 +03:00
QLIST_FOREACH(drv1, &bdrv_drivers, list) {
if (!strcmp(drv1->format_name, format_name)) {
return drv1;
}
}
2016-08-12 16:27:03 +03:00
return NULL;
}
2016-08-12 16:27:03 +03:00
BlockDriver *bdrv_find_format(const char *format_name)
{
BlockDriver *drv1;
int i;
GLOBAL_STATE_CODE();
2016-08-12 16:27:03 +03:00
drv1 = bdrv_do_find_format(format_name);
if (drv1) {
return drv1;
}
/* The driver isn't registered, maybe we need to load a module */
for (i = 0; i < (int)ARRAY_SIZE(block_driver_modules); ++i) {
if (!strcmp(block_driver_modules[i].format_name, format_name)) {
module: add Error arguments to module_load and module_load_qom improve error handling during module load, by changing: bool module_load(const char *prefix, const char *lib_name); void module_load_qom(const char *type); to: int module_load(const char *prefix, const char *name, Error **errp); int module_load_qom(const char *type, Error **errp); where the return value is: -1 on module load error, and errp is set with the error 0 on module or one of its dependencies are not installed 1 on module load success 2 on module load success (module already loaded or built-in) module_load_qom_one has been introduced in: commit 28457744c345 ("module: qom module support"), which built on top of module_load_one, but discarded the bool return value. Restore it. Adapt all callers to emit errors, or ignore them, or fail hard, as appropriate in each context. Replace the previous emission of errors via fprintf in _some_ error conditions with Error and error_report, so as to emit to the appropriate target. A memory leak is also fixed as part of the module_load changes. audio: when attempting to load an audio module, report module load errors. Note that still for some callers, a single issue may generate multiple error reports, and this could be improved further. Regarding the audio code itself, audio_add() seems to ignore errors, and this should probably be improved. block: when attempting to load a block module, report module load errors. For the code paths that already use the Error API, take advantage of those to report module load errors into the Error parameter. For the other code paths, we currently emit the error, but this could be improved further by adding Error parameters to all possible code paths. console: when attempting to load a display module, report module load errors. qdev: when creating a new qdev Device object (DeviceState), report load errors. If a module cannot be loaded to create that device, now abort execution (if no CONFIG_MODULE) or exit (if CONFIG_MODULE). qom/object.c: when initializing a QOM object, or looking up class_by_name, report module load errors. qtest: when processing the "module_load" qtest command, report errors in the load of the module. Signed-off-by: Claudio Fontana <cfontana@suse.de> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220929093035.4231-4-cfontana@suse.de> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2022-09-29 12:30:33 +03:00
Error *local_err = NULL;
int rv = block_module_load(block_driver_modules[i].library_name,
&local_err);
if (rv > 0) {
return bdrv_do_find_format(format_name);
} else if (rv < 0) {
error_report_err(local_err);
}
2016-08-12 16:27:03 +03:00
break;
}
}
module: add Error arguments to module_load and module_load_qom improve error handling during module load, by changing: bool module_load(const char *prefix, const char *lib_name); void module_load_qom(const char *type); to: int module_load(const char *prefix, const char *name, Error **errp); int module_load_qom(const char *type, Error **errp); where the return value is: -1 on module load error, and errp is set with the error 0 on module or one of its dependencies are not installed 1 on module load success 2 on module load success (module already loaded or built-in) module_load_qom_one has been introduced in: commit 28457744c345 ("module: qom module support"), which built on top of module_load_one, but discarded the bool return value. Restore it. Adapt all callers to emit errors, or ignore them, or fail hard, as appropriate in each context. Replace the previous emission of errors via fprintf in _some_ error conditions with Error and error_report, so as to emit to the appropriate target. A memory leak is also fixed as part of the module_load changes. audio: when attempting to load an audio module, report module load errors. Note that still for some callers, a single issue may generate multiple error reports, and this could be improved further. Regarding the audio code itself, audio_add() seems to ignore errors, and this should probably be improved. block: when attempting to load a block module, report module load errors. For the code paths that already use the Error API, take advantage of those to report module load errors into the Error parameter. For the other code paths, we currently emit the error, but this could be improved further by adding Error parameters to all possible code paths. console: when attempting to load a display module, report module load errors. qdev: when creating a new qdev Device object (DeviceState), report load errors. If a module cannot be loaded to create that device, now abort execution (if no CONFIG_MODULE) or exit (if CONFIG_MODULE). qom/object.c: when initializing a QOM object, or looking up class_by_name, report module load errors. qtest: when processing the "module_load" qtest command, report errors in the load of the module. Signed-off-by: Claudio Fontana <cfontana@suse.de> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220929093035.4231-4-cfontana@suse.de> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2022-09-29 12:30:33 +03:00
return NULL;
2016-08-12 16:27:03 +03:00
}
static int bdrv_format_is_whitelisted(const char *format_name, bool read_only)
{
static const char *whitelist_rw[] = {
CONFIG_BDRV_RW_WHITELIST
NULL
};