2019-04-26 10:18:00 +03:00
|
|
|
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
|
|
|
|
|
From: Wolfgang Bumiller <w.bumiller@proxmox.com>
|
2020-04-07 17:53:19 +03:00
|
|
|
Date: Mon, 6 Apr 2020 12:16:50 +0200
|
|
|
|
|
Subject: [PATCH] PVE: [Up+Config] file-posix: make locking optiono on create
|
2019-04-26 10:18:00 +03:00
|
|
|
|
|
|
|
|
Otherwise creating images on nfs/cifs can be problematic.
|
|
|
|
|
|
|
|
|
|
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
|
2019-06-06 13:58:15 +03:00
|
|
|
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
|
2019-04-26 10:18:00 +03:00
|
|
|
---
|
2021-05-27 13:43:32 +03:00
|
|
|
block/file-posix.c | 59 ++++++++++++++++++++++++++++++--------------
|
2019-04-26 10:18:00 +03:00
|
|
|
qapi/block-core.json | 3 ++-
|
2021-05-27 13:43:32 +03:00
|
|
|
2 files changed, 42 insertions(+), 20 deletions(-)
|
2019-04-26 10:18:00 +03:00
|
|
|
|
|
|
|
|
diff --git a/block/file-posix.c b/block/file-posix.c
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
index 3d60b80286..49ee1db5f9 100644
|
2019-04-26 10:18:00 +03:00
|
|
|
--- a/block/file-posix.c
|
|
|
|
|
+++ b/block/file-posix.c
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2475,6 +2475,7 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
|
2019-04-26 10:18:00 +03:00
|
|
|
int fd;
|
|
|
|
|
uint64_t perm, shared;
|
|
|
|
|
int result = 0;
|
|
|
|
|
+ bool locked = false;
|
|
|
|
|
|
|
|
|
|
/* Validate options and set default values */
|
|
|
|
|
assert(options->driver == BLOCKDEV_DRIVER_FILE);
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2515,19 +2516,22 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
|
2019-04-26 10:18:00 +03:00
|
|
|
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
|
|
|
|
|
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
|
|
|
|
|
|
|
|
|
|
- /* Step one: Take locks */
|
2019-06-06 13:58:15 +03:00
|
|
|
- result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
|
2019-04-26 10:18:00 +03:00
|
|
|
- if (result < 0) {
|
|
|
|
|
- goto out_close;
|
|
|
|
|
- }
|
|
|
|
|
+ if (file_opts->locking != ON_OFF_AUTO_OFF) {
|
|
|
|
|
+ /* Step one: Take locks */
|
2019-06-06 13:58:15 +03:00
|
|
|
+ result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
|
2019-04-26 10:18:00 +03:00
|
|
|
+ if (result < 0) {
|
|
|
|
|
+ goto out_close;
|
|
|
|
|
+ }
|
|
|
|
|
+ locked = true;
|
|
|
|
|
|
|
|
|
|
- /* Step two: Check that nobody else has taken conflicting locks */
|
|
|
|
|
- result = raw_check_lock_bytes(fd, perm, shared, errp);
|
|
|
|
|
- if (result < 0) {
|
2019-06-06 13:58:15 +03:00
|
|
|
- error_append_hint(errp,
|
|
|
|
|
- "Is another process using the image [%s]?\n",
|
|
|
|
|
- file_opts->filename);
|
2019-04-26 10:18:00 +03:00
|
|
|
- goto out_unlock;
|
|
|
|
|
+ /* Step two: Check that nobody else has taken conflicting locks */
|
|
|
|
|
+ result = raw_check_lock_bytes(fd, perm, shared, errp);
|
|
|
|
|
+ if (result < 0) {
|
2019-06-06 13:58:15 +03:00
|
|
|
+ error_append_hint(errp,
|
|
|
|
|
+ "Is another process using the image [%s]?\n",
|
|
|
|
|
+ file_opts->filename);
|
2019-04-26 10:18:00 +03:00
|
|
|
+ goto out_unlock;
|
|
|
|
|
+ }
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Clear the file by truncating it to 0 */
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2581,13 +2585,15 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
|
2019-04-26 10:18:00 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
out_unlock:
|
2019-06-06 13:58:15 +03:00
|
|
|
- raw_apply_lock_bytes(NULL, fd, 0, 0, true, &local_err);
|
2019-04-26 10:18:00 +03:00
|
|
|
- if (local_err) {
|
|
|
|
|
- /* The above call should not fail, and if it does, that does
|
|
|
|
|
- * not mean the whole creation operation has failed. So
|
|
|
|
|
- * report it the user for their convenience, but do not report
|
|
|
|
|
- * it to the caller. */
|
2019-06-06 13:58:15 +03:00
|
|
|
- warn_report_err(local_err);
|
2019-04-26 10:18:00 +03:00
|
|
|
+ if (locked) {
|
2019-06-06 13:58:15 +03:00
|
|
|
+ raw_apply_lock_bytes(NULL, fd, 0, 0, true, &local_err);
|
2019-04-26 10:18:00 +03:00
|
|
|
+ if (local_err) {
|
|
|
|
|
+ /* The above call should not fail, and if it does, that does
|
|
|
|
|
+ * not mean the whole creation operation has failed. So
|
|
|
|
|
+ * report it the user for their convenience, but do not report
|
|
|
|
|
+ * it to the caller. */
|
2019-06-06 13:58:15 +03:00
|
|
|
+ warn_report_err(local_err);
|
2019-04-26 10:18:00 +03:00
|
|
|
+ }
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
out_close:
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2612,6 +2618,7 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
|
2019-04-26 10:18:00 +03:00
|
|
|
PreallocMode prealloc;
|
|
|
|
|
char *buf = NULL;
|
|
|
|
|
Error *local_err = NULL;
|
|
|
|
|
+ OnOffAuto locking;
|
|
|
|
|
|
|
|
|
|
/* Skip file: protocol prefix */
|
|
|
|
|
strstart(filename, "file:", &filename);
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2634,6 +2641,18 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
|
2019-04-26 10:18:00 +03:00
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
+ locking = qapi_enum_parse(&OnOffAuto_lookup,
|
|
|
|
|
+ qemu_opt_get(opts, "locking"),
|
|
|
|
|
+ ON_OFF_AUTO_AUTO, &local_err);
|
|
|
|
|
+ if (local_err) {
|
|
|
|
|
+ error_propagate(errp, local_err);
|
|
|
|
|
+ return -EINVAL;
|
|
|
|
|
+ }
|
|
|
|
|
+
|
|
|
|
|
+ if (locking == ON_OFF_AUTO_AUTO) {
|
|
|
|
|
+ locking = ON_OFF_AUTO_OFF;
|
|
|
|
|
+ }
|
|
|
|
|
+
|
|
|
|
|
options = (BlockdevCreateOptions) {
|
|
|
|
|
.driver = BLOCKDEV_DRIVER_FILE,
|
|
|
|
|
.u.file = {
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -2645,6 +2664,8 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
|
2019-04-26 10:18:00 +03:00
|
|
|
.nocow = nocow,
|
2020-08-20 11:42:45 +03:00
|
|
|
.has_extent_size_hint = has_extent_size_hint,
|
|
|
|
|
.extent_size_hint = extent_size_hint,
|
2019-04-26 10:18:00 +03:00
|
|
|
+ .has_locking = true,
|
|
|
|
|
+ .locking = locking,
|
|
|
|
|
},
|
|
|
|
|
};
|
|
|
|
|
return raw_co_create(&options, errp);
|
|
|
|
|
diff --git a/qapi/block-core.json b/qapi/block-core.json
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
index e1857e7094..ddac91e8f6 100644
|
2019-04-26 10:18:00 +03:00
|
|
|
--- a/qapi/block-core.json
|
|
|
|
|
+++ b/qapi/block-core.json
|
update submodule and patches to 7.1.0
Notable changes:
* The only big change is the switch to using a custom QIOChannel for
savevm-async, because the previously used QEMUFileOps was dropped.
Changes to the current implementation:
* Switch to vector based methods as required for an IO channel. For
short reads the passed-in IO vector is stuffed with zeroes at the
end, just to be sure.
* For reading: The documentation in include/io/channel.h states that
at least one byte should be read, so also error out when whe are
at the very end instead of returning 0.
* For reading: Fix off-by-one error when request goes beyond end.
The wrong code piece was:
if ((pos + size) > maxlen) {
size = maxlen - pos - 1;
}
Previously, the last byte would not be read. It's actually
possible to get a snapshot .raw file that has content all the way
up the final 512 byte (= BDRV_SECTOR_SIZE) boundary without any
trailing zero bytes (I wrote a script to do it).
Luckily, it didn't cause a real issue, because qemu_loadvm_state()
is not interested in the final (i.e. QEMU_VM_VMDESCRIPTION)
section. The buffer for reading it is simply freed up afterwards
and the function will assume that it read the whole section, even
if that's not the case.
* For writing: Make use of the generated blk_pwritev() wrapper
instead of manually wrapping the coroutine to simplify and save a
few lines.
* Adapt to changed interfaces for blk_{pread,pwrite}:
* a9262f551e ("block: Change blk_{pread,pwrite}() param order")
* 3b35d4542c ("block: Add a 'flags' param to blk_pread()")
* bf5b16fa40 ("block: Make blk_{pread,pwrite}() return 0 on success")
Those changes especially affected the qemu-img dd patches, because
the context also changed, but also some of our block drivers used
the functions.
* Drop qemu-common.h include: it got renamed after essentially
everything was moved to other headers. The only remaining user I
could find for things dropped from the header between 7.0 and 7.1
was qemu_get_vm_name() in the iscsi-initiatorname patch, but it
already includes the header to which the function was moved.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2022-10-14 15:07:13 +03:00
|
|
|
@@ -4537,7 +4537,8 @@
|
2020-08-20 11:42:45 +03:00
|
|
|
'size': 'size',
|
|
|
|
|
'*preallocation': 'PreallocMode',
|
|
|
|
|
'*nocow': 'bool',
|
|
|
|
|
- '*extent-size-hint': 'size'} }
|
|
|
|
|
+ '*extent-size-hint': 'size',
|
2019-04-26 10:18:00 +03:00
|
|
|
+ '*locking': 'OnOffAuto' } }
|
|
|
|
|
|
|
|
|
|
##
|
|
|
|
|
# @BlockdevCreateOptionsGluster:
|