Current scheme of image fleecing looks like this:
[guest] [NBD export]
| |
|root | root
v v
[copy-before-write] -----> [temp.qcow2]
| target |
|file |backing
v |
[active disk] <-------------+
- On guest writes copy-before-write filter copies old data from active
disk to temp.qcow2. So fleecing client (NBD export) when reads
changed regions from temp.qcow2 image and unchanged from active disk
through backing link.
This patch makes possible new image fleecing scheme:
[guest] [NBD export]
| |
| root | root
v file v
[copy-before-write]<------[snapshot-access]
| |
| file | target
v v
[active-disk] [temp.img]
- copy-before-write does CBW operations and also provides
snapshot-access API. The API may be accessed through
snapshot-access driver.
Benefits of new scheme:
1. Access control: if remote client try to read data that not covered
by original dirty bitmap used on copy-before-write open, client gets
-EACCES.
2. Discard support: if remote client do DISCARD, this additionally to
discarding data in temp.img informs block-copy process to not copy
these clusters. Next read from discarded area will return -EACCES.
This is significant thing: when fleecing user reads data that was
not yet copied to temp.img, we can avoid copying it on further guest
write.
3. Synchronisation between client reads and block-copy write is more
efficient. In old scheme we just rely on BDRV_REQ_SERIALISING flag
used for writes to temp.qcow2. New scheme is less blocking:
- fleecing reads are never blocked: if data region is untouched or
in-flight, we just read from active-disk, otherwise we read from
temp.img
- writes to temp.img are not blocked by fleecing reads
- still, guest writes of-course are blocked by in-flight fleecing
reads, that currently read from active-disk - it's the minimum
necessary blocking
4. Temporary image may be of any format, as we don't rely on backing
feature.
5. Permission relation are simplified. With old scheme we have to share
write permission on target child of copy-before-write, otherwise
backing link conflicts with copy-before-write file child write
permissions. With new scheme we don't have backing link, and
copy-before-write node may have unshared access to temporary node.
(Not realized in this commit, will be in future).
6. Having control on fleecing reads we'll be able to implement
alternative behavior on failed copy-before-write operations.
Currently we just break guest request (that's a historical behavior
of backup). But in some scenarios it's a bad behavior: better
is to drop the backup as failed but don't break guest request.
With new scheme we can simply unset some bits in a bitmap on CBW
failure and further fleecing reads will -EACCES, or something like
this. (Not implemented in this commit, will be in future)
Additional application for this is implementing timeout for CBW
operations.
Iotest 257 output is updated, as two more bitmaps now live in
copy-before-write filter.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20220303194349.2304213-13-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
The new block driver simply utilizes snapshot-access API of underlying
block node.
In further patches we want to use it like this:
[guest] [NBD export]
| |
| root | root
v file v
[copy-before-write]<------[snapshot-access]
| |
| file | target
v v
[active-disk] [temp.img]
This way, NBD client will be able to read snapshotted state of active
disk, when active disk is continued to be written by guest. This is
known as "fleecing", and currently uses another scheme based on qcow2
temporary image which backing file is active-disk. New scheme comes
with benefits - see next commit.
The other possible application is exporting internal snapshots of
qcow2, like this:
[guest] [NBD export]
| |
| root | root
v file v
[qcow2]<---------[snapshot-access]
For this, we'll need to implement snapshot-access API handlers in
qcow2 driver, and improve snapshot-access block driver (and API) to
make it possible to select snapshot by name. Another thing to improve
is size of snapshot. Now for simplicity we just use size of bs->file,
which is OK for backup, but for qcow2 snapshots export we'll need to
imporve snapshot-access API to get size of snapshot.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20220303194349.2304213-12-vsementsov@virtuozzo.com>
[hreitz: Rebased on block GS/IO split]
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Add new block driver handlers and corresponding generic wrappers.
It will be used to allow copy-before-write filter to provide
reach fleecing interface in further commit.
In future this approach may be used to allow reading qcow2 internal
snapshots, for example to export them through NBD.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220303194349.2304213-11-vsementsov@virtuozzo.com>
[hreitz: Rebased on block GS/IO split]
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to complicate bitmap initialization in the further
commit. And in future, backup job will be able to work without filter
(when source is immutable), so we'll need same bitmap initialization in
copy-before-write filter and in backup job. So, it's reasonable to do
it in block-copy.
Note that for now cbw_open() is the only caller of
block_copy_state_new().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220303194349.2304213-2-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
There is a bug in handling BDRV_REQ_NO_WAIT flag: we still may wait in
wait_serialising_requests() if request is unaligned. And this is
possible for the only user of this flag (preallocate filter) if
underlying file is unaligned to its request_alignment on start.
So, we have to fix preallocate filter to do only aligned preallocate
requests.
Next, we should fix generic block/io.c somehow. Keeping in mind that
preallocate is the only user of BDRV_REQ_NO_WAIT and that we have to
fix its behavior now, it seems more safe to just assert that we never
use BDRV_REQ_NO_WAIT with unaligned requests and add corresponding
comment. Let's do so.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Denis V. Lunev <den@openvz.org>
Message-Id: <20220215121609.38570-1-vsementsov@virtuozzo.com>
[hreitz: Rebased on block GS/IO split]
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Graph modifications should be done in drained section. stream_prepare()
handler of block stream job call bdrv_set_backing_hd() without using
drained section and it's theoretically possible that some IO request
will interleave with graph modification and will use outdated pointers
to removed block nodes.
Some other callers use bdrv_set_backing_hd() not caring about drained
sections too. So it seems good to make a drained section exactly in
bdrv_set_backing_hd().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20220124173741.2984056-1-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
We want iotests pass with both the default zlib compression and with
IMGOPTS='compression_type=zstd'.
Actually the only test that is interested in real compression type in
test output is 287 (test for qcow2 compression type), so implement
specific option for it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-17-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to add filtering in _qcow2_dump_header and want all tests
use it.
The patch is generated by commands:
cd tests/qemu-iotests
sed -ie 's/$PYTHON qcow2.py "$TEST_IMG" dump-header\($\| \)/_qcow2_dump_header\1/' ??? tests/*
(the difficulty is to avoid converting dump-header-exts)
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20211223160144.1097696-15-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
If image doesn't have any compressed cluster we can easily switch to
zlib compression, which may allow to downgrade the image.
That's mostly needed to support IMGOPTS='compression_type=zstd' in some
iotests which do qcow2 downgrade.
While being here also fix checkpatch complain against '#' in printf
formatting.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20211223160144.1097696-13-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Instead of qemu_img_log("info", ..) use generic helper img_info_log().
img_info_log() has smarter logic. For example it use filter_img_info()
to filter output, which in turns filter a compression type. So it will
help us in future when we implement a possibility to use zstd
compression by default (with help of some runtime config file or maybe
build option). For now to test you should recompile qemu with a small
addition into block/qcow2.c before
"if (qcow2_opts->has_compression_type":
if (!qcow2_opts->has_compression_type && version >= 3) {
qcow2_opts->has_compression_type = true;
qcow2_opts->compression_type = QCOW2_COMPRESSION_TYPE_ZSTD;
}
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20211223160144.1097696-12-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We want iotests pass with both the default zlib compression and with
IMGOPTS='compression_type=zstd'.
Actually the only test that is interested in real compression type in
test output is 287 (test for qcow2 compression type) and it's in bash.
So for now we can safely filter out compression type in all qcow2
tests.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-11-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
The test checks different options. It of course fails if set
IMGOPTS='compression_type=zstd'. So, let's be explicit in what
compression type we want and independent of IMGOPTS. Test both existing
compression types.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-9-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
qemu_img_verbose() has a drawback of not going through generic
qemu_img_pipe_and_status(). qemu_img_verbose() is not very popular, so
update the only two users to qemu_img_log() and drop qemu_img_verbose()
at all.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-6-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Adding support of IMGOPTS (like in bash tests) allows user to pass a
lot of different options. Still, some may require additional logic.
Now we want compression_type option, so add some smart logic around it:
ignore compression_type=zstd in IMGOPTS, if test want qcow2 in
compatibility mode. As well, ignore compression_type for non-qcow2
formats.
Note that we may instead add support only to qemu_img_create(), but
that works bad:
1. We'll have to update a lot of tests to use qemu_img_create instead
of qemu_img('create'). (still, we may want do it anyway, but no
reason to create a dependancy between task of supporting IMGOPTS and
updating a lot of tests)
2. Some tests use qemu_img_pipe('create', ..) - even more work on
updating
3. Even if we update all tests to go through qemu_img_create, we'll
need a way to avoid creating new tests using qemu_img*('create') -
add assertions.. That doesn't seem good.
So, let's add support of IMGOPTS to most generic
qemu_img_pipe_and_status().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-5-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to support IMGOPTS for python iotests. Still some iotests
will not work with common IMGOPTS used with bash iotests like
specifying refcount_bits and compat qcow2 options. So we
should define corresponding unsupported_imgopts for now.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211223160144.1097696-4-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to support some addition IMGOPTS in python iotests like
in bash iotests. Similarly to bash iotests, we want a way to skip some
tests which can't work with specific IMGOPTS.
Globally for python iotests we will not support things like
'data_file=$TEST_IMG.ext_data_file' in IMGOPTS, so, forbid this
globally in iotests.py.
Suggested-by: Hanna Reitz <hreitz@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20211223160144.1097696-3-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to support IMGOPTS environment variable like in bash
tests. Corresponding global variable in iotests.py should be called
imgopts. So to not interfere with function argument, rename it in
advance.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20211223160144.1097696-2-vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We don't generate trace events for tests/ and qga/ because that it is
not simple and not necessary. We have corresponding comments in both
tests/meson.build and qga/meson.build.
Still to not miss possible future qapi code generation call, and not to
forget to enable trace events generation, let's enable it by default.
So, turn option --gen-trace into opposite --no-trace-events and use new
option only in tests/ and qga/ where we already have good comments why
we don't generate trace events code.
Note that this commit enables trace-events generation for qapi-gen.py
call from tests/qapi-schema/meson.build and storage-daemon/meson.build.
Still, both are kind of noop: tests/qapi-schema/ doesn't seem to
generate any QMP command code and no .trace-events files anyway,
storage-daemon/ uses common QMP command implementations and just
generate empty .trace-events
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220126161130.3240892-8-vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Making trace generation work for tests/ and qga/ would involve some
Meson hackery to ensure we generate the trace-events files before
trace-tool uses them. Since we don't actually support tracing there
anyway, we bypass that problem.
Let's add corresponding comments.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20220126161130.3240892-7-vsementsov@virtuozzo.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Pasto fixed, commit message punctuation tidied up]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
1. Use --gen-trace when generate qmp commands
2. Add corresponding .trace-events files as outputs in qapi_files
custom target
3. Define global qapi_trace_events list of .trace-events file targets,
to fill in trace/qapi.build and to use in trace/meson.build
4. In trace/meson.build use the new array as an additional source of
.trace_events files to be processed
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220126161130.3240892-5-vsementsov@virtuozzo.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Add trace generation disabled by default and new option --gen-trace to
enable it. The next commit will enable it for qapi/, but not for qga/
and tests/. Making it work for the latter two would involve some Meson
hackery to ensure we generate the trace-events files before trace-tool
uses them. Since we don't actually support tracing there, we'll bypass
that problem.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220126161130.3240892-4-vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Superfluous #include dropped]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
We are going to generate trace events for QMP commands. We should
generate both trace_*() function calls and trace-events files listing
events for trace generator.
So, add an output module FOO.trace-events for each FOO schema module.
Since we're going to add trace events only to command marshallers,
make the trace-events output optional, so we don't generate so many
useless empty files.
Currently nobody set add_trace_events to True, so new functionality is
disabled. It will be enabled for QAPISchemaGenCommandVisitor
in a further commit.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220126161130.3240892-2-vsementsov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
A lot of Optional[] types doesn't make code beautiful.
test_field_width defaults to 8, but that is never used in the code.
More over, if we want some default behavior for single call of
test_run(), it should just print the whole test name, not limiting or
expanding its width, so 8 is bad default.
So, just drop the default as unused for now.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211210201450.101576-1-vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
First, this permission never protected a node from being changed, as
generic child-replacing functions don't check it.
Second, it's a strange thing: it presents a permission of parent node
to change its child. But generally, children are replaced by different
mechanisms, like jobs or qmp commands, not by nodes.
Graph-mod permission is hard to understand. All other permissions
describe operations which done by parent node on its child: read,
write, resize. Graph modification operations are something completely
different.
The only place where BLK_PERM_GRAPH_MOD is used as "perm" (not shared
perm) is mirror_start_job, for s->target. Still modern code should use
bdrv_freeze_backing_chain() to protect from graph modification, if we
don't do it somewhere it may be considered as a bug. So, it's a bit
risky to drop GRAPH_MOD, and analyzing of possible loss of protection
is hard. But one day we should do it, let's do it now.
One more bit of information is that locking the corresponding byte in
file-posix doesn't make sense at all.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210902093754.2352-1-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Consider the case when the whole buffer is zero and end is unaligned.
If i <= tail, we return 1 and do one unaligned WRITE, RMW happens.
If i > tail, we do on aligned WRITE_ZERO (or skip if target is zeroed)
and again one unaligned WRITE, RMW happens.
Let's do better: don't fragment the whole-zero buffer and report it as
ZERO: in case of zeroed target we just do nothing and avoid RMW. If
target is not zeroes, one unaligned WRITE_ZERO should not be much worse
than one unaligned WRITE.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211217164654.1184218-3-vsementsov@virtuozzo.com>
Tested-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
It's unused now (except for permission handling)[*]. The only reasonable
user of it was block-stream job, recently updated to use own blk. And
other block jobs prefer to use own source node related objects.
So, the arguments of dropping the field are:
- block jobs prefer not to use it
- block jobs usually has more then one node to operate on, and better
to operate symmetrically (for example has both source and target
blk's in specific block-job state structure)
*: BlockJob.blk is used to keep some permissions. We simply move
permissions to block-job child created in block_job_create() together
with blk.
In mirror, we just should not care anymore about restoring state of
blk. Most probably this code could be dropped long ago, after dropping
bs->job pointer. Now it finally goes away together with BlockJob.blk
itself.
iotest 141 output is updated, as "bdrv_has_blk(bs)" check in
qmp_blockdev_del() doesn't fail (we don't have blk now). Still, new
error message looks even better.
In iotest 283 we need to add a job id, otherwise "Invalid job ID"
happens now earlier than permission check (as permissions moved from
blk to block-job node).
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
We are going to drop BlockJob.blk in further commit. For tests it's
enough to simply pass bs pointer.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
block-stream is the only block-job, that reasonably use BlockJob.blk.
We are going to drop BlockJob.blk soon. So, let block-stream have own
blk.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
Here we use job->blk to drop our own reference in job cleanup. Let's do
simpler: drop our reference immediately after job creation.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
We are going to drop BlockJob.blk. So let's retrieve block job context
from underlying job instead of main node.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
Add qemu-io Popen constructor wrapper. To be used in the following new
test commit.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
For qemu_io* functions support --image-opts argument, which conflicts
with -f argument from qemu_io_args.
For QemuIoInteractive use new wrapper as well, which allows relying on
default format.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
Split qemu_tool_popen() from qemu_tool_pipe_and_status() to be used
separately.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nikita Lapshin <nikita.lapshin@virtuozzo.com>
Add -j <JOBS> parameter, to run tests in several jobs simultaneously.
For realization - simply utilize multiprocessing.Pool class.
Notes:
1. Of course, tests can't run simultaneously in same TEST_DIR. So,
use subdirectories TEST_DIR/testname/ and SOCK_DIR/testname/
instead of simply TEST_DIR and SOCK_DIR
2. multiprocessing.Pool.starmap function doesn't support passing
context managers, so we can't simply pass "self". Happily, we need
self only for read-only access, and it just works if it is defined
in global space. So, add a temporary link TestRunner.shared_self
during run_tests().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211203122223.2780098-4-vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Tested-by: John Snow <jsnow@redhat.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
We are going to use do_run_test() in multiprocessing environment, where
we'll not be able to change original runner object.
Happily, the only thing we change is that last_elapsed and it's simple
to do it in run_tests() instead. All other accesses to self in
do_runt_test() and in run_test() are read-only.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211203122223.2780098-3-vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Tested-by: John Snow <jsnow@redhat.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
The only caller of nbd_do_establish_connection() that uses errp is
nbd_open(). The only way to cancel this call is through open_timer
timeout. And for this case, user will be more interested in description
of last failed connect rather than in
"Connection attempt cancelled by other operation".
So, let's change behavior on cancel to return previous failure error if
available.
Do the same for non-blocking failure case. In this case we still don't
have a caller that is interested in errp. But let's be consistent.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
It is useful when start of vm and start of nbd server are not
simple to sync.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
vnc_server_cut_text_caps() is not guaranteed to be called only once.
If it called twice, we finally call notifier_list_add() twice with same
element. Which leads to loopback QLIST. So, on next
notifier_list_notify() we'll loop forever and QEMU stuck.
So, let's only register new notifier if it's not yet registered.
Note, that similar check is used in vdagent_chr_recv_caps() (before
call qemu_clipboard_peer_register()), and also before
qemu_clipboard_peer_unregister() call in vdagent_disconnect() and in
vnc_disconnect_finish().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211110103800.2266729-1-vsementsov@virtuozzo.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Modern way is using blockdev-add + blockdev-backup, which provides a
lot more control on how target is opened.
As example of drive-backup problems consider the following:
User of drive-backup expects that target will be opened in the same
cache and aio mode as source. Corresponding logic is in
drive_backup_prepare(), where we take bs->open_flags of source.
It works rather bad if source was added by blockdev-add. Assume source
is qcow2 image. On blockdev-add we should specify aio and cache options
for file child of qcow2 node. What happens next:
drive_backup_prepare() looks at bs->open_flags of qcow2 source node.
But there no BDRV_O_NOCAHE neither BDRV_O_NATIVE_AIO: BDRV_O_NOCAHE is
places in bs->file->bs->open_flags, and BDRV_O_NATIVE_AIO is nowhere,
as file-posix parse options and simply set s->use_linux_aio.
The documentation is updated in a minimal way, so that drive-backup is
noted only as a deprecated command, and blockdev-backup used in most of
places.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
We are going to deprecate drive-backup, so use modern interface here.
In examples where target image creation is shown, show blockdev-add as
well. If target creation omitted, omit blockdev-add as well.
Reviewed-by: Kashyap Chamarthy <kchamart@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
We are going to deprecate drive-backup, so don't mention it here.
Moreover, blockdev-backup seems more correct in the context.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Make bytes argument int64_t to be consistent with modern block-layer.
Callers should be OK with it as type becomes wider.
What is inside functions?
- Conversion from int64_t to size_t. Still, we
can't have a buffer larger than SIZE_MAX, therefore bytes should not be
larger than SIZE_MAX as well. Add an assertion.
- Passing to blk_co_pwritev() / blk_co_preadv() which already has
int64_t bytes argument.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211007175243.642516-2-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: spelling fix]
Signed-off-by: Eric Blake <eblake@redhat.com>
blk_check_bytes_request is called from blk_co_do_preadv,
blk_co_do_pwritev_part, blk_co_do_pdiscard and blk_co_copy_range
before (maybe) calling throttle_group_co_io_limits_intercept() (which
has int64_t argument) and then calling corresponding bdrv_co_ function.
bdrv_co_ functions are OK with int64_t bytes as well.
So dropping the check for INT_MAX we just get same restrictions as in
bdrv_ layer: discard and write-zeroes goes through
bdrv_check_qiov_request() and are allowed to be 64bit. Other requests
go through bdrv_check_request32() and still restricted by INT_MAX
boundary.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-13-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
1. Convert bytes in BlkAioEmAIOCB:
aio->bytes is only passed to already int64_t interfaces, and set in
blk_aio_prwv, which is updated here.
2. For all updated functions the parameter type becomes wider so callers
are safe.
3. In blk_aio_prwv we only store bytes to BlkAioEmAIOCB, which is
updated here.
4. Other updated functions are wrappers on blk_aio_prwv.
Note that blk_aio_preadv and blk_aio_pwritev become safer: before this
commit, it's theoretically possible to pass qiov with size exceeding
INT_MAX, which than converted to int argument of blk_aio_prwv. Now it's
converted to int64_t which is a lot better. Still add assertions.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-11-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: tweak assertion and grammar]
Signed-off-by: Eric Blake <eblake@redhat.com>
Function is updated so that parameter type becomes wider, so all
callers should be OK with it.
Look at blk_co_copy_range() itself: bytes is passed only to
blk_check_byte_request() and bdrv_co_copy_range(), which already have
int64_t bytes parameter, so we are OK.
Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-10-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
Convert blk_pdiscard, blk_pwrite_compressed, blk_pwrite_zeroes.
These are just wrappers for functions with int64_t argument, so allow
passing int64_t as well. Parameter type becomes wider so all callers
should be OK with it.
Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().
Note also that we don't (and are not going to) convert blk_pwrite and
blk_pread: these functions return number of bytes on success, so to
update them, we should change return type to int64_t as well, which
will lead to investigating and updating all callers which is too much.
So, blk_pread and blk_pwrite remain unchanged.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-9-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
Let's drop hand-made coroutine wrappers and use coroutine wrapper
generation like in block/io.c.
Now, blk_foo() functions are written in same way as blk_co_foo() ones,
but wrap blk_do_foo() instead of blk_co_do_foo().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-8-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: spelling fix]
Signed-off-by: Eric Blake <eblake@redhat.com>
We updated blk_do_pdiscard() and its wrapper blk_co_pdiscard(). Both
functions are updated so that the parameter type becomes wider, so all
callers should be OK with it.
Look at blk_do_pdiscard(): bytes is passed only to
blk_check_byte_request() and bdrv_co_pdiscard(), which already have
int64_t bytes parameter, so we are OK.
Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-5-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
We convert blk_do_pwritev_part() and some wrappers:
blk_co_pwritev_part(), blk_co_pwritev(), blk_co_pwrite_zeroes().
All functions are converted so that the parameter type becomes wider, so
all callers should be OK with it.
Look at blk_do_pwritev_part() body:
bytes is passed to:
- trace_blk_co_pwritev (we update it here)
- blk_check_byte_request, throttle_group_co_io_limits_intercept,
bdrv_co_pwritev_part - all already have int64_t argument.
Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-4-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
For both updated functions, the type of bytes becomes wider, so all callers
should be OK with it.
blk_co_preadv() only passes its arguments to blk_do_preadv().
blk_do_preadv() passes bytes to:
- trace_blk_co_preadv, which is updated too
- blk_check_byte_request, throttle_group_co_io_limits_intercept,
bdrv_co_preadv, which are already int64_t.
Note that requests exceeding INT_MAX are still restricted by
blk_check_byte_request().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-3-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
Rename size and make it int64_t to correspond to modern block layer,
which always uses int64_t for offset and bytes (not in blk layer yet,
which is a task for following commits).
All callers pass int or unsigned int.
So, for bytes in [0, INT_MAX] nothing is changed, for negative bytes we
now fail on "bytes < 0" check instead of "bytes > INT_MAX" check.
Note, that blk_check_byte_request() still doesn't allow requests
exceeding INT_MAX.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211006131718.214235-2-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Use bdrv_new_open_driver_opts() instead of complicated bdrv_open().
Among other extra things bdrv_open() also check for white-listed
formats, which we don't want for internal node creation: currently
backup doesn't work when copy-before-write filter is not white-listed.
As well block-stream doesn't work when copy-on-read is not
white-listed.
Fixes: 751cec7a26
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=2004812
Reported-by: Yanan Fu
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210920115538.264372-5-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
- use ERRP_GUARD(): function calls error_prepend(), so it must use
ERRP_GUARD(), otherwise error_prepend() would not be called when
passed errp is error_fatal
- drop error propagation, handle return code instead
- for symmetry, do error_prepend() for the second failure
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210920115538.264372-3-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add version of bdrv_new_open_driver() that supports QDict options.
We'll use it in further commit.
Simply add one more argument to bdrv_new_open_driver() is worse, as
there are too many invocations of bdrv_new_open_driver() to update
then.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210920115538.264372-2-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
OK, that's a big rewrite of the logic.
Pre-patch we have an always running coroutine - connection_co. It does
reply receiving and reconnecting. And it leads to a lot of difficult
and unobvious code around drained sections and context switch. We also
abuse bs->in_flight counter which is increased for connection_co and
temporary decreased in points where we want to allow drained section to
begin. One of these place is in another file: in nbd_read_eof() in
nbd/client.c.
We also cancel reconnect and requests waiting for reconnect on drained
begin which is not correct. And this patch fixes that.
Let's finally drop this always running coroutine and go another way:
do both reconnect and receiving in request coroutines.
The detailed list of changes below (in the sequence of diff hunks).
1. receiving coroutines are woken directly from nbd_channel_error, when
we change s->state
2. nbd_co_establish_connection_cancel(): we don't have drain_begin now,
and in nbd_teardown_connection() all requests should already be
finished (and reconnect is done from request). So
nbd_co_establish_connection_cancel() is called from
nbd_cancel_in_flight() (to cancel the request that is doing
nbd_co_establish_connection()) and from reconnect_delay_timer_cb()
(previously we didn't need it, as reconnect delay only should cancel
active requests not the reconnection itself). But now reconnection
itself is done in the separate thread (we now call
nbd_client_connection_enable_retry() in nbd_open()), and we need to
cancel the requests that wait in nbd_co_establish_connection()
now).
2A. We do receive headers in request coroutine. But we also should
dispatch replies for other pending requests. So,
nbd_connection_entry() is turned into nbd_receive_replies(), which
does reply dispatching while it receives other request headers, and
returns when it receives the requested header.
3. All old staff around drained sections and context switch is dropped.
In details:
- we don't need to move connection_co to new aio context, as we
don't have connection_co anymore
- we don't have a fake "request" of connection_co (extra increasing
in_flight), so don't care with it in drain_begin/end
- we don't stop reconnection during drained section anymore. This
means that drain_begin may wait for a long time (up to
reconnect_delay). But that's an improvement and more correct
behavior see below[*]
4. In nbd_teardown_connection() we don't have to wait for
connection_co, as it is dropped. And cleanup for s->ioc and nbd_yank
is moved here from removed connection_co.
5. In nbd_co_do_establish_connection() we now should handle
NBD_CLIENT_CONNECTING_NOWAIT: if new request comes when we are in
NBD_CLIENT_CONNECTING_NOWAIT, it still should call
nbd_co_establish_connection() (who knows, maybe the connection was
already established by another thread in the background). But we
shouldn't wait: if nbd_co_establish_connection() can't return new
channel immediately the request should fail (we are in
NBD_CLIENT_CONNECTING_NOWAIT state).
6. nbd_reconnect_attempt() is simplified: it's now easier to wait for
other requests in the caller, so here we just assert that fact.
Also delay time is now initialized here: we can easily detect first
attempt and start a timer.
7. nbd_co_reconnect_loop() is dropped, we don't need it. Reconnect
retries are fully handle by thread (nbd/client-connection.c), delay
timer we initialize in nbd_reconnect_attempt(), we don't have to
bother with s->drained and friends. nbd_reconnect_attempt() now
called from nbd_co_send_request().
8. nbd_connection_entry is dropped: reconnect is now handled by
nbd_co_send_request(), receiving reply is now handled by
nbd_receive_replies(): all handled from request coroutines.
9. So, welcome new nbd_receive_replies() called from request coroutine,
that receives reply header instead of nbd_connection_entry().
Like with sending requests, only one coroutine may receive in a
moment. So we introduce receive_mutex, which is locked around
nbd_receive_reply(). It also protects some related fields. Still,
full audit of thread-safety in nbd driver is a separate task.
New function waits for a reply with specified handle being received
and works rather simple:
Under mutex:
- if current handle is 0, do receive by hand. If another handle
received - switch to other request coroutine, release mutex and
yield. Otherwise return success
- if current handle == requested handle, we are done
- otherwise, release mutex and yield
10: in nbd_co_send_request() we now do nbd_reconnect_attempt() if
needed. Also waiting in free_sema queue we now wait for one of two
conditions:
- connectED, in_flight < MAX_NBD_REQUESTS (so we can start new one)
- connectING, in_flight == 0, so we can call
nbd_reconnect_attempt()
And this logic is protected by s->send_mutex
Also, on failure we don't have to care of removed s->connection_co
11. nbd_co_do_receive_one_chunk(): now instead of yield() and wait for
s->connection_co we just call new nbd_receive_replies().
12. nbd_co_receive_one_chunk(): place where s->reply.handle becomes 0,
which means that handling of the whole reply is finished. Here we
need to wake one of coroutines sleeping in nbd_receive_replies().
If none are sleeping - do nothing. That's another behavior change: we
don't have endless recv() in the idle time. It may be considered as
a drawback. If so, it may be fixed later.
13. nbd_reply_chunk_iter_receive(): don't care about removed
connection_co, just ping in_flight waiters.
14. Don't create connection_co, enable retry in the connection thread
(we don't have own reconnect loop anymore)
15. We now need to add a nbd_co_establish_connection_cancel() call in
nbd_cancel_in_flight(), to cancel the request that is doing a
connection attempt.
[*], ok, now we don't cancel reconnect on drain begin. That's correct:
reconnect feature leads to possibility of long-running requests (up
to reconnect delay). Still, drain begin is not a reason to kill
long requests. We should wait for them.
This also means, that we can again reproduce a dead-lock, described
in 8c517de24a.
Why we are OK with it:
1. Now this is not absolutely-dead dead-lock: the vm is unfrozen
after reconnect delay. Actually 8c517de24a fixed a bug in
NBD logic, that was not described in 8c517de24a and led to
forever dead-lock. The problem was that nobody woke the free_sema
queue, but drain_begin can't finish until there is a request in
free_sema queue. Now we have a reconnect delay timer that works
well.
2. It's not a problem of the NBD driver, but of the ide code,
because it does drain_begin under the global mutex; the problem
doesn't reproduce when using scsi instead of ide.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210902103805.25686-5-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: grammar and comment tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
We are going to use it in nbd_channel_error(), so move it up. Note,
that we are going also refactor and rename
nbd_recv_coroutines_wake_all() in future anyway, so keeping it where it
is and making forward declaration doesn't make real sense.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210902103805.25686-3-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Don't rely on connection being totally broken in case of -EIO. Safer
and more correct is to just shut down the channel anyway, since we
change the state and plan on reconnecting.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210902103805.25686-2-vsementsov@virtuozzo.com>
[eblake: grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
When we don't have a connection and blocking is false, we return NULL
but don't set errp. That's wrong.
We have two paths for calling nbd_co_establish_connection():
1. nbd_open() -> nbd_do_establish_connection() -> ...
but that will never set blocking=false
2. nbd_reconnect_attempt() -> nbd_co_do_establish_connection() -> ...
but that uses errp=NULL
So, we are safe with our wrong errp policy in
nbd_co_establish_connection(). Still let's fix it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210906190654.183421-2-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Now that all drivers are updated by the previous commit, we can drop
the last limiter on pdiscard path: INT_MAX in bdrv_co_pdiscard().
Now everything is prepared for implementing incredibly cool and fast
big-discard requests in NBD and qcow2. And any other driver which wants
it of course.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210903102807.27127-12-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We are generally moving to int64_t for both offset and bytes parameters
on all io paths.
Main motivation is realization of 64-bit write_zeroes operation for
fast zeroing large disk chunks, up to the whole disk.
We chose signed type, to be consistent with off_t (which is signed) and
with possibility for signed return type (where negative value means
error).
So, convert driver discard handlers bytes parameter to int64_t.
The only caller of all updated function is bdrv_co_pdiscard in
block/io.c. It is already prepared to work with 64bit requests, but
pass at most max(bs->bl.max_pdiscard, INT_MAX) to the driver.
Let's look at all updated functions:
blkdebug: all calculations are still OK, thanks to
bdrv_check_qiov_request().
both rule_check and bdrv_co_pdiscard are 64bit
blklogwrites: pass to blk_loc_writes_co_log which is 64bit
blkreplay, copy-on-read, filter-compress: pass to bdrv_co_pdiscard, OK
copy-before-write: pass to bdrv_co_pdiscard which is 64bit and to
cbw_do_copy_before_write which is 64bit
file-posix: one handler calls raw_account_discard() is 64bit and both
handlers calls raw_do_pdiscard(). Update raw_do_pdiscard, which pass
to RawPosixAIOData::aio_nbytes, which is 64bit (and calls
raw_account_discard())
gluster: somehow, third argument of glfs_discard_async is size_t.
Let's set max_pdiscard accordingly.
iscsi: iscsi_allocmap_set_invalid is 64bit,
!is_byte_request_lun_aligned is 64bit.
list.num is uint32_t. Let's clarify max_pdiscard and
pdiscard_alignment.
mirror_top: pass to bdrv_mirror_top_do_write() which is
64bit
nbd: protocol limitation. max_pdiscard is alredy set strict enough,
keep it as is for now.
nvme: buf.nlb is uint32_t and we do shift. So, add corresponding limits
to nvme_refresh_limits().
preallocate: pass to bdrv_co_pdiscard() which is 64bit.
rbd: pass to qemu_rbd_start_co() which is 64bit.
qcow2: calculations are still OK, thanks to bdrv_check_qiov_request(),
qcow2_cluster_discard() is 64bit.
raw-format: raw_adjust_offset() is 64bit, bdrv_co_pdiscard too.
throttle: pass to bdrv_co_pdiscard() which is 64bit and to
throttle_group_co_io_limits_intercept() which is 64bit as well.
test-block-iothread: bytes argument is unused
Great! Now all drivers are prepared to handle 64bit discard requests,
or else have explicit max_pdiscard limits.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210903102807.27127-11-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We are going to support 64 bit discard requests. Now update the
limit variable. It's absolutely safe. The variable is set in some
drivers, and used in bdrv_co_pdiscard().
Update also max_pdiscard variable in bdrv_co_pdiscard(), so that
bdrv_co_pdiscard() is now prepared for 64bit requests. The remaining
logic including num, offset and bytes variables is already
supporting 64bit requests.
So the only thing that prevents 64 bit requests is limiting
max_pdiscard variable to INT_MAX in bdrv_co_pdiscard().
We'll drop this limitation after updating all block drivers.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210903102807.27127-10-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Now that all drivers are updated by previous commit, we can drop two
last limiters on write-zeroes path: INT_MAX in
bdrv_co_do_pwrite_zeroes() and bdrv_check_request32() in
bdrv_co_pwritev_part().
Now everything is prepared for implementing incredibly cool and fast
big-write-zeroes in NBD and qcow2. And any other driver which wants it
of course.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210903102807.27127-9-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We are generally moving to int64_t for both offset and bytes parameters
on all io paths.
Main motivation is realization of 64-bit write_zeroes operation for
fast zeroing large disk chunks, up to the whole disk.
We chose signed type, to be consistent with off_t (which is signed) and
with possibility for signed return type (where negative value means
error).
So, convert driver write_zeroes handlers bytes parameter to int64_t.
The only caller of all updated function is bdrv_co_do_pwrite_zeroes().
bdrv_co_do_pwrite_zeroes() itself is of course OK with widening of
callee parameter type. Also, bdrv_co_do_pwrite_zeroes()'s
max_write_zeroes is limited to INT_MAX. So, updated functions all are
safe, they will not get "bytes" larger than before.
Still, let's look through all updated functions, and add assertions to
the ones which are actually unprepared to values larger than INT_MAX.
For these drivers also set explicit max_pwrite_zeroes limit.
Let's go:
blkdebug: calculations can't overflow, thanks to
bdrv_check_qiov_request() in generic layer. rule_check() and
bdrv_co_pwrite_zeroes() both have 64bit argument.
blklogwrites: pass to blk_log_writes_co_log() with 64bit argument.
blkreplay, copy-on-read, filter-compress: pass to
bdrv_co_pwrite_zeroes() which is OK
copy-before-write: Calls cbw_do_copy_before_write() and
bdrv_co_pwrite_zeroes, both have 64bit argument.
file-posix: both handler calls raw_do_pwrite_zeroes, which is updated.
In raw_do_pwrite_zeroes() calculations are OK due to
bdrv_check_qiov_request(), bytes go to RawPosixAIOData::aio_nbytes
which is uint64_t.
Check also where that uint64_t gets handed:
handle_aiocb_write_zeroes_block() passes a uint64_t[2] to
ioctl(BLKZEROOUT), handle_aiocb_write_zeroes() calls do_fallocate()
which takes off_t (and we compile to always have 64-bit off_t), as
does handle_aiocb_write_zeroes_unmap. All look safe.
gluster: bytes go to GlusterAIOCB::size which is int64_t and to
glfs_zerofill_async works with off_t.
iscsi: Aha, here we deal with iscsi_writesame16_task() that has
uint32_t num_blocks argument and iscsi_writesame16_task() has
uint16_t argument. Make comments, add assertions and clarify
max_pwrite_zeroes calculation.
iscsi_allocmap_() functions already has int64_t argument
is_byte_request_lun_aligned is simple to update, do it.
mirror_top: pass to bdrv_mirror_top_do_write which has uint64_t
argument
nbd: Aha, here we have protocol limitation, and NBDRequest::len is
uint32_t. max_pwrite_zeroes is cleanly set to 32bit value, so we are
OK for now.
nvme: Again, protocol limitation. And no inherent limit for
write-zeroes at all. But from code that calculates cdw12 it's obvious
that we do have limit and alignment. Let's clarify it. Also,
obviously the code is not prepared to handle bytes=0. Let's handle
this case too.
trace events already 64bit
preallocate: pass to handle_write() and bdrv_co_pwrite_zeroes(), both
64bit.
rbd: pass to qemu_rbd_start_co() which is 64bit.
qcow2: offset + bytes and alignment still works good (thanks to
bdrv_check_qiov_request()), so tail calculation is OK
qcow2_subcluster_zeroize() has 64bit argument, should be OK
trace events updated
qed: qed_co_request wants int nb_sectors. Also in code we have size_t
used for request length which may be 32bit. So, let's just keep
INT_MAX as a limit (aligning it down to pwrite_zeroes_alignment) and
don't care.
raw-format: Is OK. raw_adjust_offset and bdrv_co_pwrite_zeroes are both
64bit.
throttle: Both throttle_group_co_io_limits_intercept() and
bdrv_co_pwrite_zeroes() are 64bit.
vmdk: pass to vmdk_pwritev which is 64bit
quorum: pass to quorum_co_pwritev() which is 64bit
Hooray!
At this point all block drivers are prepared to support 64bit
write-zero requests, or have explicitly set max_pwrite_zeroes.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210903102807.27127-8-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: use <= rather than < in assertions relying on max_pwrite_zeroes]
Signed-off-by: Eric Blake <eblake@redhat.com>
We are going to support 64 bit write-zeroes requests. Now update the
limit variable. It's absolutely safe. The variable is set in some
drivers, and used in bdrv_co_do_pwrite_zeroes().
Update also max_write_zeroes variable in bdrv_co_do_pwrite_zeroes(), so
that bdrv_co_do_pwrite_zeroes() is now prepared to 64bit requests. The
remaining logic including num, offset and bytes variables is already
supporting 64bit requests.
So the only thing that prevents 64 bit requests is limiting
max_write_zeroes variable to INT_MAX in bdrv_co_do_pwrite_zeroes().
We'll drop this limitation after updating all block drivers.
Ah, we also have bdrv_check_request32() in bdrv_co_pwritev_part(). It
will be modified to do bdrv_check_request() for write-zeroes path.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210903102807.27127-7-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We are generally moving to int64_t for both offset and bytes parameters
on all io paths.
Main motivation is realization of 64-bit write_zeroes operation for
fast zeroing large disk chunks, up to the whole disk.
We chose signed type, to be consistent with off_t (which is signed) and
with possibility for signed return type (where negative value means
error).
So, convert driver copy_range handlers parameters which are already
64bit to signed type.
Now let's consider all callers. Simple
git grep '\->bdrv_co_copy_range'
shows the only caller:
bdrv_co_copy_range_internal(), which does bdrv_check_request32(),
so everything is OK.
Still, the functions may be called directly, not only by drv->...
Let's check:
git grep '\.bdrv_co_copy_range_\(from\|to\)\s*=' | \
awk '{print $4}' | sed 's/,//' | sed 's/&//' | sort | uniq | \
while read func; do git grep "$func(" | \
grep -v "$func(BlockDriverState"; done
shows no more callers. So, we are done.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210903102807.27127-6-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
We are generally moving to int64_t for both offset and bytes parameters
on all io paths.
Main motivation is realization of 64-bit write_zeroes operation for
fast zeroing large disk chunks, up to the whole disk.
We chose signed type, to be consistent with off_t (which is signed) and
with possibility for signed return type (where negative value means
error).
So, convert driver write handlers parameters which are already 64bit to
signed type.
While being here, convert also flags parameter to be BdrvRequestFlags.
Now let's consider all callers. Simple
git grep '\->bdrv_\(aio\|co\)_pwritev\(_part\)\?'
shows that's there three callers of driver function:
bdrv_driver_pwritev() and bdrv_driver_pwritev_compressed() in
block/io.c, both pass int64_t, checked by bdrv_check_qiov_request() to
be non-negative.
qcow2_save_vmstate() does bdrv_check_qiov_request().
Still, the functions may be called directly, not only by drv->...
Let's check:
git grep '\.bdrv_\(aio\|co\)_pwritev\(_part\)\?\s*=' | \
awk '{print $4}' | sed 's/,//' | sed 's/&//' | sort | uniq | \
while read func; do git grep "$func(" | \
grep -v "$func(BlockDriverState"; done
shows several callers:
qcow2:
qcow2_co_truncate() write at most up to @offset, which is checked in
generic qcow2_co_truncate() by bdrv_check_request().
qcow2_co_pwritev_compressed_task() pass the request (or part of the
request) that already went through normal write path, so it should
be OK
qcow:
qcow_co_pwritev_compressed() pass int64_t, it's updated by this patch
quorum:
quorum_co_pwrite_zeroes() pass int64_t and int - OK
throttle:
throttle_co_pwritev_compressed() pass int64_t, it's updated by this
patch
vmdk:
vmdk_co_pwritev_compressed() pass int64_t, it's updated by this
patch
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210903102807.27127-5-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
