Qemu's bitmap APIs don't reopen the appropriate images read-write for
modification. It's libvirt's duty to reopen them via blockdev-reopen
if we wish to modify the bitmaps.

Use the new helpers to reopen the images for bitmap manipulation.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

Introduce a set of helpers to call blockdev-reopen in certain scenarios

Libvirt will use the QMP command to turn certain members of the backing
chain read-write for bitmap manipulation and we'll also want to use it
to replace/install the backing chain of a qcow2 format node.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

Introduce the monitor code for using blockdev-reopen.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

This capability will be asserted once qemu stabilizes 'blockdev-reopen'.
For now we just add the capability so that we can introduce some code
that will use the reopening call. This will show our willingness to
adopt use of reopen and help qemu developers stabilize it.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NEric Blake <eblake@redhat.com>

Signed-off-by: NGaurav Agrawal <agrawalgaurav@gnome.org>
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>

Add sample data files for validating handling of a QEMU guest started
with:

  -smp 7,maxcpus=16,sockets=2,dies=2,cores=2,threads=2
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>
Signed-off-by: NDaniel P. Berrangé <berrange@redhat.com>

The logic for querying hotpluggable CPUs needs to sort the list
of CPUs returned by QEMU. Unfortunately our sorting method failed
to use the die_id field, so CPUs were not correctly sorted.

This is seen when configuring a guest with partially populated
CPUs

  <vcpu placement='static' current='1'>16</vcpu>
  <cpu...>
    <topology sockets='4' dies='2' cores='1' threads='2'/>
  </cpu>

Then trying to start it would fail:

  # virsh -c qemu:///system start demo
  error: Failed to start domain demo
  error: internal error: qemu didn't report thread id for vcpu '0'
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>
Signed-off-by: NDaniel P. Berrangé <berrange@redhat.com>

Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reported-by: NDr. David Alan Gilbert <dgilbert@redhat.com>

We currently don't model the 'ssh' protocol properties properly and
since it seems impossible for now (agent path passed via environment
variable). To allow libguestfs to work as it used in pre-blockdev era we
must carry the properties over to the command line. For this instance we
just store it internally and format it back.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Add testing of the interpretation of the JSON pseudo-protocol backing
store into JSON structs for blockdev. This will be used to test JSON
pseudo-URIs used by libguestfs while actually also validating the output
against the QMP schema. Since libguestfs uses obsolete/undocumented
values the outputs will differ and a benefit is that modern output is
used now.

The example test case covers the fields and values used by libguestfs
when using the https driver.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The test was invoking the JSON formatter with the 'legacy' flag thus
formatting bunch of obsolete JSON blockdev definitions. We also should
test the modern ones. Add a boolean and re-run all the tests in both
cases.

Additionally for any modern invocation we should also validate that the
output conforms to the QAPI schema.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Move lookup of the schema root earlier so that multiple functions
can use it for validation.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Multiple tests require the schema. Extract the loading into a separate
variable to avoid issues with ownership of the pointer.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

libguestfs abuses a quirk of qemu's parser to accept also other variants
of the 'sslverify' field which would be valid on the command line but
are not documented in the QMP schema.

If we encounter the 'off' string instead of an boolean handle it rather
than erroring out to continue support of pre-blockdev configurations.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Add support for parsing the recently added fields from backing file
pseudo-protocol strings.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Pass in the correct fields.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Pass the alias of the secret object holding the cookie data as
'cookie-secret' to qemu.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Implement both commandline support and hotplug by adding the http cookie
handling to 'qemuBlockStorageSourceAttachData' handling functions for
it.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

QEMU's curl driver requires the cookies concatenated and allows themi to
be passed in via a secret. Prepare the value for the secret and encrypt
it.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The http cookies can have potentially sensitive values and thus should
not be leaked into the command line. This means that we'll need to
instantiate a 'secret' object in qemu to pass the value encrypted.

This patch adds infrastructure for storing of the alias in the status
XML.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Allow disabling of SSL certificate validation for HTTPS and FTPS drives
in qemu.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Upcoming patches will implement the support for sslverify, cookies,
readahead, and timeout properties. Add a test file which will collect
the cases.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Ensure that the new fields are allowed only when -blockdev is used or
when they are in the detected part of the backing chain where qemu will
handle them internally.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Some disk backends support configuring the readahead buffer or timeout
for requests. Add the knobs to the XML.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Add possibility to specify one or more cookies for http based disks.
This patch adds the config parser, storage and validation of the
cookies.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

To allow turning off verification of SSL cerificates add a new element
<ssl> to the disk source XML which will allow configuring the validation
process using the 'verify' attribute.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

There are two last callers of this function. Replace them by
qemuAliasForSecret and delete qemuDomainGetSecretAESAlias.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Originally there was only the secret for authentication so we didn't use
any suffix to tell it apart. With the introduction of encryption we
added a 'luks' suffix for the encryption secrets. Since encryption is
really generic and authentication is not the only secret modify the
aliases for the secrets to better describe what they are used for.

This is possible as we store the disk secrets in the status XML thus
only new machines will use the new secrets.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Replace qemuDomainGetSecretAESAlias by the new function so that we can
reuse qemuDomainSecretAESSetupFromSecret also for setting up other kinds
of objects.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Currently we don't have infrastructure to remember the secret aliases
for hostdevs. Since an upcoming patch is going to change aliases for
the disks, initialize the iscsi hostdevs separately so that we can keep
the alias. At the same time let's use qemuAliasForSecret instead of
qemuDomainGetSecretAESAlias when unplugging the iscsi hostdev.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

In order to be able to change the function generating the alias and thus
also the aliases itself, we must hardcode the old format for the case of
upgrading form libvirt which didn't record them in the status XML yet.

Note that this code path is tested by
'tests/qemustatusxml2xmldata/disk-secinfo-upgrade-in.xml'
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The naming of the variables was tied to what they are used for not what
the alias represents. Since we'll need to use some of the aliases for
another type of secrets fix the name so that it makes sense.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

qemuAliasForSecret is meant as a replacement qemuDomainGetSecretAESAlias
with saner API. The sub-type we are creating the alias for is passed in
as a string rather than the unflexible 'isLuks' boolean.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

We need the "$(space)" variable to contain a single whitespace
character. We do this by assigning and then appending an empty
string to the variable. Variable appends get separated by a
single whitespace historically, but GNU make 4.3 introduced a
behaviour regression.

  https://lists.gnu.org/archive/html/bug-make/2020-01/msg00057.html

[quote]
* WARNING: Backward-incompatibility!
  Previously appending using '+=' to an empty variable would
  result in a value starting with a space.  Now the initial
  space is only added if the variable already contains some
  value.  Similarly, appending an empty string does not
  add a trailing space.
[/quote]

This patch tries a new trick to get a single whitespace by
getting make to expand two non-existant variables separated
by a space.
Reviewed-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NDaniel P. Berrangé <berrange@redhat.com>

Replace it by a direct call to qemuDomainSecretAESSetupFromSecret.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Split out the lookup of the secret from the secret driver into
qemuDomainSecretAESSetupFromSecret so that we can also instantiate
secret objects in qemu with data from other sources.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Rather than passing in an empty qemuDomainSecretInfoPtr allocate it
in this function and return it. This is done by absorbing the check from
qemuDomainSecretInfoNew and removing the internals of
qemuDomainSecretInfoNew.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Use g_autofree for the ciphertext and init vector as they are not
secret and thus don't have to be cleared and use g_new0 to allocate the
iv for parity.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The comment mentioned that the function resets migration params, but
that is not true as of commit eb54cb47Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>