Before we exec() qemu we have to spawn pr-helper processes for
all managed reservations (well, technically there can only one).
The only caveat there is that we should place the process into
the same namespace and cgroup as qemu (so that it shares the same
view of the system). But we can do that only after we've forked.
That means calling the setup function between fork() and exec().
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Just like we allow users overriding path to bridge-helper
detected at compile time we can allow them to override path to
qemu-pr-helper.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Just like in previous commit, qemu-pr-helper might want to open
/dev/mapper/control under certain circumstances. Therefore we
have to allow it in cgroups.

The change virdevmapper.c might look spurious but it isn't. After
6dd84f68 any path that we're allowing in deivces CGroup is
subject to virDevMapperGetTargets() inspection. And libdevmapper
returns ENXIO for the path from subject.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

If qemu-pr-helper is compiled with multipath support the first
thing it does is open /dev/mapper/control. Since we're going
to be running it inside qemu namespace we need to create it
there. Unfortunately, we don't know if it was compiled with or
without multipath so we have to create it anyway.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

For command line we need two things:

1) -object pr-manager-helper,id=$alias,path=$socketPath
2) -drive file.pr-manager=$alias

In -object pr-manager-helper we tell qemu which socket to connect
to, then in -drive file-pr-manager we just reference the object
the drive in question should use.

For managed PR helper the alias is always "pr-helper0" and socket
path "${vm->priv->libDir}/pr-helper0.sock".
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

The capability tracks if qemu has pr-manager-helper object. At
this time don't actually detect if qemu has the capability. Not
just yet. Only after the code is written the feature will be
enabled.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Couple of reasons for that:

a) there's no monitor command to change path where the pr-helper
connects to, or
b) there's no monitor command to introduce a new pr-helper for a
disk that already exists.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

This is a definition that holds information on SCSI persistent
reservation settings. The XML part looks like this:

  <reservations enabled='yes' managed='no'>
    <source type='unix' path='/path/to/qemu-pr-helper.sock' mode='client'/>
  </reservations>

If @managed is set to 'yes' then the <source/> is not parsed.
This design was agreed on here:

https://www.redhat.com/archives/libvir-list/2017-November/msg01005.htmlSigned-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Rather than have virJSONValueArraySize return a -1 when the input
is not an array and then splat an error message, let's check for
an array before calling and then change the return to be a size_t
instead of ssize_t.

That means using the helper virJSONValueIsArray as well as using a
more generic error message such as "Malformed <something> array".
In some cases we can remove stack variables and when we cannot,
those variables should be size_t not ssize_t. Alter a few references
of if (!value) to be if (value == 0) instead as well.

Some callers can already assume an array is being worked on based
on the previous call, so there's less to do.
Signed-off-by: NJohn Ferlan <jferlan@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The pointer to the qemu driver is already included in domain object's
private data, so does not need to be passed as yet another parameter
when the domain object is already passed.

Also removes parameter 'driver' from functions which had it just because of
qemuBlockJobUpdate.
Signed-off-by: NRoland Schulz <schullzroll@gmail.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>

Signed-off-by: NLin Ma <lma@suse.com>

Signed-off-by: NLin Ma <lma@suse.com>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Currently the VSH_OT_ARGV options don't support complete, But some of
VSH_OT_ARGV options are gonna support complete in upcoming patches.

Once applied the upcoming completion patches for VSH_OT_ARGV options, If
we don't ignore VSH_OT_ARGV here, The vshReadlineOptionsGenerator will
be called, Hence complete output will consist of the result by command
completer + the result by option completer, It's confusing.
e.g.
    $ virsh domstats --domain <TAB><TAB>
    --backing     --interface      --list-paused      --perf      --vcpu
    --balloon     leap42.3         --list-persistent  --raw       win10
    --block       --list-active    --list-running     sles12sp3
    --cpu-total   --list-inactive  --list-shutoff     sles15
    --enforce     --list-other     --list-transient   --state

After this patch and the upcoming completion patches:
    $ virsh domstats --domain <TAB><TAB>
    leap42.3    sles12sp3    sles15    win10
Signed-off-by: NLin Ma <lma@suse.com>

It's helpful for users while they type certain kind of VSH_OT_ARGV options.
e.g.

$ virsh domstats --domain sles12sp3 --d<TAB>
Signed-off-by: NLin Ma <lma@suse.com>

These VSH_OT_STRING "domain" options support domain name completion now.
Signed-off-by: NLin Ma <lma@suse.com>

Signed-off-by: NLin Ma <lma@suse.com>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

The first entry in the returned array is the substitution for TEXT. It
causes unnecessary output if other commands or options share the same
prefix, e.g.

$ virsh des<TAB><TAB>
des      desc     destroy

or

$ virsh domblklist --d<TAB><TAB>
--d        --details  --domain

This patch fixes the above issue.
Signed-off-by: NLin Ma <lma@suse.com>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Signed-off-by: NLin Ma <lma@suse.com>

centralize the definition of macro VIRSH_COMMON_OPT_DOMAIN_FULL to virsh.h
to avoid unnecessary duplicated definition
Signed-off-by: NLin Ma <lma@suse.com>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Unused since commit <a7424faf>.
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Use the new helper when checking that the VM needs to be tainted as a
host-cdrom passthrough.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NLaine Stump <laine@laine.org>

It's only required on el5 which we don't support anymore. Everywhere
else it's not used for anything useful

https://fedoraproject.org/wiki/RPMGroupsSigned-off-by: NCole Robinson <crobinso@redhat.com>

None of currently supported distributions need that.
Last one was EL5 which is EOL for a while.
Signed-off-by: NIgor Gnatenko <ignatenkobrain@fedoraproject.org>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

None of currently supported distributions need that.
It was needed last for EL5 which is EOL now
Signed-off-by: NIgor Gnatenko <ignatenkobrain@fedoraproject.org>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

Remove my name and email from these files
Reviewed-by: NDaniel P. Berrangé <berrange@redhat.com>
Signed-off-by: NDavid Kiarie <davidkiarie4@gmail.com>

STRPREFIX takes only two arguments, but the code it was adapted from
used function with 3 arguments.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>

Make sure that 'host_device' is generated for type='block'.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

The test cases would correspond to the following -drive command lines:

dir-fat-cache.xml:
-drive file=fat:/var/somefiles,if=none,id=drive-dummy,readonly=on,cache=directsync
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=off

file-backing_basic-cache-directsync.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow2,if=none,id=drive-dummy,cache=directsync
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=off

file-backing_basic-cache-none.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow2,if=none,id=drive-dummy,cache=none
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=on

file-backing_basic-cache-unsafe.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow2,if=none,id=drive-dummy,cache=unsafe
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=on

file-backing_basic-cache-writeback.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow2,if=none,id=drive-dummy,cache=writeback
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=on

file-backing_basic-cache-writethrough.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow2,if=none,id=drive-dummy,cache=writethrough
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=off

network-qcow2-backing-chain-cache-unsafe.xml:
-drive file=rbd:rbdpool/rbdimg:id=testuser-rbd:auth_supported=cephx\;none:
        mon_host=host1.example.com\;host2.example.com,
        file.password-secret=node-a-s-secalias,format=qcow2,
        if=none,id=drive-dummy,cache=directsync
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=off
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

The test cases would correspond to the following -drive command lines:

file-backing_basic-detect.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow,if=none,id=drive-dummy,detect-zeroes=on
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-backing_basic-unmap-detect.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow,if=none,id=drive-dummy,discard=unmap,detect-zeroes=unmap
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-backing_basic-unmap-ignore.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow,if=none,id=drive-dummy,discard=ignore,detect-zeroes=on
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-backing_basic-unmap.xml:
-drive file=/var/lib/libvirt/images/a,format=qcow,if=none,id=drive-dummy,discard=unmap
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

iscsi and rbd support authentication of the connection. Combine it with
encryption of qcow2.

The top level disk image would generate the following '-drive' cmdline:

-drive file=rbd:rbdpool/rbdimg:id=testuser-rbd:auth_supported=cephx\;none:
            mon_host=host1.example.com\;host2.example.com,
            file.password-secret=node-a-s-secalias,encrypt.format=luks,
            encrypt.key-secret=node-b-f-encalias,format=qcow2,
            if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Add tests for backing chain handling, including a very long chain which
is fully specified in the XML and an unterminated chain.

The top level disk image would generate the following '-drive':

file-qcow2-backing-chain-encryption.xml:
-drive file=/var/lib/libvirt/images/a,encrypt.format=luks,
    encrypt.key-secret=node-b-f-encalias,format=qcow2,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-qcow2-backing-chain-noopts.xml:
-drive file=/var/lib/libvirt/images/rhel7.3.1507297895,format=qcow2,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-qcow2-backing-chain-unterminated.xml:
-drive file=/var/lib/libvirt/images/rhel7.3.1507297895,format=qcow2,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Apart from adding test data add a function which sets up fake secrets
for the test.

The top level disk image would generate the following '-drive' cmdline:

-drive file=/path/luks.img,key-secret=test1-encalias,format=luks,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Test that the 'aio' option is applied correctly for the 'file' protocol
backend and across the backing chain.

The top level disk image would generate the following '-drive' cmdline:

file-backing_basic-aio_threads:
-drive file=/var/lib/libvirt/images/a,format=qcow,if=none,id=drive-dummy,aio=threads
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-raw-aio_native:
-drive file=/path/to/i.img,format=raw,if=none,id=drive-dummy,cache=none,aio=native
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy,write-cache=on
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Test mapping of the 'FAT' disk format to 'vvfat' in qemu.

The top level disk image would generate the following '-drive' cmdline:

dir-fat-readonly.xml:
-drive file=fat:/var/somefiles,if=none,id=drive-dummy,readonly=on
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

dir-fat-floppy.xml
-drive file=fat:floppy:/var/somefiles,if=none,id=drive-dummy,readonly=on
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Formats supporting backing chain such as qed, vmdk, don't have any other
parameters than the backing store and 'qcow' has only encryption params
which will be tested extra. Add this test case so they are covered since
any further test cases will mainly care about 'qcow2' and 'raw'.

The top level disk image would generate the following '-drive' cmdline:

-drive file=/var/lib/libvirt/images/a,format=qed,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Similarly to the 'raw' case add tests for bochs, cloop, dmg, ploop, vdi
vhd, and vpc. Covering all supported non-backing formats.

Note that the JSON name for 'ploop' maps to 'parallels' and 'vhd' maps
to 'vhdx'.

Files added here would result in the followint configs:

file-bochs-noopts.xml:
-drive file=/path/to/i.img,format=bochs,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-cloop-noopts.xml:
-drive file=/path/to/i.img,format=cloop,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-dmg-noopts.xml:
-drive file=/path/to/i.img,format=dmg,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-ploop-noopts.xml:
-drive file=/path/to/i.img,format=ploop,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-vdi-noopts.xml:
-drive file=/path/to/i.img,format=vdi,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-vhd-noopts.xml:
-drive file=/path/to/i.img,format=vhd,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy

file-vpc-noopts.xml:
-drive file=/path/to/i.img,format=vpc,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Test the JSON props generator with a very simple 'raw' image with no
other options. The node-names for the image are 31 bytes long so that we
validate our node name detector.

The top level disk image would generate the following '-drive' cmdline:

-drive file=/var/lib/libvirt/images/i.img,format=raw,if=none,id=drive-dummy
-device virtio-blk-pci,scsi=off,drive=drive-dummy,id=dummy
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Add a test infrastructure that will allow testing the JSON object
generator used for generating data to use with blockdev-add.

The resulting disk including the backing chain is validated to conform
to the QAPI schema and the expected output files.

The first test cases make sure that libvirt will not allow nodenames
exceeding 31 chars.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

The function will be reused in the test code where we don't care much
that the gluster debug level can't be populated from the qemu config.

Set the level only when 'cfg' is passed.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

It will be used in the qemublocktest.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>