… (monotonic interface naming) (#12826)

* KVM: assign explicit PCI slot when hot-plugging NIC to ensure sequential naming

When hot-plugging a NIC to a running VM, libvirt auto-assigns the next
free PCI slot. Since non-NIC devices (virtio-serial, disk, balloon,
watchdog) occupy slots immediately after existing NICs, the hot-plugged
NIC gets a much higher slot number (e.g. 0x09 instead of 0x05), causing
the guest to see non-sequential interface names (ens9 instead of ens5).

This fix queries the domain XML to find all used PCI slots and assigns
the next free slot after the highest existing NIC slot. This matches
the approach already used by LibvirtReplugNicCommandWrapper which
preserves PCI slots during re-plug operations.

Fixes #12825

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(kvm): NPE in PlugNic when libvirt domain XML is unavailable

LibvirtPlugNicCommandWrapper.findNextAvailablePciSlot calls
vm.getXMLDesc(0) and pipes the result straight into Pattern.matcher,
which NPEs if libvirt returned null (or, in the
LibvirtComputingResourceTest.testPlugNicCommandNoMatchMack unit test,
when the Domain mock isn't stubbed for getXMLDesc). Reported by
@DaanHoogland after the SL packaging run on #12826.

Defensive null check returns null from findNextAvailablePciSlot when
the domain XML can't be parsed, which falls through to libvirt's
auto-assignment of the PCI slot — same behaviour as before this PR
when nextSlot is null.

Also stubs Domain.getXMLDesc(0) in testPlugNicCommandNoMatchMack with
a minimal <domain> XML that exercises the parser path (rather than
just relying on the null-fallback), so the test continues to cover
the happy path of the new logic.

* address review (#12826): parse PCI addresses with an XML parser, split slot selection into helpers, add unit tests

- getUsedPciSlots() parses the domain XML with the safer DocumentBuilderFactory
  and only considers <address type='pci'> elements, replacing the regex.
- getHighestNicSlot() and getFirstFreeSlotAbove() are separate methods.
- The javadoc now states the guarantee precisely: deterministic and monotonic
  after the last NIC, not contiguous when other devices sit in between.
- LibvirtPlugNicCommandWrapperTest covers parsing, selection and the fallbacks.

Signed-off-by: James Peru <jmsperu@gmail.com>

---------

Signed-off-by: James Peru <jmsperu@gmail.com>
Co-authored-by: James Peru <jamesperu@Jamess-Mac-mini.local>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: jmsperu <jmsperu@users.noreply.github.com>

Template/ISO for a VM could be in deleted state therefore links should not be shown for them in the VM info-card.

Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>

…eUtil (#13809)

* Improve date/timestamp handling performance in GenericDaoBase and DateUtil

* address review comment

…4025)

…4169)

* lock on only domains that have finite limits for checked reservation

* fix domain lock set to reflect tag->untagged fallback and limit inheritance

listRowsToLockForLimitCheck previously only locked ancestor domains that
owned an explicit finite resource_limit row for the exact tag, so it
missed domains that inherit a finite limit from an ancestor's row, and
domains that fall back to the untagged limit when no tag-specific limit
is configured (per findCorrectResourceLimitForDomain). Either gap let
concurrent reservations bypass domain-level serialization.

Replace the set-based approximation with an in-memory walk of the
account's domain chain that replicates findCorrectResourceLimitForDomain's
nearest-row-with-fallback semantics, using bulk queries (covered by
existing indexes) to fetch the chain's resource_limit rows and to resolve
the final resource_count rows to lock.

---------

Co-authored-by: Mark Armstrong <marms@apple.com>

)

* server: scope IPv6 security group member rules to the exact host

When a security group rule references another security group, each member VM
should be authorized as an exact host. The IPv4 address is correctly pinned to
a /32, but the IPv6 address was expanded to /64, opening the whole subnet the
member sits in rather than just that member. Pin the IPv6 member to /128 to
match the IPv4 behaviour.

* server: scope the Impl2 IPv6 security group member rule to the exact host