Extract test logic from test_memory_overhead into a shared
_test_memory_overhead wrapper function in tests_wrappers.rs.
The custom memory size is set in the parent test case via
with_memory(). The wrapper uses default_kernel_cmdline()
and default_memory() for setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_pci_bar_reprogramming into
a shared _test_pci_bar_reprogramming wrapper function in
tests_wrappers.rs. The kernel_path selection for aarch64
(edk2) is kept in the parent test case. The wrapper uses
default_kernel_cmdline() for kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_direct_kernel_boot_noacpi into
a shared _test_direct_kernel_boot_noacpi wrapper function in
tests_wrappers.rs. The kernel cmdline modification (acpi=off)
is kept in the parent test case. The wrapper uses
default_kernel_cmdline() for kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_console_file into a shared
_test_console_file wrapper function in tests_wrappers.rs.
Update the parent test case to use the basic_regular_guest
macro. The wrapper uses default_kernel_cmdline() for
kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_virtio_console into a shared
_test_virtio_console wrapper function in tests_wrappers.rs.
Update the parent test case to use the basic_regular_guest
macro. The wrapper uses default_kernel_cmdline() for
kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_multiple_network_interfaces into
a shared _test_multiple_network_interfaces wrapper function in
tests_wrappers.rs. Update the parent test case to use the
basic_regular_guest macro. The wrapper uses
default_kernel_cmdline() for kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Extract test logic from test_serial_off into a shared
_test_serial_off wrapper function in tests_wrappers.rs.
Update the parent test case to use the basic_regular_guest
macro. The wrapper uses default_kernel_cmdline() for
kernel/cmdline setup.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add a CVM variant of the DMI OEM strings test in
integration_cvm.rs. This test creates a confidential
guest using the basic_cvm_guest! macro and invokes the
shared _test_dmi_oem_strings() wrapper to validate DMI
OEM strings behavior under CVM.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move the DMI OEM strings test logic from integration.rs
into a shared _test_dmi_oem_strings() function in
tests_wrappers.rs. The original test in integration.rs
now uses the basic_regular_guest! macro and delegates to
this shared function, enabling reuse by CVM tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add a CVM variant of the DMI UUID test in
integration_cvm.rs. This test creates a confidential
guest using the basic_cvm_guest! macro and invokes the
shared _test_dmi_uuid() wrapper to validate DMI UUID
behavior under CVM.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move the DMI UUID test logic from integration.rs into a
shared _test_dmi_uuid() function in tests_wrappers.rs.
The original test in integration.rs now uses the
basic_regular_guest macro and delegates to this shared
function, enabling reuse by CVM tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_api_http_create_boot, chaining with_cpu(4) for the
required CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_api_http_pause_resume, chaining with_cpu(4) for the
required CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_api_http_delete, chaining with_cpu(4) for the
required CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_api_http_shutdown, chaining with_cpu(4) for the
required CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_virtio_vsock_hotplug. The aarch64 variant chains
with_kernel_path for EDK2 firmware support.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_virtio_vsock for consistency and reduced boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_direct_kernel_boot for consistency and reduced
boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_pci_multiple_segments for consistency and reduced
boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_virtio_net_ctrl_queue for consistency and reduced
boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_pci_msi for consistency and reduced boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_power_button for consistency and reduced boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_virtio_queue_affinity, chaining with_cpu(4) for the
required CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_cpu_affinity, chaining with_cpu(2) for the required
CPU count configuration.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Replace manual UbuntuDiskConfig and GuestFactory guest
creation with the basic_regular_guest! macro in
test_multi_cpu for consistency and reduced boilerplate.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move the DMI serial number test logic from integration.rs
into a shared _test_dmi_serial_number() function in
tests_wrappers.rs. The original test now delegates to
this shared function, enabling reuse by CVM tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add a CVM variant of the split IRQ chip test in
integration_cvm.rs. This test creates a confidential
guest and invokes the shared _test_split_irqchip()
wrapper to validate split IRQ chip behavior under CVM.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move the split IRQ chip test logic from integration.rs
into a shared _test_split_irqchip() function in
tests_wrappers.rs. The original test now delegates to
this shared function, enabling reuse by CVM tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add a CVM variant of the virtio block dynamic VHDX expand
test in integration_cvm.rs. This test creates a
confidential guest and invokes the shared
_test_virtio_block_dynamic_vhdx_expand() wrapper to
validate VHDX expansion under CVM.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move the virtio block dynamic VHDX expand test logic from
integration.rs into a shared _test_virtio_block_dynamic_
vhdx_expand() function in tests_wrappers.rs. The original
test in integration.rs now calls this shared function,
enabling reuse by CVM tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Refactor test_focal_ovmf to use the basic_regular_guest!
macro with the with_kernel() builder instead of manually
constructing UbuntuDiskConfig and setting kernel_path.
This also removes leftover unused disk_config in
test_focal_hypervisor_fw.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Introduce a basic_regular_guest! macro in integration.rs
to reduce boilerplate when creating regular (non-CVM)
guest instances. Also add a with_kernel() builder method
to Guest in test_infra, allowing fluent configuration of
the kernel path.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Introduce a basic_cvm_guest! macro in integration_cvm.rs
to reduce boilerplate when creating confidential VM guest
instances. This replaces repetitive UbuntuDiskConfig and
GuestFactory::new_confidential_guest_factory() calls
across multiple CVM test functions.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Move all CVM (Confidential VM) integration tests from
integration.rs into a dedicated integration_cvm.rs file.
This separation improves code organization and makes it
easier to manage CVM-specific test cases independently
from regular integration tests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This summarizes the latest state all relevant parties have agreed on
in the last meeting.
This commit was rephrased and improved with the help of Claude Code
using the Sonnet 4.6 model.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Mosts tests used 4GB of RAM, although the VM is mostly idling. In CI, we
experienced OOM issues on the ARM runners. If we reduce the VM memory of
the parallel live migration tests to 1.5GB RAM, we still have enough
capacity in the VM so that everything succeeds while reducing resource
usage.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This adds two new integration tests for the new functionality:
- VM under load, downtime=1ms, timeout=1s, timeout_strategy=cancel
- VM under load, downtime=1ms, timeout=1s, timeout_strategy=force
By using a short downtime and timeout plus adding a stress worker in the
guest, we can prevent quick migration. Therefore, we can nicely test the
timeout_strategy.
Testing for a specific downtime is cumbersome to do and highly depends
on CPU/host utilization. To prevent flakiness, there is no such test
integration test. I did, however, manual testing of that functionality.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Wire the new `downtime_ms`, `timeout_s`, and `timeout_strategy` fields
from `VmSendMigrationData` into the precopy loop, replacing the previous
hard-coded 5-iteration cap.
Each iteration now evaluates three convergence criteria in order:
- no dirty pages remain;
- the estimated final-iteration downtime is within the configured budget
- or the overall migration timeout has elapsed.
On timeout, `TimeoutStrategy::Cancel` aborts and keeps the VM live on
the source, while `TimeoutStrategy::Force` proceeds regardless of the
downtime target. The convergence callback is updated to return a Result
to propagate the cancel error cleanly up the call stack.
With the recent changes [0], it is fairly easy to implement the new
checks and operate on actual metrics.
These changes are inspired by [1] but differ significantly in details.
[0] https://github.com/cloud-hypervisor/cloud-hypervisor/pull/7799
[1] https://github.com/cloud-hypervisor/cloud-hypervisor/pull/7033
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Management software needs fine-grained control over live migration to
meet QoS requirements for VM guests. Add `downtime_ms`, `timeout_s`, and
`timeout_strategy` fields to `VmSendMigrationData`, exposed via API.
This commit contains the API changes only; the VMM does not yet act on
these values. This follows in the next commit.
For the JSON API, downtime and timeout are represented as plain integers
(downtime_ms and timeout_s) to make the units explicit. Using Duration
directly would require custom (de)serialization logic, so instead the
internal raw integers are exposed as Duration via getters. This
introduces minor conversion overhead but keeps the Rust API clear and
unambiguous.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This change prepares upcoming options (following commit) that are added
to VmSendMigrationData.
VmSendMigrationData is a special case as it is currently the only
"rich configuration" type that lives outside `config.rs`, as it is
purely API-facing. Therefore, it isn't integrated into the existing
OptionParser infrastructure. We therefore introduce a `parse()` method
to use that in `ch-remote` in the following.
In `ch-remote`, we remove `--local` for `send-migration` and switch to
the new option string parsing constructor (breaking change!). This
prepares the addition of downtime and timeout options in the following
and streamlines the `ch-remote` command line interface with other
commands, such as `ch-remote add-net`.
Lastly, this commit updates the integration tests.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
When configuring multiple queues for a virtio device, the guest can
activate between 1 and the configured amount of queues. The firmware,
for example, may activate only one queue, while a Linux guest would
likely activate all available queues.
The constructor of virtio-blk initializes the `paused_sync` barrier
using the configured queue count (plus one for the main thread). This
can be wrong if the guest enable a different number of queues at
activation time, which can make pause hang. Thus, we now recompute the
barrier size from the queues that are actually activated.
On-behalf-of: SAP sebastian.eydam@sap.com
Signed-off-by: Sebastian Eydam <sebastian.eydam@cyberus-technology.de>
When configuring multiple queues for a virtio device, the guest can
activate between 1 and the configured amout of queues. The firmware,
for example, may activate only one queue, while a Linux guest would
likely activate all available queues.
The constructor of virtio-net initializes the `paused_sync` barrier
using the configured queue count (plus one for the main thread). This
can be wrong if the guest enables a different number of queues at
activation time, which can make pause hang. Thus, we now recompute the
barrier size from the queues that are actually activated.
On-behalf-of: SAP sebastian.eydam@sap.com
Signed-off-by: Sebastian Eydam <sebastian.eydam@cyberus-technology.de>
Include the number of segments found in the request payload in
the TooManySegments error variant so the logged message shows
both the actual and maximum values.
Suggested-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Introduce DISCARD_WZ_MAX_PAYLOAD as the precomputed product of
DISCARD_WZ_SEG_SIZE and MAX_DISCARD_WRITE_ZEROES_SEG. Use it in the
DISCARD and WRITE_ZEROES segment count checks instead of repeating
the multiplication inline.
Suggested-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The libc crate provides all functionality provided by the net_gen crate.
Removing the net_gen crate reduces the maintenance burden.
The switch to libc required some fixes, most notably the switch from a
`Vec<u8>` to a `CString` for the `net_util::Tap.if_name` field.
On-behalf-of: SAP julian.schindel@sap.com
Signed-off-by: Julian Schindel <julian.schindel@cyberus-technology.de>
