As well as rejecting writes to sector 0 in the case of raw files where
the user hasn't specified the image_type also reject virtio requests of
type discard and write_zeroes.
Signed-off-by: Rob Bradford <rbradford@meta.com>
If the disk image was autodetected to raw (not specified with image_type
= 0) then in the virtio-block subsystem generate errors for writes to
block 0 (treat as if read-only). This gives an immediate error vs using
the image implementations in the block subsystem.
Signed-off-by: Rob Bradford <rbradford@meta.com>
Add an image_type to DiskConfig to specify the image type. If none is
specified autodetect the image type but disable potentially unsafe
behaviour in the QCOW2 backend by disabling the backing file support.
If the image type is autodetected then fix it in the config so that it
will be persistant across reboots and migrations/snapshot & restores.
This also handles the case where the image type was not specified as
part of the disk configuration.
Signed-off-by: Rob Bradford <rbradford@meta.com>
Verify that opening a QCOW2 image with a backing file reference
through QcowDiskSync with backing_files=off produces the user-facing
BackingFilesDisabled error rather than MaxNestingDepthExceeded.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
When a QCOW2 image has a backing file but backing_files=on is not set,
the error was MaxNestingDepthExceeded which gives no indication that
this is a policy decision or how to resolve it.
Add a BackingFilesDisabled error variant whose message indicates that
backing file support is disabled and references the backing_files
option. The translation from MaxNestingDepthExceeded to
BackingFilesDisabled happens at the QcowDiskSync boundary where the
policy decision is made, preserving the original error for genuine
recursive depth exhaustion.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The 0.6.7 version of the mshv crates introduced a new version of
make_default_partition_create_arg inside `struct Mshv`. This version
queries the available processor features on the host and gives the same
feature set to the guests.
Move Cloud Hypervisor to this new function.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
test_guest_numa_generic_initiator was missing the #[test] attribute and
did not match the VFIO CI filter pattern "vfio::test_nvidia"
Add #[test] and rename to test_nvidia_guest_numa_generic_initiator so
the existing CI infrastructure picks it up on the vfio-nvidia runner.
Fixes: #7718
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
The device_id field was added to both NumaConfig and NumaNode as part
of the Generic Initiator support, but create_numa_nodes() change
was missed when the commits were reorganized.
As a result, node.device_id is never propogated from the config to
the runtime node and the ACPI SRAT Type 5 (Generic Initiator Affinity)
entries were never emitted.
Add the missing propogation so that create_srat_table() can resolve
the device and emit the correct affinity structure
Fixes: #7717
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Add performance tests for standalone qcow2 images without backing
files - uncompressed, zlib and zstd compressed. Each variant
includes single queue and multiqueue tests for sequential
read, random read and warmed up sequential read.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add multiqueue num_queues=4 performance tests for qcow2 overlay
images with both qcow2 and raw backing files - sequential read,
random read, and warm read variants.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The backing_files option defaults to false, so qcow2 overlay
tests fail with MaxNestingDepthExceeded. Pass backing_files=on
when the test file is an overlay.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The focal image checksums have been moved in the -common
sha1sums file. Use the correct file for metrics.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Unlike most virtio feature bits, VIRTIO_BLK_F_RO is not optional.
It indicates that the host is refusing to permit write operations, and
the guest must not be allowed to override it.
However, the block device currently does not enforce this. If the guest
does not negotiate VIRTIO_BLK_F_RO, the block device will think the
device is writable and forward write requests to the backend.
This is not a security problem right now because the backing device of a
read-only device is always opened read-only. The kernel will thus
reject the write operations with EBADF. If support is added for
receiving the backing device file descriptor via SCM_RIGHTS (#7704),
it will be possible to have a read-only block device backed by a
writable file descriptor. This would make the bug a genuine security
vulnerability.
Fix the bug by explicitly checking if VIRTIO_BLK_F_RO was offered but
not negotiated. In this case, log a warning and proceed as if the guest
did acknowledge the feature. This always indicates a guest driver bug.
Fixes: #7697
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
OVMF sends FLUSH requests to read-only virtio-block devices. Refusing
these requests prevents OVMF from accessing the EFI System Partition and
therefore makes VMs unable to boot. Accept these requests instead.
them.
Ignoring these requests is possible, but inconsistent with fsync(2)
which honors them.
Fixes: #7698
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
Add handling for GHCB_INFO_SPECIAL_DBGPRINT VMG exit in the SEV-SNP
guest exit handler. This exit occurs when the guest sends debug print
requests through the GHCB interface.
Without this handler, SEV-SNP guests fail to boot when debug output
is triggered, such as when a debugger is attached to the guest image.
The handler acknowledges the exit without printing to avoid performance
degradation from frequent debug print requests.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The --net help text documented fd as fd=<fd1,fd2...>, but
comma-separated FD lists in option values must be bracketed to avoid
top-level option splitting.
Update NetConfig::SYNTAX to use fd=<[fd1,fd2,...]>, matching parser
behavior and existing net parsing tests:
`cargo test -p vmm test_net_parsing`
On-behalf-of: SAP leander.kohler@sap.com
Signed-off-by: Leander Kohler <leander.kohler@cyberus-technology.de>
Add comprehensive integration tests for DISCARD and WRITE_ZEROES:
Multiqueue stress tests verify concurrent operations across queues,
testing scattered writes with simultaneous fstrim, and write/discard
races that stress refcount table locking.
Format specific tests verify QCOW2 deallocates clusters after DISCARD,
raw files create holes using fallocate, and unsupported formats VHD
and VHDX correctly reject DISCARD requests.
Tests for sparse=off verify raw files preallocate full disk size and
QCOW2 uses zero flag instead of deallocating clusters.
Add helper functions to verify sparse files, count QCOW2 zero flagged
regions using qemu-img map, and verify guest reads zeros from
discarded regions.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add comprehensive tests for DISCARD and WRITE_ZEROES operations:
QCOW2 zero flag test validates the complete workflow: allocate
cluster, DISCARD it, verify reads return zeros, write new data,
verify cluster reallocated.
QcowSync tests verify punch_hole and write_zeroes with Arc<Mutex<>>
sharing, including tests for cache consistency with multiple async
I/O operations.
RawFileSync tests verify punch_hole and write_zeroes using
fallocate.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Implement punch_hole() and write_zeroes() for raw file backends using
io_uring and fallocate.
punch_hole() uses FALLOC_FL_PUNCH_HOLE to deallocate storage.
write_zeroes() uses FALLOC_FL_ZERO_RANGE to write zeros efficiently.
Both use FALLOC_FL_KEEP_SIZE to maintain file size.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Implement punch_hole and write_zeroes for QcowSync backend by
delegating to QcowFile::punch_hole which triggers cluster
deallocation. write_zeroes delegates to punch_hole as unallocated
clusters read as zeros in QCOW2.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add VIRTIO_BLK_T_DISCARD and VIRTIO_BLK_T_WRITE_ZEROES request types.
Parse discard/write_zeroes descriptors (sector, num_sectors, flags),
convert to byte offsets, and call punch_hole/write_zeroes on the disk
backend. Mark as unsupported in sync mode.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Implement DISCARD using QCOW2 zero flag (bit 0 of L2 entries) with
sparse aware behavior.
When sparse=true - fully deallocate clusters by decrementing
refcount, clearing L2 entry, and reclaiming storage via punch_hole
when refcount reaches zero.
When sparse=false - use zero flag to keep storage allocated while
marking as reading zeros. Only works when cluster is not shared.
Shared clusters are fully deallocated.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
When sparse=false is configured, preallocate the entire raw disk file
at startup using fallocate(). This provides space reservation and
reduces fragmentation.
Only applies to raw disks. QCOW2/VHD/VHDX formats manage their own
allocation.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add sparse parameter to QcowFile constructors and propagate it from
device_manager through QcowDiskSync. This makes the sparse configuration
available throughout the QCOW2 implementation for controlling allocation
and deallocation behavior.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add supports_zero_flag() to DiskFile trait to indicate whether a disk
format can mark clusters/blocks as reading zeros without deallocating
storage.
QCOW2 supports this via the zero flag in L2 entries. VHDX also has
PAYLOAD_BLOCK_ZERO state for this, though it's not yet implemented in
cloud-hypervisor.
This enables DISCARD to be advertised even with sparse=false for formats
with zero-flag support, since they can mark regions as zeros (keeps
storage allocated) instead of requiring full deallocation.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add sparse boolean configuration option to DiskConfig with a default
value of true to control disk space allocation behavior.
When sparse is true, the disk uses sparse allocation where deallocated
blocks are returned to the filesystem, and the DISCARD feature is
advertised to the guest.
When sparse is false, disk space is kept fully allocated and DISCARD
is not advertised.
WRITE_ZEROES is always advertised when the backend supports it,
regardless of the sparse setting.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add capability query to DiskFile trait to check backend
support for sparse operations (punch hole, write zeroes,
discard). Only advertise VIRTIO_BLK_F_DISCARD and
VIRTIO_BLK_F_WRITE_ZEROES when the backend supports these
operations.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add functions to probe whether a file or block device actually
supports PUNCH_HOLE and ZERO_RANGE operations at runtime. The
probe is performed at file open time by testing the operations
at EOF with a zero-length range, which is a safe no-op.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add punch_hole() and write_zeroes() methods to the AsyncIo trait
with stub implementations for all backends. These will be used to
support DISCARD and WRITE_ZEROES operations.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Decompose the monolithic `new_from_memory_manager` function into
smaller, focused helper methods to improve code readability,
maintainability, and testability.
Changes:
- Extract `should_force_iommu()` to determine IOMMU requirements for
confidential computing (TDX/SEV-SNP)
- Extract `should_stop_on_boot()` to check debug pause configuration
- Extract `create_cpu_manager()` to encapsulate CPU manager creation
and CPUID population
- Extract `init_tdx_if_enabled()` for TDX-specific VM initialization
- Extract `create_device_manager()` to encapsulate device manager setup
- Extract `hypervisor_specific_init()` to orchestrate initialization
sequences for different hypervisors (KVM, MSHV, SEV-SNP)
- Extract `init_sev_snp()` for SEV-SNP confidential VM setup
- Extract `init_mshv()` for MSHV hypervisor initialization
- Extract `init_kvm()` for KVM hypervisor initialization
- Extract `create_fw_cfg_if_enabled()` for fw_cfg device creation
This refactoring replaces complex nested `cfg_if!` blocks with cleaner
conditional method calls, providing clear separation between hypervisor-
specific initialization paths while preserving existing functionality.
No functional changes intended.
Issue: https://github.com/cloud-hypervisor/cloud-hypervisor/issues/7598
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Update .lychee.toml to exclude the following patterns:
- ARM domains (developer.arm.com, infocenter.arm.com) which return 403
Forbidden due to anti-bot protections in CI.
- Local TCP addresses (192.168.1.10) which are unsupported by the
link-checker tool.
- The .lychee.toml file itself, to prevent the tool from recursively
checking its own regex exclusion patterns as valid URLs.
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Document device_id parameter in NumaConfig, automatic
guest_numa_id assignment, default NUMA distances and
restrictions on Generoc Initiator NUMA nodes
Add numa configuration examples with GPU device and distance
relationships.
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Add test_guest_numa_generic_initiator to validate ACPI Generic
Initiator Affinity (SRAT Type 5) support for VFIO device.
The test verifies the following :
- Guest VM boots with a VFIO device associated with a {cpu,
memort}-less NUMA node
- Guest Kernel correctly detects Generic Initiator through
ACPI tables SRAT, SLIT
- NUMA topology in the guest includes the device-only node
with correct distances
Invoked via :
./scripts/dev_cli.sh tests --integration -- --hypervisor kvm \
--test-filter test_guest_numa_generic_initiator
The test requires a real VFIO device bound to vfio-pci driver and
skips gracefully if hardware is unavailable.
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Update FDT generation to skip NUMA properties when Generic Initiator
nodes are present, preventing conflicts between FDT and ACPI NUMA
information. FDT cannot represent Generic Initiator nodes, so ACPI
(via SRAT Type 5) becomes the authoritative source for the entire
NUMA topology when Generic Initiators exist.
Skip FDT numa-node-id properties in CPU and memory nodes
when Generic Initiator is present
Distance map bug fix : iterate over actual NUMA node IDs instead
of 0..len()
Use distance symmetry to derive distance when forward config is
missing
Default to distance cost 20 when neither direction specified
Only create memory nodes if NUMA node has memory region
Added unit tests
ARM64 boot protocol:
https://docs.kernel.org/arch/arm64/booting.html
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Support ACPI Generic Initiator Affinity to associate
PCI devices with NUMA proximity domains
Add GenericInitiatorAffinity struct
Add from_pci_bdf() to encode PCI Segment:Bus:Device.Function
Add from_acpi_device() for ACPI device handles (future use)
Generate SRAT Type 5 entries for nodes with device_id
Improve create_slit_table() to check distance symmetry when
forward distance is missing
Track device ID to BDF mappings in DeviceManager
Includes comprehensive unit tests
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Validate device_id in numa config is mutually
exclusive with cpus and memory_zones
Add NumaConfig::validate() and modify NumaConfig::parse()
Add ValidationError::InvalidNumaConfig for detailed error
messages
Include unit tests covering valid and invalid configs
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Add an optional device_id string field to NumaConfig for identifying
PCI devices associated with a NUMA node. This is used by the Generic
Initiator support to map devices to their proximity domain.
Update OpenAPI spec (cloud-hypervisor.yaml) to include the
new device_id field in the NumaConfig schema.
The device_id is optional and parsed from the --numa parameter:
--numa "device_id=<device_id>,distances=[...],..."
The optional field is accepted but not used.
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
The documentation says guest_numa_id is required to be unique and
therefore the parser() giving default value for non-existing
guest_numa_id with .unwrap_or(0) is dangerous.
Return a validation error if guest_numa_id is not provided instead
of silently defaulting to 0.
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Verify live resize of QCOW2 disks works via the API, including
resizing that requires L1 table growth.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add support for live resizing QCOW2 images. This enables growing
the virtual size of a QCOW2 disk while the VM is running.
Key features:
- Growing the image automatically expands the L1 table if needed
- Shrinking is not supported
- Resizing for images with backing files is not supported
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Provide a stub implementation for save_data_tables() to unblock pause
functionality. Without this, pausing a VM causes Cloud Hypervisor to
panic due to the unimplemented!() macro. This unblocks the
test_api_http_pause_resume testcase. We don't need to save any state
just to pause and resume.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Instead of closing a file descriptor that belongs to the vhost-user
frontend, drop the vu_common_ctrl::VhostUserHandle and the
vhost::vhost_user::Frontend it contains. This causes the destructor to
drop the file descriptor.
This breaks the last DPDK test, so disable it. See #7689.
Fixes: #7163
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
When a VFIO device with multiple MMIO regions is hot-unplugged, each
region must be individually matched and removed from the DeviceManager's
mmio_regions list. Compare per-region rather than building an aggregate
across all regions, which would never match any individual entry.
Also remove the now-unused HashSet import.
Signed-off-by: Damian Barabonkov <dbctl@pm.me>
Change has_matching_slots() to compare two MmioRegion instances
directly rather than requiring callers to construct an intermediate
HashSet of slot numbers. Remove the now-unused
user_memory_region_slots() method and HashSet import.
Signed-off-by: Damian Barabonkov <dbctl@pm.me>
Replace `clone()` with `take()` when retrieving device configurations
from `DeviceManager.config`.
This avoids unnecessarily copying the device configuration lists (e.g.,
`disks`, `net`, `fs`) when they are being processed and subsequently
moved out of the configuration. This optimization improves performance
by reducing memory allocations and cloning overhead.
Signed-off-by: Changyuan Lyu <changyuanl@google.com>
Since kernel commit 6693731487a8 ("vsock/virtio: Allocate nonlinear SKBs
for handling large transmit buffers"), a large vsock packet can be split
into multiple descriptors.
If we encounter such TX packets, pull the content into an owned buffer.
Fixes: #7672
Signed-off-by: Wei Liu <liuwe@microsoft.com>
When running manual tests locally, it is sometimes necessary to
generate a cloud-init file at a custom path instead of defaulting
to /tmp. This is useful for developers and higher-level management
layers where files in /tmp may be cleaned up automatically.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Backing files (e.g. for QCOW2) interact badly with landlock since they
are not obvious from the initial VM configuration. Only enable their use
with an explicit option.
Signed-off-by: Rob Bradford <rbradford@meta.com>
Test reading from overlay at offsets beyond backing file returns
zeros. Covers reads within backing range, beyond backing, and
boundary spanning.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
When an overlay QCOW2 image is larger than its backing file, reads
from offsets beyond the backing file virtual size would previously
fail with an I/O error.
The backing file virtual size is determined at open time and stored
for bounds checking during read operations:
- If the entire read is beyond the backing size, return all zeros
- If the read spans the boundary, read available data from backing and
fill the remainder with zeros
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Set HV_PARTITION_CREATION_FLAG_SMT_ENABLED_GUEST when the guest
topology has more than one thread per core. This allows the
hypervisor to schedule guest VPs correctly on SMT-enabled hosts.
Without this flag, the hypervisor schedules guest VPs incorrectly,
causing SMT unusable.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The kernel allows madvise on shared memory if
/sys/kernel/mm/transparent_hugepage/shmem_enabled is set.
Always try and configure THP via madvise when
the user requests THP be enabled.
If this fails, only a warning log is emitted and THP won't be enabled.
Signed-off-by: Champ-Goblem <cameron@northflank.com>
When restoring from snapshot with shared=false, write access to the
backing file is not required. Opening it read-only allows restore to
succeed on read-only media and overlay lower layers while preserving
MAP_PRIVATE semantics.
Signed-off-by: Rowen-Ye <rowenye1@gmail.com>
This patch adds the skeleton of the CVM test
support and modify existing scripts and test framework
to enable such scenario. Split the sha1sum to support both
regular and CVM guest. Add one test case for CVM. Will further
add more test cases.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
X64_64 image download steps is being used for both
regular and CVM guest. Keeping the steps withing a function
in the test-util.sh
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Now Guest struct has an option to set timeout.
No need to pass timeout while booting the guest.
If no timeout is set, default is used.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Modify Guest struct to keep some test specific
data so that test cases could be shared between
regular guest and CVM.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Testing generic vhost-user devices will require virtiofsd to support the
--tag option, which v1.8.0 does not support.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
Replace generic WritingHeader error with specific SyncingHeader
error for header fsync operations. This provides more precise
error reporting when syncing QCOW2 header changes to disk fails.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Wrap QcowFile in Arc<Mutex<>> to ensure thread safety when multiple
virtio queues access the same QCOW2 image concurrently.
Previously, each queue received its own QcowSync instance via
new_async_io() that shared the underlying QcowFile through Clone.
However, cloned QcowFile instances share internal mutable state
(L2 cache, reference counts, file seek position) without
synchronization, leading to data corruption under concurrent I/O.
This change serializes all QCOW2 operations through a mutex, which
ensures correctness at the cost of parallelism. A more performant
solution would require separating metadata locking from actual I/O
operations, tracked in #7560.
Related: #7560
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
TL;DR: Would reduce CI pressure by cancelling more "unnecessary" runs
but I can't verify without running a merge queue.
A common development pattern is to push a change and then immediately
check CI results. Follow-up fix pushes are quite common, which leads to
multiple CI runs being queued for the same pull request.
In Cloud Hypervisor, the size and cost of the CI matrix means that
several consecutive pushes (for example 3-4 in a short time) put
significant pressure on CI runners and noticeably increase feedback
latency.
In practice, concurrency handling is especially tricky for the merge
queue. From personal experience: If one does not take special care, CI
runs triggered by a `merge_group` can cancel each other, as in a merge
queue there are two runs for each job by default: one for the normal PR
and one for the merge commit. This is easy to run into, also because the
available documentation and best practices for this feature are not very
good.
At the same time, our workflows do not run on `push` events, but only
on `pull_request` and `merge_group`. Because of this, using
`${{ github.ref }}` alone as a concurrency key is not very meaningful,
and in practice only few runs are actually cancelled for successive PR
updates. Therefore, we should improve the usage of this feature.
This change tries to improve the situation by refining the concurrency
group key. The goal is to keep cancellation for multiple PR pushes,
while at the same time preventing unintended cancellations in the merge
queue by separating `merge_group` runs from regular PR runs.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
In kata-containers we use the api_client crate, but it's currently
failing our cargo deny check due to missing license, and there aren't
any license files within the crate, so I haven't found a good way to
work around this.
Alternatively I'd be happy to add the license to the workspace crate
and then reference it here, but that seems to clash with the direction
of the project in #7525.
Signed-off-by: stevenhorsman <steven@uk.ibm.com>
QCOW2 v3 autoclear_features field contains bits for features whose
metadata becomes invalid when the image is modified by software that
doesn't understand them. Defined bits:
- Bit 0: Bitmaps extension
- Bit 1: Raw external data
Cloud-hypervisor doesn't support bitmaps or external data files, so
all autoclear bits are cleared on writable open. This signals other
tools that these features' data may be stale.
Readonly opens preserve autoclear bits unchanged.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
This adds some missing features that are useful. In particular it adds
VIRTIO_F_RING_INDIRECT_DESC which gives a performance improvement.
Signed-off-by: Rob Bradford <rbradford@meta.com>
Reported-by: Daniel Farina <daniel@ubicloud.com>
TDX builds its own ACPI tables in `create_acpi_tables_tdx` so it will
return None in the standard `create_acpi_tables` function and the
assertion for `rsdp_addr` will fail.
Signed-off-by: Zhibin Li <banlu.lzb@antgroup.com>
Add tests for corrupt bit behavior during I/O operations.
- Unaligned L2 table address triggers corrupt bit on read
- Unaligned cluster address triggers corrupt bit on read and write
- Normal operations do not set the corrupt bit
- V2 images work correctly without feature bits
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
MSHV doesn't present an ITS to guests. So, /proc/interrupts would never
have "ITS-PCI-MSIX".
Instead, a Gicv2m frame is presented to guests. So expect
"GICv2m-PCI-MSIX" in testcases.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Add integration tests for QCOW2 corrupt bit handling. Verify that
images with the corrupt bit set are rejected for writable access but
allowed for read-only access with a warning.
Helper functions are added to read and modify the corrupt flag in the
QCOW2 v3 header.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Validate that L2 table offsets and refcount block offsets are cluster
aligned. Set the corrupt bit when unaligned offsets are detected, as
this indicates corrupted L1 or refcount table entries.
Validate that data cluster offsets from L2 entries are cluster aligned
during both reads and writes to existing clusters. Set the corrupt bit
when unaligned data cluster offsets are detected.
Prevent allocation of clusters at offset 0, which contains the QCOW2
header and should never be allocated. This catches corruption in the
available clusters list. Set the corrupt bit when this condition is
detected.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Set the QCOW2 corrupt bit when internal inconsistencies are detected
that indicate image metadata may be corrupted:
- Decompression decode failure, meaning compressed cluster data is
invalid
- Decompression size mismatch, where decompressed data doesn't match
expected cluster size
- Partial write after decompression, where L2 table was updated but
data cluster not fully written, leaving metadata inconsistent
- Invalid refcount index, where cluster address is outside valid
refcount table range, indicating a corrupted L2 entry
- Dirty L2 with zero L1 address, where L2 table is marked dirty but
L1 has no address for it
Note: Marking decompression failures as corrupt is more conservative
than QEMU, which returns EIO without setting the corrupt bit. This is
debatable since corrupted compressed data doesn't necessarily indicate
metadata corruption, but it provides a stronger safety guarantee by
preventing further writes to potentially damaged images.
Once set, the image can only be opened read-only until repaired with
qemu-img check -r.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add comprehensive tests for the corrupt bit handling. Cover writable
rejection, read-only access, persistence, and dirty bit
coexistence.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Implement proper handling of the QCOW2 corrupt bit (incompatible feature
bit 1) according to the specification:
- Add Error::CorruptImage for rejecting writable opens of corrupt images
- Add CORRUPT to SUPPORTED features (handled specially, not rejected)
- Add QcowHeader::set_corrupt_bit() to mark images as corrupt
- Add QcowHeader::is_corrupt() helper method
- Reject writable opens of corrupt images with Error::CorruptImage
- Allow readonly opens of corrupt images with a warning
The corrupt bit indicates that image metadata may be inconsistent. Per
spec, such images must not be written to until repaired by external
tools like qemu-img. Read-only access is permitted to allow data
recovery.
Users can open corrupt images read-only using:
--disk path=/path/to/image.qcow2,readonly=on
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
VsockPacket::hdr holds a raw pointer to the address of the VSock packet
header, which is in guest memory. It opens the door to double-fetch
(or TOCTOU) race conditions. Therefore, VSockPacket::hdr content can't
be trusted since it can be arbitrarily changed by the guest, at any
time.
To mitigate this, we can copy the header content to an array in VMM's
memory that the guest can't modify.
Signed-off-by: Thomas Leroy <thomas.leroy.mp@gmail.com>
Update QcowHeader and other related places to use BeUint methods
internally for reading/writing header fields.
This removes the byteorder dependency from mod.rs and consolidates
all big-endian file I/O through the shared BeUint trait.
Suggested-by: Rob Bradford <rbradford@rivosinc.com>
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add a read_be() method to the BeUint trait and make it pub(super)
so it can be used across the qcow module. Change BeUint::write_be()
to take Self instead of u64, providing type safety through TryFrom
conversion.
Suggested-by: Rob Bradford <rbradford@rivosinc.com>
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add tests to verify dirty bit is set while VM runs and cleared on
clean shutdown. As part of it, ensure graceful shutdown when OS
disk verification requires consistent image state.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Verify dirty bit is set on open and cleared on close for v3 images.
Ensure v2 and read-only files are not affected. Update existing
tests.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add support for the dirty bit (bit 0 of incompatible_features) which
indicates the image was not closed cleanly. This improves data
integrity by allowing detection of potentially corrupted images.
On open:
- If dirty bit is already set, log a warning and trigger
refcount rebuild
- Set the dirty bit and write it to disk immediately
- Sync to ensure persistence before any writes
- Skip dirty bit and refcount rebuild for readonly files
On clean close:
- Clear the dirty bit in the header
- Write it to disk and sync
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Test all refcount_order values (0-6):
- Basic open for each width
- Write/read roundtrip
- Overwrite and multi-cluster allocation
- L2 cache eviction under memory pressure
- Sub-byte and byte-aligned max value handling
- Overflow error detection
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Reject refcount values exceeding the maximum for the image's
refcount_order. This prevents silent truncation when storing
refcounts in narrow widths (e.g., 1-bit max is 1, 4-bit max is 15,
etc.).
Returns RefcountOverflow error with the attempted value, maximum,
and bit width. Propagates as EINVAL to the guest.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
QCOW2 v3 specifies refcount_order 0-6 with
refcount_bits = 1 << refcount_order. Previously only 16-bit (order 4)
was supported.
Changes:
- RefcountBytes trait handles byte-aligned types (8/16/32/64-bit)
- Generic pack/unpack for sub-byte widths (1/2/4-bit)
- Function pointers for read/write selected at open time
- Internal refcount type widened from u16 to u64
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Parse the feature name table header extension to provide descriptive
error messages when unsupported incompatible features are detected.
Currently only the compression bit (bit 3, zstd) is supported.
This prevents opening qcow2 images with features that would cause
incorrect behavior or data corruption (e.g., dirty bit, corrupt bit,
external data file, extended L2 entries).
Feature names are defined as follows:
1. The image's feature name table header extension (if present)
2. Hardcoded fallback names for known features
3. Generic "unknown feature bit N" for undefined features
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Co-developed-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Implement read support for bit 0 in QCOW2 L2 table entries.
When this flag is set, the cluster reads as zeros without accessing
disk. This improves compatibility with QCOW2 images that use this
optimization.
According to the QCOW2 specification, bit 0 of the standard cluster
descriptor indicates that the cluster reads as zeros. Unlike
l2_entry == 0 indicating a completely unallocated entry, bit 0 can
be set on an allocated cluster.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
build_edk2 was leaving behind .built markers
even when compilation failed.
Gate creation of.built marker to occur only on
successful build
Modify build_edk2() to exit with error code
when arm64 firmware artifact : CLOUDHV_EFI.fd
is not produced
Fixes#7608
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
build_edk2() module in scripts/common-aarch64.sh
does not produce the UEFI firmware for aarch64 as
the commits used to assemble sources for acpica,
edk2-platforms and edk2 do not compile after GCC
version upgraded from 11.4.0 to 13.3.0 in the
developer container (ubuntu 22.04 to 24.04)
Apply minimum upgrade to EDK2_REPO and ACPICA_REPO
required to compile with GCC 13.3.0
while still assuring guest VM boot for all
integration tests
BaseTools: Brotli compression submodule that was
previously failing has been fixed following commit
bump
Developers can now produce UEFI firmware for
aarch64 using the following commands
```
./scripts/dev_cli.sh shell
source scripts/test-util.sh
source scripts/common-aarch64.sh
build_edk2
```
Update docs/uefi.md
Fixes#7608
Signed-off-by: Saravanan D <saravanand@crusoe.ai>
Recent changes related to arm64 support in MSHV exposed
inconsistencies in the VM initialization and CVM boot paths.
The VM creation flow currently diverges across multiple scenarios,
including regular MSHV, CVM, and arm64, with each path performing
guest initialization steps in a different order.
Certain platform-specific requirements further constrain the ordering
of operations, such as the timing of address space creation,
IGVM loading, interrupt controller setup, and payload loading. For
CVM case address-space creation must be done after IGVM loading, and
PSP measurement. For Regular and arm64 this memory initialization
must be done early. For MSHV, vm.init() and sev_snp.init() are called in
different order which is run time and build time conditionally checked.
Additionally, while the KVM initialization path differs slightly
from MSHV, it shares common logic that is currently split across
separate conditional and build-time code paths, contributing to
fragmentation of the overall flow.
This change restructures the VM creation and initialization sequence
to better align shared logic, enforce scenario-specific ordering
constraints, and ensure consistent and correct behavior across all
supported configurations. In doing so, it restores proper CVM boot
behavior and improves the maintainability of the initialization code.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The spec says simply that that an empty payload should be returned on
error. Be slightly more helpful by adding a warning.
Signed-off-by: Rob Bradford <rbradford@meta.com>
Based upon the discussion and in
https://github.com/rust-vmm/vhost/issues/29#issue-830820820 and the QEMU
behaviour the get_config offset should be zero. This was not caught by
our integration tests as the vhost-user-blk backend as implemented in
this repository does not use the offset.
Fixes: #7615
Signed-off-by: Rob Bradford <rbradford@meta.com>
Since the mshv integration workflow has been stable for a long time,
make the workflows no longer optional.
Signed-off-by: Aastha Rawat <aastharawat@microsoft.com>
Add warmup_iterations field to run iterations before measuring
performance. This complements existing cold start tests
by separating cache effects from steady state throughput.
New tests with 2 warmup iterations:
- block_qcow2_backing_qcow2_read_warm_MiBps
- block_qcow2_backing_raw_read_warm_MiBps
Results show warm cache is much faster and more consistent:
- QCOW2: 1766 MiB/s (4% variance) vs cold 960 MiB/s (73% variance)
- RAW: 1822 MiB/s (6% variance) vs cold 1300 MiB/s (55% variance)
RAW backing is 3% faster than QCOW2 in steady state.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The write_to() function is used by test code to create qcow2 files for
testing. For v3 headers with extended header_size (>104), it needs to:
1. Write the mandatory compression_type field at bytes 104-111
2. Write the header extension end marker at the header_size offset
3. Seek to backing_file_offset before writing the backing file path
Additionally, create_for_size_and_path() must set backing_file_offset
to account for the 8 byte extension end marker in v3 files, so the
backing file path doesn't overwrite the extension area.
Add unit tests for read_header_extensions() covering backing format
parsing (raw/qcow2), unknown extensions, and error cases (invalid
formats, invalid UTF-8). These tests depend on the header writing fixes
to create properly formatted v3 test files.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add support for parsing QCOW v3 header extensions to read the
backing file format. The QCOW v3 spec allows optional header
extensions between the fixed header and the backing file name.
Implement read_header_extensions() to parse the extension area,
which starts at the header_size offset. At the moment it is
used to read the backing file format. Further extension
processing is open in folow up implementations.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add support for raw backing files in addition to qcow2 backing
files. This enables QCOW2 overlays to use raw images as their
backing store.
The backing file format is auto-detected when not specified,
using the existing detect_image_type() function.
Add backing_file_format field to QcowHeader to store the format
type, which will be populated from header extensions by a
subsequent patch.
Modify new_from_backing() to accept a backing_format parameter,
consolidating support for both raw and qcow2 backing files in a
single function. The backing_file_size parameter allows overlay
creation without opening the backing file multiple times.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
In most of the cases, special registers don't change after emulations,
but current code sets them back unconditionally, and although some of
them are set over the register page, others require a system call and a
hypervisor to be updated, which is a waste it there were not changes.
Introduce and CPU update method for Microsoft Hypervisor emulator and
set special registers only when they were changed. This change reduces
guest boot time by 4% for a single VP guest boot (in L1VH partition) in
my experiments.
Signed-off-by: Stanislav Kinsburskii <skinsburskii@linux.microsoft.com>
This is a precursor change to overall ioctl and hypercall reduction
effort. The old (current) CPU state can be compared to the new to
determine what has changed and avoid unnecessary register updates.
Signed-off-by: Stanislav Kinsburskii <skinsburskii@linux.microsoft.com>
Minor modifications were made to make the sentences sound more natural.
Also fixed some parameter usage issues in bash code block.
Signed-off-by: Yi Wang <foxywang@tencent.com>
There are some syntax and format issues in tdx/sev documents.
Make some modification to make the description more natural.
And the link of SEV-SNP is invalid, fix it.
Signed-off-by: Yi Wang <foxywang@tencent.com>
There are some minor syntax and command issues in debug-port document.
As commit 5febdec81a (vmm: Enable `gdbstub` on AArch64) supported
aarch64, the docs should keep consistent.
Signed-off-by: Yi Wang <foxywang@tencent.com>
Fix some minor syntax issues in api/building document to make
the sentences more fluent and easy to read.
Signed-off-by: Yi Wang <foxywang@tencent.com>
Some description in the device document were inconsistent with the
source code. Also fix some syntax issues to make the sentences more
fluent.
Signed-off-by: Yi Wang <foxywang@tencent.com>
CI reports:
In scripts/test-util.sh line 216:
cleanup() {
^-- SC2329 (info): This function is never invoked. Check usage (or ignored if invoked indirectly).
The shellcheck can't trace calling in trap, so we need add hint
to make it happy.
Signed-off-by: Yi Wang <foxywang@tencent.com>
The IORT table's ID mapping uses a 256-ID offset per PCI segment to
ensure unique device IDs across all segments. This partitioning scheme
(output_base = 256 * segment_id) must match the device ID encoding used
in KVM MSI routing configuration [1].
This mapping assumes one bus per PCI segment, and supports up to 256 PCI
segments in the system.
[1] c9374d87ac
Signed-off-by: Bo Chen <bchen@crusoe.ai>
The IORT specification (Revision E.b, Table 12) defines the ITS Group
Node structure with an ITS Identifiers array following the node header.
Although the IORT table is zero-initialized, this commit adds an
explicit write of the ITS identifier value (0) for clarity and spec
compliance.
This ITS identifier must match the `translation_id` field in the MADT
GIC ITS structure to ensure proper interrupt routing on ARM platforms.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
The current IORT table implementation is based on IORT Spec revision E.b
[1], as evidenced by:
* The PCI root complex node revision being set to `3`
* The code being updated in late 2021 [2] when revision E.b was the
latest version
This patch ensures the IORT table is properly generated according to
this specification revision, fixing three issues:
1. The IORT table revision should be `3` rather than `2` (see Table 2 in
the spec [1])
2. The GIC ITS group node revision should be `1` rather than `0`
(see Table 12 in the spec [1])
3. The "Memory access properties" and "ATS Attribute" fields of the PCI
root complex node was set incorrectly - specifically the MAF (Memory
Access Flags) including CPM and DACS bits (see Tables 14, 15, and 17
in the spec [1])
[1] https://developer.arm.com/documentation/den0049/eb/?lang=en
[2] https://github.com/cloud-hypervisor/cloud-hypervisor/pull/3356
Signed-off-by: Bo Chen <bchen@crusoe.ai>
It should always succeed and is apparently implicitly called by libc or
some dependency somewhere.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
x86::__cpuid is safe on Rust ≥1.94 but unsafe on older versions. This
causes unused_unsafe warnings when compiling with Rust ≥1.94. However,
on earlier Rust versions, the code won’t compile if the unsafe blocks
are absent.
Work around this by adding #[allow(unused_unsafe)] where needed to
suppress the warnings.
See #7588 for more discussion.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
On MSHV, exposing multithreaded CPU topologies requires setting the
PROCESSORS_PER_SOCKET partition property so that CPUID.0xB reports
correct logical processor counts and topology levels to the guest.
This property must be set after all vCPUs are configured, as the
hypervisor uses the complete vCPU layout to derive and report CPU
topology information.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
This change is a prerequisite for live disk resizing. Before this
commit, the epoll-handler threads just got a copy of the sector
size which we cannot update during runtime.
On-behalf-of: SAP thomas.prescher@sap.com
Signed-off-by: Thomas Prescher <thomas.prescher@cyberus-technology.de>
Add sequential and random read performance tests for QCOW2 overlays
with QCOW2 backing files.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Introduce a new BlockControl struct to encapsulate fio operation
parameters. This replaces the tuple-based fio_control with a more
extensible structure that includes:
- fio_ops: The FIO operation type
- bandwidth: Whether to measure bandwidth or IOPS
- test_file: The file path to test against
This refactoring enables reusing performance_block_io with different
test files.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Virtio PCI devices are created in a set of nested functions. In each
of this functions a vector is created to add created devices to, only
to be appended to the vector of the higher nesting level. Those nested
vectors are unnecessary as we can directly write to the member of
`DeviceManager`.
Signed-off-by: Pascal Scholz <pascal.scholz@cyberus-technology.de>
On-behalf-of: SAP pascal.scholz@sap.com
The TILE data state of AMX may require 8KB+ space, calling the legacy
KVM_GET_XSAVE will encounter an error since KVM_GET_XSAVE only can get
4KB space. This patch adds KVM_GET_XSAVE2 support to allow snapping more
data.
Fixes: #7533
Signed-off-by: Songqian Li <sionli@tencent.com>
Refactor write_pointer_table to accept iterators instead of requiring
materialized vectors, eliminating temporary allocations in L1 table
sync operations.
Changes:
- Modified write_pointer_table() to take Iterator<Item = &T> and
dereference internally before passing owned values to the callback
- Added write_pointer_table_direct() convenience wrapper for cases
without value transformation
- Updated sync_caches() to use l1_table.iter() directly instead of
.get_values().iter().copied()
- Implemented Deref<Target = [T]> for VecCache to enable direct .iter()
Performance impact:
- Eliminates L1 table allocation during sync (~2KB per 100GB disk)
- L2 and refcount table writes already used slices, no change there
- Zero performance overhead: iterator dereferencing is equivalent to
.copied() and optimizes identically
The L1 sync previously collected entries into a Vec to apply the
OFLAG_COPIED flag. The new iterator+callback pattern computes this
on-the-fly, avoiding the allocation entirely.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add Deref<Target = [T]> implementation for VecCache<T> to allow direct
slice operations without explicitly calling get_values(). This enables
cleaner code patterns like cache.iter() instead
of cache.get_values().iter().
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add fcntl to virtio_block_thread_rules to allow try_clone() on file
descriptors. The try_clone() method uses fcntl(fd, F_DUPFD_CLOEXEC)
to duplicate file descriptors, which is needed for efficient QCOW2
L1 table sync that avoids temporary allocations.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Creating a vdpa_net with an associated MAC address and setting the right
amount of queues in order to fix the integration test related to
vdpa_net.
Fixes: #5756
Signed-off-by: Sebastien Boeuf <seb@rivosinc.com>
In [0] we refactored some Arc<Mutex<T>> parameters to &Mutex<T>> to
satisfy clippy's needless_pass_by_value lint. Nevertheless, this is also
not so idiomatic, so as a follow-up, we put the responsibility to lock
objects to the caller side (only where this is not strictly needed by
the callee).
While on it, I also tried to pass vm_config directly into
pre_create_console_devices() which would clean up some code, but then
we have interleaving mutable and immutable borrows of the Vmm, which
are denied by the borrow checker.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Our CI test execution system finds some mergeable pages in the system
directly on boot since a recent update. Remove the broken assumption in
the test_memory_mergeable integration test and only check if the guest
VMs influence the shared pages count.
Signed-off-by: Stefan Nürnberger <stefan.nuernberger@cyberus-technology.de>
https://github.com/cloud-hypervisor/cloud-hypervisor/pull/7294 adjusted
the checks for read-only requests made to virtio-blk devices and started
rejecting VIRTIO_BLK_T_GET_ID requests. These requests do not perform
any writes and are needed in order to access device serials from within
the guest.
Signed-off-by: Connor Brewster <cbrewster@hey.com>
`cargo rustc` is incompatible with virtual manifests, so the CI needs to
use cargo build instead. However, passing `RUSTFLAGS="-D warnings"` via
the environment would propagate to all dependencies, and some of them
currently fail to build under ``-D warnings` due to issues like [0]:
```
error: creating a mutable reference to mutable static
--> src/temp.rs:97:5
|
97 | DIRS.pop()
| ^^^^^^^^^^ mutable reference to mutable static
```
To resolve this, apply ``-D warnings` only to the `cargo clippy`
commands (which apply to our workspace only) and avoid enforcing it for
the entire cargo build.
[0]: https://github.com/cloud-hypervisor/cloud-hypervisor/actions/runs/19962283528/job/57245376263?pr=7525
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
TL;DR: cargo clippy|check|... now runs on whole workspace by default.
## Steps
- add new workspace member `cloud-hypervisor`
- move `./src` to new workspace member
- move `./tests` to new workspace member
- move relevant parts from Cargo.toml to new workspace member
- kept necessary parts in main Cargo.toml, such as profile
configurations
## About
The main Cargo.toml historically mixes workspace and crate definitions
for cloud-hypervisor and ch-remote. This makes it hard to read and
requires `--workspace` to run cargo clippy or cargo test on all
workspace members, which is counter-intuitive.
This patch separates the workspace from the crate definition in the main
Cargo.toml file. After this, cargo clippy, cargo test, etc., work on the
whole workspace naturally, giving a smoother developer experience. The
Cargo.toml without a package definition is also called a virtual
workspace or virtual manifest by Cargo [0].
Backporting is not a concern: CHV no longer backports, but the affected
files are rarely modified anyway.
[0] https://doc.rust-lang.org/cargo/reference/workspaces.html#virtual-workspace
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This will also prevent some useless rebuilds. Using `--verbose` we can
observe that the build.rs causes frequent useless rebuilds - having
less is a good thing. They come from the dependency of `build.rs` to
the local git repository.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
The BufWriter must be flushed explicitly to handle errors
properly. Without explicit flush, errors during the implicit
drop flush are ignored.
This is the same issue fixed for write_pointer_table
in commit 85556951a.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The MSR enumeration logic in MSHV does not accurately
account for the capabilities of the underlying
hardware and hypervisor. Previously, integration tests
passed because only the default MSRs—universally supported
across platforms—were involved. However, recent feature
additions in MSHV have introduced MSRs that are not
supported on Intel hardware, causing failures.
This patch retrieves the MSRs from the VM specific
API call that filters out the unsupported MSRs.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add set_cluster_refcount_track_freed() helper to consolidate the
common pattern of setting a cluster refcount and tracking freed
refblocks. This reduces code duplication and improves readability.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The image passed for the guest construction is copied. Previously,
check-img has been checking the unchanged image from the workspace dir,
which is supposed to be error free.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Freed clusters correctly have refcount=0. Remove the assertion that
expected no clusters with zero refcount, as it was validating the
buggy behavior.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
When converting a compressed cluster to standard during write
operations, the old compressed cluster's refcount was never
decremented, causing leak warnings by `qemu-img check ..`
`Leaked cluster X refcount=N reference=M`
Additionally, compressed data can span multiple physical clusters,
not just one. The compressed cluster address and size are encoded
in the L2 entry, and the data may cross cluster boundaries.
The proper handling is implemented as follows:
- Extract compressed cluster address and size before overwriting
L2 entry
- Identify all clusters occupied by the compressed data
- Decrement refcount for each cluster in the range
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
When a refcount block is evicted from cache and replaced with a new
one, the old refcount block cluster was added to unref_clusters but
its refcount was never decremented to 0 on disk. This left the cluster
with refcount=1 while no metadata referenced it, causing errors in
qemu-img check
`Leaked cluster X refcount=1 reference=0`
This fix recursively calls set_cluster_refcount(freed_cluster, 0) to
properly decrement the freed refcount block's refcount on disk. The
recursion handles cascading replacements where freeing one refcount
block may trigger the replacement of another.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
"Signed-off-by" is the only variant that is accepted. So we
should remove the inconsistency to prevent:
- user forgets this at all
- CI complains
- user adds "Signed-Off-By"
- CI still complains because of the wrong format
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
We enable AMX tile state components via the hypervisor (as introduced
in the previous commit) instead of doing this inline in the body of
`CpuManager::new`.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
Currently when the user configures AMX the corresponding state
components get dynamically enabled directly inside the function body of
vmm::cpu::CpuManager::new.
With our ongoing work on CPU templates/profiles, there will (likely) be
one more binary crate for producing CPU profiles that also needs to do
this (without creating a CpuManager) and it may also be the case that
we will need to call this function prior to `CpuManager::new` during
live migrations.
We thus add a method for enabling the AMX tile state components on the
hypervisor trait that may be called wherever necessary. We argue that
this is beneficial for code clarity independently of the upcoming CPU
templates/profiles PR that we are working on.
The astute reader will notice that the logic introduced here is not 100%
the same as what is done inside the vmm::cpu::Cpumanager::new method. We
claim that our approach is more in-line with the official documentation.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
This patch updates the documentation to reflect the newly added
nested CPU feature option in the CLI.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add an option in the CLI to enable nested support.
Adding an option --cpu nested=on|off to the CPU
argument to enable nested support for Microsoft
Hypervisor. By default it is enabled for both KVM
and MSHV.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Create HypervisorVmConfig early and pass the
struct to VM creation API in the vmm crate. Getting
rid of multiple conditional parameter.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The OFLAG_COPIED bit (bit 63) indicates a cluster's refcount is exactly
1 and doesn't need copy-on-write. This bit must be set in L1 entries
when their referenced L2 clusters have refcount=1.
Previously, L1 entries were always written as raw addresses without the
OFLAG_COPIED bit, violating the QCOW2 specification and causing qemu-img
check to report errors like
`ERROR OFLAG_COPIED L2 cluster: l1_index=X .... refcount=1`
The implementation queries each L2 cluster's refcount in sync_caches()
and sets OFLAG_COPIED appropriately when writing the L1 table. This
ensures QCOW2 images are specification compliant and maintain correct
COW semantics to avoid data corruption.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Previously the code relies on the implicit flush when BufWriter is
dropped. That's not safe.
Per BufWriter's document:
```
It is critical to call flush before BufWriter<W> is dropped. Though
dropping will attempt to flush the contents of the buffer, any errors
that happen in the process of dropping will be ignored. Calling flush
ensures that the buffer is empty and thus dropping will not even attempt
file operations.
```
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The MSRs are constant at boot time so rather than creating a vector in
the boot_msr_entries() method instead reaturn a reference to static MSR
array data.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Previously, state that we accumulated during the migration process in
the receiver was kept in `mut Option` variables or HashMaps. The
problem is that it is unclear in the code when these variables can be
safely used. It's also difficult to add new state, such as the state
we need to handle the upcoming feature for performing the migration
using multiple connections.
To solve this, I've refactored the code to use the state machine
pattern. Each state carries the data it needs to. Any state that
didn't arrive yet (memory_files, memory_manager) cannot be accessed
until we are in the proper state.
Some benefits that fall out of this:
- We now respond to all requests, even invalid ones, at least with an
error message.
- Any error handling a request will result in an error message being
sent.
- There is only a single place where responses are sent and thus it's
very hard to mess this up in the code.
- The main protocol state machine fits on a screen.
I would argue that especially the error cases are now much more
consistent. There is still a lot to be done. There is still state
transfer via self.vm and similar. In an ideal world, this would also
be carried by the state machine. I also want to see better handling of
payloads, which are still handled all over the place, but this change
is already too big. :)
Co-authored-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP julian.stecklina@sap.com
On-behalf-of: SAP philipp.schuster@sap.com
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
... and nuke some Option<> while I was there. Given that HashMap has a
usable default and we end up passing an empty HashMap anyway, just get
rid of the Option.
On-behalf-of: SAP julian.stecklina@sap.com
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
During the lifecycle of a confidential VM, the guest may revoke
previously shared pages via an attribute-intercept VM exit.
When this happens, the host-side cache must be updated so that any
subsequent access by the VMM triggers a fresh request for the guest
to re-share the page.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This is a follow-up of [0].
# Advantages
- This saves dozens of unneeded clone()s across the whole code base
- Makes it much easier to reason about how parameters are used
(often we passed owned Arc/Rc versions without actually needing
ownership)
# Exceptions
For certain code paths, the alternatives would require awkward or overly
complex code, and in some cases the functions are the logical owners of
the values they take. In those cases, I've added
#[allow(clippy::needless_pass_by_value)].
This does not mean that one should not improve this in the future.
[0] 6a86c157af
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Fix the condition to correctly identify a new PTY connection.
Fixes: 287887c99 (vmm: fix console IO safety)
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Previously, `into_raw_fd()` transferred fd ownership to epoll while
keeping separate clones in `reader` and `writer`, causing leaks when
the stream was closed. Now, `reader` owns the fd and epoll borrows it.
When the FD is closed in the kernel, reader will be reset to None.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
The Intel Granite Rapids processors include more AMX related features
that are advertised in leaf 0x7 subleaf 0x1. If the VM is not
configured to support AMX (the default) then these feature bits need
to be masked out.
Furthermore Tile information and TMUL information
in leaves 0x1d and 0x1e respectively are also purely related to AMX
and should also be zeroed whenever AMX support is disabled.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP <oliver.anderson@sap.com>
When enabling amx feature, we should call arch_prctl to request
permission to use tile data for guest. The permission should be
requested before the first vcpu is created, so we need to call
arch_prctl in vmm thread. This patch adds the arch_prctl syscall for
vmm_thread_rules.
Fixes: #7516
Signed-off-by: Songqian Li <sionli@tencent.com>
This helps to uncover expensive and needless clones in the code base.
For example, I prevented extensive clones in the snapshot path where
(nested) BTreeMap's have been cloned over and over again. Further,
the lint helps devs to much better reason about the ownership of
parameters.
All of these changes have been done manually with the necessary
caution. A few structs that are cheap to clone are now `copy` so that
this lint won't trigger for them.
I didn't enable the lint so far as it is a massive rabbit hole and
needs much more fixes. Nevertheless, it is very useful.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
We've encountered issues that Cloud Hypervisor corrupts disk images
after use. Those issues may not be immediately obvious until the
corrupted images are used again.
Run consistency checks over the disk images in the test cases to catch
issues as early as possible.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
It is a common use case to run a subset of tests locally to verify
certain functionalities.
The default behaviour for nextest is to error out if no tests are run.
That causes the test scripts to return a non-zero value (failure). Pass
`--no-tests=pass` to nextest to match what `cargo test` does if no tests
are run.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This is continuing the work from [0]. Some places in CHV
"misuse" Result::map_err for side-effects. inspect_err is
a better suited alternative for exactly that use-case.
Unfortunately, there is no clippy lint for this yet.
[0] f02745a7ed
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This commit is the first in a series of similar commits to clean up
obsolete `extern crate` statements
Since Rust 1.30, normal macros can be imported via `use`, and with Rust
1.31 and edition 2018 this has become the preferred approach.
`extern crate` is only needed for `alloc` in `no_std` crates, which does
not apply here.
By dropping these (often redundant or odd) `extern crate` lines, we
expose the actual dependencies more clearly and reduce technical debt.
## Auto-generation of the series
Most of this series was produced automatically:
1. Removed all "extern crate" references
2. Run the script [0] to add missing `use` statements
3. Run `cargo +nightly fmt --all`
4. Fix the remaining problems manually
The treewide changes were then split into per-folder commits.
[0]
```python
import os
import re
# Mapping of macro/function usage to imports
MACRO_IMPORTS = {
"info!": "use log::info;\n",
"debug!": "use log::debug;\n",
"error!": "use log::error;\n",
"trace!": "use log::trace;\n",
"warn!": "use log::warn;\n",
"event!": "use event_monitor::event;\n",
"anyhow!(": "use anyhow::anyhow;\n",
"bitflags!(": "use bitflags::bitflags;\n",
"ioctl_ior_nr!": "use vmm_sys_util::{ioctl_ior_nr};\n",
"ioctl_iow_nr!": "use vmm_sys_util::{ioctl_iow_nr};\n",
}
# Regex for finding the first use statement
USE_REGEX = re.compile(r"^\s*(use|pub use) .+?;")
def process_file(path):
with open(path, "r", encoding="utf-8") as f:
lines = f.readlines()
content = "".join(lines)
existing_imports = set(lines)
needed_imports = set()
# Check macros/functions against mapping, only add if not already present
for key, import_stmt in MACRO_IMPORTS.items():
if key in content and import_stmt not in existing_imports:
needed_imports.add(import_stmt)
if not needed_imports:
print(f"Unmodified {path} (no new imports needed)")
return # Nothing to do
# Find first use or pub use statement
for i, line in enumerate(lines):
if USE_REGEX.match(line):
insertion_index = i + 1
break
else:
print(f"Unmodified {path} (no use or pub use statement found)")
return # No use statement found, skip file
# Insert imports
lines[insertion_index:insertion_index] = list(needed_imports)
# Write back file
with open(path, "w", encoding="utf-8") as f:
f.writelines(lines)
print(f"Modified {path}, added imports: {''.join(needed_imports).strip()}")
for root, _, files in os.walk("."):
for file in files:
if file.endswith(".rs"):
process_file(os.path.join(root, file))
```
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Add tests:
- zlib: test_virtio_block_qcow2_zlib()
- zstd: test_virtio_block_qcow2_zstd()
Both these tests use zlib- and zstd-compressed images as OS image.
Modify test_virtio_block_qcow2_backing_file() test: it is practical
to test qcow2 file-backing with compression, so use zlib-compressed
image as a backing file.
Signed-off-by: Eugene Korenevsky <ekorenevsky@aliyun.com>
Add support of reading and writing compressed clusters.
Support zlib and zstd compressions.
L2 cache: store entire L2 entries, not only standard cluster addresses.
Read path. Offsets of compressed clusters cannot be determined,
therefore replace QcowFile.file_offset_read() with QcowFile.file_read().
This method reads the cluster, decompresses it if necessary and returns
the data to the caller.
Write path. QcowFile.file_offset_write(): since writing to compressed
clusters is not generally possible, allocate a new standard
(non-compressed) cluster if compressed L2 entry is encountered; then
decompress compressed cluster into new cluster; then return offset
inside new cluster to the caller. Processing of standard clusters is
not changed.
Signed-off-by: Eugene Korenevsky <ekorenevsky@aliyun.com>
As well as saving the MSRs as it is currently does ensure that the KVM
capability is enabled along with keeping the internal state updated.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Co-authored-by: Chengyu Fu <chengyu.fu@linux.alibaba.com>
The granularity has significant implications in typical cloud
deployments with network storage. The Linux kernel will sync advisory
locks to network file systems, but these backends may have different
policies and handle locks differently. For example, Netapp speaks a NFS
API but will treat advisory OFD locks for the whole file as mandatory
locks, whereas byte-range locks for the whole file will remain
advisory [0].
As it is a valid use case to prevent multiple CHV instances from
accessing the same disk but disk management software (e.g., Cinder in
OpenStack) should be able to snapshot disks while VMs are running, we
need special control over the lock granularity. Therefore, it is a valid
use case to lock the whole byte range of a disk image without
technically locking the whole file - to get the best of both worlds.
This also brings CHVs behavior in line with QEMU [1].
Whole-file locks remain a valid use case and could be supported later.
This patch only provides the necessary groundwork; making it
configurable is out of scope for now.
[0] https://kb.netapp.com/on-prem/ontap/da/NAS/NAS-KBs/How_is_Mandatory_Locking_supported_for_NFSv4_on_ONTAP_9
[1] <qemu>/util/osdep.c::qemu_lock_fcntl()
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This should be guaranteed by GuestMemory and GuestMemoryRegion, but
those traits are currently safe, so add checks to guard against
incorrect implementations of them.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
It takes a pointer to a userspace address that it accesses, so it should
be marked unsafe. This was missed earlier.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
Callers of get_host_address_range() rely on it returning a pointer to at
least size bytes of memory. mem.get_host_address() is an overrideable
method of a safe trait, so it is better for safe code to not rely on its
correctness for safety. Instead, use mem.get_slice(), which returns a
VolatileSlice whose invariants guarantee that it points to a sufficient
amount of memory. If mem.check_range() succeeds but mem.get_slice()
returns a slice that is too small, this means that there is either a
logic error or a situation the code cannot support yet, so panic.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
The kernel will validate that the size is page-aligned. The file offset
is always zero, so the kernel will also validate that the offset is
page-aligned.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
This API passes a u64 to a kernel API that treats the u64 as a userspace
address. Therefore, it should be marked unsafe, but it currently is not
[1]. Wrap the call in an unsafe block to document that invariants must
be upheld to avoid undefined behavior. This causes a compiler warning,
so suppress the warning with #[allow(unused_unsafe)].
[1]: https://github.com/rust-vmm/vfio/issues/100
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
I believe that its only caller used it safely, but it is still better to
mark the code as unsafe. Also add additional validity checks.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
Also drop support for building the TDX code for 32-bit targets. All
CPUs with TDX support are 64-bit so supporting 32-bit targets is not
needed.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
To ensure that struct sizes are the same on 32-bit and 64-bit, various
kernel APIs use __u64 (Rust u64) to represent userspace pointers.
Userspace is expected to cast pointers to __u64 before passing them to
the kernel, and cast kernel-provided __u64 to a pointer before using
them. However, various safe APIs in Cloud Hypervisor took
caller-provided u64 values and passed them to syscalls that interpret
them as userspace addresses. Therefore, passing bad u64 values would
cause memory disclosure or corruption.
Fix the bug by using usize and pointer types as appropriate. To make
soundness of the code easier to reason about, the PCI code gains a new
MmapRegion abstraction that ensures the validity of pointers. The rest
of the code already has an MmapRegion abstraction it can use. To avoid
having to reason about whether something is keeping the MmapRegion
alive, reference counting is added. MmapRegion cannot hold references
to other objects, so the reference counting cannot introduce cycles.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
To ensure that struct sizes are the same on 32-bit and 64-bit, various
kernel APIs use __u64 (Rust u64) to represent userspace pointers.
Userspace is expected to cast pointers to __u64 before passing them to
the kernel, and cast kernel-provided __u64 to a pointer before using
them. However, various safe APIs in Cloud Hypervisor took
caller-provided u64 values and passed them to syscalls that treat them
as userspace addresses. Therefore, passing bad u64 values would cause
memory disclosure or corruption. The memory region APIs are one example
of this, so mark them as unsafe.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
When booting a Linux guest in SMP configuration, on sapphire rapids
and granite rapids the following kernel warning can be observed:
[Firmware Bug]: CPUID leaf 0x1f subleaf 1 APIC ID mismatch 1 != 0
[Firmware Bug]: CPUID leaf 0x1f subleaf 2 APIC ID mismatch 1 != 0
The reason is that we announce the presence of the extended topology
leaf, but fail to announce the x2apic ID in EDX for each subleaf.
Signed-off-by: Thomas Prescher <thomas.prescher@cyberus-technology.de>
On-behalf-of: SAP thomas.prescher@sap.com
as of rust 1.90, writes to unix sockets use the sendto syscall. This
affects the vcpu threads when vfio_user devices are accessed.
Signed-off-by: Stefan Nürnberger <stefan.nuernberger@cyberus-technology.de>
This is the first commit in a series of commits to improve the Code
Quality in Cloud Hypervisor in a sustainable way. These are the
default rules from `clippy::all` but written here to be more explicit.
`clippy::all` refers to all "default sensible" lints, not all
existing lints.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This better aligns with the rest of the code and makes it clearer
that these tests can run "as is" in a normal hosted environments
without the special test environment.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
TL;DR: Massive quality of life improvement for devs
Cloud Hypervisor uses the Cargo test framework for multiple tests:
- normal unit tests
- unit tests requiring special environment (the Tap device tests)
- integration tests requiring a special environment
This prevented the execution of `cargo test --workspace`, which results
in a very poor developer experience. Although
`./scripts/run_unit_tests.sh` exists, there are valid reasons why devs
cannot or even don't want to use it.
By adding a new `chv_testenv` rustc config, we can conditionally only
activate tests when the `./scripts/` magic runs them. This improves
the general developer experience by a lot.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Commit 5ec47d4883 was intended to patch ebx bits 23-16 in cpuid leaf
0x1, but it was not working as expected, as in rust, operator << has a
stronger precedence than & [1]. Later commit b6667f948e fixed the
operator precedence clippy warning, but did not fix the actual issue. As
a result, the current code is not changing ebx,
```
cpu_ebx |= ((dies_per_package as u32) * (cores_per_die as u32) * (threads_per_core as u32))
& (0xff << 16);
```
Since the total number of logical processors is generally less than
65536, the right hand side of the expression is 0 in most cases.
[1] https://doc.rust-lang.org/reference/expressions.html#expression-precedence
Fixes: 5ec47d4883 ("arch: x86_64: enable HTT flag")
Signed-off-by: Changyuan Lyu <changyuanl@google.com>
A major improvement to the developer experience of clippy in
Cloud Hypervisor.
1. Make `cargo clippy` just work with the same lints we use in CI
2. Simplify adding new lints
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
... by just passing the iterator along. For large VMs this bitmap is
gigantic. A 12TB VM has 384MB of dirty bitmap.
With all these optimizations from the previous commits in place, we
see quite the improvement when it comes to scanning the dirty bitmap.
For a bitmap with 1% bits (randomly) set, dirty_log() takes:
Original code: 2166ms (100.0%)
New code: 382ms ( 17.6%)
on my system. The sparser the dirty bitmap the faster. Scanning an
empty bitmap is 100x faster. For a 5% populated bitmap we are still 3x
faster.
If someone wants to play with this, there is a benchmark harness here:
https://github.com/blitz/chv-bitmap-bench
On-behalf-of: SAP julian.stecklina@sap.com
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
This would be a good opportunity to optimize another pointless vector
away, but I don't have a good way to test this at the moment. But
maybe someone else gives it a shot.
On-behalf-of: SAP julian.stecklina@sap.com
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Adding itertools as dependency improves the iteration code in the
following significantly.
With this change, we don't need a copy of the vector. Just something
that can be coerced into an iterator. We also use the bit position
iterator to make the code somewhat clearer. The new code is much
faster, because it will not iterate over every bit, just each 1 bit in
the input.
The next commit will complete this optimization and have some concrete
numbers.
On-behalf-of: SAP julian.stecklina@sap.com
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Due to the fw_cfg test being disabled on MSHV this results in no tests
being runnable which results in an error in nextest. Reduce that error
to a warning use --no-tests=warn
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
cargo nextest is an improved test runner that allows retries as well a
reporting the times for the test runs.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When using nextest for running tests each test is run in its own process
so the old solution of using a static variable for the guest ID (used to
determine the network segment) no longer works.
Instead use a text file on the filesystem protected with an exclusive
lock. The test process will read from it and then write back the next ID
that can be used. It wraps around at the limit of u8 and skips ID 0.
This function intentionally panics rather than propagate errors as it
should only be called for testing purposes and there the panic handler
will give a useful backtrace and cleanup.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This is required to support exclusive locking on files which is needed
for safe test ID generation when using nextest (since it runs each test
as a separate process.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
as of rust 1.90, writes to unix socket streams use send_with_flags
instead of write, so it uses a sendto syscall instead of write.
Signed-off-by: Matt Moriarity <matt@mattmoriarity.com>
To get that list, I've used
```
git log | grep --fixed-strings -- "-by:" | head -n 100000 | sort | less
```
on the Linux kernel's git repository.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
The set of extensions supported by a RISC-V system needs to be exposed
to the guest - currently that is a fixed, minimal set of extensions.
These extensions are not sufficient to boot Ubuntu 25.10 which now has a
mininimum requirement of RVA23S64 (which is a minimum set of extensions
that make sense for server use cases.)
The easiest way to convey the extensions that the guest should use is to
copy those that the host kernel understands (and thus includes in the
/proc/cpuinfo) data.
However since nested virtualisation is not currently possible - exclude
the "H" (Hypervisor) extension from the list of short (single letter)
extensions.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
After updating the Linux kernel to 6.19.6 the second segment (segment=1)
is now under the 2nd IOMMU group (which it a more logical setup) and as
such the added device which is on that segment is in that second IOMMU
group.
The same check is made in test_vdpa_block so also test there.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Resignal every 10ms the thread if it has not acknowledged the signal via
setting the atomic when the vCPU thread was acknowledged. Further, avoid
an infinite loop by generating an error if it takes more than 1000ms to
interrupt the thread.
The retry helps mitigate a race condition where the signal is received
between checking the pause atomic and entering KVM_RUN ioctl when
pausing. Hitting this race condition would leave the
wait_untial_signal_acknowledged() method spinning indefinitely.
The timeout error avoids the VMM process being blocked indefinitely.
See: #7427
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Dump kernel logs after running the tests in the MSHV workflow to help
debug failures.
In addition to getting the kernel logs using `dmesg` also use AzCli to
retrieve the serial console logs. If the VM is hung or panicked, the
workflow would be unable to SSH into it and execute `dmesg`. In this
case the serial console logs would be helpful.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Replace the use of `guestmount and `guestunmount` with standard tools
(losetup, mount) to modify cloud disk image. This eliminates
dependency on guestmount which requires /dev/kvm and is not available
in MSHV root partition.
Signed-off-by: Aastha Rawat <aastharawat@microsoft.com>
Refactor MSHV hypervisor selection logic to enable aarch64 support
instead of exiting. Use `--features mshv` to conditionally build &
test when hypervisor is mshv replacing hard exit for aarch64 on mshv.
Signed-off-by: Aastha Rawat <aastharawat@microsoft.com>
There are several copy-pasted code fragments in impl QcowFile. All of
them add L2 entry to the cache and one of them (in file_offset_write())
does also allocating new L2 entry if necessary.
Fold all these code fragments (except of one in l2_table() which does
error handling in special way) into cache_l2_cluster() method without
changing the logic.
This will make the code more compact and clean.
Signed-off-by: Eugene Korenevsky <ekorenevsky@aliyun.com>
With the current syntax docker gives 'docker: invalid reference format'
error. Also during parsing /xxx:/yyy in process_volumes_args
with \" inside variable i.e arr_vols=("${arg_vols//#/ }")
gives wrong output.
Example:
scripts/dev_cli.sh tests --integration --volumes /mshv:/mshv
Error: The volume /mshv /mshv does not exist.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The API documentation tells users to expect a message when Cloud
Hypervisor is launched, this message was removed in commit 13724db
This change updates the documentation to reflect how the program
actually functions, which is to say no message.
Signed-off-by: Ariel Chenet <achenet@fastmail.com>
Consuming `&Arc<T>` as argument is almost always an antipattern as it
hides whether the callee is going to take over (shared) ownership
(by .clone()) or not. Instead, it is better to consume `&dyn T` or
`Arc<dyn T>` to be more explicit. This commit cleans up the code.
The change is very mechanic and was very easy to implement across the
code base.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Disable test_snapshot_restore_hotplug_virtiomem for mshv. It is failing
frequently in the CI. It needs to be stabilized before enabling again.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
The ivshmem tests involving snapshot/restore & live migration are
failing in the MSHV CI with this error:
Could not get vCPU state GetMsrEntries(
Hypercall 80 failed with 0x5 : InvalidParameter ...<snip>)
This needs more investigation. It is worth noting that the general live
migration tests are also not run in the CI for MSHV.
Disable these tests for MSHV for the time being.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
The ivshmem tests are all failing in the CI for MSHV because Cloud
Hypervisor is built without the mshv feature.
Error: Cloud Hypervisor exited with the following chain of errors:
0: Failed to open hypervisor interface (is hypervisor interface
available?)
1: Failed to create the hypervisor
2: no supported hypervisor
Modify the build command to include the mshv feature.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
The existing concurrency configuration causes workflow runs for one Pull
Request to cancel in-progress runs for another Pull Request. When the
workflow is triggered by the `pull_request_target event`, `github.ref`
is always set to the base branch, regardless of which branch raised the
PR. This causes every new pull request to share the same concurrency key
and thus cancel the running CI of any other open pull request.
This commit updates the `concurrency.group` to use
`github.event.pull_request.number` for unique isolation when running on
PR, falling back to `github.ref` for non-PR push (e.g., merge group).
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
Disable the following common parallel and common sequential tests for
mshv since these are failing consistently in the CI.
Common parallel:
- test_tpm
- test_cpu_topology_421
- test_cpu_topology_142
- test_cpu_topology_262
- test_cpu_hotplug
Common sequential:
- test_snapshot_restore_basic
- test_snapshot_restore_with_fd
- test_snapshot_restore_pvpanic
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
test_fw_cfg is frequently failing in the CI for MSHV. Exclude it for
now. It needs further investigation. See issue #7434 for details.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Move the GICv2M MSI frame SPI base from 32 to 128 to avoid conflicts
with platform device interrupts.
While at it, rename the constants to make it explicit that they are
associated with the GICv2m MSI Frame.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
This was added in 7be69edf51 to deal with
changes to the KVM bindings that made run() and set_immediate_exit()
take &mut self. Instead adopt a Box<> value in Vcpu allowing the removal
of this internal Mutex.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The virtio spec defines the feature bit range and 4 x 32-bit pages.
However no features are currently defined with bits > 63. However Linux
has now started acking features in those higher pages. Since this is
harmless (we drop the acks to those feature pages) and quietly return 0
for available features in those higher pages this warning can be
removed.
Fixes: #7414
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Fix a race condition that happens in resume()-pause() cycles.
It is odd that for pause(), the CpuManager waited via `state.paused`
for the vCPU thread to ACK the state change but not for `resume()`.
In the `resume()` case, oddly CpuManager "owned" the state change
in `state.paused`. This commit changes this so that the vCPU ACKs
its state change itself in `state.paused` when it transitions from
pause->run. Further, `CpuManager::resume()` now gracefully waits for
the vCPU to be resumed.
More technical:
This change ensures proper synchronization and prevents situations
in that park() follows right after unpark(), causing deadlocks and
other weird behavior due to race conditions.
Calling resume() now takes slightly longer, very similar to pause().
This is, however, even for 254 vCPUs in the range of less than 10ms,
and ultimately we now have correct behaviour.
## Reproducer
Since [0] is merged, the underlying problem can be tested without this
commit by modifying the pause() API call to run `CpuManager::pause()`
and `CpuManager::resume()` in a loop a thousand times.
`ch-remote --api-socket ... pause`
```patch
diff --git a/vmm/src/vm.rs b/vmm/src/vm.rs
index d7bba25cc..35557d58f 100644
--- a/vmm/src/vm.rs
+++ b/vmm/src/vm.rs
@@ -2687,6 +2687,10 @@ impl Pausable for Vm {
MigratableError::Pause(anyhow!("Error activating pending virtio devices: {:?}", e))
})?;
+ for _ in 0..1000 {
+ self.cpu_manager.lock().unwrap().pause()?;
+ self.cpu_manager.lock().unwrap().resume()?;
+ }
self.cpu_manager.lock().unwrap().pause()?;
self.device_manager.lock().unwrap().pause()?;
```
## Outlook
Decades of experience in VMM development showed us that using
many AtomicBools is a footgun. They are not synchronized with each
other at all. On the long term, we might want to refactor things to
have a single shared AtomicU64 with different bits having different
meanings.
[0] https://github.com/cloud-hypervisor/cloud-hypervisor/pull/7290
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
These bindings actually refer to atomic bool shared across all
vCPUs to instruct single vCPUs with their next action. As there
are already enough Arc<AtomicBool>, this helps while debugging
things to see that different bindings refer to the same atomic
bool.
In other words: This naming really confused us while debugging.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This gives a reliable way of identifying if the VM has booted as well as
a reliable way to validate the addition of the network interface.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Add a second L1 network interface definition to the guest Cloud Init
configuration, including an additional host IP. Do this by splitting the
network range into two /25s. For clarity the network struct members have
also been renamed.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Previous RISC-V ACPI related change introduced MSIS which will be
rejected by typo check, let's extend the typos to accept it.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Add necessary definitions and RISC-V ACPI tables to enable ACPI feature.
More specifically, this commit add MADT definitions for RISC-V AIA
interrupt chips.
Signed-off-by: Can Zhang <icloud9957@gmail.com>
Rename the MSHV CI workflow to better align with the existing naming
scheme and also make the arch explicit.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Add step-level timeout for running integration tests so that the
workflow doesn't get stuck on any test.
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
We are actually using v0.3.0 of vm-fdt since main branch of that crate
already includes everything. Let's pin vm-fdt to v0.3.0 instead.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Pull requests from forked repos cannot access GitHub secrets which
results in failure of MSHV CI. Switching to pull_request_target
resolves this. It allows the workflow to run with access to repo secrets
and ensures that code from the base branch is used instead of forked
code, preventing potential security risks.
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
Add workflow to run integration tests on mshv. It calls the azure
infra setup workflow and executes integration tests in the
provisioned environment.
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
Add workflow to setup Azure infra to validate MSHV.
This is used to provision the environment before executing tests
on it.
Signed-off-by: AASTHA RAWAT <aastharawat@microsoft.com>
This provides clearer life-cycle management of resources around vfio,
and aligns better with the kernel behavior as reported below with vfio
legacy mode (with vfio container/group).
Fixes: #7328
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Yi Wang <foxywang@tencent.com>
We recently moved many of our tests to use focal to jammy as the guest
images, including rate-limiter tests (#7367). We forgot to update the
rate-limiter scripts to reflect such change.
Our CI pipeline failed to report such error because our self-host runner
happened to be not working at the time we land the changes (see #7405).
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Only use 75% of the available threads - this will reduce dislk and
memory pressure. Reducing the chance of flaky tests.
See: #7405
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
While an IGVM is provided validation fails as there is
no kernel or firmware. This patch fixes the sev_snp boot
failure.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
MSHV (Microsoft Hypervisor) doesn't expose an ITS to guests. Instead it
exposes a GICv2m MSI Frame for mapping MSIs to SPIs.
Update the MADT generation code to create the GIC MSI Frame structure
instead of the ITS structure whenever the guest vGIC is using a GICv2m
frame. This allows booting guests via firmware on MSHV.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
In preparation for removing focal support - start by using a more
generic "disk_config" variable name for the DiskConfig struct vs
embedding the image type.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Rust's aliasing rules do not permit mutating a `usize` through a shared
reference. Unsafe blocks and FFI are no exception.
This PR fixes such a violation detected in the AMX support checks by
passing a raw mutable pointer instead of a shared reference in the
syscall.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP <oliver.anderson@sap.com>
For a graceful resource management of externally provided FDs in Cloud
Hypervisor, corresponding FDs need to be closed on a device removal.
This is the case for virtio-net devices using external FDs, for example.
With the fix introduced in this commit, we allow management software to
properly clean up resources, e.g., libvirt can clean up tap devices.
PS: CHV uses "added" and "removed", which has the same meaning as
hot device attach/hotplug and hot device detach/unplug.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Fix clippy warning `uninlined_format_args` reported by rustc rustc
1.89.0 (29483883e 2025-08-04).
```console
warning: variables can be used directly in the `format!` string
--> block/src/lib.rs:649:17
|
649 | info!("{} failed to create io_uring instance: {}", error_msg, e);
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#uninlined_format_args
= note: `#[warn(clippy::uninlined_format_args)]` on by default
help: change this to
|
649 - info!("{} failed to create io_uring instance: {}", error_msg, e);
649 + info!("{error_msg} failed to create io_uring instance: {e}");
|
```
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
"Update configs must have a unique combination of 'package-ecosystem',
'directory', and 'target-branch'. Ecosystem 'cargo' has overlapping
directories for target branch 'main'."
Drop `target-branch` in rust-vmm group to make those two unique.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
## TL;DR
This unifies error handling, implementation, and logging of config
objects that are populated with additional FDs received as part of
the HTTP API request from the management software. This prevents current
and future code in repetition.
In the future, if we want to support let's say virtio-blk with external
FDs, the AddBlk API handler simply needs a call to `attach_fds_to_cfg()`
and gets all the magic for free.
## Motivation
The motivation of the new abstraction is to provide a verbose, solid,
and bulletproof solution for a complex domain. The interaction between
- the management layer,
- the passing of file descriptors over UNIX domain sockets via
SCM_RIGHTS,
- the attachment of configurations to those FDs,
- and the ability to give new developers clear insights into what
happens under the hood
is not trivial. These factors justify encapsulating the complexity
behind a convenient and well-documented abstraction, making the system
both robust in production and approachable for new developers.
In addition, it allows us to perform unit testing. Further, We get rid
of existing partial code duplication and inconsistencies.
Finally, while this approach may initially result in more code, every
new handler that accepts FDs benefits from reduced duplication and a
correct implementation by relying on the shared abstraction. This will
also enable future functionality, such as virtio-blk devices backed by
FDs which can then be integrated with ease.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Deserializing values as `-1` makes sense to prevent errors, so let's
keep it. However, serializing them differently adds confusion. For
example, a `ch-remote info` call should not report `-1` but the actual
FDs.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
(1) The old messages are missing the "why" part. With this change, users
of Cloud Hypervisor have somehow more context and people looking at
the code perfectly know what's going on.
(2) Using warn! implies that the user should take action, but in this
case, there’s nothing the user can do. If the API is used correctly
and file descriptors are passed via an SCM_RIGHTS message over a
UNIX domain socket, everything works as intended. In that case,
there's no need to issue a warning — debug! is sufficient.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Live migration, state save/resume, and hotplug are not trivial when it
comes to virtio-net devices backed by externally provided FDs. As the
mechanism behind it can be considered as quite "multi-step magic" even
for experienced programmers, it makes sense to thoroughly document this
to ease debugging and to improve the mental model of developers working
on this in the future.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This change enables easier integration with third-party
tools by removing the requirement for a dummy IP address
when configuring tap devices. The modification applies to
both CLI and API interactions.
Previously, cloud-hypervisor would automatically set a
default static IP address (192.168.249.1) if none was provided.
This could lead to:
* multiple devices without explicit IP configurations
would end up with the same default IP
* unnecessary inclusion of this IP in firewall rules
* the IP address could clash with host networking and
routing
This introduces a new constraint:
When providing an IP, the mask must also be provided.
Removes warnings introduced in #7179.
Closes issue #7083.
Signed-off-by: Maximilian Güntner <code@mguentner.de>
The function `read_local_stream_port` had no proper handling for
unexpected or incomplete input.
When the control socket of the VSOCK device was closed without sending
the expected `CONNECT <PORT>\n` statement completely, the thread
got stuck in an infinite loop as it attempted to read from a closed
socket over and over again which never returned any data.
This resulted in the thread responsible for `epoll` being completely
blocked. New VSOCK connections could not be established and existing
ones became defunct, effectively leading to a Denial of Service of
the entire VSOCK device.
The issue can be reproduced by opening a socket and immediately
closing it.
```
socat - UNIX-CONNECT:/socket.vsock
<Ctrl-C>
```
Instead of applying a quick fix by handling the `EPOLLHUP` event before
reading, the function is refactored to remove the error-prone `while`
loop and multiple `read`s.
Notably, we now check if the number of bytes read is zero, which occurs
when `event_set == EPOLLHUP | EPOLLIN`, indicating that the socket has
been closed by the client.
Additionally, the actual parsing code is now extracted into a dedicated
function that is tested.
Fixes: #6798
Signed-off-by: Maximilian Güntner <code@mguentner.de>
For MSHV customers don't want to make everything
default during partition creation. For example
nested support, some synthetic features could be
controlled from CLI through platform argument.
Create_vm API getting messy after adding more flags.
This patch introduces common data struct to be passed
from vmm crate to hypervisor crate during partition creation.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Handle MSHV reset intercepts on arm64 such that guests can gracefully
shutdown/reboot instead of panicking due to unhandled intercept.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Reference link to set_device_attribut for aarch64
is invalid. Looks like the code browsing does not
have the reference anymore.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This message makes it easier to confirm a successful shutdown when
scanning logs during development and in production. If the message is
missing and the last log entry is not an error, the most likely cause
is a livelock (e.g. contention on a lock) introduced while developing
new Cloud Hypervisor features.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This commit includes all simple clippy fixes excluding the
collapsing of nested ifs using the let-chains feature. This
follows in the next commit.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Bump mshv-ioctls and mshv-bindings to 0.6.0. Most notably, this version
contains fixes and new bindings for arm64 guests.
Bump the vfio crates too so that they point to the latest mshv crates.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
Virtio console is activated much later in boot. The output it spits out
lacks the initial CPU configuration of the guest.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
These can differ between platforms, so it's better to use centralized
definitions of them. We can't currently do this for the KVM and VFIO
ioctls, because the corresponding crates don't publicly expose the
ioctl numbers.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
The previously linked file was converted to ReStructuredText, then
later removed from the kernel, with part of the justification being
"Most of what booting-without-of.rst contains is now in the DT
specification", so point to that instead.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Since the user has to explicitly ask for Landlock to be enabled in
Cloud Hypervisor, it's surprising that, even when they've done that,
Landlock will silently not be enabled if the kernel doesn't support
it.
To prevent accidental absence of a desired security protection, exit
with an error if Landlock, or the one feature of it we use in Cloud
Hypervisor (file access) is not supported.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
arch::aarch64::fdt::create_cpu_nodes will always look at this if it
exists. (If it doesn't exist, this is a no-op —
add_rule_with_access() won't add rules for paths that don't exist.)
Fixes: b3e5738b4 ("vmm: Introduce ApplyLandlock trait")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
MSHV now supports movable pages i.e VA backed guest.
Also with more features and stability in the MSHV
Kernel drives MSHV now supports more test scenario.
This patch enables more integration test cases.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Without write access to the directory the socket will go in, it's not
possible to create the socket.
I've tested outgoing connections from the VM, and they don't seem to
need read permissions on that directory to connect to a socket on the
host.
Fixes: b3e5738b4 ("vmm: Introduce ApplyLandlock trait")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
path_beneath_rules() just needs the paths given to it to be
AsRef<Path>, so there's no need to create new PathBufs for it.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Raise the max number of supported (v)CPUs on kvm x86_64 hosts
to 8192 (the max allowed value of CONFIG_NR_CPUS in the Linux kernel).
Other platfroms keep their existing CPU limits pending further
development and testing.
The change has been tested on Intel and AMD hosts.
Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Neel Natu <neelnatu@google.com>
Signed-off-by: Ofir Weisse <oweisse@google.com>
Signed-off-by: Peter Oskolkov <posk@google.com>
These are now used by Cloud Hypervisor when linked with Glibc 2.42.
These values should be correct for all currently supported Cloud
Hypervisor platforms, although they are not for all Linux platforms.
Closes: https://github.com/cloud-hypervisor/cloud-hypervisor/issues/7276
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This commit removes the SGX support from cloud hypervisor. SGX support
was deprecated in May as part of #7090.
Signed-off-by: Shubham Chakrawar <schakrawar@crusoe.ai>
This patch gives user an option to override the
default migratable version to any later release.
This option makes MSHV specific tests suitable for
tests since MSHV is stable after some breaking changes.
This patch is also necessary for MSHV CI.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This step was done manually by searching for "unsafe fn" in
the code base and adding corresponding Safety sections.
`clippy::missing_safety_doc` only works for public functions
but none of the corresponding functions is public.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Implement the batch submission function for raw disk, default it is
enabled. After parsing the requests this method is
called for better IO latency and bandwidth.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Cache and batch IO requests after parsing all
items in the queue, improving performance—especially
for small block sizes—by reducing per-request overhead.
Introduced two methods in the AsyncIo trait for batch
submission, with implementation in the raw disk backend.
This method should be called during/after parsing all block IO requests
in the available queue. If the batch submission is not enabled, by
default it does the old way of submitting requests.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Instead of returning boolean return an struct of completion status
so that it can be cached for batch submission.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This patch changes the read-only check using acked features bit, which
will help to check more features.
Signed-off-by: Songqian Li <sionli@tencent.com>
Also tweak extended interrupt handlings, as needed.
Most credit should go to Neel Natu, who figured out the magic
bits needed to make things work and provided detailed comments.
This patch is still NOOP, as VM config allows only up to
254 vCPUs on x86_64.
Note: changes in this and related previous patches/PRs have
only been tested on Linux hosts running on Intel x86_64 hardware.
Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Neel Natu <neelnatu@google.com>
Signed-off-by: Ofir Weisse <oweisse@google.com>
Signed-off-by: Peter Oskolkov <posk@google.com>
test_direct_kernel_boot_bzimage runs only on x86, so the cfg!() branch
for selecting grep_cmd is unnecessary. Remove it for clarity.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
One can call `to_vec()` anyway if one needs an owned copy. This change
further helps to prevent needless copies in upcoming changes.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
On aarch64 and RISC-V, calling load_firmware() through load_kernel()
provides no benefit and only duplicates checks already performed in
load_payload(). load_payload() now directly invokes load_firmware() or
load_kernel(), removing unnecessary indirection and redundancy.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Both functions are defined separately for the two architecture with
minor differences.
* `load_firmware()`: call `arch::uefi::load_uefi` which are available on
both architecture;
* `load_kernel()`: manually align to `arch::layout::KERNEL_START` 2MB
for both architecture (e.g. no-op for `aarch64`);
Signed-off-by: Bo Chen <bchen@crusoe.ai>
With 254 vCPUs, pausing now takes ~4ms instead of >254ms. This
improvement is visible when running `ch-remote pause` and is
particularly important for live migration, where every millisecond
of downtime matters.
For the wait logic, it is fine to stick to the approach of
sleeping 1ms on the first missed ACK as:
1) we have to wait anyway
2) we give time to the OS, enabling it to schedule a vCPU thread next
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Enable kvm build test and clippy test on RISC-V 64-bit platform to
ensure whole projects builds properly.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
MP table is a legacy device that is incompatible
with x2apic CPU IDs exceeding 254. The Linux kernel
is perfectly happy without MP table in these cases.
Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Neel Natu <neelnatu@google.com>
Signed-off-by: Ofir Weisse <oweisse@google.com>
Signed-off-by: Peter Oskolkov <posk@google.com>
Currently, the following scenarios are supported by Cloud Hypervisor to
bootstrap a VM:
1. provide firmware
2. provide kernel
3. provide kernel + cmdline
4. provide kernel + initrd
5. provide kernel + cmdline + initrd
As the difference between `--firmware` and `--kernel` is not very clear
currently, especially as both use/support a Xen PVH entry, adding this
helps to identify the cause of misconfiguration.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This is necessary to use the let-chains feature in a
follow-up. After upgrading to Rust edition 2024, clippy
wants to collapse various if's with let-chains.
Update image to 20250815-0 since MSRV in Dockerfile is updated.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This patch introduces the inter-vm shared memory(ivshmem) device
to share a memory region between multiple processes running
different guests and the host.
This patch supports the basic ivshmem functions like ivshmem-plain
in QEMU[1].
[1] https://www.qemu.org/docs/master/specs/ivshmem-spec.html
Signed-off-by: Yi Wang <foxywang@tencent.com>
Signed-off-by: Songqian Li <sionli@tencent.com>
Move UserspaceMapping to vm-device to avoid redefinition since
UserspaceMapping is used by both `virtio-devices` and `device`
crate.
Signed-off-by: Songqian Li <sionli@tencent.com>
It is always called with topology provided, so there is no
need to pass topology as an Option. Simplifying the signature
makes further topology-related changes to arc/src/x86_64 module
simpler.
Signed-off-by: Peter Oskolkov <posk@google.com>
This is the second patch in a series intended to let Cloud Hypervisor
support more than 255 vCPUs in guest VMs; the first patch/commit is
https://github.com/cloud-hypervisor/cloud-hypervisor/pull/7231
At the moment, CPU topology in Cloud Hypervisor is using
u8 for components, and somewhat inconsistently:
- struct CpuTopology in vmm/src/vm_config.rs uses four components
(threads_per_core, cores_per_die, dies_per_package, packages);
- when passed around as a tuple, it is a 3-tuple of u8, with
some inconsistency:
- in get_x2apic_id in arch/src/x86_64/mod.rs the three u8
are assumed to be (correctly)
threads_per_core, cores_per_die, and dies_per_package, but
- in get_vcpu_topology() in vmm/src/cpu.rs the three-tuple is
threads_per_core, cores_per_die, and packages (dies_per_package
is assumed to always be one? not clear).
So for consistency, a 4-tuple is always passed around.
In addition, the types of the tuple components is changed from u8 to
u16, as on x86_64 subcomponents can consume up to 16 bits.
Again, config constraints have not been changed, so this patch
is mostly NOOP.
Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Neel Natu <neelnatu@google.com>
Signed-off-by: Ofir Weisse <oweisse@google.com>
Signed-off-by: Peter Oskolkov <posk@google.com>
This is the first change to Cloud Hypervisor in a series of changes
intended to increase the max number of supported vCPUs in guest VMs,
which is currently limited to 255 (254 on x86_64).
No user-visible/behavior changes are expected as a result of
applying this patch, as the type of boot_cpus and related
fields in config structs remains u8 for now, and all configuration
validations remain the same.
Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Neel Natu <neelnatu@google.com>
Signed-off-by: Ofir Weisse <oweisse@google.com>
Signed-off-by: Peter Oskolkov <posk@google.com>
This allows us to enable/disable the fw_cfg device via the cli
We can also now upload files into the guest vm using fw_cfg_items
via the cli
Signed-off-by: Alex Orozco <alexorozco@google.com>
This allows the fw_cfg device to be recognized by the guest linux
kernel. This becomes more relavnt in the following cl where I add
the option to load files into the guest via fw_cfg. The Linux kernel
already has a fw_cfg driver that will automatically load these files
under /sys when CONFIG_FW_CFG_SYSFS is enabled in the kernel config
For arm we must add fw_cfg to the devices tree
Signed-off-by: Alex Orozco <alexorozco@google.com>
We pass a reference to the guest memory when we create the device
in DeviceManager. This allows us to access the guest memory for DMA.
Signed-off-by: Alex Orozco <alexorozco@google.com>
The acpi tables are created in the same place the acpi tables would be
created for the regular bootflow, except here we add them to the
fw_cfg device to be measured by the fw and then the fw will put the
acpi tables into memory.
Signed-off-by: Alex Orozco <alexorozco@google.com>
The kernel and initramfs are passed to the fw_cfg device as
file references. The cmdline is passed directly.
Signed-off-by: Alex Orozco <alexorozco@google.com>
Here we add the fw_cfg device as a legacy device to the device manager.
It is guarded behind a fw_cfg flag in vmm at creation of the
DeviceManager. In this cl we implement the fw_cfg device with one
function (signature).
Signed-off-by: Alex Orozco <alexorozco@google.com>
This partially reverts
ed8f347fe6 from #7183 and
6277d7d5f2 from #7201.
# Output how it was merged for v47 (#7066)
```
Error: Cloud Hypervisor exited with the following chain of errors:
0: Error booting VM
1: The VM could not boot
2: Error manipulating firmware file
3: No such file or directory (os error 2)
Debug Info: VmBoot(VmBoot(FirmwareFile(Os { code: 2, kind: NotFound, message: "No such file or directory" })))
```
# Output after #7183 and #7201
```
cloud-hypervisor: 31.385730ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:27 -- Error: Cloud Hypervisor exited with the following chain of errors:
cloud-hypervisor: 31.417961ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:39 -- 0: Error booting VM
cloud-hypervisor: 31.448078ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:39 -- 1: The VM could not boot
cloud-hypervisor: 31.486711ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:39 -- 2: Error manipulating firmware file
cloud-hypervisor: 31.513331ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:39 -- 3: No such file or directory (os error 2)
cloud-hypervisor: 31.548037ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:44 --
cloud-hypervisor: 31.568045ms: <main> ERROR:/home/pschuster/dev/cloud-hypervisor/src/lib.rs:45 -- Debug Info: VmBoot(VmBoot(FirmwareFile(Os { code: 2, kind: NotFound, message: "No such file or directory" })))
```
The "proper logger" has indeed the advantage that messages can
be gracefully redirected to log files etc. However, this makes the
error message hardly readable.
Therefore, I propose to use error!() only for runtime errors messages
but not a pretty-printed version of those.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Error::UefiLoad is required for load_firmware to propagate errors
encountered, define it for riscv64.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
uefi_flash field in memory_manager is required for uefi loading and
booting, enable it for riscv64 architecture.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The option parser did not allow quoting any special characters, which
means that some strings could not be passed via command-line options.
Fix this by allowing double quotes to quote all the special characters,
such as '@', ',', '[', and ']'. '"' can be escaped by doubling it. The
implementation uses some tricks to stay modular.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
rustc 1.90.0-beta.1 (788da80fc 2025-08-04) suggests using library
feature `unsigned_is_multiple_of`. It is stabled in Rust 1.87.0.
Update image to 20250807-0 since MSRV in Dockerfile is updated.
Signed-off-by: Songqian Li <sionli@tencent.com>
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Now 'create_acpi_tables_internal()' can generate ACPI tables with
different dsdt offset, so that it can be reused for generating ACPI
tables for `fw_cfg`.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Alex Orozco <alexorozco@google.com>
Optimize lychee workflow to check only changed
files in pull requests, avoiding excessive API
calls to prevent rate limits.
Fixes#7056
Signed-off-by: Shubham Chakrawar <schakrawar@crusoe.ai>
In case of CVM guest rsdp is set to none. Unwrapping it
make the vmm crashed. Don't call configure system if the
rsdb address is none.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Many of the workspace members in the Cloud-hypervisor workspace share
common dependencies. Making these workspace dependencies reduces
duplication and improves maintainability.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
The current documentation only mentions Windows
server support. Since only minimal adjustments
are needed for Windows 11 support, these were
added where applicable.
Also contains whitespace fixes, adds missing `\`
in example code and adds more cross references.
Signed-off-by: Lucas Grosche <lucas.grosche@cyberus-technology.de>
The other workspace members in the Cloud-hypervisor workspace
currently declare log version 0.4.22, but the tpm crate
has an older version. This inconsistency is addressed by this PR
which opens the door for declaring log as a workspace dependency.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
Running one or two tests in a tight loop can cause the download
functions to quickly hit GitHub's API rate limit. That causes the test
script to fail for no apparent reason.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Generate the data disk under a temporary directory so that multiple
instances of the test suites can run at the same time.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The other workspace members in the Cloud-hypervisor workspace
currently declare libc crate version 0.2.167, but the tpm crate
has an older version. This inconsistency is addressed by this PR
which opens the door for declaring libc as a workspace dependency.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
The other workspace members in the Cloud-hypervisor workspace
currently declare anyhow version 1.0.94, but the tpm crate
has an older version. This inconsistency is addressed by this PR
which opens the door for declaring anyhow as a workspace dependency.
Signed-off-by: Oliver Anderson <oliver.anderson@cyberus-technology.de>
On-behalf-of: SAP oliver.anderson@sap.com
It shouldn't be necessary to lock the file for the adaptor. This removes
two layers of indirection for QcowDiskSync and VhdxDiskSync.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The PL031 RTC provides two features: a real-time counter and an alarm
interrupt. To use the alarm, the driver normally writes a time value
into the match register RTCMR, and when the counter reaches that value
the device triggers the interrupt.
At the moment the implementation ignores programming of the alarm, as
the feature seems rarely used in VMs. However the interrupt is still
triggered arbitrarily when the guest writes to registers, and the line
is never cleared. This really confuses the Linux driver, which loops in
the interrupt handler until Linux realizes that no one is dealing with
the interrupt (200000 unanswered calls) and disables the handler.
One way to fix this would be implementing the alarm function properly,
which isn't too difficult but requires adding some async timer logic
which probably won't ever get used. In addition the device's interrupt
is level-triggered and we don't support level interrupts at the moment,
though we could probably get away with changing this interrupt to edge.
The simplest fix, though, is to just disable the interrupt logic
entirely, so that the alarm function still doesn't work but the guest
doesn't see spurious interrupts.
Add a default() implementation to satisfy clippy's new_without_default
check, since Rtc::new() doesn't take a parameter after this change.
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
The read_exact() call was introduced in 82ac114b8 ("virtio-devices:
vsock: handle short read in muxer") to solve a crash when a connection
disconnected without sending any data, but it introduced a problem of
its own: because the socket is non-blocking, read_exact() may read
some data, then return ErrorKind::WouldBlock. In that case, the data
it read will be discarded. So for example if it read "CONNECT ",
and then nothing else was available to read yet, "CONNECT " would be
discarded, and so the next time this function was called, when epoll
triggered again for the socket, only the following data would end up
in command.buf, causing an error due to just a port number being an
invalid command.
Contrary to that commit message, this code was actually designed to
handle short reads just fine — in the case of a short read, it stores
the data it has read in command, and returns
Error::UnixRead(ErrorKind::WouldBlock), which is ignored by the
caller, and the function gets called again when there is more data to
read, building up command potentially over the course of several
reads. The only thing it didn't handle correctly, as far as I can
tell, was a 0-byte read, which happens when a client disconnects from
the socket without writing anything. All that's needed to fix this is
to avoid an invalid subtraction in that case, so this change reverts
82ac114b8, fixing the issue with partial commands being discarded, and
instead handles the 0-byte read by using slice::get, and treating an
empty command as an incomplete command, which of course it is.
Fixes: 82ac114b8 ("virtio-devices: vsock: handle short read in muxer")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This makes it possible to run cargo test just for the virtio-devices
crate (as long as either KVM or MSHV is specified).
Signed-off-by: Alyssa Ross <hi@alyssa.is>
`std::mem::offset_of` could be used for calculating nested fields, use
this feature to shorten aarch64 reg offset calculation.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`std::mem::offset_of` is stabilized since Rust 1.77, let's use
implementation provided by std instead of manual implementation.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Manually implemented `_offset_of` and `offset_of` in
`arch/riscv64/mod.rs` are not used now, remove them.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`std::mem::offset_of` supports calculating offset of nested structures,
let's use implementation provided by std instead of manual
implementation.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Other lines are already logged using `log::error!` and
`env_logger` is initialized before calling
`start_net_backend` in `main()`.
Signed-off-by: Maximilian Güntner <code@mguentner.de>
Unify log formatting and printing as `eprintln!` and `log::error!`
would be used alongside each other.
When using e.g. `env_logger` lines printed with `eprintln!` would
lack formatting / colors.
Currently only relevant in `ch-remote` + `cli_print_error_chain`.
Note that the replaced messages now also end up in the logfile of
`cloud-hypervisor` when configured and not any longer in stderr.
Signed-off-by: Maximilian Güntner <code@mguentner.de>
Until now all messages generated using `log::level!`
(e.g., `warn!`) have not been printed as `ch-remote` did not
register a logger.
Furthermore, replace all `eprintln!` with `error!`
to align formatting for consistency.
Signed-off-by: Maximilian Güntner <code@mguentner.de>
Recently vfio crates have moved to crates.io, thus we should start
consuming the crate from crates.io instead git url.
This results in better versioning instead of tracking some git commit
sha.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Since action-rs/cross is deprecrated, thus move to
houseabsolute/actions-rust-cross.
We should pin the cross-version to the latest version to fix the build
issues with virtio-bindings crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
When invoking the script chown shows a warning.
chown: warning: '.' should be ':': ‘1000.1000’
From `info coreutils 'chown invocation'`.
Some older scripts may still use ‘.’ in place of the ‘:’ separator.
POSIX 1003.1-2001 (*note Standards conformance::) does not require
support for that, but for backward compatibility GNU ‘chown’ supports
‘.’ so long as no ambiguity results, although it issues a warning and
support may be removed in future versions. New scripts should avoid the
use of ‘.’ because it is not portable, and because it has undesirable
results if the entire OWNER‘.’GROUP happens to identify a user whose
name contains ‘.’.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
That feature was dropped when consolidating the UUID dependency because
somehow building the whole project worked. The CI system was happy.
However, building the block crate alone is broken. The vhdx code uses
Uuid::new_v4, which requires `v4` to be enabled.
Add the feature back.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
When a proper Rust backtrace is printed, the Rust std wants to use the
SYS_getcwd(79) system call to prettify some paths while printing. In
Cloud Hypervisor, this is at least relevant for printing panics or if
a `anyhow::Error` value is printed using `{e:?}` (but not `{e:#?}`).
The syscall cause can be found in `impl fmt::Display for Backtrace {}`
in `library/std/src/backtrace.rs`.
Without this addition, the seccomp violation of the SYS_getcwd (79)
hinders the proper error message including a full backtrace from showing
up. This annoying behaviour already delayed many debugging efforts. With
this fix, things just work. The new syscall itself should be pretty
harmless for normal operation.
```
thread 'vmm' panicked at virtio-devices/src/rng.rs:224:9:
Yikes, things went horribly wrong!
==== Possible seccomp violation ====
Try running with `strace -ff` to identify the cause and open an issue: https://github.com/cloud-hypervisor/cloud-hypervisor/issues/new
[1] 287683 invalid system call (core dumped) RUST_BACKTRACE=full cargo run --bin cloud-hypervisor -- --api-socket --kerne
```
```
thread 'vmm' panicked at virtio-devices/src/rng.rs:224:9:
Yikes, things went horribly wrong!
stack backtrace:
0: 0x557d91286b62 - std::backtrace_rs::backtrace::libunwind::trace::hc20b48b31ee52608
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library/std/src/../../backtrace/src/backtrace/libunwind.rs:117:9
1: 0x557d91286b62 - std::backtrace_rs::backtrace::trace_unsynchronized::h5d207cd20f193d88
at /rustc/17067e9ac6d7ecb70e50f92c1944e545188d2359/library/std/src/../../backtrace/src/backtrace/mod.rs:66:14
...
67: 0x0 - <unknown>
Error: Cloud Hypervisor exited with the following error:
Failed to join on VMM thread: Any { .. }
Debug Info: ThreadJoin(Any { .. })
```
- add any panic, for example into the create or drop function of a
device
- add --seccomp=true|log to analyze the situation
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
The virtio_features_to_tap_offload() defined in ctrl_queue.rs
is duplicated. Remove it and use the one defined in lib.rs
instead.
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Checksumming a type that has padding would use the padding in the
checksum, which is definitely wrong and is often unsound.
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com>
Cargo fuzz build report an error:
error: lifetime flowing from input to output with different syntax can be confusing
--> /home/runner/work/cloud-hypervisor/cloud-hypervisor/block/src/lib.rs:747:13
|
747 | fn file(&mut self) -> MutexGuard<F>;
| ^^^^^^^^^ ------------- the lifetime gets resolved as `'_`
| |
| this lifetime flows to the output
error: lifetime flowing from input to output with different syntax can be confusing
--> /home/runner/work/cloud-hypervisor/cloud-hypervisor/block/src/async_io.rs:68:11
|
68 | fn fd(&mut self) -> BorrowedDiskFd;
| ^^^^^^^^^ -------------- the lifetime gets resolved as `'_`
| |
| this lifetime flows to the output
Signed-off-by: Yi Wang <foxywang@tencent.com>
Remote server errors are transferred as raw HTTP body. This way,
we lose the nested structured error information.
This is an attempt to retrieve the errors from the HTTP response
and to align the output with the normal error output.
For example, this produces the following chain of errors. Note
that everything after level 0 was retrieved from the HTTP server
response:
```
Error: ch-remote exited with the following chain of errors:
0: http client error
1: Server responded with InternalServerError
2: Error from API
3: The disk could not be added to the VM
4: Failed to validate config
5: Identifier disk1 is not unique
Debug Info: HttpApiClient(ServerResponse(InternalServerError, Some("Error from API<br>The disk could not be added to the VM<br>Failed to validate config<br>Identifier disk1 is not unique")))
```
In case the JSON can't be parsed properly, ch-remote will print:
```
Error: ch-remote exited with the following chain of errors:
0: http client error
X: Can't get remote's error messages from JSON response: EOF while parsing a value at line 1 column 0: body=''
Debug Info: HttpApiClient(ServerResponse(InternalServerError, Some("")))
```
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
With the foundations of each error type implementing std::error::Error,
we can now nicely walk the `.source()` chain and print an error trace.
This commit introduces improved user-facing error printing when:
- Cloud Hypervisor fails with an error
- ch-remote fails (client error)
- ch-remote fails (remote error)
The additional context is a clear improvement in UX for both users and
developers. In the following example, the new behaviour is shown for
a direct invocation of Cloud Hypervisor leading to a failure. This
looks similar for ch-remote.
```
Old Style
`target/release/cloud-hypervisor --api-socket /tmp/chv2.sock --kernel /etc/bootitems/linux/kernel_minimal/stable.bzImage --cmdline console=ttyS0 --serial tty --console off --disk path=img.raw --initramfs /etc/bootitems/linux/initrd_minimal/default`
Error booting VM: VmBoot(LockingError(BlockError(LockDiskImage(AlreadyLocked)))
```
```
`target/release/cloud-hypervisor --api-socket /tmp/chv2.sock --kernel /etc/bootitems/linux/kernel_minimal/stable.bzImage --cmdline console=ttyS0 --serial tty --console off --disk path=img.raw --initramfs
/etc/bootitems/linux/initrd_minimal/default`
Error: Cloud Hypervisor exited with the following chain of errors:
0: Error booting VM
1: The VM could not boot
2: Error locking disk images: Another instance likely holds a lock
3: Cannot lock images of all block devices
4: Failed to get Write lock for disk image: ./img.raw
5: The file is already locked
Debug Info: VmBoot(VmBoot(LockingError(DiskLockError(LockDiskImage { error: AlreadyLocked, lock_type: Write, path: "./raw_disk.bin" })))
```
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
The changes were mostly automatically applied using the following
Python script:
```python
import os, re
for root, _, files in os.walk("."):
for f in files:
if not f.endswith(".rs"):
continue
p = os.path.join(root, f)
with open(p, "r", encoding="utf-8") as file:
lines = file.readlines()
changed = False
for i in range(len(lines) - 1):
if re.search(r'#\[error\(".*: \{0[^}]*\}"\)\]', lines[i]) and "#[source]" in lines[i + 1].strip():
lines[i] = re.sub(r': \{0[^}]*\}"\)\]', '")]', lines[i])
changed = True
if changed:
with open(p, "w", encoding="utf-8") as file:
file.writelines(lines)
print("Fixed:", p)
```
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
# Conflicts:
# vmm/src/api/http/mod.rs
Having them sorted alphabetically makes more sense since there are
already many, and the list is growing. This improves the UX for users
and developers.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Instead of exiting on IO errors, report the errors to the guest with
VIRTIO_BLK_S_IOERR. For example, the guest kernel will log something
similar to this if the nbd behind /dev/vdc is unexpectedly disconnected:
[ 166.033957] I/O error, dev vdc, sector 264 op 0x1:(WRITE) flags 0x9800 phys_seg 1 prio class 2
[ 166.035083] Aborting journal on device vdc-8.
[ 166.037307] Buffer I/O error on dev vdc, logical block 9, lost sync page write
[ 166.038471] JBD2: I/O error when updating journal superblock for vdc-8.
[...]
[ 174.234470] EXT4-fs (vdc): I/O error while writing superblock
In case the rootfs is not located on the affected block device, this
will not crash the guest.
Fixes: #6995
Signed-off-by: Gauthier Jolly <contact@gjolly.fr>
This was caught by the nightly compiler during cargo fuzz build.
error: lifetime flowing from input to output with different syntax can be confusing
--> /home/runner/work/cloud-hypervisor/cloud-hypervisor/hypervisor/src/arch/x86/emulator/mod.rs:493:26
|
493 | pub fn new(platform: &mut dyn PlatformEmulator<CpuState = T>) -> Emulator<T> {
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ----------- the lifetime gets resolved as `'_`
| |
| this lifetime flows to the output
|
= note: `-D mismatched-lifetime-syntaxes` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(mismatched_lifetime_syntaxes)]`
help: one option is to remove the lifetime for references and use the anonymous lifetime for paths
|
493 | pub fn new(platform: &mut dyn PlatformEmulator<CpuState = T>) -> Emulator<'_, T> {
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
cargo fuzz build complaints about some un-used function in the
instruction emultator. Silence the warning by allowing dead code
generation.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Since it is used by multiple components at this point, it is better to
move it to workspace level dependency.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
There are a little many cpuid logs now. When starting a vm with
64 vcpu, we can get more than four thousand INFO messages:
cat vm1.log |grep 'arch/src/x86_64/mod.rs:891' |wc -l
4352
Signed-off-by: Yi Wang <foxywang@tencent.com>
This commit adds the warning to deprecate the SGX support with the
intention to remove the support from code base in two release cycles.
See: #6960
Signed-off-by: Bo Chen <bchen@crusoe.ai>
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the Error implementation in the Cloud Hypervisor code
base to match the remaining parts so that everything follows the agreed
conventions. These are leftovers missed in the previous commits.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
When booting a Linux guest in SMP configuration,
the following kernel warning can be observed:
[Firmware Bug]: CPUID leaf 0xb subleaf 1 APIC ID mismatch 1 != 0
The reason is that we announce the presence of the extended topology
leaf, but fail to announce the x2apic ID in EDX.
Signed-off-by: Thomas Prescher <thomas.prescher@cyberus-technology.de>
On-behalf-of: SAP thomas.prescher@sap.com
No sane guest OS will allow hotplugging all cpus.
However, the REST API currently allows specifying
`ch-remote resize --cpus 0`.
On Linux, we can then observe the following error in the kernel log:
processor cpu0: Offline failed.
Subsequent resize commands via ch-remote will then fail with
VcpuPendingRemovedVcpu because the removal of cpu0 was never
successful.
Fix this by disallowing resizing to zero vCPUs.
Signed-off-by: Thomas Prescher <thomas.prescher@cyberus-technology.de>
On-behalf-of: SAP thomas.prescher@sap.com
For use in QEMU, I would like GuestMemoryRegion to return a BitmapSlice
instead of a &Bitmap. This adds some flexibility that QEMU needs in
order to support a single global dirty bitmap that is sliced by the
various GuestMemoryRegions.
However, this removes access to the methods of AtomicBitmap, and in
particular reset() and get_and_reset(). Fortunately, cloud-hypervisor
always uses GuestMemoryMmap, and therefore `region` is known to be a
&GuestRegionMmap. Dereferencing it returns the MmapRegion to which the
bitmap is attached, thus calling MmapRegion::bitmap(); this has the
same effect as `<GuestRegionMmap as GuestRegion>::bitmap()`, and works
both with or without https://github.com/rust-vmm/vm-memory/pull/324.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
This streamlines the code base to follow best practices for
error handling in Rust: Each error struct implements
std::error::Error (most due via thiserror::Error derive macro)
and sets its source accordingly.
This allows future work that nicely prints the error chains,
for example.
So far, the convention is that each error prints its
sub error as part of its Display::fmt() impl.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Allow tap interfaces to be configured with an IPv6 address. The change
is fairly straightforward: we need to update the API types and CLI
parsing to accept either an IPv6 or IPv4 and then match on the IP
address type when the tap device is configured.
For IPv6 addresses, the netmask (prefix) must be provided at the same
time as the address itself (in the SIOCSIFADDR ioctl). They cannot be
configured separately. So we remove the separate "set_netmask" function
and convert "set_ip_addr" to also accept a netmask. For IPv4 addresses,
the IP address and netmask were already always set together, so this
should have no functional impact for users of IPv4 addresses.
Signed-off-by: Gregory Anders <ganders@cloudflare.com>
The DeviceManagerError type is among the types missing the Error trait
so far. To streamline the code and to simplify usage on higher levels
of this error type, this type now implements Display and Error.
As not all variant values are Error yet, `#[source]` is not everywhere
where it could be. This is done in the next commits.
The high level goal is: Enable future work to improve the error output
of cloud hypervisor.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
The new locking behavior uncovered that unfortunate test situation:
Many tests running in parallel access the same disk image with
rw permissions. Luckily, none of the tests actually writes to
the disk. Therefore, we can set it to readonly=true. In case this
changes, the test needs to be moved to the sequential test module.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
# What
This commit introduces file-based advisory locking for the files backing
up the block devices by using the fcntl() syscall with OFD locks. The
per-open-file-descriptor (OFD) locks are more robust than traditional
POSIX locks (F_SETLK) as they are not tied to process IDs and avoid
common issues in multithreaded or multi-fd scenarios [1]. Therefore,
we don't use `std::fs::File::try_lock()`, which is backed by F_SETLKW.
The locking mechanism is aware of the `readonly` property and allows
`n` readers or `1` writer (exclusive mode).
As the locks are advisory, multiple cloud-hypervisor processes can
prevent themselves from writing to the same file. However, this is not
a system-wide file-system level locking mechanism preventing to open()
a file.
The introduced new locking mechanism does not cover vhost-user devices.
# Why
To prevent misconfiguration and improve safety, it is good practice to
protect disk image files with a locking mechanism. Experience and common
best practices suggest that advisory locks are preferable over mandatory
locks due to better compatibility and fewer pitfalls (in fs space).
The introduced functionality is aligned with the approach taken by
QEMU [0], and is also recommended in [1].
# Implementation Details
We need to ensure that not only normal operation keeps working but also
state save/resume and live-migration. Especially for live migration,
it is crucial that the sender VMM releases the locks when the VM stops
so the receiver VMM can acquire them right after that.
Therefore, the locking and releasing happen directly on the block
device struct. The device manager knows all block devices and can
forward requests to these types.
Last but not least, this commit uses on explicit lock acquiring
but implicit lock releasing (FD close). It only explicitly releases
the locks where this integrates more smoothly into the existing
code.
# Testing
I tested
- normal operation
- state save/resume,
- device hot plugging,
- and live-migration
with read/shared and write/exclusive locks.
One can use the `fcntl-tool` to test if locks are actually acquired
or released [2].
# Links
[0] 825b96dbce/util/osdep.c (L266)
[1] https://apenwarr.ca/log/20101213
[2] https://crates.io/crates/fcntl-tool
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
To support PCIe P2P between VFIO devices, we populate IOMMU mappings for
the non-emulated MMIO regions of all VFIO devices via
`VFIO_IOMMU_MAP_DMA` (f0c1f8d), but the patch did not properly update
the IOMMU mappings with BAR reprogramming.
Fixes: #7027
Signed-off-by: Bo Chen <bchen@crusoe.ai>
The way how we handle PCI configuration space for vfio and vfio-user
devices are different from the rest of PCI devices. Besides accesses to
BAR registers (trapped to access the shadowing PCI config space we
maintained), accesses to other registers (including the COMMAND
register) are handled directly by the underline vfio or vfio-user
device.
This patch adds the proper handling of pending BAR reprogramming for
vfio and vfio-user devices.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
The Memory Space Enable (MSE) bit from the COMMAND register in the
PCI configuration space controls whether a PCI device responds to memory
space accesses, e.g. read and write cycles to the device MMIO regions
defined by its BARs. The MSE bit is used by the device drivers to ensure
the correctness of BAR reprogramming. A common workflow is, the driver
first clears the MSE bit, then writes new values to the BAR registers,
and finally set the MSE bit to finish the BAR reprogramming.
This patch changes how we handle BAR reprogramming for all PCI
devices (e.g. virtio-pci, vfio, vfio-user, etc.), so that we follow the
same convention, e.g. moving PCI BARs when its MSE bit is set.
Note that some device drivers (such as edk2) only clear and set MSE once
while reprogramming multiple BARs of a single device. To support such
behavior, this patch adds support for multiple pending BAR reprogramming.
See: https://github.com/cloud-hypervisor/cloud-hypervisor/issues/7027#issuecomment-2853642959
Signed-off-by: Bo Chen <bchen@crusoe.ai>
A BAR reprogramming of a PCI device will only happen when the (guest)
kernel write to its PCI config space, e.g. the detection of bar
reprogramming (`detect_bar_repgraomming()`) can be embedded to the PCI
config space write (`write_config_register()`). It simplifies APIs
exposed by the `struct PciConfiguration` and `trait PciDevice`. It also
prepares for easier handling of pending bar reprogramming when the MSE
bit of the COMMAND register is not enabled at the time of changing BAR
registers.
See: https://github.com/cloud-hypervisor/cloud-hypervisor/issues/7027#issuecomment-2853642959
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Fixing the following clippy issue using `cargo clippy --fix`:
error: variables can be used directly in the `format!` string
--> build.rs:25:27
|
25 | version.push_str(&format!("-{}", extra_version));
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#uninlined_format_args
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Currently a lot of functions are stubbed out with unimplemented feature
tag. Add the missing implementation to successfully boot ARM64 guests on
MSHV.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Num SPI and Base SPI nodes should be added before the end node to be
included in the device tree generation.
Fixes: eac44e6 (arch: Extend FDT for GICv2M device for ARM64 on MSHV)
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
It seems like address allocation has been spread into different files
and different location for x86 vs ARM. This makes it hard to follow the
code. Thus, unify it a single location which satisfies all the
requirement.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
It just simplifies code and improves the code read-ability without much
affecting the boot performance.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This will satisfy the requirement of MSHV i.e., setting the GICD base
address before initializing the VM.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
For MSHV guests, we would need interrupt controller to be initialized
before the VM gets initialized. This is because we are registering the
base address of GIC distributor with MSHV as part of interrupt
controller initialization workflow. And MSHV mandates that this property
is set before we initialize the VM.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
For ARM64 guests we need to set three important partition property:
1) PPI interrupt ID for timer interrupt
2) PPI interrupt ID for PMU interrupts.
3) Hiding LPI support from the guest because MSHV does emulate ITS for
the guest.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Since this would be used in other places inside the hypervisor and
hypervisor crate cannot take a dependency on arch crate, as that creates
cyclic dependency.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Currently PPI interrupt ID are hardcoded as numbers, it would be ideal
to define them as constants and could be reused in other parts of the
hypervisor crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Along with also bump the vfio-bindings crates to use the latest
mshv-bindings.
There is a breaking change in the new mshv crate which requires an
additional step to initialize vm after creating it.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
On x86 MSHV guests only used to support MSI based interrupts via IOAPIC
but ARM64 guests uses legacy interrupt for its functioning. Thus, extend
the logic to create routing entry to support legacy interrupts for ARM64
guests on MSHV.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
For MSHV arm64 guest, there is an in-hypervisor GICv2M emulation and for
that to work, it needs to be enlightened with the base address of GIC
redistributor exposed to guest via FDT.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
GICv2M requires two additional properties to be exposed via FDT:
1) Base SPI number and 2) Total number of SPIs. SPIs in general starts
from 32 and goes upto 1019. But currently we are limiting the range to
96 as that should be good enough for any normal Linux guest to function.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Since there are not a wide availability of RISC-V boards available at
the moment, it is easiest to develop with a QEMU virtual machine. I had
a hard time setting one up, but with the assistance of Ruoqing, I was
able to get one running. These are the steps I took to do so.
Signed-off-by: abm-77 <andrewmiller77@protonmail.com>
Ubuntu 24.04 LTS (Noble Numbat) is tested against cloud-hypervisor v45.0
release and it is working, document status in README.md.
Signed-off-by: Bingxin Li <bl497@cam.ac.uk>
ARM64 system register constants are not 1:1 mapped to MSHV definition of
those registers so we need a small helper function to translate that
mapping before retrieving those system registers.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
There are other potential users of these registers definitions in the
hypervisor crate. And hypervisor crate cannot use definitions from arch
crate because it creates cyclic dependency.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
In order to be on-pair with what's we're using from micro-http, let's
also add the proper status code here as well (as it will be used by
`ch-remote`).
Signed-off-by: Fabiano Fidêncio <fidencio@northflank.com>
Although the CPU manager gives us a quite descriptive error, on the
application side (the part calling Cloud Hypervisor) we have absolutely
no way to distinguish such error from any other error that may happen
when resizing a VM.
With this in mind, let's be more specific and return a TooManyRequests
(429) error, allowing the caller to have a chance to decide whether they
want to retry the operation or not.
https://datatracker.ietf.org/doc/html/rfc6585#section-4
Signed-off-by: Fabiano Fidêncio <fidencio@northflank.com>
As the coming patches in this series will take advantage of a status
code that was recently added there.
Signed-off-by: Fabiano Fidêncio <fidencio@northflank.com>
New MSHV version updates the get_msr_list output as
vector instead of fam-wrapper. It avoids unnecessary
conversions.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Latest mshv crates contains some IOCTL changes that
enhances VM creation and configures the features in correct
way. Also adds some features that improves register access.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
MSHV allows VMM to map the VP register page into root.
This feature helps VMM to faster process most of the frequent
used registers. This patch uses the VP register page for port
handling in CPU run method.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
As part of this configure the program counter, pstate and X0 registers.
Program counter will point to the start address of the kernel/firmware
in the guest memory. X0 will point to start of the FDT.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
As part of this configuration, two things are being done:
1. Setting up the base address of GIC Distributor
2. Setting up the base address of GIC Interrupt Translator
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Initial PSTATE value would be same for both KVM and MSHV. Thus, move it
to common register definition pool.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
There are a lot of internal functions that are not and probably should
not be called from other places.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
If a device returns 0xff as a capability pointer bad things happen.
The code before the previous commits would crash in debug builds due
to integer overflow. With the two lowest bits masked out, it sends the
code into an endless loop.
Be more robust by at least handling the case where the capability
appears to point to itself.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
The PCI standard mandates that the lower bits of the capability
pointer are masked out before using the pointer. See PCI Local Bus
Specification 3.0 Chapter 6.7 "Capabilities List".
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Currently, the code tries to follow the PCI capabilities list in
offset 0x34 in the config space regardless of whether the status
registers says this is valid. Fix by adding the appropriate check.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
OSDev has cranked up its bot protection. The following link works for
me locally after clicking the "I'm a human" button. I guess the CI
fails this check...
Without this exception the CI fails the link check stage:
* [403] [https://wiki.osdev.org/IOAPIC](https://wiki.osdev.org/IOAPIC) | Network error: Forbidden
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Use the context from Unmapped Gpa exit from the hypervisor to initialize
the MshvEmulatorContext and later call the emulator to decode the
instruction.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Currently it would be using the syndrome register for instruction
decoding which is what KVM has been using in-kernel to decode
instructions for ARM64 guests. In future, it could be extended with an
actual instruction emulator if required. But most Linux guests works
well with the instruction decoder using syndrome register.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This helps in implementing an instruction decoder for MSVH which does
not support in-kernel instruction decoding like KVM.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Some tests are specifically designed for KVM hypervisor platform. Thus,
guard them using appropriate feature flags.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
There are a bunch of unused variables as of now on the MSHV side and
compiler warns about them. Thus, mark them as unused for the time being.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
MSHV does not emulate a GICv3-ITS for guests to support MSI interrupts,
instead it exposes a GICv2m device. Currently adding a skeleton code
which would be modified later on with complete implementation.
With this we can start compiling cloud-hypervisor for MSHV on ARM64.
This will make sure that we don't regress in future in terms of basic
compilation test.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
`zerocopy` is bumped to 0.8.x after 0.3.4 of igvm crates, bump to rev
d062818 to capture `zerocopy` upgrade, but we should bump to 0.3.5
later.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Manually bump zerocopy to 0.8.24 since our dependabot could not perform
the upgrade properly.
Manually bump acpi_tabls as well since it's depending on zerocopy.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The dependency `bitfield-struct` 0.10.x of `igvm` 0.3.5 requires MSRV
1.83.0, bump to catch up.
Update image to 20250412-0 because MSRV in Dockerfile is updated.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
This aligns with our CLI syntax. The correctness of `DiskConfig` will be
ensured via `VmConfig::validate()`, e.g. `path` and `socket` are
mutually exclusive.
Fixes: #7016
Signed-off-by: Bo Chen <bchen@crusoe.ai>
Use the builtin function instead of using `==` operator.
Warning from the beta compiler:
error: use `std::ptr::eq` when comparing raw pointers
--> pci/src/vfio.rs:1616:24
if host_addr == libc::MAP_FAILED {
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
help: try: `std::ptr::eq(host_addr, libc::MAP_FAILED)`
= help: for further information visit
= https://rust-lang.github.io/rust-clippy/master/index.html#ptr_eq
= note: `-D clippy::ptr-eq` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::ptr_eq)]`
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Use the builtin function to improve the readability of the code.
Warning from beta compiler:
error: manual arithmetic check found
--> vm-allocator/src/address.rs:151:30
|
|let adjust = if alignment > 1 { alignment - 1 } else { 0 };
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
| help: replace it with: `alignment.saturating_sub(1)`
|
= help: for further information visit
https://rust-lang.github.io/rust-clippy/master/index.html#implicit_saturating_sub
= note: `-D clippy::implicit-saturating-sub` implied by `-D warnings`
= help: to override `-D warnings` add`#[allow(clippy::implicit_saturating_sub)]`
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Rust has a new way of constructing other error and clippy complains if
we are still using the older way to construct error message. Thus,
migrate to the new approach suggested by the clippy.
Warning from beta compiler:
error: this can be `std::io::Error::other(_)`
--> block/src/vhdx/mod.rs:142:17
|
| / std::io::Error::new(
| | std::io::ErrorKind::Other,
| | format!("Failed to update VHDx header: {e}"),
| | )
| |_________________^
|
= help: for further information visit
https://rust-lang.github.io/rust-clippy/master/index.html#io_other_error
help: use `std::io::Error::other`
std::io::Error::other(
format!("Failed to update VHDx header: {e}"),
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
`serde_json` crate is referenced by multiple components, centralize it
to workspace to better manage this crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
KVM supports GICv3-ITS emulation and the current GicState is modelled
around the KVM implementation. We should refactor this to accomodate
other hypervisor requirements. For example, MSHV only support GICv2M
emulation for guests for delivering MSI interrupts instead of GICv3-ITS.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Add `riscv.md` to guide developers/users on preparing guest kernel,
image, and direct boot VM on riscv64 platforms; document support status
and known limitations.
Co-authored-by: Bo Chen <bchen@crusoe.ai>
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The larger the VM memory, the greater the memory pressure, and the
greater the stop_dirty_log() overhead. Moving stop_dirty_log() outside
the downtime period can reduce downtime.
Signed-off-by: Jinrong Liang <cloudliang@tencent.com>
`riscv64_set_one_reg_to_vcpu` macro is used to set value of specific
RISC-V `$reg_name` stored in `state` to KVM Vcpu.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`riscv64_get_one_reg_from_vcpu` macro is used to extract RISC-V register
data from KVM Vcpu according to `$reg_name` provided to `state`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`igvm` crates are referenced by multiple components, centralize them to
workspace to better manage those crates.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Certain MSHV ioctls are only available on x86 architecture. Thus,
conditionally compile seccomp filter for x86 and in general enable
seccomp filters when compiling for MSHV on aarch64.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The CLI has grown to a big variety of options. clap prints them in the
help message (--help) in the order they were defined. We now are at a
point where grouping things logically together doesn't work well.
Further, there is no support by clap for logical grouping and the
current code base wasn't consistent. Therefore, this commit introduces
two changes:
- a new structure to define arguments (all in an array)
- an alphabetical ordering of the arguments
No other changes have been made. No options have been altered.
This significantly improves:
- code maintainability and extensibility
- readability of the --help output
A unit test ensures they stay sorted. A better approach to check if the
list of arguments (known at build time) is sorted would be a compile
time check (`const`), but this currently isn't possible in stable Rust.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
On-behalf-of: SAP philipp.schuster@sap.com
Use the definitions from the rust-vmm/mshv crate for various
datastructures such as StandardRegisters, RegList, VcpuInit etc.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
We are moving the metrics and rate-limiter workers to a bare-metal
system, so that we can have more consistent results (particularly for
the block device metrics)
Signed-off-by: Bo Chen <bchen@crusoe.ai>
v0.5.0 of `seccompiler` has supported riscv64, let's bump from v0.4.0 to
v0.5.0 to capture that support.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
In addition to the BDF information on the guest, reporting the device
path of the VFIO device on the host is more useful when it comes to
debug DMA map/unmap errors particularly ones caused by failing hardware
on the host.
Signed-off-by: Bo Chen <bchen@crusoe.ai>
`test_snapshot_restore_with_fd` uses unsafe file descriptors and with
rust 1.82.0 it errors with:
```
fatal runtime error: IO Safety violation: owned file descriptor already
closed
```
so has been skipped for now.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
We are having complains from Rust 1.86.0-beta.1 (f0cb41030 2025-02-17)'
clippy, which suggests us to replace `repeat().take()` with
`repeat_n()`. While `repeat_n()` is stablized in Rust 1.82.0.
Update image to 20250307-2 because MSRV in Dockerfile is updated.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
While executing `rustup toolchain add --profile minimal
1.82.0-x86_64-unknown-linux-musl`, it errors out with:
```
error: toolchain '1.82.0-x86_64-unknown-linux-musl' may not be able to
run on this system
note: add the `--force-non-host` flag to install the toolchain anyway
```
Add `--force-non-host` to install minimal musl toolchain for x86_64.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Currently we are just storing the StandardRegisters in the Vcpu state
which would be required for saving and restoring the ARM64 guest on
MSHV.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The reason `test_vfio_user` fails is as @likebreath pointed: our ARM
host does not support SVE, while the nvme_tgt binary built from the
container image requires it. As a result, we encountered a SIGILL when
running the nvme_tgt binary. This also explains why this is not
happening when the container is built on the same host itself.
And quote from @rbradford:
When a job is run on one of the workers it looks to see if there is a
container locally matching the name as specified in the dev_cli.sh
script - if there is then it uses it. Otherwise it will try and download
it from the container registry - if that fails then it will built
locally. For the x86-64 workers started dynamically it will never have a
local version as they are a fresh VM. But on the ARM64 builder is a
local container image cache.
This can lead to an issue where if the image is build with one version
(a handcrafted datestamp) and then the Dockerfile is changed without
changing the timestamp then an old version may be fetched from the cache
or server. It is there for essential to always bump the datestamp (there
is a number after the - that can be used for this.)
However there is also the added complexity that image that is build and
uploaded to the container registry is not the same as the built locally
and thus used for the initial testing of the Dockerfile change. This
leads to the issue we have seen where different CPU compiler flags (from
-march=native) from the QEMU cross build in the hosted GHA action and
the local ARM64 build. Resulting in a binary in the remotely built
container not working locally.
We end up specifying TARGET_ARCHITECTURE="armv8.2-a" for building spdk,
and put built `python/spdk/` folder into `/usr/local/bin/spdk-nvme`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`arm64` build in ubuntu:22.04 errors out with `error processing package
libc-bin`. This issue is a known issue between the binfmt (running
different architectures via QEMU) and the libc ldconfig binary running
in container. We're "suddenly" having issues as ubuntu-latest (which is
the OS version we run the GH action container with) was recently changed
from 22.04 to 24.04 and hence why upgrading the container userspace from
22.04 to 24.04 solves the problem.
Removed deprecated package `python3-distutils`.
Update image name from `20250111-0` to `20250222-0`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Currently a bunch of KVM specific interfaces are leaked into the vmm
crate which should ideally does not contain any hypervisor specific data
structures.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Use `lychee` to check availability of links in cloud-hypervisor.
The urls explictly excluded in config file are manually checked.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
We have lost track to releases before v27.0 since these projects no
longer exists. Delete links to those projects.
Update links to a detailed view specific to each group of release.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Previous repo hosts details of `Rust Style` is removed, point `Rust
Style` to `style-guide` in `rust-lang/rust` repo.
Link to provide illustration on `signed-off-by language` is also
removed, use a snapshot found in web archive [1] instead.
Format `CONTRIBUTING.md` using `mdformat` with GitHub Flavored Markdown
(GFM).
[1] https://web.archive.org
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
TDX homepage was moved to elsewhere, and `sgx` support is upstreamed
since v5.11 kernel.
Format `intel_sgx.md` using `mdformat` with GitHub Flavored Markdown
(GFM).
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
TDX homepage was moved to elsewhere, and `tdx-tools` repo was removed.
Provide a valid link of TDX homepage and change all reference to
`tdx-tools` to `tdx-linux`.
Format `intel_tdx.md` using `mdformat` with GitHub Flavored Markdown
(GFM).
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
There is a link referencing `rate-limiter` module of `firecracker`, but
that module no longer exsits.
Point the link to a commit with the same date in `firecracker` when this
commit was merged to `cloud-hypervisor`.
Format `io_throttling.md` using `mdformat` with GitHub Flavored Markdown
(GFM).
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`docs/arm64.md` was removed and splited into `README.md`, `building.md`
and `uefi.md` in #4991.
Let's point it to
`8ab15b9a98/docs/arm64.md`
the commit right before `docs/arm64.md` was removed in main branch.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Previous link `docs/logging` is not valid, replacing `docs/logging` with
`docs/logging.md`.
Format `logging.md` using `mdformat` with GitHub Flavored Markdown
(GFM).
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Use this to add any extra versioning information to the binary. It is
useful when packaging Cloud Hypervisor.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This is a typo here, `interrupt-controller` attribute of each CPU node
should not have a `#` prepended.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Replace the use of a reference kernel configuration file from this
repository with the use of a defconfig from the linux fork.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The IMSIC attr of RISC-V AIA is wrongly configured to start from 0, which
would error out with `os error 22` (invalid argument).
```console
Error booting VM: VmBoot(DeviceManager(CreateInterruptController(CreateAia(CreateVaia(Vaia error SetDeviceAttribute(SetDeviceAttribute(Invalid argument (os error 22))))))))
```
`riscv_imsic_attr_of` should shift `cpu_index` by 1 here to produce
correct IMSIC attr.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
For more control over updating the guest kernel use a fixed tag name
rather than fetching the latest.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Currently, Cloud Hypervisor does not set a VIRTIO_IOMMU_F_INPUT_RANGE
feature bit for the VirtIO IOMMU device, which, according to spec[1],
means that the guest may use the whole 64-bit address space is for
IOMMU purposes:
>If the feature is not offered, virtual mappings span over the whole
>64-bit address space (start = 0, end = 0xffffffff ffffffff)
As far as I am aware, there are currently no host platforms on
the market capable of addressing the whole 64-bit address space.
For example, I am currently working with a host platform that reports
39-bit address space for IOMMU purposes:
>DMAR: Host address width 39
When running a VFIO pass-through guest on such a platform, NVIDIA
driver in guest gets DMA mapping failures when working with large data,
and this results in Cloud Hypervisor exiting with the following error:
>cloud-hypervisor: 1501.220535s: <__iommu>
>ERROR:virtio-devices/src/thread_helper.rs:53 -- Error running worker:
>HandleEvent(Failed to process request queue : ExternalMapping(Custom
>{ kind: Other, error: "failed to map memory for VFIO container, iova
>0x7fff00000000, gpa 0x24ce25000, size 0x1000: IommuDmaMap(Error(22))"
>}))
Passing "--platform iommu_address_width=39" to Cloud Hypervisor built
with this change fixes this.
[1]: https://docs.oasis-open.org/virtio/virtio/v1.3/csd01/
virtio-v1.3-csd01.html#x1-5420006
Signed-off-by: Nikolay Edigaryev <edigaryev@gmail.com>
When VM restore fails, the VMM state is left with some side-effects,
such as a VM being created. It would prevent the VMM from creating and
booting a new VM or restoring from a VM snapshot.
To fix this issue, this patch explicitly handles the side effects to the
VMM state when VM restore fails, e.g. clear the VmConfig and shutdown
the VM being created.
Fixes: #6869
Signed-off-by: Bo Chen <bo.arvin.chen@gmail.com>
With VM restore, the VMM is always re-creating a VM based on the
restored `VmConfig`. We should always re-generate the 'console_info'
from the `Vmm` struct to stay consistent with the new VM being created.
Signed-off-by: Bo Chen <bo.arvin.chen@gmail.com>
Instead of reinventing this mock infrastructure in the upcoming fuzzer,
reuse the one that is already available.
However this change makes Clippy complain that TestBackend and
TestContext don't implement Default. This is just test code, we can
suppress Clippy in this case.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Updating the kernel to v6.12 has shown up a flaw in the workflow for our
binary kernel releases. The CI job that builds the binary kernel in the
cloud-hypervisor/linux repository fetches the config from the main
branch of the cloud-hypervisor/cloud-hypervisor repository. However the
CI job to update the kernel version to use is in the cloud-hypervisor
repository.
As a workaround - update the kernel config and version in the
cloud-hypervisor repository to point to v6.12 and use the ability to
build the kernel during the CI run. Once merged to main a new release
can be made in the linux respository which will build a binary asset
using the new config. After that release the CI jobs on the
cloud-hypervisor repository can changed back to using the binary kernel
assets.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Remove an erroneous optimisation that used the page size mask to reduce
the range to iterate through on the set of mappings. This doesn't work
as the virtio-iommu ranges are larger than a single page. This may have
worked in the past when the mappings were limited to a single page.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
No change to the Dockerfile but I observed that the 20251022-0 image was
not available in the repository.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
MSHV's SEV-SNP implementation calls ioeventfds whenever there is an
event.
This change removes the need frequent allocation and deallocation of a
vector, while at the same time makes sure other call sites are
unaffected.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This avoids ambiguity of parameters:
error: ambiguous reference to positional arguments by number in a tuple variant; change this to a named argument
--> block/src/qcow/mod.rs:48:48
|
48 | #[error("File larger than max of {}: {0}", MAX_QCOW_FILE_SIZE)]
| ^^^^^^^^^^^^^^^^^^
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This will become useful when we build the fuzzing target for the
instruction emulator, because there is no need to pull in the rest of
the hypervisor crate in that situation.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The fastfmt feature and VEX support use techniques that appear to leak
memory in the eye of LLVM's address sanitizer.
While at it, disable a bunch of instruction set decoding support we
never intend to support.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The size was set to one because without VIRTIO_BLK_F_SEG_MAX, the guest
only used one data descriptor per request.
The value 32 is empirically derived from booting a guest. This value
eliminates all SmallVec allocations observable by DHAT.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This allows the guest to put in more than one segment per request. It
can improve the throughput of the system.
Introduce a new check to make sure the queue size configured by the user
is large enough to hold at least one segment.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
When the main fuzzer function returns (), it is equivalent to
returning Corpus::Keep.
In some of the return paths, we want to reject the input so that the
libfuzzer won't spend more time mutating them.
The should make fuzzing more efficient. No functional change intended.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The checksum field in the original buffer should be zeroed.
The code was zeroing a temporary buffer. That's wrong.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The original code was buggy. It always attempted to update the header,
even when the file was opened as read-only. That led to an error.
The specification states that the headers should be updated when the
first user visible write happens. We can just drop the incorrect code.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The ASYNC flag forces requests to go to worker threads. Worker threads
are expensive. Let the kernel decide what to do.
With this change, I no longer see an excessive amount of io_uring worker
threads.
Quote from the manual for io_uring_sqe_set_flags(3):
```
IOSQE_ASYNC
Normal operation for io_uring is to try and issue an sqe
as non-blocking first, and if that fails, execute it in an
async manner. To support more efficient overlapped
operation of requests that the application knows/assumes
will always (or most of the time) block, the application
can ask for an sqe to be issued async from the start. Note
that this flag immediately causes the SQE to be offloaded
to an async helper thread with no initial non-blocking
attempt. This may be less efficient and should not be
used liberally or without understanding the performance
and efficiency tradeoffs.
```
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Instead of silently ignoring the error, return an error to the callers.
This in practice should never happen, because the submission queue size
(ring depth) is the same as the virtio queue size. Virtio queue won't
push more requests than there are submission queue entries.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The original code relied on the default `read_vectored` or
`write_vectored` implementations from the standard library.
The default implementation of those functions only uses the first
non-empty buffer. That's not correct when there are more than one
buffers.
Fixes: #6876
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This system is erroring out on jobs due to insufficient memory - reduce
parallelism to allow CI jobs to complete.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Updated the live migration documentation to include instructions for
performing cross-host live migrations over TCP connections.
Signed-off-by: Jinrong Liang <cloudliang@tencent.com>
Add support for cross-host live migration over TCP, which
significantly improves performance in cross-host live migration
scenarios compared to the Unix socket forwarding method.
Signed-off-by: Jinrong Liang <cloudliang@tencent.com>
Bump vm-memory from 0.16.0 to 0.16.1 to include the implementations of
ReadVolatile and WriteVolatile for TcpStream.
Signed-off-by: Jinrong Liang <cloudliang@tencent.com>
While non-Intel CPU architectures don't have a special concept of IO
address space, support for PCI I/O regions is still needed to be able
to handle PCI devices that use them.
With this change, I'm able to pass through an e1000e device from QEMU
to a cloud-hypervisor VM on aarch64 and use it in the cloud-hypervisor
guest. Previously, it would hit the unimplemented!().
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Add a bitmap to MshvVM struct for caching the pages
that the VMM got shared access from the guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Pass memory size from vmm crate to hypervisor VM
creation API so use for creating Bitmap for SEV-SNP
guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
For SEV-SNP VM on MSHV we need to request page access during
IO, we want to avoid such request for the page that have already
been requested. In order to maintain the bitmap we need the memory size
during bitmap creation.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
VMM needs to handle VMG exit for guest request. This patch
removes the old method that uses gpa_read/write(IOCTL and hypercall),
which is expensive and update the GHCB page using
mapped(root partition) struct.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
VMM needs to handle VMG exit for MMIO. This patch
removes the old method that uses gpa_read/write(IOCTL and hypercall),
which is expensive and update the GHCB page using
mapped(root partition) struct.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
VMM needs to handle VMG exit for IO Port. This patch
removes the old method that uses gpa_read/write(IOCTL and hypercall),
which is expensive and update the GHCB page using
mapped(root partition) struct.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
VMM needs to handle VMG exit for doorbell page. This patch
removes the old method that uses gpa_read/write(IOCTL and hypercall),
which is expensive and update the GHCB page using
mapped(root partition) struct.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
After handling the VMG exit vmm needs to clear the exitinfo1
into the GHCB page. This patch replaces the old
method(gpa_write) and clear the exitinfo1 using mapped GHCB struct.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
For confidential VM on MSHV, GHCB page is the communication
method between guest and host. All the CVM exits, VMM
needs to read and write to the GHCB page. MSHV provides
an option to remap the page to the root partition. This
way VMM could directly read and write to the page and
skip extra IOCTL and hypercall. This improvement makes
the IO 10% faster.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
TPM support on RISC-V is not scheduled in this stage, we will come back
when it's ready and we have workloads to verify it on RISC-V.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Currently `devices` module cannot be solely built, by specifying
`arch/kvm` to turn on the features required for its dependency - `arch`
crate to build. Thus enabling `devices` crate to be built and tested
with command:
```sh
cargo build -p devices --features kvm
```
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`arch` module are required for devices module to build. Enabling `arch`
module in target specific dependencies section.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Provide implementation for GSI allocator to work on riscv64
architecture, and doc comment for riscv64 as well.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Currently `vm-allocator` module cannot be solely built, by add
`features` section and specifying `arch/kvm` to turn on the features
required for its dependency - `arch` module to build. Thus enabling
`vm-allocator` crate to be built and tested with command:
```sh
cargo build -p vm-allocator --features kvm
```
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`arch` module are required to provide definitions for riscv64 system irq
and memroy allocation, enabling `arch` module in target specific
dependencies section.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Integrate fdt setup and layout for `riscv64` to `arch` crate, provide
definitions of `riscv64` platform specific error and reuse
`MmioDeviceInfo`, `PciSpaceInfo` structures and `DeviceInfoForFdt` impl
block for `riscv64`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Reused layout set for `aarch64` as much as possible to provide a concise
and compatible (for devices) layout for RISC-V 64-bit.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`fdt_parser` and `vm-fdt` crates are required to setup fdt used to boot
riscv64 virtual machines, enabling the two crates in target specific
dependencies section.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
It turns out we need to setup `a0`, `pc` and `a1` for all vcpus before
we run them, remove predicates used to set `pc` and `a1` for `vcpu0`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Previous `Vaia` implementation uses types as it is, thus the property
string generated for fdt setup requires additional type conversion.
Change the types used in the methods of `Vaia` trait to provide a
concise fdt setup process.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The rate-limiter worker was moved to use small Azure VMs (#6731) and now
requires more time to complete.
Increasing its timeout to stablize this worker.
Signed-off-by: Bo Chen <bo.arvin.chen@gmail.com>
With 0.5.0 release of `rust-hypervisor-firmware`, `aarch64` binary were
added to assets, which causes the `FW_URL` to have multiple download url
separated by a white space, thus our integration tests would fail.
Constrain `FW_URL` to `hypervisor-fw` to resolve this.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The previous docs were very developer centric and have led several
people to believe that cross-machine migration is not supported at
all.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
I've added newlines between paragraphs and code blocks for easier
reading. I've also changed the code blocks to use the correct
highlighting.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
Replace `map_or()` on false condition with `is_some_and` to provide
better readability, as suggestted by v1.84.0-beta.1 `cargo clippy`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Adapt the existing code to transparently the MemorySlotAllocator. The
MemoryManager is the canonical holder of the these values with them
turned into a MemoryAllocator on demand.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Add an allocator for KVM memory slots, a finite resource in the kernel
- these need to be manipulated across different crates with a common
dependency on this crate.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
`prepare_linux` is capable of determining whether we need to invoke
`build_custom_linux` for building linux from source or `download_linux`
for downloading pre-built.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`prepare_linux` checks if a `--build-guest-kernel` option is present,
and build kernel from `cloud-hypervisor/linux.git`. Otherwise, it will
invoke `download_linux` to use pre-built kernel.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
If any VM argument (e.g. --disk) is provided require some payload (e.g.
--kernel or --firmware) when parsing the command line arguments.
See: #6831
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Add unit-test to make sure get_regs and set_regs on riscv64 architecture
work as expected, effectively avoiding typos in register names.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Complement `create_standard_regs` implementation on RISC-V platform to
work with `From` trait of `kvm_riscv_core`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Incorporates riscv64 register interaction and AIA creation to kvm
module. Complete `Vcpu` trait on RISC-V platform.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Integrate `aia` module into `riscv64` module, and enable `riscv64`
module if target architecture is RISC-V 64-bit.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`fuzz` is a separate workspace, enable consistency check on rust-vmm
crates in `fuzz` to keep stuffs in sync with root workspace.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
`get_device_attr` in 0.19.0 kvm-ioctls is marked as unsafe, wrap
invokation of `get_device_attr` with `unsafe` block.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
- Bump kvm-bindings from 0.9.1 to 0.10.0
- Bump kvm-ioctls from 0.18.0 to 0.19.0
- Bump vm-memory from 0.15.0 to 0.16.0
- Bump linux-loader from 0.12.0 to 0.13.0
- Bump virtio-bindings from 0.2.1 to 0.2.4
- Bump virtio-queue from 0.13.0 to 0.14.0
- Pin mshv-bindings to 0.3.1
- Pin mshv-ioctls to 0.3.1
- Pin vhost to rev "d983ae0"
- Pin vhost-user-backend to rev "d983ae0"
Since vhost 0.12.0 and vhost-user-backend 0.16.0 are going to be yanked,
temporarily pin these two to "d983ae0", which are expected to be
replaced by 0.13.0 vhost and 0.17.0 vhost-user-backend after released.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Add `package-consistency-check.py` script to prevent #6809 and #6815
from happening. This script takes a string present in the repository
field of packages to identify pacakges from a specific source.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
This reverts commit b264aebcde.
Pulls in new vm-memory so should be done as part of a global bump of
Rust-VMM crates.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: the following explicit lifetimes could be elided: 'a
--> hypervisor/src/arch/x86/emulator/mod.rs:492:6
|
492 | impl<'a, T: CpuStateManager> Emulator<'a, T> {
| ^^ ^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_lifetimes
= note: `#[warn(clippy::needless_lifetimes)]` on by default
help: elide the lifetimes
|
492 - impl<'a, T: CpuStateManager> Emulator<'a, T> {
492 + impl<T: CpuStateManager> Emulator<'_, T> {
|
warning: the following explicit lifetimes could be elided: 'a
--> hypervisor/src/mshv/x86_64/emulator.rs:19:6
|
19 | impl<'a> MshvEmulatorContext<'a> {
| ^^ ^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_lifetimes
help: elide the lifetimes
|
19 - impl<'a> MshvEmulatorContext<'a> {
19 + impl MshvEmulatorContext<'_> {
|
warning: the following explicit lifetimes could be elided: 'a
--> hypervisor/src/mshv/x86_64/emulator.rs:65:6
|
65 | impl<'a> PlatformEmulator for MshvEmulatorContext<'a> {
| ^^ ^^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_lifetimes
help: elide the lifetimes
|
65 - impl<'a> PlatformEmulator for MshvEmulatorContext<'a> {
65 + impl PlatformEmulator for MshvEmulatorContext<'_> {
|
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
--> test_infra/src/lib.rs:1307:25
|
1307 | let child = self
| _________________________^
1308 | | .command
1309 | | .stderr(Stdio::piped())
1310 | | .stdout(Stdio::piped())
1311 | | .spawn()
1312 | | .unwrap();
| |_________________________^
|
= note: consider calling `.wait()`
= note: not doing so might leave behind zombie processes
= note: see https://doc.rust-lang.org/stable/std/process/struct.Child.html#warning
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#zombie_processes
= note: `#[warn(clippy::zombie_processes)]` on by default
The API caller should ensure that they call .wait() or equivalent on the
spawned child to reap it.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The snapshot/restore test didn't wait on the child being spawned:
warning: spawned process is never `wait()`ed on
--> performance-metrics/src/performance_tests.rs:495:25
|
495 | let mut child = GuestCommand::new(&guest)
| _________________________^
496 | | .args(["--api-socket", &api_socket_source])
497 | | .args([
498 | | "--cpus",
... |
507 | | .spawn()
508 | | .unwrap();
| |_____________________^
|
= note: consider calling `.wait()`
= note: not doing so might leave behind zombie processes
= note: see https://doc.rust-lang.org/stable/std/process/struct.Child.html#warning
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#zombie_processes
= note: `#[warn(clippy::zombie_processes)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
We know what we are doing within `tracer` crate, disable lint on these
shared reference to mutable static.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Functions marked as `pub` in test module need to be documented. Since
these are literally helper functions and there is no need to be `pub`,
remove these `pub` keywords.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
As clippy of rust-toolchain version 1.83.0-beta.1 suggests, remove
redundant `return` statement to keep code clean.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
As clippy of rust-toolchain version 1.83.0-beta.1 suggests, use `c""` to
construct nul-terminated `CStr`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
As clippy of rust-toolchain version 1.83.0-beta.1 suggests, remove
manual implementation of `is_power_of_two` to improve readability.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
As clippy of rust-toolchain version 1.83.0-beta.1 suggests, replace
manually implemented `div_round_up!` and the like with `div_ceil` from
std.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
This system has a lot of cores (80) resulting in all the tests being
spawned simultaneously and leading to exhaustion of the available
memory. Instead limit the number of threads.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
After updating the OS image to the latest jammy version this test now
requires more memory to run
Fixes: #6782
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When running with the newer jammy OS image error messages from Rust
Hypervisor Firmware about lack of ram when rebooting where observed.
Increase the RAM allocation to allow it to boot after a reboot.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Currently `arch` crate cannot be built, by specifying `hypervisor/kvm`
to turn on the features required for its dependency - `hypervisor` crate
to build. Thus enabling `arch` crate to be built with command:
```sh
cargo build -p arch --features kvm
```
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
AddressManager::move_bar() acquires the device_tree mutex.
The function is called from PciConfigIo::config_space_write()/
PciConfigMmio::config_space_write() while the pci_bus mutex
is acquired.
The functions DeviceManager::pci_resources()/eject_device()
acquire these mutexes in reverse order, which leads to a deadlock.
Fixes: #6775
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
This lint disallows asserttions on is_ok()/is_err() in favor of either
using unwrap (so that at least if the test fails, we the failure message
will contain the actual failure reason instead of just "was not
ok/err"), or actually matching the specific variant.
Inspired by and quoted from @roypat.
Enable this lint for `build.yaml` and `quality.yaml`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Asserting on .is_ok()/.is_err() leads to hard to debug failures (as if
the test fails, it will only say "assertion failed: false". We replace
these with `.unwrap()`, which also prints the exact error variant that
was unexpectedly encountered (we can to this these days thanks to
efforts to implement Display and Debug for our error types). If the
assert!((...).is_ok()) was followed by an .unwrap() anyway, we just drop
the assert.
Inspired by and quoted from @roypat.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
This is based on a revert of commit
ce49a6f4b8 with some modifications for a
different worker configuration.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Only download the kernel binaries from the github release if the remote
file is newer (avoids multiple copies accumulating in the download
directory.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Cargo's features are additive. Disabling default features in individual
packages but not on the workspace level makes no sense.
This fixes the several build warnings.
Fixes: 5a70d7ec69 (build: Centralize rust-vmm crates to workspace)
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Since cloning Option<Arc<T>> will clone for Arc<T>, this patch fixes the
follow warning:
warning: this call to `as_ref.map(...)` does nothing
--> vmm/src/lib.rs:872:13
|
872 | self.console_resize_pipe.as_ref().map(Arc::clone),
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: try: `self.console_resize_pipe.clone()`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#useless_asref
= note: `#[warn(clippy::useless_asref)]` on by default
Signed-off-by: Songqian Li <sionli@tencent.com>
This patch removes pub import vm_config in config.rs to eliminate
the ambiguity of vm_comfig reference.
Signed-off-by: Songqian Li <sionli@tencent.com>
By introducing `imports_granularity="Module"` format strategy,
effectively groups imports from the same module into one line or block,
improving maintainability and readability.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Enforcing group_imports="StdExternalCrate" requires using cargo fmt from
the nightly toolchain. Create a new workflow that runs on nightly to run
cargo fmt.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Historically the Cloud Hypervisor coding style has been to ensure that
all imports are ordered and placed in a single group. Unfortunately
cargo fmt has no support for ensuring that all imports are in a single
group so if whitespace lines were added as part of the import statements
then they would only be odered correctly in the group.
By adopting "group_imports="StdExternalCrate" we can enforce a style
where imports are placed in at most three groups for std, external
crates and the crate itself. Choosing a style enforceable by the tooling
reduces the reviewer burden.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Serial device doesnt support resize semantics. Setting up the
console resize pipe in the Serial device setup path, overwrites
the setup done as part of virtio-console.
Signed-off-by: BharatNarasimman <bharatn@microsoft.com>
This patch removes locks in VmCreate request and VmInfo response
since we needn't use a lock here and should ensure that internal
implementation is transparent to the runtime.
Signed-off-by: Songqian Li <sionli@tencent.com>
Killing process with SIGKILL will miss the information since CLH process
needs to end normally to export code coverage information.
Signed-off-by: Songqian Li <sionli@tencent.com>
Add release and target params to `cargo test` since we collect
the code coverage reports from `xx/$BUILD_TARGET/release/`.
Signed-off-by: Songqian Li <sionli@tencent.com>
The construction of `FilePair` in `virtio_devices` component has changed
in 287887c, wrapping the parameters with `Arc` to fix fuzz build.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Modify `Cargo.toml` in each member crate to follow the dependencies
specified in root `Cargo.toml` file.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
For a VM with virt-console enabled, when a reboot is requested, the
console devices are closed during the shutdown path. As part of this
the sigwinch listener process and the console resizer pipe are closed.
For the new incarnation of the VM, fresh set of console devices are
setup and a new console resizer pipe is created. The new VM should
be setup to use the newly created console devices including the console
resizer pipe.
Reading from the older console resizer pipe results in unexpected eof
error and terminates the cloud hypervisor process.
Signed-off-by: BharatNarasimman <bharatn@microsoft.com>
Rebooting a VM fails with the following error when debug assertions
are enabled:
fatal runtime error: IO Safety violation: owned file descriptor already closed
This happens because FromRawFd::from_raw_fd is used on RawFds stored
in ConsoleInfo every time a VM begins to boot, so the second
time (after a reboot, or if the first attempt to boot via the API
failed), the fd will be closed. Until this assertion is hit, the code
is operating on either closed file descriptors, or new file
descriptors for something completely different. If debug assertions
are disabled, it will just continue doing this with unpredictable
results.
To fix this, and prevent the problem reocurring, ownership of the
console file descriptors needs to be properly tracked, using Rust's
type system, so this commit refactors the console code to do that.
The file descriptors are now passed around with reference counts, so
they won't be closed prematurely. The obvious way to do this would be
to just have each member of ConsoleInfo be an Arc<File>, but we need
to accomodate that serial console file descriptors can also be
sockets. We can't just store an OwnedFd and convert it when it's
used, because we only get a reference from the Arc, so we need to
store the descriptors as their concrete types in an enum. Since this
basically duplicates the ConsoleOutputMode enum from the config, the
ConsoleOutputMode enum is now not used past constructing the
ConsoleInfo.
So that ownership can be represented consistently, the debug console's
tty mode now uses its own stdout descriptor.
I'm still using .try_clone().unwrap() (i.e. dup()) to clone file
descriptors for Endpoint::FilePair and Endpoint::TtyPair, because I
assume there's a reason for them not just to hold a single file
descriptor.
I've also retained the existing behaviour of having serial manager
ignore the tty file descriptor passed to it (which is stdout), and
instead using stdin. It looks a lot weirder now, because it has to
explicitly indicate it's ignoring the fd with an underscore binding.
Fixes: 52eebaf6 ("vmm: refactor DeviceManager to use console_info")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
The assignment of console_resize_pipe in the TTY case seems to have
been accidentally deleted. I've put it back, but since this is adding
code, I used the new safe API for checking whether a file is a
terminal, introduced in Rust 1.70.0. We should probably use that
everywhere, but that's out of scope of this bug fix.
Fixes: 52eebaf6 ("vmm: refactor DeviceManager to use console_info")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
I've moved this so that it's just after the enum definition, which
will hopefully make it less easy to miss if events are added/removed
again in future.
Fixes: 6d1077fc ("vmm: Unix socket backend for serial port")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Complete the removal of the DAX support by removing the use of
non-standard messages. These messages have since been removed from the
vhost_user crate (rust-vmm/vhost#246) and so need to be removed from our
implementation since that would otherwise block updating to a newer
version of the crate.
The ability to enable DAX support in Cloud Hypervisor has been disabled
some time ago but this code was residual with no way to enable it.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
And modify to code to use the updated interfaces.
Arguments for map_guest_memory, get_dirty_bitmap, vp.run(),
import_isolated_pages, modify_gpa_host_access have changed.
Update these to use the new interfaces, including new MSHV_*
definitions, and remove some redundant arguments.
Update seccomp IOCTLs to reflect interface changes.
Fix irq-related definitions naming.
Bump vfio-ioctls to support mshv v0.3.0.
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
`GetRegList` variant will be referenced on both Aarch and RISC-V. Fixed
comment to generalize this error variant.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Since RISC-V has its own definition of `CoreRegister`, expand the Aarch
variant to avoid collision of `HypervisorCpuError`.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The error message of `SetRegister` and `GetRegister` fail to describe
the reason as the error variant suggests. Fixed error message
accordingly.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
The rate-limiter worker now is running on Azure VMs whose performance
are more prone to be fluctuate, particularly on block performances. This
commit relaxes the limit check for group rate limiter tests to make them
less flaky.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Port I/O is only supported on x86_64 - use inverted conditional logic to
match the other architectures rather than calling them out specifically.
This will be relevant when RISC-V support is added.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The steps with predicates of `x86_64` targets would never turn on
`use-cross`, removing them from quality check.
Signed-off-by: Ruoqing He <heruoqing@iscas.ac.cn>
Bound checks for virtio-mem and ACPI memory hotplug are off by
one and two, respectively. This prevents users to fully use the reserved
memory hotplug size.
For ACPI, if we specific `--memory size=2G,hotplug_size=4G` and run
`ch-remote resize --memory 6G`, cloud-hypervisor will report the
following error because of the incorrect bound check:
`<vmm> ERROR:vmm/src/lib.rs:1631 -- Error when resizing VM:
MemoryManager(InsufficientHotplugRam)`
Similarly, for virtio-mem, cloud-hypervisor will fail the incorrect
bound check and abort the resize. The VM will see the following error
in dmesg:
`virtio_mem virtio3: unknown error, marking device broken: -22`
This patch has fixed both bound checks and ensure that users can
hot add memory up to the reserved hotplug size.
Signed-off-by: Yuhong Zhong <yz@cs.columbia.edu>
Since the underline bare-metal system is going to retire, we are moving
the metrics and rate-limiter workers to use Azure VMs.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Otherwise, you just see output like "Error running command: Server
responded with an error: InternalServerError", which isn't very
helpful. The response body used to be include with the message, but
was removed when the Error enum was converted to thiserror.
Fixes: 5d0d56f5 ("api_client: Use thiserror for errors")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
With cd0cdac ("virtio-devices: Fix seccomp rules for SevSnp guest"), the
sys_ioctl rule that was applied in virtio_thread_common, would override
previously specified sys_ioctl rules for individual thread type. This
causes the SevSnp guest to crash with seccomp violation.
Fixes: cd0cdac ("virtio-devices: Fix seccomp rules for SevSnp guest")
Signed-off-by: Purna Pavan Chandra <paekkaladevi@microsoft.com>
The Linux kernel fork repository for Cloud Hypervisor now produces
prebuilt x86-64 and aarch64 binaries. Speed up the CI by using those
binaries.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
An example warning output is:
error: first doc comment paragraph is too long
--> virtio-devices/src/lib.rs:158:1
|
158 | / /// Convert an absolute address into an address space (GuestMemory)
159 | | /// to a host pointer and verify that the provided size define a valid
160 | | /// range within a single memory region.
161 | | /// Return None if it is out of bounds or if addr+size overlaps a single region.
| |_
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#too_long_first_doc_paragraph
= note: `-D clippy::too-long-first-doc-paragraph` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::too_long_first_doc_paragraph)]`
Signed-off-by: Bo Chen <chen.bo@intel.com>
In case of SEV-SNP guests on MSHV, any host component including VMM
trying to access any guest memory region, needs to call gain_access API
to acquire access to that particular memory region. This applies to all
the virtqueue buffers as well which the VMM is using to use to perform
DMA into the guest and in order to facilitate device emulation.
While creating various virtqueues we are already using access platform
hook to translate the addresses but currently we are missing the size
arguments which would be required by the SevSnpAccessPlatformProxy to
gain access to the memory of these queues.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This would be used in case of SEV-SNP guest because we need to accquire
access to these virtqueue segments before accessing them.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
DeviceManager::add_virtio_console_device used to create the console
resize pipe and assign it to self.console_resize_pipe, but when this
was changed to use console_info, that was deleted without replacement.
This meant that, even though the console resize pipe was created by
pre_create_console_devices, the DeviceManager never found out about
it, so console resize didn't work (at least for pty consoles).
To fix this, the console resize pipe needs to be passed to the Vm
initializer, which is already supported, it was just previously not
used for new VMs.
Since DeviceManager already stores the console resize pipe in an Arc,
and Vmm also needs a copy of it, the sensible thing to do is change
DeviceManager::new to take Arc, and then we don't need to dup the file
descriptor, which could fail.
Fixes: 52eebaf6 ("vmm: refactor DeviceManager to use console_info")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
When the rate limit was reached it was possible for the notification to
the guest to be lost since the logic to handle the notification was
tightly coupled with processing the queue. The notification would
eventually be triggered when the rate limit pool was refilled but this
could add significant latency.
Address this by refactoring the code to separate processing queue and
signalling - the processing of the queue is suspended when the rate
limit is reached but the signalling will still be attempted if needed
(i.e. VIRTIO_F_EVENT_IDX is still considered.)
Signed-off-by: wuxinyue <wuxinyue.wxy@antgroup.com>
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Previously we didn't handle extended guest requests at all and always
returned an error. This lead to issues with some guests that expected
extended requests to succeed. Instead, handle extended requests like
normal requests and write zeros to the extended area to signal to the
guest that we don't want to supply any additional certificate data.
Signed-off-by: Tom Dohrmann <erbse.13@gmx.de>
In case of ARM64 kvm_regs are considered as StandardRegister which is no
longer required since we have defined architecture independent
StandardRegisters in hypervisor crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Use the StandardRegisters defined in the hypervisor crate instead of
re-defining it from MSHV/KVM crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This change is along the lines of x86 platform which defined similar
convinence macros to access the StandardRegisters.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
With commit 1e967697c ("vmm: pass AccessPlatform implementation for
SEV-SNP guest"), we started performing one additional ioctl to gain
access to the guest memory before accessing those regions inside
virtio-device emulation code path. This additional ioctl is not part of
the current seccomp filter, which is causing the SevSnp guest to crash
in this scenario with seccomp violation.
Fixes: 1e967697c ("vmm: pass AccessPlatform implementation for SEV-SNP
guest")
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This fixes a issue of running vm compiled in debug with Rust
1.80.0 or later, where this check was introduced.
Signed-off-by: Wenyu Huang <huangwenyuu@outlook.com>
Implement get_guest_debug_hw_bps() for mshv and simply return 0 for now.
This unblocks the usage of the crashdump feature with mshv. If left
unimplemented, Cloud Hypervisor built with mshv and guest_debug features
crashes immediately upon start due to unimplemented!() macro.
Signed-off-by: Anirudh Rayabharam <anrayabh@microsoft.com>
With this we are removing the CloudHypervisor definition of
StandardRegisters instead using an enum which contains different
variants of StandardRegisters coming from their bindigs crate.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Currently we are redefining StandardRegisters instead of using the ones
coming from bindings. With this we can remove the unnecessary
construction of global structure which contains fields from different
hypervisor dependent structs.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The index is derived from the access offset, so it is controlled by the
guest. Check it before accessing internal data structures.
Since Rust enforces strict bound check even in release builds, the VMM
process will crash if the guest misbehaves. There is no security issue
since the guest can only DoS itself.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Use the IOCTL numbers directly from mshv-ioctls instead of hardcoding
them in the seccomp filters.
Remove seccomp rules for unused ioctls:
MSHV_GET_VERSION_INFO,
MSHV_ASSERT_INTERRUPT.
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
CVM guests can configure GHCB page multiple times during it's
lifetime depending on it's requirement. For example a Linux CVM guest
configures a different GHCB page during compressed kernel boot and sets
up a new one after decompressing the kernel. As a cleanup step, VMM
should unset the previous GHCB page before registering a new one for
a particular vcpu thread.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
DeviceFd::get_device_attr should be marked as unsafe, because it
allows writing to an arbitrary address. I have opened a kvm-ioctls
PR[1] to fix this. The hypervisor crate was using the function
unsafely by passing it addresses of immutable variables. I noticed
this because an optimisation change[2] in Rust 1.80.0 caused the
kvm::aarch64::gic::tests::test_get_set_icc_regs test to start failing
when built in release mode.
To fix this, I've broken up the _access functions into _set and _get
variants, with the _get variant using a pointer to a mutable variable.
This has the side effect of making these functions a bit nicer to use,
because the caller now has no need to use references at all, for
either getting or setting.
[1]: https://github.com/rust-vmm/kvm-ioctls/pull/273
[2]: d2d24e395a
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This fixes the vsock::device::tests::test_virtio_device test with Rust
1.80.0 or later, where this check was introduced.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Passing AccessPlatform trait to virtio-device for requesting
restricting page access during IO for SEV-SNP guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Implement AccessPlatform for SEV-SNP guest to access
restricted page using IO. VMM calls MSHV api to get access
of the pages, MSHV requests guest to release the access.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add a structure to hold the reference of the Vm trait
from Hypervisor crate to access of restricted page
from SEV-SNP guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Pvmemcontrol provides a way for the guest to control its physical memory
properties, and enables optimizations and security features. For
example, the guest can provide information to the host where parts of a
hugepage may be unbacked, or sensitive data may not be swapped out, etc.
Pvmemcontrol allows guests to manipulate its gPTE entries in the SLAT,
and also some other properties of the memory map the back's host memory.
This is achieved by using the KVM_CAP_SYNC_MMU capability. When this
capability is available, the changes in the backing of the memory region
on the host are automatically reflected into the guest. For example, an
mmap() or madvise() that affects the region will be made visible
immediately.
There are two components of the implementation: the guest Linux driver
and Virtual Machine Monitor (VMM) device. A guest-allocated shared
buffer is negotiated per-cpu through a few PCI MMIO registers, the VMM
device assigns a unique command for each per-cpu buffer. The guest
writes its pvmemcontrol request in the per-cpu buffer, then writes the
corresponding command into the command register, calling into the VMM
device to perform the pvmemcontrol request.
The synchronous per-cpu shared buffer approach avoids the kick and busy
waiting that the guest would have to do with virtio virtqueue transport.
The Cloud Hypervisor component can be enabled with --pvmemcontrol.
Co-developed-by: Stanko Novakovic <stanko@google.com>
Co-developed-by: Pasha Tatashin <tatashin@google.com>
Signed-off-by: Yuanchu Xie <yuanchu@google.com>
BusDevice trait functions currently holds a mutable reference to self,
and exclusive access is guaranteed by taking a Mutex when dispatched by
the Bus object. However, this prevents individual devices from serving
accesses that do not require an mutable reference or is better served
with different synchronization primitives. We switch Bus to dispatch via
BusDeviceSync, which holds a shared reference, and delegate locking to
the BusDeviceSync trait implementation for Mutex<BusDevice>.
Other changes are made to make use of the dyn BusDeviceSync
trait object.
Signed-off-by: Yuanchu Xie <yuanchu@google.com>
After moving landlock config to VMConfig, there is no need to start
destination VM with landlock cmdline options in
test_live_migration_with_landlock test.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
There were some scenarios where we are not clearing SW_EXIT_INFO1 to
indicate that there were no error while handling the GHCB exit.
Recently, new Linux guests got stricter with checking the value of
SW_EXIT_INFO1 after coming back from VMGEXIT and started crashing. Fix
this behavior by clearing out SW_EXIT_INFO1 in case of no error.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Support event idx feature for virtio-blk device.
This feature could improve disk IO performance by suppressing
notifications from guest to host and interrupts from host to
guest, which has been already supported in virtio-net and
vhost-user devices.
To achieve this, virtqueue's event-idx-related API is
leveraged for avail_event field update and needs_notification
check.
Fixes: #6580
Signed-off-by: wuxinyue <wuxinyue.wxy@antgroup.com>
HV APIC(i.e., synthetic APIC controller exposed by Microsoft Hypervisor)
does not support one-shot operation using a TSC deadline value. Due to
which we see the following backtrace inside the guest when running with
hypervisor-fw/OVMF:
[ 0.560765] unchecked MSR access error: WRMSR to 0x832 (tried to
write 0x00000000000400ec) at rIP: 0xffffffff8f473594
(native_write_msr+0x4/0x30)
[ 0.560765] Call Trace:
[ 0.560765] ? native_apic_msr_write+0x2b/0x30
[ 0.560765] __setup_APIC_LVTT+0xbc/0xe0
[ 0.560765] lapic_timer_set_oneshot+0x27/0x30
[ 0.560765] clockevents_switch_state+0xaf/0xf0
[ 0.560765] tick_setup_periodic+0x47/0x90
[ 0.560765] tick_setup_device.isra.0+0x7c/0x110
[ 0.560765] tick_check_new_device+0xce/0xf0
[ 0.560765] clockevents_register_device+0x82/0x170
[ 0.560765] clockevents_config_and_register+0x2f/0x40
[ 0.560765] setup_APIC_timer+0xe1/0xf0
[ 0.560765] setup_boot_APIC_clock+0x5f/0x66
[ 0.560765] native_smp_prepare_cpus+0x1d6/0x286
[ 0.560765] kernel_init_freeable+0xcf/0x255
[ 0.560765] ? rest_init+0xb0/0xb0
[ 0.560765] kernel_init+0xe/0x110
[ 0.560765] ret_from_fork+0x22/0x40
Also, if this feature is exposed guest would not finish booting and get
stuck right before unpacking the root filesystem.
Fixes: 06e8d1c40 ("hypervisor: mshv: fix topology for Intel HW on MSHV")
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This is a preparatory work to make space for ARM64 emulator.
There is no functional change in this commit, just moving code from one
file to another.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
As per VirtIO spec 1.2 section 5.2.6, the `status` field is a byte, not
u32. cloud-hypervisor writes an `u32` to guest memory, which
accidentally zeros out the following 3 bytes, and may corrupt guest OS
internal state.
Signed-off-by: Changyuan Lyu <changyuanl@google.com>
The assertion that only code emulation requires GVA to GPA translation
is wrong.
Allow the caller of `translate` to pass in permission flags directly.
Provide a new method `read_memory_flags` so that we can add the EXECUTE
permission flag where necessary.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
This method will return the existing Response if the status is
successful (Status::Ok) otherwise issue a command to abandon the
migration and return the desired error.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This requires stashing the config values in `struct Vmm`. The configs
should be validated before before creating the VMM thread. Refactor the
code and update documentation where necessary.
The place where the rules are applied remain the same.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Add file/dir paths from landlock-rules arguments to ruleset. Invoke
apply_landlock on VmConfig to apply config specific rules to ruleset.
Once done, any threads spawned by vmm thread will be automatically
sandboxed with the ruleset in vmm thread.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Introduce ApplyLandlock trait and add implementations to VmConfig
elements with PathBufs. This trait adds config specific rules to
landlock ruleset.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Users can use this parameter to pass extra paths that 'vmm' and its
child threads can use at runtime. Hotplug is the primary usecase for
this parameter.
In order to hotplug devices that use local files: disks, memory zones,
pmem devices etc, users can use this option to pass the path/s that will
be used during hotplug while starting cloud-hypervisor. Doing this will
allow landlock to add required rules to grant access to these paths when
cloud-hypervisor process starts.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Users can use this cmdline option to enable/disable Landlock based
sandboxing while running cloud-hypervisor.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
landlock syscalls are required by event_monitor, signal_handler,
http-server and vmm threads. Rest of the threads are spawned by the vmm
thread and they automatically inherit the ruleset from the vmm thread.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
We will start receiving GPA intercepts. For our use cases they are
handled the same way as UNMAPPED GPA intercepts.
Put in some logging to distinguish the two cases.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Previously we required the hypervisor to give us a valid instruction
stream. That worked well enough because we never hit any edge conditions
(such as when the instruction stream crosses page boundary).
Now that MSHV can deal with partial or empty instruction stream, we can
remove that requirement.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Original we checked for R and W, but that code path never got executed.
It is now understood that we can only get here when we execute code. Fix
the permission flags.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The default value of IP is zero. If the decoder's state not set
properly, then the guest state is going to be wrong.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
TIOCGWINSZ modifies its argument, so it needs to mutably borrow it.
Unfortunately, ioctl()'s signature is not able to enforce this, and
the write happens in the kernel, so I don't think anything like miri,
valgrind, UBSan, etc. would have been able to catch this.
The UB passing an immutable reference caused resulted, for me, in
get_win_size() returning (0, 0) since LLVM commit
9a09c737a052 ("[BasicAA] Make isNotCapturedBeforeOrAt() check for
calls more precise (#69931)").
I've had a look through the other ioctl() calls in Cloud Hypervisor,
and I don't think any others have the same problem.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Comparing RAX with RIP makes no logical sense other than RIP happens to
be the correct value. Use `target_rax` instead.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
cargo-cyclonedx rejects the current license key in Cargo.toml, because
it doesn't comply with the SPDX standard:
Package cloud-hypervisor has an invalid license expression (LICENSE-APACHE & LICENSE-BSD-3-Clause), using as named license: Invalid Lax SPDX expression: unknown term
Fix by using the names from the SPDX list:
https://spdx.github.io/license-list-data/
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
In 42e9632c53 a fix was made to address a
typo in the taplo configuration file. Fixing this typo indicated that
many Cargo.toml files were no longer adhering to the formatting rules.
Fix the formatting by running `taplo fmt`.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Killing CLH by SIGKILL will cause inaccurate code coverage
information. This patch changes the signal to SIGTERM.
Fixes: #6507
Signed-off-by: Songqian Li <sionli@tencent.com>
Since vdpa device does not support pause/resume [1], it does not make
sense to restore on paused state.
[1] 099cdd2af8
Signed-off-by: Bo Chen <chen.bo@intel.com>
Impose a limit on the maximum nesting of file formats that can open more
files. For example, a qcow2 file can have a backing file, which could be
another qcow2 file with a backing file (or even the same file as the
original), potentially causing unbounded recursion.
This commit is based on crosvm implementation:
eb1640e301Fixes: #6472
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
While checking if the console device is a tty use the cloned fd instead
of libc::STDOUT_FILENO.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Console devices are created after vm_config is received and the created
devices are passed Vm during vm_receive_state.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
During vm_shutdown or vm_snapshot, all the console devices will be
closed. When this happens stdout (FD #2) will also be closed as the
console device using these FD is closed. If the VM were to be started
later, FD#2 can be assigned to a different file. But
pre_create_console_devices looks for FD#2 while opening tty device,
which could point to any file.
To avoid this problem, the STDOUT FD is duplicated when being
assigned to a console device. Even if the console devices were to be
closed, the duplicated FD will be closed and FD#2 will continue to
point to STDOUT.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
While adding console devices, DeviceManager will now use the FDs in
console_info instead of creating them.
To reduce the size of this commit, I marked some variables are unused
with '_' prefix. All those variables are cleaned up in next commit.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Use pre_create_console_devices method to create and populate console
device FDs into console_info in Vmm Object.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
With this change all the information to manage console devices is now
available within Vmm Object.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
Introduce ConsoleInfo struct. This struct will be used to store FDs of
console devices created in pre_create_console_devices and passed to
vm_boot.
Move set_raw_mode, create_pty methods to console_devices.rs to
consolidate console management methods into a single module.
Lastly, copy the logic to create and configure console devices into
pre_create_console_devices method.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
If the KVM version is older than v6.6, KVM_GET_SUPPORTED_CPUID will omit
the L1 cache information in CPUID function 0x8000_0005. Fall back to
the host L1 cache information if it is omitted by KVM.
Signed-off-by: Sean Banko <sbanko@crusoe.ai>
Currently by default each core is allocated it's own socket. Basically
it is n socket 1 core 1 thread/core kind of a structure as witnessed
from within the guest.
CPU(s): 8
On-line CPU(s) list: 0-7
Thread(s) per core: 1
Core(s) per socket: 1
Socket(s): 8
NUMA node(s): 1
This is not a good default topology because resources are distributed
across multiple sockets. For example, a Linux guest with multi socket
configuration will have to calibrate TSC per socket due to which it
might observe a higher amount of boot time than usual.
A better idea for default topology would be 1 socket n core 1
thread/core which ensure better resource locality.
After this change topology would change to:
CPU(s): 8
On-line CPU(s) list: 0-7
Thread(s) per core: 1
Core(s) per socket: 8
Socket(s): 1
NUMA node(s): 1
Fixes: #6497
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Populating these cpuid with default values so that
CLH can patch with topology information. Otherwise it gets
skipped while setting some topology information.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Rather than calling cpuid and then updating the APIC ID field - use the
existing common CPUID data which already includes CPUID data for eax=1
(aka function = 1). This removes the need to call cpuid per vCPU thread
created. This has a positive impact on boot time with multiple vCPUs as
the cpuid instruction is serialising.
Fixes: #5646
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The cloud-hypervisor package does not have a lib target, so it is
invalid. Cargo just ignores that line.
Dropping it removes a warning.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The original code gave an owned fd to UnixListener. That made the same
fd wrapped into two owned files.
When the files were dropped, the same fd would be closed more than once.
A newly introduced check in Rust's stdlib caught that error.
A newly cloned fd should be given to UnixListener.
Fixes: #6485
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Whenever the file descriptor table is full, Linux expands it by doubling
it's size.
The filesystem code that does this uses RCU synchronization to ensure
all pre-existing RCU read-side critical sections have completed. The
latency induced by this synchronization is a big part of the total time
required to restore a snapshot.
The kernel has an optimization in code, where it doesn't call
synchronize_rcu() if there is only one thread in the process. We can
take advantage of this optimization by expanding the descriptor table at
the application start, when it has only one thread.
This commit tries to expand the table to 4096 entries, this way we avoid
any expansion that could take place later.
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
Certain VMEXITs can only happen for x86 guests, thus reduce the scope to
x86_64 at the compilation stage.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
For MSHV we always create frozen partition, so we
resume the VM during boot. Also during pause and resume
VM events we call hypervisor specific API.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Create a partition frozen always, then unfreeze the partition
during boot phase or resume phase. We also freeze the
partition during pause event. Time is freeze during the
time between freeze and unfreeze.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add Pause/Resume functions for VM trait. For KVM it
will be empty implementations. For MSHV it needs to freeze
and unfreeze the partition.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Currently, the watchdog configuration in systemd is reloaded by
reboot. We can cancel this reboot operation by restarting the
systemd daemon process.
Signed-off-by: Songqian Li <sionli@tencent.com>
tests_snapshot_restore* have been earlier removed from common_parallel
due to the falkiness they add testsuite. Running them sequentially would
eliminate the flakiness. Hence, add the tests back to testsuite but into
common_sequential module.
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
test_snapshot_restore_* tests often have transient failures and add to
overall flakiness of the integration testsuite. Hence, remove them from
common_parallel. However, these tests need to be added back to
common_sequential
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
VM is created with FDs explicitly passed to CH via --net parameter
and snapshotted. New net FDs are passed in turn during restore.
Boilerplate code from _test_snapshot_restore().
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
Add a section about restoring VM with new Net FDs explicitly passed to
ch-remote via 'net_fds' parameter
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
Enable restore command the ability to send file descriptors along with
HTTP request. This is useful when restoring a VM with explicit FDs
passed to NetConfig(s).
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
Consume FDs passed via SCM_RIGHTs to VmRestore API and assign them
appropriately to RestoredNetConfig's fds field.
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
'NetConfig' FDs, when explicitly passed via SCM_RIGHTS during VM
creation, are marked as invalid during snapshot. See: #6332.
So, Restore should support input for the new net FDs. This patch adds
new field 'net_fds' to 'RestoreConfig'. The FDs passed using this new
field are replaced into the 'fds' field of NetConfig appropriately.
The 'validate()' function ensures all net devices from 'VmConfig' backed
by FDs have a corresponding 'RestoreNetConfig' with a matched 'id' and
expected number of FDs.
The unit tests provide different inputs to parse and validate functions
to make sure parsing and error handling is as per expectation.
Fixes#6286
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
Co-authored-by: Bo Chen <chen.bo@intel.com>
The list is gathered from going through various code paths in the code
base.
No functional change intended.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Per the KVM API document, that capability is only valid with in-kernel
irqchip that handles MSIs.
Through out the code base, there is no call to KVM_IOCTL_SIGNAL_MSI.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The compiler is now able to warn if an invalid attribute (e.g like a
feature) is not available.
See https://blog.rust-lang.org/2024/05/06/check-cfg.html for more
details.
Add build.rs files in the crates that use #cfg(fuzzing) to add fuzzing
to the list of valid cfg attributes.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The ability to control the rustc flags (required for adding new
attributes to the allowed list of #[cfg(..)]) requires bumping the MSRV
to 1.77.0
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Code in this crate is conditional on this feature so it necessary to
expose as a new feature and use that feature as a dependency when the
feature is enabled on the vmm crate.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Enable the use of the vmm crate with the "guest_debug" feature and make
the code that exercises that in the fuzzer unconditional on
"guest_debug" as a feature (as that is not specified as a feature in the
fuzz workspace itself.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The "dhat-heap" feature needs to be enabled inside the vmm crate as a
depenency from the top-level as there is build time check for that
feature inside the vmm crate.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
These tests use relatively large memory allocations and if they are
allowed to run in parallel can result in the container encountering an
OOM situation.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Update the vhost-user-backend crate version used along with related
crates (vhost and virtio-queue.) This requires minor changes to the
types used for the memory in the backends with the use of the
BitmapMmapRegion type for the Bitmap implementation.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Replace the specific Bitmap implementation from the type signature used
for functions that take memory. This allows more flexibility when using
these functions in particular when these functions are used by the
vhost-user-block backend. An updated vhost-user-backend crate requires
extra constraints on the Bitmap implementation used (it must support
BitmapReplace which AtomicBitmap does not.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Replace the specific Bitmap implementation from the type signature used
for functions that take memory. This allows more flexibility when using
these functions in particular when these functions are used by the
vhost-user-net backend. An updated vhost-user-backend crate requires
extra constraints on the Bitmap implementation used (it must support
BitmapReplace which AtomicBitmap does not.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This is to resolve the inconsistencies from our openapi specification,
as default values do not make sense for required fields.
Reported-by: James O. D. Hunt <james.o.hunt@intel.com>
Signed-off-by: Bo Chen <chen.bo@intel.com>
Change from default 4G to 2G to reduce the memory usage. This will help
avoid "No disk space left" errors when all snapshot tests run at once.
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
The snapshot_dir is not needed anymore fter the guest is restored.
Hence, remove it. This will potentially avoid transient failures caused
due to insufficient disk space.
Signed-off-by: Purna Pavan Chandra <paekkaladevi@linux.microsoft.com>
The mem_size field is not needed in TestContext. Drop it.
Make sure guest_evvq is read once. Clippy cannot figured out that it was
used.
While at it, add an extra assert for the spurious rxvq event test, too.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Clippy complained one use of the char_array macro.
Instead of fixing that use case, the observation is that the macro was
needed only because MP table types mixed c_char and c_uchar for no
particular reason.
Only use c_uchar in those types, and drop char_array.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The maintainer can then replace the plaintext body (derived from the
tag) with a markdown version and then publish.
As part of 6c4144e943 creating a draft
release was accidentally dropped.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
As MSHV also implements set/get_clock data, this patch
removes the KVM feature guard and make it x86_64 only and
both for KVM and MSHV.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The contents of this crate may change and cause conflicts - re-exporting
the contents is unnecessary.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This is required as the VcpuFd::run and VcpuFd::set_immediate_exit
methods now take a &mut self. I explored alternative solutions:
1. Using RefCell for runtime interior mutability - the Vcpu trait is
Sync and it's not possible to use RefCell with types that are Sync.
2. Using UnsafeCell - the mutable reference nature of ::run and and
::set_kvm_immediate_exit was added for a reason so it unwise to
bypass this.
3. Adjusting the trait interface to expose the &mut self - this requires
an Arc<Mutex<>> around the hypervisor::vcpu::Vcpu object and was very
intrusive.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The members for {Io, Mmio}{Read, Write} are unused as instead exits of
those types are handled through the VmOps interface. Removing these is
also a prerequisite due to changes in the mutability of the
VcpuFd::run() method.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
On Microsoft Hypervisor, we need to save/restore five
VP state components which are as follows:
1. Local APIC
2. Xsave
3. Synthetic Message Page
4. Synthetic Event Flags Page
5. Synthetic Timers
In the MSHV crate we created a single struct for all the
components and API to get/set the states.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
When using multiple PCI segments, the 32-bit and 64-bit mmio
aperture is split equally between each segment. Add an option
to configure the 'weight'. For example, a PCI segment with a
`mmio32_aperture_weight` of 2 will be allocated twice as much
32-bit mmio space as a normal PCI segment.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
In accordance with reuse requirements:
- Place each license file in the LICENSES/ directory
- Add missing SPDX-License-Identifier to files.
- Add .reuse/dep5 to bulk-license files
Fixes: #5887
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Replace actions: actions-rs/toolchain, upload-release-asset.
actions-rs/toolchain is unmaintained, for cross-compilation replace it
with houseabsolute/actions-rust-cross
Action upload-release-asset is unmaintained and has been archived since
Mar 4, 2021. An issue with upload-release-asset is that it creates
separate release for each job of matrix strategy. Thus, to solve these
issues, replace unmaintained upload-release-asset with upload-artifact
See: #5929
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
This removes the requirement to ensure that we land PRs that update the
Dockerfile (and the appropraite dev_cli.sh change) in a specific time
frame.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Assigning KVM_IRQFD (when unmasking a guest IRQ) after
KVM_SET_GSI_ROUTING can avoid kernel panic on the guest that is not
patched with commit a80ced6ea514 (KVM: SVM: fix panic on out-of-bounds
guest IRQ) on AMD systems.
Meanwhile, it is required to deassign KVM_IRQFD (when masking a guest
IRQ) before KVM_SET_GSI_ROUTING (see #3827).
Fixes: #6353
Signed-off-by: Yi Wang <foxywang@tencent.com>
Signed-off-by: Bo Chen <chen.bo@intel.com>
The caller shouldn't pass in an &str that's too long. This is a
precaution if something goes wrong in the caller.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Properly detach a device from a domain if that device is already
attached to another domain on an attach request (following section
5.13.6.3.2 of the virtio-iommu spec). Resolves nested virtualization
reboot.
Signed-off-by: Andrew Carp <acarp@crusoeenergy.com>
Ensures that any endpoints already attached to the domain are properly
mapped to a new endpoint on said endpoint's attach request. This is done
by search for all previous mappings in the domain and then issuing map
requests for the newly attached endpoint.
Signed-off-by: Andrew Carp <acarp@crusoeenergy.com>
Add infrastructure to lookup the host address for mmio regions on
external dma mapping requests. This specifically resolves vfio
passthrough for virtio-iommu, allowing for nested virtualization to pass
external devices through.
Fixes#6110
Signed-off-by: Andrew Carp <acarp@crusoeenergy.com>
VfioUserDmaMapping is already in the pci crate, this moves
VfioDmaMapping to match the behavior. This is a necessary change to
allow the VfioDmaMapping trait to have access to MmioRegion memory
without creating a circular dependency. The VfioDmaMapping trait
needs to have access to mmio regions to map external devices over
mmio (a follow-up commit).
Signed-off-by: Andrew Carp <acarp@crusoeenergy.com>
The memory region that is associated with the hotpluggable part of
a virtio-mem zone isn't backed by the file specified in the
MemoryZoneConfig. The file is used only for the fixed part of the
zone. When you try to restore a snapshot with virtio-mem, the
backing file is used for all its regions. This results in the
following error:
VmRestore(MemoryManager(GuestMemoryRegion(MappingPastEof)))
This patch sets backing_file only for the fixed part of a virtio-mem
zone.
Fixes: #6337
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
HV_PAGE_SIZE is defined as a usize in mshv_bindings.
Remove the redefinition, and perform casts where necessary.
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
Add IOCTL number for generic hypercall ioctl (MSHV_ROOT_HVCALL).
Update IOCTL numbers for set/get vp state.
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
This reverts commit fcf229a33a.
The virtio-queue version needs to stay the same as the rest of the
source tree.
Signed-off-by: Bo Chen <chen.bo@intel.com>
The 'NetConfig' may contain FDs which can't be serialized correctly, as
FDs can only be donated from another process via a Unix domain socket
with `SCM_RIGHTS`. To avoid false use of the serialized FDs, this patch
explicitly set 'NetConfig' FDs as invalid for (de)serialization.
See: #6286
Signed-off-by: Bo Chen <chen.bo@intel.com>
Prefix field with an _ to indicate that this is intentionally unused.
This resolved a nightly compiler check issue due to the unusued field.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When restoring a VM, the VirtioPciCfgCapInfo struct is not properly
initialized. All fields are 0, including the offset where the
capabibility starts. Hence, when you read a PCI configuration register
in the range [0..length(VirtioPciCfgCap)] you get the value 0 instead of
the actual register contents.
Linux rescans the whole PCI bus when adding a new device. It reads the
values vendor_id and device_id for every device. Because these are
stored at offset 0 in pci configuration space, their value is 0 for
existing devices. As such, Linux considers that the devices have been
unplugged and it removes them from the system.
Fixes: #6265
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
Current MSHV emulator only works for x86 instruction decoder. So, let's
restrict its usage to x86.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
According to the virtio iommu spec (section 5.13.6.6), all mappings
within the entire range from virt_start to virt_end in an unmap
request must be removed. This change adds this functionality,
iterating through all mappings that fall within an unmap request
for that domain and removing them.
Signed-off-by: Andrew Carp <acarp@crusoeenergy.com>
Current instruction emulator can only decode x86 instructions. Thus,
restrict the exit handling for just x86 guests.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
APIC controller is only available on x86 machine. ARM uses a different
interrupt controller so those exit messages won't happen for ARM guests.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
MSR and CPUID are limited to x86 architecture so, reduce the visbility
of these two members inside struct MshvVcpu to just x86 architecture.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
error: the item `io` is imported redundantly
Error: --> devices/src/legacy/uart_pl011.rs:468:9
|
467 | use super::*;
| -------- the item `io` is already imported here
468 | use std::io;
| ^^^^^^^
error: the item `Arc` is imported redundantly
Error: --> devices/src/legacy/uart_pl011.rs:469:21
|
467 | use super::*;
| -------- the item `Arc` is already imported here
468 | use std::io;
469 | use std::sync::{Arc, Mutex};
| ^^^
error: could not compile `devices` (lib test) due to 8 previous errors
Error: warning: build failed, waiting for other jobs to finish...
Error: The process 'cross' failed with exit code 101
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
error: the item `read_le_u32` is imported redundantly
Error: --> devices/src/legacy/gpio_pl061.rs:342:17
|
341 | use super::*;
| -------- the item `read_le_u32` is already imported here
342 | use crate::{read_le_u32, write_le_u32};
| ^^^^^^^^^^^
|
= note: `-D unused-imports` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(unused_imports)]`
error: the item `write_le_u32` is imported redundantly
Error: --> devices/src/legacy/gpio_pl061.rs:342:30
|
341 | use super::*;
| -------- the item `write_le_u32` is already imported here
342 | use crate::{read_le_u32, write_le_u32};
| ^^^^^^^^^^^^
error: the item `Arc` is imported redundantly
Error: --> devices/src/legacy/gpio_pl061.rs:343:9
|
341 | use super::*;
| -------- the item `Arc` is already imported here
342 | use crate::{read_le_u32, write_le_u32};
343 | use std::sync::Arc;
| ^^^^^^^^^^^^^^
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
clippy was flagging this up as a mixture of mixed attributes but that
attribute is no longer necessary
warning: item has both inner and outer attributes
--> hypervisor/src/arch/x86/emulator/mod.rs:769:1
|
769 | / #[cfg(test)]
770 | | mod tests {
771 | | #![allow(unused_mut)]
| |_________________________^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#mixed_attributes_style
= note: `#[warn(clippy::mixed_attributes_style)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
error: the item `kvm_bindings` is imported redundantly
Error: --> hypervisor/src/kvm/aarch64/gic/mod.rs:9:18
|
9 | use crate::kvm::{kvm_bindings, KvmVm};
| ^^^^^^^^^^^^ the item `kvm_bindings` is already defined by prelude
|
= note: `-D unused-imports` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(unused_imports)]`
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: assigning the result of `Clone::clone()` may be inefficient
--> vmm/src/device_manager.rs:4188:17
|
4188 | id = child_id.clone();
| ^^^^^^^^^^^^^^^^^^^^^ help: use `clone_from()`: `id.clone_from(child_id)`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#assigning_clones
= note: `#[warn(clippy::assigning_clones)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: `devices` (lib) generated 1 warning (run `cargo clippy --fix --lib -p devices` to apply 1 suggestion)
warning: assigning the result of `Clone::clone()` may be inefficient
--> virtio-devices/src/transport/pci_device.rs:1073:9
|
1073 | self.bar_regions = bars.clone();
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use `clone_from()`: `self.bar_regions.clone_from(&bars)`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#assigning_clones
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: assigning the result of `Clone::clone()` may be inefficient
--> devices/src/pvpanic.rs:213:9
|
213 | self.bar_regions = bars.clone();
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use `clone_from()`: `self.bar_regions.clone_from(&bars)`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#assigning_clones
= note: `#[warn(clippy::assigning_clones)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Model Specific Registers (MSRs) are usually available on x86
architecture. So, let's reduce the visibility of MSRs to just x86
architecture.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
A deadlock can happen from the destination VM of live upgrade or
migration due to waiting on paused device worker threads. For example,
when a serialization error happens after the `DeviceManager` struct is
restored (where all virtio device worker threads are spawned but in
paused/parked state), a deadlock will happen from
`DeviceManager::drop()`, as it blocks for waiting worker threads to
join.
This patch ensures that we wake up all device (mostly virtio) worker
threads before we block for them to join.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Current version can't boot a SNP guest while using
more than one VCPU. It turns out that there is an
issue in the Snp AP creation process. We should be writing
to Software exitinfo1 instead of exitinfo2. This patch fixes
the issue and we can boot multi vcpu SNP guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Currently, we only support injecting NMI for KVM guests but we can do
the same for MSHV guests as well to have feature parity.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Microsoft Hypervisor raises this request to change the guest page
visibility in case of SEV-SNP guest.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
Kernel can panic when get a unknown nmi when set unknown_nmi_panic,
we can use this feature to test if nmi injected to guest.
Signed-off-by: Yi Wang <foxywang@tencent.com>
When running on the merge group this workflow is run twice - once for
the create event (merge queue creates a new branch) and once for the
merge_group event. Unfortunately the second event would cause the first
to be cancelled - unfortunately sometimes that second event is the
create event where the job in the workflow only runs if it is also a
tag.
By creating distinct concurrency groups for each event type then the
cross cancellation can be avoided.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The workers share a common public IP address and often GitHub will
reject attempts to access the API due to exceeding the anonymous rate
limit threshold.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When a bare-metal worker is canceled, its workspace can be left with
files owned by the root user as a result of running tests from our
container. This patch add a step to fix workspace permissions for such
case before checking out code.
Signed-off-by: Bo Chen <chen.bo@intel.com>
This commit ensures that the HttpApi thread flushes all the responses
before the application shuts down. Without this step, in case of a
VmmShutdown request the application might terminate before the
thread sends a response.
Fixes: #6247
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
Because of double quotes the current value is passed
as a single argument with a space in it to cargo.
This commit changes it to an array so each element
is passed as a different arguments.
Signed-off-by: Alexandru Matei <alexandru.matei@uipath.com>
On platforms where PCIe P2P is supported, inject a PCI capability into
NVIDIA GPU to indicate support.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
The DCO tool doesn't understand merge_groups but we still need to have a
valid status check to allow the merge group to proceed.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
And clean up some of the whitespace formatting so that the "name" and
"on" are grouped away from the "jobs".
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Most of our CI workers are now running form GitHub actions, so we are
ready to disable Jenkins CI workers.
See: #6231
Signed-off-by: Bo Chen <chen.bo@intel.com>
cloud-hypervisor, ch-remote, vhost-user-block and vhost-user-net all
need at least one argument to do anything useful, so printing command
help is helpful when they are run without arguments or a subcommand.
Use clap::Command::arg_required_else_help(true) to do this.
Signed-off-by: Chris Webb <chris@arachsys.com>
ch-remote crashes when run with --api-socket but no subcommand:
$ target/release/ch-remote --api-socket /tmp/api
thread 'main' panicked at src/bin/ch-remote.rs:509:14:
internal error: entered unreachable code
Use clap::Command::subcommand_required(true) to yield a more friendly
error in this case.
Signed-off-by: Chris Webb <chris@arachsys.com>
Host data that is passed to the hypervisor. Then
the firmware includes the data in the attestation report.
The data might include any key or secret that the SevSnp guest
might need later.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
The host data provided at launch. Data is passed
to the hypervisor during the completion of the
isolated import.
Host Data provided by the hypervisor during guest launch.
The firmware includes this value in all attestation
reports for the guest.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
To reduce issues caused by flaky tests split the musl and glibc jobs
into separate jobs. This means fewer jobs will need to be restarted for
flaky tests. This will also increase CI throughput since the musl builds
account for ~40% of the total CI time when run together with glibc.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Since we don't register ioevents in case of SEV-SNP guests. Thus, we
should not unregister it as well.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
With the nightly toolchain (2024-02-18) cargo check will flag up
redundant imports either because they are pulled in by the prelude on
earlier match.
Remove those redundant imports.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When a guest running on a terminal reboots, the sigwinch_listener
subprocess exits and a new one restarts. The parent never wait()s
for children, so the old subprocess remains as a zombie. With further
reboots, more and more zombies build up.
As there are no other children for which we want the exit status,
the easiest fix is to take advantage of the implicit reaping specified
by POSIX when we set the disposition of SIGCHLD to SIG_IGN.
For this to work, we also need to set the correct default exit signal
of SIGCHLD when using clone3() CLONE_CLEAR_SIGHAND. Unlike the fallback
fork() path, clone_args::default() initialises the exit signal to zero,
which results in a child with non-standard reaping behaviour.
Signed-off-by: Chris Webb <chris@arachsys.com>
Allow cloud-hypervisor to direct boot the bzImage kernel format using
the regular 32 bit entry point. This can share the memory and vcpu
setup with the regular PVH boot code, but requires the setup of the
'zero page'.
Signed-off-by: Stefan Nuernberger <stefan.nuernberger@cyberus-technology.de>
For SevSnp guest IO events are handled by GHCB protocol.
While we get the notification we have to notify via eventfd.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
In case of SEV-SNP guest devices use sw-iotlb to gain access guest
memory for DMA. For that F_IOMMU/F_ACCESS_PLATFORM must be exposed in
the feature set of virtio devices.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
According to PCIe specification, a 64-bit MMIO BAR should be
naturally aligned. In addition to being more compliant with
the specification, natural aligned BARs are mapped with
the largest possible page size by the host iommu driver, which
should speed up boot time and reduce IOTLB thrashing for virtual
machines with VFIO devices.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
The 'generate_ram_ranges' function currently hardcodes the assumption
that there are only 2 E820 RAM entries. This is not flexible enough to
handle vendor specific memory holes. Returning a Vec is also more
convenient for users of this function.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
Since the ACPI tables are generated inside the IGVM file in case of
SEV-SNP guest. So, we don't need to generate it inside the cloud
hypervisor.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
... enabled VMs. IOEvents are not supported in case of SEV-SNP VMs. All
the IO events are delievered via GHCB protocol.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This will help in identify whether a VM supports sev-snp and based on
that disable/enable certain features.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Implement workflow to to run static analysis and linting of all shell
scripts by using shfmt and shellcheck.
Fixes: #5396
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
We occasionally saw cloud-hypervisor crashed due to seccomp violations. The
coredumps showed the HTTP API thread crashing after it attempted to call
sched_yield(). The call came from rust stdlib's mpmc module, which calls
sched_yield() if several attempts to busy-wait for a condition to fulfil fall
short.
Since the system call is harmless and it comes from the stdlib, I opted to allow
all threads to call it.
Signed-off-by: Peteris Rudzusiks <rye@stripe.com>
Currently the only way to set the affinity for virtio block threads is
to boot the VM, search for the tid of each of the virtio block threads,
then set the affinity manually. This commit adds an option to pin virtio
block queues to specific host cpus (similar to pinning vcpus to host
cpus). A queue_affinity option has been added to the disk flag in
the cli to specify a mapping of queue indices to host cpus.
Signed-off-by: acarp <acarp@crusoeenergy.com>
Traditional way to configure vcpu don't work for sev-snp guests. All the
vCPU configuration for SEV-SNP guest is provided via VMSA.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This also fixes the following clippy warning on nightly build from cargo
fuzz:
warning: struct `RegionEntry` is never constructed
--> /home/chenb/project/cloud-hypervisor/cloud-hypervisor/block/src/vhdx/vhdx_header.rs:357:8
|
357 | struct RegionEntry {
| ^^^^^^^^^^^
|
= note: `RegionEntry` has derived impls for the traits `Debug` and `Clone`, but these are intentionally ignored during dead code analysis
= note: `#[warn(dead_code)]` on by default
Signed-off-by: Bo Chen <chen.bo@intel.com>
For all VFIO devices, map all non-emulated MMIO regions to
the vfio container to allow PCIe P2P between all VFIO devices
on the virtual machine. This is required for a wide variety of
advanced GPU workloads such as GPUDirect P2P (DMA between two
GPUs), GPUDirect RDMA (DMA between a GPU and an IB device).
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
Currently kernel and firmware are checked as a payload.
IGVM should be checked as well. Otherwise, it hangs indefinitely.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Unaccepted GPA is usually thrown by Microsoft hypervisor in case of
mismatch between GPA and GVA mappings. This is a fatal message from the
hypervisor perspective so we would need to error out from the vcpu run
loop. Along with add some debug message to identify the broken mapping
between GVA and GPA.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
On guests with large amounts of memory, using the `prefault` option can
lead to a very long boot time. This commit implements the strategy
taken by QEMU to prefault memory in parallel using multiple threads,
decreasing the time to allocate memory for large guests by
an order of magnitude or more.
For example, this commit reduces the time to allocate memory for a
guest configured with 704 GiB of memory on 1 NUMA node using 1 GiB
hugepages from 81.44134669s to just 6.865287881s.
Signed-off-by: Sean Banko <sbanko@crusoeenergy.com>
The `ByteValued` trait implementations for the data structures from the
'linux_loader' crate are no longer needed, and hence their wrappers can
be removed.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Fix beta clippy issue:
error: file opened with `create`, but `truncate` behavior not defined
--> src/main.rs:624:26
|
624 | .create(true)
| ^^^^^^^^^^^^- help: add: `.truncate(true)`
|
= help: if you intend to overwrite an existing file entirely, call `.truncate(true)`
= help: if you instead know that you may want to keep some parts of the old file, call `.truncate(false)`
= help: alternatively, use `.append(true)` to append to the file instead of overwriting it
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#suspicious_open_options
= note: `-D clippy::suspicious-open-options` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::suspicious_open_options)]`
Signed-off-by: Bo Chen <chen.bo@intel.com>
The embedded error in the enum will be read on debug output of the
error.
Fixes beta clippy issue:
warning: field `0` is never read
--> vhost_user_block/src/lib.rs:64:23
|
64 | CreateKillEventFd(io::Error),
| ----------------- ^^^^^^^^^
| |
| field in this variant
|
= note: `#[warn(dead_code)]` on by default
help: consider changing the field to be of unit type to suppress this warning while preserving the field numbering, or remove the field
|
64 | CreateKillEventFd(()),
| ~~
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The embedded error in the enum will be read on debug output of the
error.
Fixes beta clippy issue:
warning: field `0` is never read
--> performance-metrics/src/performance_tests.rs:25:11
|
25 | Infra(InfraError),
| ----- ^^^^^^^^^^
| |
| field in this variant
|
= note: `#[warn(dead_code)]` on by default
help: consider changing the field to be of unit type to suppress this warning while preserving the field numbering, or remove the field
|
25 | Infra(()),
| ~~
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: field `0` is never read
--> block/src/qcow/vec_cache.rs:139:21
|
139 | struct NumCache(pub u64);
| -------- ^^^^^^^
| |
| field in this struct
|
= note: `#[warn(dead_code)]` on by default
help: consider changing the field to be of unit type to suppress this warning while preserving the field numbering, or remove the field
|
139 | struct NumCache(());
| ~~
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
These structs directly embed another struct and then implement
ByteValued on that struct to implement ByteValued for the inner struct.
As such the inner struct is never directly accessed so to avoid the dead
code analysis mark this as allowed.
Beta clippy fix:
warning: field `0` is never read
--> arch/src/x86_64/mod.rs:129:32
|
129 | struct MemmapTableEntryWrapper(hvm_memmap_table_entry);
| ----------------------- ^^^^^^^^^^^^^^^^^^^^^^
| |
| field in this struct
|
= note: `MemmapTableEntryWrapper` has a derived impl for the trait `Clone`, but this is intentionally ignored during dead code analysis
= note: `#[warn(dead_code)]` on by default
help: consider changing the field to be of unit type to suppress this warning while preserving the field numbering, or remove the field
|
129 | struct MemmapTableEntryWrapper(());
|
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Beta clippy fix
warning: useless use of `vec!`
--> tests/integration.rs:5845:23
|
5845 | let kernels = vec![direct_kernel_boot_path()];
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: you can use an array directly: `[direct_kernel_boot_path()]`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#useless_vec
= note: `#[warn(clippy::useless_vec)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Beta clippy fix:
warning: initializer for `thread_local` value can be made `const`
--> vmm/src/sigwinch_listener.rs:27:40
|
27 | static TX: RefCell<Option<File>> = RefCell::new(None);
| ^^^^^^^^^^^^^^^^^^ help: replace with: `const { RefCell::new(None) }`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#thread_local_initializer_can_be_made_const
= note: `#[warn(clippy::thread_local_initializer_can_be_made_const)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Beta clippy fix:
warning: this call to `as_ref.map(...)` does nothing
--> vmm/src/device_manager.rs:1234:9
|
1234 | self.console_resize_pipe.as_ref().map(Arc::clone)
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: try: `self.console_resize_pipe.clone()`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#useless_asref
= note: `#[warn(clippy::useless_asref)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
According to crates.io the crc-any crate is actively maintained which
avoids issues with the crc32c crate and the nightly compiler.
Fixes: #6168
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When we import a page, we have a page with
some data or empty, empty does not mean there is no data,
it rather means it's full of zeros. We can skip writing the
data as guest memory of the page is already zeroed.
A page could be partially filled and the rest of the content is zero.
Our IGVM generation tool only fills data here if there is some data
without zeros. Rest of them are padded. We only write data
without padding and compare whether we complete writing
the buffer content. Still it's a full page and update the variable
with length of the full page.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
When the HTT flag CPUID.1.EDX[HTT] is 0, it indicates that there is
only a single logical processor in the package. When HTT is 1, it
indicates that CPUID.1.EBX[23:16] contains the number of logical
processors in the package.
When this information is not included in CPUID leaf 0x1, some cpu
topology enumeration software such as hwloc are known to crash.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
This commit adds the debug-console (or debugcon) device to CHV. It is a
very simple device on I/O port 0xe9 supported by QEMU and BOCHS. It is
meant for printing information as easy as possible, without any
necessary configuration from the guest at all.
It is primarily interesting to OS/kernel and firmware developers as they
can produce output as soon as the guest starts without any configuration
of a serial device or similar. Furthermore, a kernel hacker might use
this device for information of type B whereas information of type A are
printed to the serial device.
This device is not used by default by Linux, Windows, or any other
"real" OS, but only by toy kernels and during firmware development.
In the CLI, it can be configured similar to --console or --serial with
the --debug-console parameter.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Set the SEV control register so we know where to
start running. This register configures the
SEV feature control state on a virtual processor.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This register configures the SEV feature control
state on a virtual processor.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Consistent with the other data structures and constants used in TDX
support code import the necessary structures from the kernel for
accessing the vmcall structure.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
These Default implementations either don't produce valid configs, are
no longer used outside of tests, or both.
For the tests, we can define our own local "default" values that make
the most sense for the tests, without worrying about what's
a (somewhat) sensible "global" default value.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Bumping anyhow crate from 1.0.75 to 1.0.79 will cause seccomp
failures through integration tests. Newly added backtrace support
relies on readlink and many other syscalls.
Issue noticed with test_api_http_pause_resume test, where second time
of VM PAUSE or VM RESUME prints error and causes panic.
Noticed that panic message in a thread which is not allowed to write
output triggered the issue.
So implementing Display trait for HttpError and ApiError enums to avoid
adding many syscalls to seccomp filter section.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
On hosts with >256 cpus, setting the cpu affinity to a host cpu index
>255 will return an error because type of `host_cpu` is `u8`.
This commit changes the type of `host_cpu` to `usize` to remove this
limitation.
Signed-off-by: Sean Banko <sbanko@crusoeenergy.com>
Move the duplicated logic to download OVMF into a helper function.
Explicitly specify the OVMF_FW_TAG instead of downloading the latest
so that new OVMF versions can be easily tested.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
Testing instructions are very old and type=dpdkvhostuser is
deprecated from DPDK.
logs from /var/log/openvswitch/ovs-vswitchd.log
|netdev_dpdk|WARN|dpdkvhostuser ports are considered deprecated;
please migrate to dpdkvhostuserclient ports
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Uses of the old ApiRequest enum conflated two different concerns:
identifying an API request endpoint, and storing data for an API
request. This led to ApiRequest values being passed around with junk
data just to communicate a request type, which forced all API request
body types to implement Default, which in some cases doesn't make any
sense — what's the "default" path for a vhost-user socket? The
nonsensical Default values have led to tests relying on being able to
use nonsensical data, which is an impediment to adding better
validation for these types.
Rather than having API request types be represented by an enum, which
has to carry associated body data everywhere it's used, it makes more
sense to represent API request types as trait objects. These can have
an associated type for the type of the request body, and this makes it
possible to pass API request types and data around as siblings in a
type-safe way without forcing them into a single value even where it
doesn't make sense. Trait objects also give us dynamic dispatch,
which lets us get rid of several large match blocks.
To keep it possible to fuzz the HTTP API, all the Vmm methods called
by the HTTP API are pulled out into a trait, so the fuzzer can provide
its own stub implementation of the VMM.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
With the change in 7bc764d to expose the SVM bit for nested
virtualisation test_vfio can be re-enabled on the AMD workers.
Fixes: #5895
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
When using amd topology, the svm feature flag on cpuid leaf
0x8000_0001.ecx is overwritten. We update the amd cpu topology
logic to use the flag values that originated in
KVM_GET_SUPPORTED_CPUID ioctl and override as necessary.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
warning: this URL is not a hyperlink
--> arch/src/aarch64/layout.rs:114:58
|
114 | ...in https://www.kernel.org/doc/Documentation/arm64/booting.txt.
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: use an automatic link instead: `<https://www.kernel.org/doc/Documentation/arm64/booting.txt.>`
|
= note: bare URLs are not automatically turned into clickable links
= note: `#[warn(rustdoc::bare_urls)]` on by default
I also noticed that it looks like this comment was supposed to be
applied to FDT_MAX_SIZE, not FDT_START, so I moved it.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
The VIRTIO specification[1] says:
> The upper 32 bits of the CID are reserved and zeroed.
We should therefore not allow the user to supply a VSOCK CID with
those bits set. To accomplish this, limit the public API of the
virtio-vsock device to only accept 32-bit CIDs, while still using
64-bit CIDs internally since that's how virtio-vsock works.
[1]: https://docs.oasis-open.org/virtio/virtio/v1.2/csd01/virtio-v1.2-csd01.html#x1-4400004
Signed-off-by: Alyssa Ross <hi@alyssa.is>
I accidentally ran a VM with CID 2 (VMADDR_CID_HOST), and very strange
and difficult to debug behavior ensued. I don't think a virtio-vsock
device should be allowed to have any of the special CIDs
(VMADDR_CID_ANY, VMADDR_CID_HYPERVISOR, VMADDR_CID_LOCAL, VMADDR_CID_HOST).
Signed-off-by: Alyssa Ross <hi@alyssa.is>
The socket is nonblocking, so it's not guaranteed that it will be
possible to read the whole connect command in a single iteration of
the event loop. To reproduce:
(echo -n 'CONNECT '; sleep 1; echo 1234; cat) | socat STDIO UNIX-CONNECT:vsock.sock
This would produce the error:
cloud-hypervisor: 5.509209s: <_vsock4> INFO:virtio-devices/src/vsock/unix/muxer.rs:446 -- vsock: error adding local-init connection: UnixRead(Os { code: 11, kind: WouldBlock, message: "Resource temporarily unavailable" })
To fix this, if we only get a partial command, we need to save it for
future iterations of the event loop, and only proceed once we've read
a complete command.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Complete the isolated import, telling the
Microsoft hypervisor that import is done so that
MSHV can issue SNP_LAUNCH_FINISH command.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Use right and exact size 32 bytes for host data field
for completing the isolated import. This way OOB
can be avoided during a function call.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Import all the isolated pages after parsing is
done on the iGVM file. Hypervisor adds those
pages for PSP measurement(part of the hashing).
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Current instructions are incorrect and there is now a new profile called
dev-opt to build the debug version of TD-SHIM.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Under the fuzzer this code appears dead:
error: field `0` is never read
--> /home/rob/src/cloud-hypervisor/arch/src/x86_64/mod.rs:128:32
|
128 | struct MemmapTableEntryWrapper(hvm_memmap_table_entry);
| ----------------------- ^^^^^^^^^^^^^^^^^^^^^^
| |
| field in this struct
|
= note: `MemmapTableEntryWrapper` has a derived impl for the trait `Clone`, but this is intentionally ignored during dead code analysis
= note: `-D dead-code` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(dead_code)]`
help: consider changing the field to be of unit type to suppress this warning while preserving the field numbering, or remove the field
|
128 | struct MemmapTableEntryWrapper(());
| ~~
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Add a 'rate_limit_groups' field to VmConfig that defines a set of
named RateLimiterGroups.
When the 'rate_limit_group' field of DiskConfig is defined, all
virtio-blk queues will be rate-limited by a shared RateLimiterGroup.
The lifecycle of all RateLimiterGroups is tied to the Vm.
A RateLimiterGroup may exist even if no Disks are configured to use
the RateLimiterGroup. Disks may be hot-added or hot-removed from the
RateLimiterGroup.
When the 'rate_limiter' field of DiskConfig is defined, we construct
an anonymous RateLimiterGroup whose lifecycle is tied to the Disk.
This is primarily done for api backwards compatability. Importantly,
the behavior is not the same! This implementation rate_limits the
aggregate bandwidth / iops of an individual disk rather than the
bandwidth / iops of an individual queue of a disk.
When neither the 'rate_limit_group' or the 'rate_limiter' fields of
DiskConfig is defined, the Disk is not rate-limited.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
Add a 'rate_limiter/group' module that defines the RateLimiterGroup
and a RateLimiterGroupHandle types.
The RateLimiterGroupHandle can be used in place of a RateLimiter to
limit the aggregate bandwidth and/or ops of multiple virtio-blk or
virtio-net queues.
Each RateLimiterGroup has an associated worker thread that broadcasts
an event to each RateLimiterGroupHandle when a RateLimiter is unblocked.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
CI reports errors:
error: writing `&Vec` instead of `&[_]` involves a new object where a slice will do
--> arch/src/x86_64/mod.rs:1351:19
|
1351 | epc_sections: &Vec<SgxEpcSection>,
| ^^^^^^^^^^^^^^^^^^^ help: change this to: `&[SgxEpcSection]`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#ptr_arg
= note: `-D clippy::ptr-arg` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::ptr_arg)]`
Signed-off-by: Yi Wang <foxywang@tencent.com>
CI reports clippy errors:
error: argument to `Path::join` starts with a path separator
--> tests/integration.rs:4076:58
|
4076 | let serial_socket = guest.tmp_dir.as_path().join("/tmp/serial.socket");
| ^^^^^^^^^^^^^^^^^^^^
|
= note: joining a path starting with separator will replace the path instead
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#join_absolute_paths
Signed-off-by: Yi Wang <foxywang@tencent.com>
CI reports clippy errors:
error: in a `match` scrutinee, avoid complex blocks or closures with blocks; instead, move the block or closure higher and bind it with a `let`
--> test_infra/src/lib.rs:93:51
|
93 | match (|| -> Result<(), WaitForBootError> {
| ___________________________________________________^
94 | | let listener =
95 | | TcpListener::bind(listen_addr.as_str()).map_err(WaitForBootError::Listen)?;
96 | | listener
... |
145 | | }
146 | | })() {
| |_________^
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#blocks_in_conditions
= note: `-D clippy::blocks-in-conditions` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::blocks_in_conditions)]`
Signed-off-by: Yi Wang <foxywang@tencent.com>
This PR addresses a bug in which the cpu topology of a guest
with non power-of-two number of cores is incorrect. For example,
in some contexts, a virtual machine with 2-sockets and 12-cores
will incorrectly believe that 16 cores are on socket 1 and 8
cores are on socket 2. In other cases, common topology enumeration
software such as hwloc will crash.
The root of the problem was the way that cloud-hypervisor generates
apic_id. On x86_64, the (x2) apic_id embeds information about cpu
topology. The cpuid instruction is primarily used to discover the
number of sockets, dies, cores, threads, etc. Using this information,
the (x2) apic_id is masked to determine which {core, die, socket} the
cpu is on. When the cpu topology is not a power of two
(e.g. a 12-core machine), this requires non-contiguous (x2) apic_id.
Signed-off-by: Thomas Barrett <tbarrett@crusoeenergy.com>
The following tests have been temporarily disabled:
1. Live upgrade/migration test with ovs-dpdk (#5532);
2. Disk hotplug tests on windows guests (#6037);
This patch has been tested with PR #6048.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Signed-off-by: Michael Zhao <michael.zhao@arm.com>
Tested-by: Bo Chen <chen.bo@intel.com>
For SEV-SNP guests we need to provide the extended memory. It follows a
very simple layout and very similar to other x86 guests.
First segment: [HIGH_RAM_START - MEM_32BIT_RESERVED_START]
PCI hole: [MEM_32BIT_RESERVED_START - RAM_64BIT_START]
Second segment: [RAM_64BIT_START - RAM_END]
Fixes#5993
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The 'test_vfio_user' is prone to fail when the system is under high
workloads with errors:
```
Error while connecting to /var/tmp/spdk.sock
Is SPDK application running?
Error details: Invalid or non-existing address: '/var/tmp/spdk.sock'
```
This is because SPDK is not fully functional before we request to
create a nvme device using the vfio_user protocol. This patch stabilize
this test with allowing retires to execute host commands.
Signed-off-by: Bo Chen <chen.bo@intel.com>
This patch defines a new function 'generate_ram_ranges', to generate
usable physical memory ranges for the guest based on the existing guest
memory managed by VMM. This function is also made public, so that it can
be reused, say by the IGVM loader in the future [1].
No functional change.
See: #6020
Signed-off-by: Bo Chen <chen.bo@intel.com>
This patch adds missing new lines after functions,
fixes few typos in the comments, adds few missing
comments to SNP related functions.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
There is no requirement to call copy_from_slice, since all the member
variables are identical and we can directly assign them value.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
There is a mix of v2, v3 and v4 in the codebase. Let's move to v4
everywhere because v2 seems to be using a deprecated version of nodejs.
This is throwing warnings when the Github action CI is running.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Currently there are some inconsistencies in Cargo.toml which is causing
the following warnings during the build process:
Error parsing Cargo.toml manifest, fallback to caching entire file:
Invalid TOML document: expected key-value, found comma
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
vmm: Add igvm module and loader module
Add a separate module named igvm to the vmm crate
with definitions to parse and load igvm to the guest memory.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add necessary API to retrieve cpuid leaf on MSHV.
This API is used to update cpuid information
during the parsing of the igvm file.
Microsoft hypervisor does not provide common
CpuID like KVM. That's why we need to call this API
during the IGVM parsing.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This patch adds igvm to the Vm config and params as well as
the command line argument to pass igvm file to load into
guest memory. The file must maintain the IGVM format.
The CLI option is featured guarded by igvm feature gate.
The IGVM(Independent Guest Virtual Machine) file format
is designed to encapsulate all information required to
launch a virtual machine on any given virtualization stack,
with support for different isolation technologies such as
AMD SEV-SNP and Intel TDX.
At a conceptual level, this file format is a set of commands created
by the tool that generated the file, used by the loader to construct
the initial guest state. The file format also contains measurement
information that the underlying platform will use to confirm that
the file was loaded correctly and signed by the appropriate authorities.
The IGVM file is generated by the tool:
https://github.com/microsoft/igvm-tooling
The IGVM file is parsed by the following crates:
https://github.com/microsoft/igvm
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Random failures of downloading `cloud-hypervisor-static-aarch64` with
`wget` were seen. The commit applies a workaround to retry the download
for a few times.
Signed-off-by: Michael Zhao <michael.zhao@arm.com>
This commit changes existing behavior of named TAP interfaces.
When booting a VM with configuration for a named TAP interface,
cloud-hypervisor will create the interface and apply a given
IP configuration to that interface. If the named interface
already exists on the system, the configuration is NOT overwritten.
Setting the ip and netmask fields in a tap interface configuration
for a named tap interface now works by handing this configuration
to the virtio_devices::Net object when it is created with a name.
This commit also touches net_util to make sure that the ip configuration
of existing TAP interfaces is not modified with ip or netmask handed to
open_tap.
Signed-off-by: Markus Sütter <markus.suetter@secunet.com>
The wget has been causing frequent CI failure for the past few days,
while it can't be reproduced manually. Let's show the wget details in
our CI pipeline to understand better future errors.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Running on host where vdpa_sim_blk module is not correctly loaded
test_vdpa_block passes.
"test common_parallel::test_vdpa_block ... ok"
This commit fixes the vdpa_sim_blk test to fail in that case.
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Valid title format:
component1[, component2, componentN]: submodule: summary
Title should have at least one component
Components are separated by comma+space: ", "
Components are validated to be in valid_components
Components list is ended by a colon
Submodules are not validated
See: #5846
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
We can skip the CI worker build if the changes are only from gitlint, or
the hidden files in the root folder, such as .github, .gitignore,
.gitlint, .rustfmt.toml, and .typos.toml.
Signed-off-by: Bo Chen <chen.bo@intel.com>
This is particularly useful for commits with long component names. Of
course, it is better to have concise subject that is less than 50 char.
Signed-off-by: Bo Chen <chen.bo@intel.com>
As part SMP bringup for a SEV-SNP guest, BSP sets up the VMSA page for
each AP threads and informs hypervisor about the same using a VMGEXIT.
Thus, extend the current GHCB interface to handle this scenario.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
SEV-SNP guest can request AMD's secure co-processor i.e., PSP to
generate an runtime attesation report. During this process guest needs
to inform PSP about the request and response GPAs where that report
would be generated by the PSP. This is handled via a VMGEXIT request.
Thus, extend the current GHCB handling to add support for it.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
We found that it's slow to load JSON when reading snap files. As
described in [1], using from_slice instead of from_reader can fix
this.
Also, fix the error type being returned.
1. https://github.com/serde-rs/json/issues/160
Signed-off-by: Yi Wang <foxywang@tencent.com>
Since the 'write()' to the event file was moved to its own thread
(see #5633), we have no reliable way to read the latest contents of
the event file from our integration tests, since we can't ensure the
'read()' from our test always happen after 'write()' is completed from
Cloud Hypervisor. This is also why we started to see random failures on
snapshot_restore tests (particularly when the system workload is high).
This patch adds a 1s sleep before reading the event file to mitigate the
random failures.
Signed-off-by: Bo Chen <chen.bo@intel.com>
error: use of a fallible conversion when an infallible one could be used
Error: --> virtio-devices/src/vhost_user/vu_common_ctrl.rs:206:51
|
206 | let actual_size: usize = queue.size().try_into().unwrap();
| ^^^^^^^^^^^^^^^^^^^ help: use: `into()`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#unnecessary_fallible_conversions
= note: `-D clippy::unnecessary-fallible-conversions` implied by `-D warnings`
= help: to override `-D warnings` add `#[allow(clippy::unnecessary_fallible_conversions)]`
error: could not compile `virtio-devices` (lib) due to previous error
Error: warning: build failed, waiting for other jobs to finish...
error: could not compile `virtio-devices` (lib test) due to previous error
Error: The process '/home/runner/.cargo/bin/cargo' failed with exit code 101
Signed-off-by: Bo Chen <chen.bo@intel.com>
When the script is invoked via dev_cli.sh, it always gets AUTH_DOWNLOAD_TOKEN
from the environment. The original test always returns true.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Instead of relying on "wc" and "grep", this patch provides helper
functions for checking line counts and searching/counting keywords.
To understand assertion failures better, it also generate logs for the
L1/L2 VM commands when checks fail.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Add a test that supports configuring serial and console as TTY mode
at the same time. With this configuration, the VM can set up a legacy
serial device as an early printk console device, and then change to a
virito console device after the virito console device is initialized.
In this case, we can capture the logs printed by legacy serial on early
boot, and later by the virtio console.
Signed-off-by: Yong He <alexyonghe@tencent.com>
Cloud Hypovrisor supports legacy serial device and virito console device
for VMs. Using legacy serial device, CH can capture full VM console logs,
but its implementation is based on KVM PIO emulation and has poor
performance. Using the virtio console device, the VM console logs will
be sent to CH through the virtio ring, the performance is better, but CH
will only capture the VM console logs after the virtio console device is
initialized, the VM early startup logs will be discarded.
This patch provides a way to enable both the legacy serial device and the
virtio console device as a TTY mode by setting the leagcy serial port as
the VM's early printk device and setting the virtio console as the VM's
main console device.
Then CH can capture early boot logs from the legacy serial device and
capture later logs from the virito console device with better performance.
Signed-off-by: Yong He <alexyonghe@tencent.com>
The script "run_integration_tests_vfio.sh" requires to be ran on a
specific machine with specific hardware (e.g. the "vfio" bera-metal
worker equipped with a Nvidia Tesla T4 card). So the provisioning of the
running machine is out of the scope of this script, including the custom
guest image with Nvidia drivers installed, and properly configured
Nvidia Tesla T4 card.
See: #4405, #5509
Signed-off-by: Bo Chen <chen.bo@intel.com>
With #4324 being resolved, the nested VFIO test (e.g. "test_vfio") is
now a part of the general Azure VM-based workers. No need to run it on
the bare-metal worker.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Now, default values for vcpu topology are 0s, that is not correct and may
lead to bug. Fix it by setting default value to 1s. Also add check in
case one or more of these values are zero.
Fixes: #5892
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
virtiofsd must be built with the default gnu toolchain so clear the
environment variables that may poisoned by the alternative C library
support.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The seccompiler v0.4.0 started to use `seccomp` syscall instead of the
`prctl` syscall. Also, threads for virtio-deivces should not need any of
these syscalls anyway.
Signed-off-by: Bo Chen <chen.bo@intel.com>
MMIO emulation is also performed via VMGEXIT in case of SEV-SNP guest.
Emulation is done in a very similar way like a regular guest. Just need
to make sure that guest memory is access via read/write GPA hypercall
instead of directly accessing it.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Legacy port emulation requires reading RAX register from GHCB page for
SEV-SNP guest. This is the major difference between a regular guest and
SEV-SNP enabled guest.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Currently MSHV does not support fetching extended guest report and thus
return an appropriate error stating the NAE event is not valid.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Nuno Das Neves <nudasnev@microsoft.com>
As part of this handling there are 4 different operations:
1. Getting the hypervisor preffered doorbell page GPA.
2. Informing hypervisor about the doorbell page chosen by the guest
3. Querying the GPA of the doorbell page
4. Clearing the GPA of the doorbell page from hypervisor
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
As part of handling this request, hypervisor is expected to three
things:
1. Maximum GHCB protocol version supported.
2. Minimum GHCB protocol version supported.
3. SEV-page table encryption bit number.
If the guest cannot support the protocol range supplied by the
hypervisor, it should terminate
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
SEV-SNP guest allocates a GHCB page and in order to update hypervisor
about the same, there is a vmgexit which allows registering GHCB page
with the hypervisor.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
A VMGEXIT exit occurs for any of the listed NAE events in the GHCB
specification [1] (e.g. CPUID, RDMSR/WRMSR, MMIO, port IO, etc.). Some
of these events are handled by hypervisor while other are handled by
VMM. Currently, we are adding support for one such request i.e.,
report supported SEV-SNP features by hypervisor.
[1] GHCB protocol specification:
https://www.amd.com/content/dam/amd/en/documents/epyc-technical-docs/specifications/56421.pdf
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This is the function that needs to be called by the VMM
to inform the MSHV that isolation is complete and inform
PSP about this completion.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Add hypervisor VM specific API to import the isolated
pages. Hypervisor adds those pages for PSP measurement.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This patch fixes following warnings:
error: boolean to int conversion using if
--> vmm/src/vm.rs:866:42
|
| .create_vm_with_type(if sev_snp_enabled.into() {
| __________________________________________^
| | 1 // SEV_SNP_ENABLED
| | } else {
| | 0 // SEV_SNP_DISABLED
| | })
| |_____________________^ help: replace with from: `u64::from(sev_snp_enabled.into())`
|
= note: `-D clippy::bool-to-int-with-if` implied by `-D warnings`
= note: `sev_snp_enabled.into() as u64` or `sev_snp_enabled.into().into()` can also be valid options
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#bool_to_int_with_if
error: useless conversion to the same type: `bool`
--> vmm/src/vm.rs:866:45
|
| .create_vm_with_type(if sev_snp_enabled.into() {
| ^^^^^^^^^^^^^^^^^^^^^^ help: consider removing `.into()`: `sev_snp_enabled`
|
= note: `-D clippy::useless-conversion` implied by `-D warnings`
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#useless_conversion
error: could not compile `vmm` due to 2 previous errors
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Partially revert 111225a2a5
and add the new dbus and pvpanic arguments.
As we are switching back to clap observe the following changes.
A few examples:
1. `-v -v -v` needs to be written as`-vvv`
2. `--disk D1 --disk D2` and others need to be written as `--disk D1 D2`.
3. `--option value` needs to be written as `--option=value.`
Change integration tests to adapt to the breaking changes.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
There is no need to set them in the test scripts while the main script
already has them.
The consolidates how things are done.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
There is no need to reconstruct it from within the scripts since the
main script already constructed it once.
Drop the previously useless setting of BUILD_TARGET from various
scripts. The value was always overwritten at a later point.
No functional change intended.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Sev-Snp on MSHV uses igvm crate to parse the igvm file.
igvm crate needs minimum rust version 1.65 to build.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
This struct contains all configuration fields that controls the way how
we generate CPUID for the guest on x86_64. This allows cleaner extension
when adding new configuration fields.
Signed-off-by: Bo Chen <chen.bo@intel.com>
For a SEV-SNP enabled partition on MSHV, some of the VMGEXITS are
offloaded for Hypervisor to handle while the rest are handled by VMM.
By setting this additional partition property hypervisor is informed
about the VMGEXITs it needs to take care off, rest all would be handled
by the CloudHypervisor.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The lock to `vm_config` is held for accessing `cpus.kvm_hyperv` passing
as a reference to `arch::generate_common_cpuid()`, so acquiring the same
lock again while calling to the same function is a deadlock.
Fixes: 3793ffe888
Reported-by: Yi Wang <foxywang@tencent.com>
Signed-off-by: Bo Chen <chen.bo@intel.com>
As part of this initialization for a SEV-SNP VM on MSHV, it is required
that we transition the guest state to secure state using partition
hypercall. This implies all the created VPs will transition to secure
state and could access the guest encrypted memory.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
When performing integration testing on the Ampere One server, an error occurred when compiling spdk and not recognizing the CPU ID.
The latest spdk already contains fixes.
Signed-off-by: dom.song <dom.song@amperecomputing.com>
It's a requirement that a SEV-SNP enabled guest on MSHV must have
isolation policy set before launching the guest.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Cloud-Hypervisor takes a path for Unix socket, where it will listen
on. Users can connect to the other end of the socket and access serial
port on the guest.
"--serial socket=/path/to/socket" is the cmdline option to pass to
cloud-hypervisor.
Users can use socat like below to access guest's serial port once the
guest starts to boot:
socat -,crnl UNIX-CONNECT:/path/to/socket
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
The 'derive' feature of `zerocopy` crate now is optional and requires to
be enabled explicitly [1]. Also, a version bump on `acpi_tables` is
needed to reply on a single version of `zerocopy` to avoid compilation
errors.
[1] https://github.com/google/zerocopy/pull/176
Signed-off-by: Bo Chen <chen.bo@intel.com>
Include the TSC frequency as part of the KVM state so that it will be
restored at the destination.
This ensures migration works correctly between hosts that have a
different TSC frequency if the guest is running with TSC as the source
of timekeeping.
Fixes: #5786
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
The bare metal machines are now not reachable from our Jenkins
controller, and causing CI pipeline timeout for the past few days.
Signed-off-by: Bo Chen <chen.bo@intel.com>
The cumulative average formula [1] requires to use signed integers
for proper calculations, while calculated result (e.g. cumulative
average) is always positive. This patch reflects the above requirements
in our code.
[1] https://en.wikipedia.org/wiki/Moving_average#Cumulative_averageFixes: #5745
Signed-off-by: Bo Chen <chen.bo@intel.com>
Fix make error:
$ make -C BaseTools -j `nproc`
/bin/sh: line 1: python: command not found
make[1]: *** [GNUmakefile:11: test] Error 127
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Bump these crates need to be bumped together to avoid build issues from
shared structures.
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
EntryPoint had an optional entry_addr, but there is no usage of this
struct that makes it necessary that the address is optional.
Remove the Option to avoid being able to express things that are not
useful.
Signed-off-by: Julian Stecklina <julian.stecklina@cyberus-technology.de>
This fixes all typos found by the typos utility with respect to the config file.
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
There is a "LATENCY_SCALE" being used for calculating cumulative average
latency, so it should also be used for the latency of the first op.
See: #5712
Signed-off-by: Bo Chen <chen.bo@intel.com>
Update to the latest vm-memory and all the crates that also depend upon
it.
Fix some deprecation warnings.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
We need to make sure that SEV-SNP and TDX are not enabled at the same
time. As these two features belong to mutually exclusive hardware
vendors. So, we should make sure that these two features are not enabled
at the same. Thus, a compile time check for it.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
Creating some brief documentation for SEV-SNP, summarizing the links on
where to find more information about SEV-SNP, as well as how to build
and run Cloud Hypervisor on it.
This document is a work in progress and will be updated in future PRs
as we add support for it.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This feature flag gates the development for SEV-SNP enabled guest.
Also add a helper function to identify if SNP should be enabled for the
guest.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
The worker is running focal not bionic - this fixes build issues on that
worker machine due to the bionic version not being installable.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This is necessary for O_DIRECT based use of raw block devices which may
require access at a larger block size than that of a sector (512 bytes.)
Fixes: #5722
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
It doens't make sense to limit this to the async_io module and import
across modules so place in default create module.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Assume rax is 0xfee003e0 and the displacement is negative 0x60. The effective
address is then 0xfee00380. This is perfectly valid.
Example instruction:
c7 40 a0 00 10 00 00 movl $0x1000,-0x60(%rax)
Signed-off-by: Philipp Schuster <philipp.schuster@cyberus-technology.de>
Logically until we have handled the first operation the latency is
infinite; this logic was applied to the minimum latency originally but
this patch extends that logic to the maximum and average latency.
To prevent the initial average latency being skewed by the inclusion of
infinity the average value is initally seeded with the first measured
latency.
Fixes: #5704
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
In the documentation of function check_latest_events_exact use same events
argument name as in the implementation
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Since kernel v6.3 the vsock packet is not split over two descriptors
and is instead included in a single one.
This change is based on the discovery and fix identified by Stefano
Garzarella for the vm-virtio vsock implementation and adapted for our
very different codebase.
Fixes: #5691
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This commit builds on top of the `Monitor::subscribe` function and
makes it possible to broadcast events published from `event-monitor`
over D-Bus.
The broadcasting functionality is enabled if the D-Bus API is enabled
and users who wish to also enable the file based `event-monitor` can do
so with the CLI arg `--event-monitor`.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
There is no need for this struct to be public and since it is used in
this module the #[allow(dead_code)] invocation can be removed.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> vmm/src/device_manager.rs:1908:35
|
1908 | fn set_raw_mode(&mut self, f: &mut dyn AsRawFd) -> vmm_sys_util::errno::Result<()> {
| ^^^^^^^^^^^^^^^^ help: consider changing to: `&dyn AsRawFd`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
= note: `#[warn(clippy::needless_pass_by_ref_mut)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> src/bin/ch-remote.rs:397:52
|
397 | fn dbus_api_do_command(toplevel: &TopLevel, proxy: &mut DBusApi1ProxyBlocking<'_>) -> ApiResult {
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: consider changing to: `&DBusApi1ProxyBlocking<'_>`
|
= note: this is cfg-gated and may require further changes
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
= note: `#[warn(clippy::needless_pass_by_ref_mut)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> virtio-devices/src/transport/pci_common_config.rs:100:17
|
100 | queues: &mut [Queue],
| ^^^^^^^^^^^^ help: consider changing to: `&[Queue]`
|
= warning: changing this function will impact semver compatibility
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
= note: `#[warn(clippy::needless_pass_by_ref_mut)]` on by default
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> hypervisor/src/arch/x86/emulator/instructions/mod.rs:22:15
|
22 | platform: &mut dyn PlatformEmulator<CpuState = T>,
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: consider changing to: `&dyn PlatformEmulator<CpuState = T>`
|
= note: this is cfg-gated and may require further changes
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> net_util/src/queue_pair.rs:167:14
|
167 | tap: &mut Tap,
| ^^^^^^^^ help: consider changing to: `&Tap`
|
= warning: changing this function will impact semver compatibility
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: this argument is a mutable reference, but not used mutably
--> block/src/async_io.rs:68:28
|
68 | fn query_block_size(f: &mut File, block_size_type: BlockSize) -> std::io::Result<u64> {
| ^^^^^^^^^ help: consider changing to: `&File`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_pass_by_ref_mut
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
warning: boolean expression will never evaluate to 'true'
--> block/src/vhdx/vhdx_metadata.rs:136:20
|
136 | if disk_spec.block_size < BLOCK_SIZE_MIN && disk_spec.block_size > BLOCK_SIZE_MAX {
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= note: since `BLOCK_SIZE_MIN` < `BLOCK_SIZE_MAX`, the expression evaluates to false for any value of `disk_spec.block_size`
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#impossible_comparisons
= note: `#[warn(clippy::impossible_comparisons)]` on by default
&& was used when || should have been used
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Similar to balloon inflation, memory allocation is also constrained to
align with the page size. Therefore, memory is allocated in units of the
host page size, one page at a time, until all host pages that the memory
range requested by the guest are managed. If the requested size is
smaller than the page size, the entire page will still be allocated
because smaller allocations are not possible due to the page size
limitation.
Fixes: cloud-hypervisor#5369
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Currently, virtio-balloon can't work well with page size other than 4k.
The virtio-balloon always works in units of 4kiB (BALLOON_PAGE_SIZE), but
we can only actually discard memory in units of the host page size.
We get some idea from [1] to solve this issue.
What has been done in this commit:
For balloon inflation:
A bitmap is employed to track the memory range to be released in 4k
granularity. Once it accumulates to one host page size, the corresponding
page is released, and the bitmap is cleared to handle the next record.
This process continues until all the memory range is managed. Memory will
only be released when a consecutive set of balloon request entries from
the same host page reaches the full host page size. If a balloon request
entry from a different host page is encountered, the bitmap and the base
host page address will be reset. Consequently, memory is released in
units of the page size, ensuring efficient memory management. That's say
if memory range length to be released smaller than page size or if the
guest scatters requests each of whose size is smaller than page size
across different host pages no memory will be released.
[1] https://patchwork.kernel.org/project/qemu-devel/patch/20190214043916.22128-6-david@gibson.dropbear.id.au/
Fixes: cloud-hypervisor#5369
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Add pending removed vcpu check according to VcpuState.removing, which
can avoid cloud hypervisor hangup during continual vcpu resize.
Fix#5419
Signed-off-by: Yi Wang <foxywang@tencent.com>
These had to be previously disabled because a released binary has to be
used in course of the test scenario. The necessary functionality is now
supported with the release 34.0, thus these tests should now pass.
Fixes: #5660
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
We fixed the L2 and L3 level cache sharing issues and confirmed that the
L2 level cache is independent, while the L3 level cache is shared per-socket.
See:#5505
Signed-off-by: zhongbingnan <zhongbingnan@bytedance.com>
This patch modifies `event_monitor` to ensure that concurrent access to
`event_log` from multiple threads is safe. Previously, the `event_log`
function would acquire a reference to the event log file and write
to it without doing any synchronization, which made it prone to
data races. This issue likely went under the radar because the
relevant `SAFETY` comment on the unsafe block was incomplete.
The new implementation spawns a dedicated thread named `event-monitor`
solely for writing to the file. It uses the MPMC channel exposed by
`flume` to pass messages to the `event-monitor` thread. Since
`flume::Sender<T>` implements `Sync`, it is safe for multiple threads
to share it and send messages to the `event-monitor` thread.
This is not possible with `std::sync::mpsc::Sender<T>` since it's
`!Sync`, meaning it is not safe for it to be shared between different
threads.
The `event_monitor::set_monitor` function now only initializes
the required global state and returns an instance of the
`Monitor` struct. This decouples the actual logging logic from the
`event_monitor` crate. The `event-monitor` thread is then spawned by
the `vmm` crate.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
These tests require a binary from a previous release to perform a
migration. That binary doesn't yet have the shutdown fixes and thus
these tests are supposed to fail. The tests are to be re-enabled after
the next release.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
The fixup_msix_region() function added in
a718716831
made the assumption that MSI-X was always available. This is the case
with many VFIO devices and all our virtio devices but created regression
with MSI devices.
Simply return the existing region size if MSI-X is not supported by the
device.
Fixes: #5649
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
With the addition of the spinning waiting for the exit event to be
received in the CMOS device a regression was introduced into the CMOS
fuzzer. Since there is nothing to receive the event in the fuzzer and
there is nothing to update the bit the that the device is looping on;
introducing an infinite loop.
Use an Option<> type so that when running the device in the fuzzer no
Arc<AtomicBool> is provided effectively disabling the spinning logic.
Fixes: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=61165
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
As the container image can be used on both Intel and AMD,
ensure the SPDK binaries are compatible. This implies -march=skylake
passed to the underlaying toolchain.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Now feature "mshv" can be built together with "kvm". There is no need to
use "--no-default-features" any more.
Fixes: #5647
Signed-off-by: Bo Chen <chen.bo@intel.com>
The reset system is asynchronous with an I/O event (PIO or MMIO) for
ACPI/i8042/CMOS triggering a write to the reset_evt event handler. The
VMM thread will pick up this event on the VMM main loop and then trigger
a shutdown in the CpuManager. However since there is some delay between
the CPU threads being marked to be killed (through the
CpuManager::cpus_kill_signalled bool) it is possible for the guest vCPU
that triggered the exit to be re-entered when the vCPU KVM_RUN is called
after the I/O exit is completed.
This is undesirable and in particular the Linux kernel will attempt to
jump to real mode after a CMOS based exit - this is unsupported in
nested KVM on AMD on Azure and will trigger an error in KVM_RUN.
Solve this problem by spinning in the device that has triggered the
reset until the vcpus_kill_signalled boolean has been updated
indicating that the VMM thread has received the event and called
CpuManager::shutdown(). In particular if this bool is set then the vCPU
threads will not re-enter the guest.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Currently if container registry is inaccessible the image will be built
locally and that takes time. This patch adds support to use mirror
registry. To use a different registry CTR_IMAGE environment variable
must be set. For example:
CTR_IMAGE="registry/cloud-hypervisor" scripts/dev_cli.sh
or
export CTR_IMAGE="registry/cloud-hypervisor"
scripts/dev_cli.sh
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Split interrupt source group restore into two steps, first restore
the irqfd for each interrupt source entry, and second restore the
GSI routing of the entire interrupt source group.
This patch will reduce restore latency of interrupt source group,
and in a 200-concurrent restore test, the patch reduced the
average IOAPIC restore time from 15ms to 1ms.
Signed-off-by: Yong He <alexyonghe@tencent.com>
The VFIO tests based on the NVMe emulation framework require cause a
high number of file descriptors to be opened on the system. On systems
with a low limit on opened files, tests like test_vfio_user can possibly
fail with "too many open files" exeption. This issue is fixed by raising
the corresponding limit.
A tricky detail here is, that the limit has to be changed in the test
container and not on the host. As the the container is executed in the
privileged mode, the setting in it will override the host value.
Related: #5426
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Add integration test of coredump with no need pause.
As file of coredump has been tested in test_coredump(), so this
patch only test vm state after coredump.
Signed-off-by: Yi Wang <foxywang@tencent.com>
If the VMM is not already paused then pause the VM prior to executing
the coredump and then resume it after. If the VM is already paused then
the original state is maintained.
Signed-off-by: Yi Wang <foxywang@tencent.com>
As the commit 2f15d027c05f (KVM: x86: Properly handle APF vs disabled
LAPIC situation) has been merged into kernel 5.13, so update the minimum
performance host kernel version to that.
Signed-off-by: Yi Wang <foxywang@tencent.com>
The commit b92fe648e9 (vmm: cpu: Disable KVM_FEATURE_ASYNC_PF_INT in
CPUID) disabled APF (Asynchronous Page Fault) mechanism to address
problem that makes vcpu thread spin 100%. As the actual issue is in
KVM, which has been merged in commit 2f15d027c05f (KVM: x86: Properly
handle APF vs disabled LAPIC situation) since 2021, so it's okay to
re-enable APF now.
Signed-off-by: Yi Wang <foxywang@tencent.com>
Add MSHV_CREATE_DEVICE, MSHV_SET_DEVICE_ATTR ioctls to filters. These
ioctls are required to passthrough PCI devices on mshv.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
The pause of vcpu is async now, which makes the vm pause is not
synchronised. As virtio device calls paused_sync wait() to make
sure device_manager pause synchronously, if we make vcpu pause
synchronously, the vm pause can be synchronously then. After
vm.pause() returns the vm is really paused now.
This patch adds a AtomicBool variable to mark vcpu paused state,
to make sure the pause of CpuManager is synchronised.
Signed-off-by: Yi Wang <foxywang@tencent.com>
This test case creates a new qcow2 file using the image of ubuntu as
its backing file, and boot a virtual machine with this image file.
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This preserves any data that the backing file had on a cluster when
doing a write to a subset of that cluster. These writes cause a
performance penalty on creating new clusters if a backing file is
present.
This commit is based on crosvm implementation:
5ad3bc3459
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
Reads to qcow files with backing files will fall through to the backing
file if there is no allocated cluster. As of this change, a write will
still trash the cluster and hide any data already present.
This commit is based on crosvm implementation:
d8144a56e2
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This commit allows opening qcow with a backing file, which supports any
type implementing `BlockBackend`.
This commit is based on crosvm implementation:
9ca6039b03
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This commit introduces the trait `BlockBackend` with generic ops
including read, write and seek, which can be used for common I/O
interfaces for the block types without using `DiskFile` and `AsyncIo`.
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This commit merges crates `qcow`, `vhdx` and `block_util` into the
crate `block`, which can allow `qcow` to use functions from `block_util`
without introducing a circular crate dependency.
This commit is based on crosvm implementation:
f2eecc4152
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This commit introduces a generic type `FixedVhd` for its sync and async
disk I/O types, and the disk I/O types will use it directly instead of
maintaining a file themselves.
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
Using this over the LocalApic supports APIC IDs (and hence number of
vCPUs) above 256 (size increases from u8 to u32.)
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
If the VM has been configured but not yet booted, all we need to do to
support removing a device is to remove it from the config, so it will
never be created.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This fixes the valid VM config unit tests, which would otherwise fail
to deserialize their expected JSON config due to the missing "gdb" field.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
error: private item shadows public glob re-export
Error: --> hypervisor/src/mshv/mod.rs:42:27
|
42 | CpuIdEntry, FpuState, LapicState, MsrEntry, SpecialRegisters, StandardRegisters,
| ^^^^^^^^^^
|
note: the name `LapicState` in the type namespace is supposed to be publicly re-exported here
--> hypervisor/src/mshv/mod.rs:16:9
|
16 | pub use mshv_bindings::*;
| ^^^^^^^^^^^^^^^^
note: but the private item here shadows it
--> hypervisor/src/mshv/mod.rs:42:27
|
42 | CpuIdEntry, FpuState, LapicState, MsrEntry, SpecialRegisters, StandardRegisters,
| ^^^^^^^^^^
= note: `-D hidden-glob-reexports` implied by `-D warnings`
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
warning: usage of `Arc<T>` where `T` is not `Send` or `Sync`
--> virtio-devices/src/vsock/device.rs:376:22
|
376 | backend: Arc::new(RwLock::new(backend)),
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
= help: consider using `Rc<T>` instead or wrapping `T` in a std::sync type like `Mutex<T>`
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#arc_with_non_send_sync
= note: `#[warn(clippy::arc_with_non_send_sync)]` on by default
The vsock backend may be shared between threads, so the type `B` in
`Vsock` should be `VsockBackend` and `Sync`.
Considering that `api_receiver` and `gdb_receiver` are only used in vmm
threads, the `Arc` can be replaced by `Rc`.
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
warning: useless use of `vec!`
--> test_infra/src/lib.rs:111:30
|
111 | let mut events = vec![epoll::Event::new(epoll::Events::empty(), 0); 1];
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: you can use an array directly: `[epoll::Event::new(epoll::Events::empty(), 0); 1]`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#useless_vec
= note: `#[warn(clippy::useless_vec)]` on by default
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
warning: casting raw pointers to the same type and constness is unnecessary (`*const protocol::MemoryRange` -> `*const protocol::MemoryRange`)
--> vm-migration/src/protocol.rs:280:17
|
280 | self.data.as_ptr() as *const MemoryRange as *const u8,
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: try: `self.data.as_ptr()`
|
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#unnecessary_cast
= note: `#[warn(clippy::unnecessary_cast)]` on by default
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
This gives users the chance to reduce the number of dependencies
included, which is generally good practice and also reduces code size.
Furthermore, `io_uring` specifically is a strong contender for something
one may wish to disable due to the syscall API's many security issues[1]
[1]: https://security.googleblog.com/2023/06/learnings-from-kctf-vrps-42-linux.html
Signed-off-by: Manish Goregaokar <manishsmail@gmail.com>
test_cpu_affinity needs the number of host CPUs. Since it is possible
for the host to have more than 255 CPUs; increase the size of the
datatype used for parsing the string to accomodate this.
Signed-off-by: dom.song <dom.song@amperecomputing.com>
Add integration test for pvpanic, by two methods:
- the vendor id and device id of pci device in guest
- triggering a guest panic and check event-monitor.
Also, to support pvpanic-pci driver, add pvpanic config
in resources.
Signed-off-by: Yi Wang <foxywang@tencent.com>
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Implemetion of BusDevice and PciDevice trait for pvpanic device.
Pvpanic device can be implemented as an ISA device or as a
PCI device, we choose PCI device for now.
Signed-off-by: Yi Wang <foxywang@tencent.com>
Introduce emulation of pvpanic device to allow cloud hypervisor to get
the notify from guest's pvpanic driver when guest kernel crash.
Signed-off-by: Yi Wang <foxywang@tencent.com>
This reverts commit efb579b224.
This PR was mistakenly merged due to the confusion around nightly
builds. The vm-memory update needs to be done in both the root workspace
and fuzz workspace together.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
On x86-64, when the underlying hypervisor platform is KVM, no
instruction emulator is necessary. KVM handles instruction boundaries
internally.
This change allows to skip the iced-x86 dependency on KVM, improving
build times, prunes the dependency graph and reduces network traffic
during the initial build.
For Hyper-V, the emulator is still necessary on x86-64, so nothing
changes there.
Signed-off-by: Christian Blichmann <cblichmann@google.com>
Otherwise, CONFIG_DEBUG_INFO will be enabled automatically bloating the
size of the kernel image.
Now the size of kernel image is reduced from 635M to 60M on x86_64.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Remove clippy:all and update it with correct clippy lint,
so that it gives proper information.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Between musl and glibc there is a difference in the signature of the
ioctl libc function. Use an anonymous cast to force the type coversion.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
According the std docs implementing From<..> is preferred since it
gives you Into<..> for free where the reverse isn’t true.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Add a Default implementation that delegates to new, so that clippy
can be removed.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Fix these warnings:
SC2086 info: Double quote to prevent globbing and word splitting.
DL3047 info: Avoid use of wget without progress bar. Use `wget --progress=dot:giga <url>`. Or consider using `-q` or `-nv` (shorthands for `--quiet` or `--no-verbose`).
SC2006 style: Use $(...) notation instead of legacy backticks `...`.
Ignore these warning cause they break the build or they do not apply:
DL3008 warning: Pin versions in apt get install. Instead of `apt-get install <package>` use `apt-get install <package>=<version>`
DL3003 warning: Use WORKDIR to switch to a directory
SC2016 info: Expressions don't expand in single quotes, use double quotes for that.
SC2046 warning: Quote this to prevent word splitting.
DL4006 warning: Set the SHELL option -o pipefail before RUN with a pipe in it. If you are using /bin/sh in an alpine image or if your shell is symlinked to busybox then consider explicitly setting your SHELL to /bin/ash, or disable this check
SC2155 warning: Declare and assign separately to avoid masking return values.
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
hadolint error:
DL3015 info: Avoid additional packages by specifying `--no-install-recommends`
Add required packages ca-certificates unzip iproute2 dbus
Without these packages build or integration tests fail if
--no-install-recommends is specified. Previously these packages were
installed as part of recommended dependency packages.
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Setting environment variable
DEBIAN_FRONTEND=noninteractive
Makes unnecessary setting it in all apt-get install commands
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Some parts of Dockerfile are indented with spaces others with tabs.
Unify indentation to use spaces.
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Test was failing due to regression caused by commit
d5558aea2a
Failing command:
sudo /mnt/ch-remote --api-socket /tmp/ch_api.sock resize --memory=1073741824"
Fixes#5190
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
This does the same thing as df2a7c17 ("vmm: Ignore and warn TAP FDs
sent via the HTTP request body"), but for the vm.create endpoint,
which also previously would accept file descriptors in the body, and
try to use whatever fd occupied that number as a TAP device.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Port of df2a7c17 ("vmm: Ignore and warn TAP FDs sent via the HTTP
request body"), but for the vm.create endpoint, which would previously
accept file descriptors in the body, and try to use whatever fd
occupied that number as a TAP device.
Since I had to move the wrapping of the net config in an Arc until
after it was modified, I made the same change to all other endpoints,
so the style stays consistent.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Using the data from sysfs forward the host host cache layout to the
guest using the FDT tables.
TEST=The host cache layout (from sysfs) can be seen in inside the guest
using lscpu.
Signed-off-by: zhongbingnan <zhongbingnan@bytedance.com>
Currently, vfio device fails to initialize as the msix-cap region in BAR
is mapped as RW region.
To resolve the initialization issue, this commit avoids mapping the
msix-cap region in the BAR. However, this solution introduces another
problem where aligning the msix table offset in the BAR to the page
size may cause overlap with the MMIO RW region, leading to reduced
performance. By enlarging the entire region in the BAR and relocating
the msix table to achieve page size alignment, this problem can be
overcomed effectively.
Fixes: #5292
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
To avoid code duplication extract page related functions to their
own module and add utility functions for manipulating addresses
related to page sizes
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
This patch clarifies the assumptions we have regarding the guest address
space layout while creating memory mapping in E820 on x86_64 and fdt on
aarch64. It also explicitly checks on these assumptions and report
errors if these assumptions do not hold.
Signed-off-by: Bo Chen <chen.bo@intel.com>
This commit renames `ram_region_sub_size` to `ram_region_available_size`
and make its value align down to the default page size or hugepage
size of the current memory zone, which can prevent the memory zone from
being split into misaligned parts. And if the available size of ram
region is zero, this region will be marked as consumed even it has
unused space.
Note that there is two methods to use hugepages.
1. Specify `hugepages` for `memory` or `memory-zone`, if the
`hugepage_size` is not specified, the value can be got by `statfs`
for `/dev/hugepages`.
2. Specify a `file` in hugetlbfs for `memory-zone`, the hugepage size
can also be got by `statfs` for the file.
The value for alignment will be the hugepage size if this memory zone
is using hugepages, otherwise the value will be default page size of
system.
Fixes: #5463
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
The previous `arch_memory_regions` function will provide some memory
regions with the specified memory size and fill all the previous
regions before using the next one, but sometimes there may be no need
to fill up the previous one, e.g., the previous one should be aligned
with hugepage size.
This commit make `arch_memory_regions` function not take any
parameters and return the max available regions, the memory manager
can use them on demand.
Fixes: #5463
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
The original codes did not consider that the previous memory region
might not be full and always set it to the maximum size.
This commit fixes this problem by creating memory mappings based on
the actual memory details in both E820 on x86_64 and fdt on aarch64.
Fixes: #5463
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
The device manager will set tty or pty to raw mode, all the `\n` will
be LF without CR, which makes the output difficult to read.
This commit solves it by using `write` with `\r\n` instead of
`writeln`, which can print CR and LF explicitly.
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
Remove "enum_variant_names" clippy. Enumeration variant names should
specify their variant, not repeat the enumeration name.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Download firmware and focal image to fix issue of running vfio tests
with empty worlkoads directory:
$ scripts/dev_cli.sh tests --integration-vfio
+ cp /root/workloads/focal-server-cloudimg-amd64-custom-20210609-0.raw /root/workloads/vfio
cp: cannot stat '/root/workloads/focal-server-cloudimg-amd64-custom-20210609-0.raw': No such file or directory
+ cp /root/workloads/hypervisor-fw /root/workloads/vfio
cp: cannot stat '/root/workloads/hypervisor-fw': No such file or directory
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Program the APIC ID (CPUID leaf 0x1 EBX) with the CPU id. This resolves
an issue where the EDKII firmware expects the APIC ID to vary per-CPU.
Fixes: #5475
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Script `create-cloud-init.sh" uses the same cloud init data as the the CI
but this means it is somewhat overloaded with unhelpful network configuration
entries and scripts that are only needed for the CI.
For local testing data kept in test_data/cloud-init/ubuntu/local folder.
This contains minimum configuration data for user to test it locally.
Script assigns default IP address using "network-config" details
with --net "tap=" option. The default network interface will be "ens4".
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Script `create-cloud-init.sh" uses the same cloud init data as the the CI
but this means it is somewhat overloaded with unhelpful network configuration
entries and scripts that are only needed for the CI.
CI related data moved to test_data/cloud-init/ubuntu/ci folder.
Fixes: #4773
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
SerialBuffer uses VecDeque::extend, which calls realloc, which a
maximum buffer size of 1 MiB. Starting at allocation sizes of
128 KiB, musl's mallocng allocator will use mremap for the allocation.
Since this was not permitted by the seccomp rules, heavy write load
could crash cloud-hypervisor with a seccomp failure. (Encountered
using virtio-console, but I don't see any reason it wouldn't happen
for the legacy serial device too.)
Signed-off-by: Alyssa Ross <hi@alyssa.is>
In this way, our integration tests exercise the same set of build
features (e.g. "kvm,mshv") being used for releases.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Bump to the latest rust-vmm crates, including vm-memory, vfio,
vfio-bindings, vfio-user, virtio-bindings, virtio-queue, linux-loader,
vhost, and vhost-user-backend,
Signed-off-by: Bo Chen <chen.bo@intel.com>
Implemented a `TargetApi` enum to make the process of implementing
tests for the D-Bus and HTTP API more convenient.
Refactored `test_api_{create_boot, shutdown, pause_resume, delete}` tests
with the `TargetApi` enum to also implement tests for the D-Bus API.
Added a new test named `test_api_dbus_and_http_interleaved` that uses
both the HTTP and D-Bus API at the same time.
Modified integration test scripts to enable the `dbus_api` feature when
compiling and start a dbus-session when integration tests are run.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
Introduces three new CLI options, `dbus-service-name`,
`dbus-object-path` and `dbus-system-bus` to configure the DBus API.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
This commit applies the previously created seccomp filter
to the `DbusApi` thread.
Also encloses the main loop of the `DBusApi` thread using
`std::panic::catch_unwind` and `AssertUnwindSafe` in order to mirror
the behavior of the HTTP API.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
This commit adds support for graceful shutdown of the DBusApi thread
using `futures::channel::oneshot` channels. By using oneshot channels,
we ensure that the thread has enough time to send a response to the
`VmmShutdown` method call before it is terminated. Without this step,
the thread may be terminated before it can send a response, resulting
in an error message on the client side stating that the message
recipient disconnected from the message bus without providing a reply.
Also changes the default values for DBus service name, object path
and interface name.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
This commit introduces three new dependencies: `zbus`, `futures`
and `blocking`. `blocking` is used to call the Internal API in zbus'
async context which is driven by `futures::executor`. They are all
behind the `dbus_api` feature flag.
The D-Bus API implementation is behind the same `dbus_api` feature
flag as well.
Signed-off-by: Omer Faruk Bayram <omer.faruk@sartura.hr>
Rust documentation comments follows 80 char per line. But some of
these help options doesn't have space in between. Longlines of
text is starting from beginning of the next line instead of
proper format.
Before fix:
Options:
--cpus boot=<boot_vcpus>,max=<max_vcpus>,topology=<threads_per_core>:<cores_per_die>:<dies_per_package>:<packages>,kvm_hyperv=on|off,max_phys_bits=<maximum_number_of_physical_bits>,affinity=<list_of_vcpus_with_their_associated_cpuset>,features=<list_of_features_to_enable>
After fix:
Options:
--cpus boot=<boot_vcpus>, max=<max_vcpus>,
topology=<threads_per_core>:<cores_per_die>:<dies_per_package>:<packages>,
kvm_hyperv=on|off,
max_phys_bits=<maximum_number_of_physical_bits>,
affinity=<list_of_vcpus_with_their_associated_cpuset>,
features=<list_of_features_to_enable>
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Commit 21d40d7 ("main: reset tty if starting the VM fails") changed
start_vmm() to join the vmm thread if an error happens after the vmm
thread is started. The implementation put all the error-prone code that
is run after the vmm is started in a closure, to be able to always join
the vmm thread, regardless of any error happening. However, it missed
propagating the error that might happen inside the closure back to the
main function, after joining the vmm thread.
For some cmd line options, the above issue inhibits proper error
reporting when starting a VM with invalid commands, as many parameters
are parsed after the vmm is started, thus if such parsing fails, no
error will be reported back to the user.
See: #5435
Fixes: 21d40d7 ("main: reset tty if starting the VM fails")
Signed-off-by: Rafael Mendonca <rafaelmendsr@gmail.com>
The refactoring on deferring address space allocation (#5169) broke TDX,
as TDX initialization needs to access guest memory for encryption and
measurement of guest pages.
Signed-off-by: Bo Chen <chen.bo@intel.com>
In current implementation, memory region used in vfio is assumed to
align to 4k which may cause error when the PAGE_SIZE is not 4k, like on
Arm, it can be 16k and 64k.
Remove this assumption and align memory resource used by vfio to
PAGE_SIZE then vfio can run on host with 64k PAGE_SIZE.
Fixes: #5292
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
The current implementation of breadth first traversal for device tree
uses a temporary vector, therefore causes unnecessary memory copy.
Remove it and do it within vector nodes.
Signed-off-by: Hao Xu <howeyxu@tencent.com>
Unlike KVM, there's no internal handling for topoolgy under MSHV. Thus,
if no topology has been passed during the CH launch, use the boot CPUs
count to construct the topology struct.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Passing the CPUID leafs with the topology is integrated into the common
mechanism of setting and patching CPUID in Cloud Hypervisor. All the
CPUID values will be passed to the hypervisor through the register
intercept call.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
Originally the AML only accepted one hex number for PCI segment
numbering. Change it to accept two numbers. That makes it possible to
add up to 256 PCI segments.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
When I refactored this to centralise resetting the tty into
DeviceManager::drop, I tested that the tty was reset if an error
happened on the vmm thread, but not on the main thread. It turns out
that if an error happened on the main thread, the process would just
exit, so drop handlers on other threads wouldn't get run.
To fix this, I've changed start_vmm() to write to the VMM's exit
eventfd and then join the thread if an error happens after the vmm
thread is started.
Fixes: b6feae0a ("vmm: only touch the tty flags if it's being used")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Previously, we used two different functions for configuring ttys.
vmm_sys_util::terminal::Terminal::set_raw_mode() was used to configure
stdio ttys, and cfmakeraw() was used to configure ptys created by
cloud-hypervisor. When I centralized the stdio tty cleanup, I also
switched to using cfmakeraw() everywhere, to avoid duplication.
cfmakeraw sets the OPOST flag, but when we later reset the ttys, we
used vmm_sys_util::terminal::Terminal::set_canon_mode(), which does
not unset this flag. This meant that the terminal was getting mostly,
but not fully, reset.
To fix this without depending on the implementation of cfmakeraw(),
let's just store the original termios for stdio terminals, and restore
them to exactly the state we found them in when cloud-hypervisor exits.
Fixes: b6feae0a ("vmm: only touch the tty flags if it's being used")
Signed-off-by: Alyssa Ross <hi@alyssa.is>
In particular the Std::write() API requires that the value implements
AsBytes and copies the slice representation into the table data. This
avoids unaligned writes which can cause a panic with the updated
toolchain.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
For structures that are used in SDT ACPI tables it is necessary for them
to implement this trait for the newly safe Std::write() API.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This is used on older kernels where close_range() is not available.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Fixes: 505f4dfa ("vmm: close all unused fds in sigwinch listener")
On KVM this is provided by an ioctl, on MSHV this is constant. Although
there is a HV_MAXIMUM_PROCESSORS constant the MSHV ioctl API is limited
to u8.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Don't import via glob to avoid (unused) objects colliding in the
namespace. This fixes a beta clippy issue.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This fixes the following tests that have been consistently failing on
the CI:
[2023-04-22T07:00:53.760Z] failures:
[2023-04-22T07:00:53.760Z] common_parallel::test_focal_hypervisor_fw
[2023-04-22T07:00:53.760Z] common_parallel::test_focal_ovmf
I'm not sure of the origin of this check but it obviously dependent on
the underlying platform as the guest OS has not changed. Since it
depends on the host environment it doesn't make sense to assert for it.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
In this way, we can cover the scenario where a VM with hotplugged net
device using FDs can work properly with reboot.
Signed-off-by: Bo Chen <chen.bo@intel.com>
The custom 'clone' duplicates 'preserved_fds' so that the validation
logic can be safely carried out on the clone of the VmConfig.
The custom 'drop' ensures 'preserved_fds' are safely closed when the
holding VmConfig instance is destroyed.
Signed-off-by: Bo Chen <chen.bo@intel.com>
Preserved FDs are the ones that share the same life-time as its holding
VmConfig instance, such as FDs for creating TAP devices.
Preserved FDs will stay open as long as the holding VmConfig instance is
valid, and will be closed when the holding VmConfig instance is destroyed.
Signed-off-by: Bo Chen <chen.bo@intel.com>
When neither serial nor console are connected to the tty,
cloud-hypervisor shouldn't touch the tty at all. One way in which
this is annoying is that if I am running cloud-hypervisor without it
using my terminal, I expect to be able to suspend it with ^Z like any
other process, but that doesn't work if it's put the terminal into raw
mode.
Instead of putting the tty into raw mode when a VM is created or
restored, do it when a serial or console device is created. Since we
now know it can't be put into raw mode until the Vm object is created,
we can move setting it back to canon mode into the drop handler for
that object, which should always be run in normal operation. We still
also put the tty into canon mode in the SIGTERM / SIGINT handler, but
check whether the tty was actually used, rather than whether stdin is
a tty. This requires passing on_tty around as an atomic boolean.
I explored more of an abstraction over the tty — having an object that
encapsulated stdout and put the tty into raw mode when initialized and
into canon mode when dropped — but it wasn't practical, mostly due to
the special requirements of the signal handler. I also investigated
whether the SIGWINCH listener process could be used here, which I
think would have worked but I'm hesitant to involve it in serial
handling as well as conosle handling.
There's no longer a check for whether the file descriptor is a tty
before setting it into canon mode — it's redundant, because if it's
not a tty it just won't respond to the ioctl.
Tested by shutting down through the API, SIGTERM, and an error
injected after setting raw mode.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
If the VM is shut down, either it's going to be started again, in
which case we still want to be in raw mode, or the process is about to
exit, in which case canon mode will be set at the end of main.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Having PMU in guests isn't critical, and not all hardware supports
it (e.g. Apple Silicon).
CpuManager::init_pmu already has a fallback for if PMU is not
supported by the VCPU, but we weren't getting that far, because we
would always try to initialise the VCPU with KVM_ARM_VCPU_PMU_V3, and
then bail when it returned with EINVAL.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Cloud Hypervisor's vhost-user implementation will reconnect if it gets
disconnected from the backend. That means connections happen inside
the vhost-user seccomp sandbox, so all syscalls used in reconnecting
have to be allowed in that sandbox.
clock_nanosleep is used by Glibc, and nanosleep is used by musl.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Recently generated mshv-bindings has most of the registers
renamed. This patch renames some of the MSHV registers.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Currently we are overwriting the RUST_BACKTRACE if set
explicitly by user while running performance metrics tests
using dev_cli.sh
This change will allow user to set the RUST_BACKTRACE
while running performance metrics tests with dev_cli.sh
which invokes run_metrics.sh to run the performance binary.
We will set RUST_BACKTRACE to 1 if not set explicitly.
Signed-off-by: smit-gardhariya <gardhariya.smit@gmail.com>
Previously, we were only using it for PTYs, because for PTYs there's
no alternative. But since we have to have it for PTYs anyway, if we
also use it for TTYs, we can eliminate all of the code that handled
SIGWINCH for TTYs.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Now that the SIGWINCH listener has fallbacks for older kernels, we
don't expect it to routinely fail, so if there's an error setting it
up, we want to know about it.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This will allow the SIGWINCH listener to run on kernels older than
5.5, although on those kernels it will have to make 64 syscalls to
reset all the signal handlers.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
The PTY main file descriptor had to be introduced as a parameter to
start_sigwinch_listener, so that it could be closed in the child.
Really the SIGWINCH listener process should not have any file
descriptors open, except for the ones it needs to function, so let's
make it more robust by having it close all other file descriptors.
For recent kernels, we can do this very conveniently with
close_range(2), but for older kernels, we have to fall back to closing
open file descriptors one at a time.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
It seems like these examples were always intended to be doctests,
since there are lines marked with "#" so that they are excluded from
the generated documentation, but they were not recognised as doc tests
because they were not formatted correctly.
The code needed some adjustments so that it would actually compile and
run as doctests.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
When doctests are built, the crate is built with itself as a
dependency via --extern. This causes a compiler error if using a
module with the name same as the crate, because it's ambiguous whether
it's referring to the module, or the extern version of the crate, so
it's necessary to disambiguate when using the hypervisor module here.
Fixes running cargo test --doc --workspace.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Doc comments are Markdown, and can include HTML tags. Anything in
angle brackets will therefore be inserted as an HTML tag into
rustdoc's output. If that's not intentional, the left angle bracket
needs to be escaped.
I haven't fixed the doc comments in src/main.rs, because argh doesn't
understand the escaping, so the backslashes would show up in the
--help output. I've opened https://github.com/google/argh/issues/159
about that.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
These need to be //! comments, because they apply to the module as a
whole, not to whatever directly follows the comment. Using ///
comments here resulted in documentation being attached to the wrong
thing, or not rendered at all.
I've also checked the Markdown formatting of these comments as
rendered by rustdoc, and fixed it where appropriate.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
This doesn't need to be rendered in the HTML API documentation, and
wouldn't be formatted correctly if it was.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
It was not possible to build just hypervisor with Cargo's -p flag,
because it was not properly specifying the features it requires from
vfio-ioctls.
Signed-off-by: Alyssa Ross <hi@alyssa.is>
balloon_free_page_reporting test case should not
work as expected. The reason is that MSHV pins
all the pages during the memory map for the guest.
Those pages can not be altered without unpinning the pages.
MSHV does not support modifying the pages during the guest
life cycle. This test case can be enabled once we add
VA backed VM support.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Unfortunately the build had a transient error and so needed to be
restarted which generated a build with a different date tag.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This commit will add timeout parameter to performance metrics
testcases as runtime argument. Testcases are facing the timeout
sometimes when we run with MSHV hypervisor. This change will
provide mechanism to pass the timeout as per the need while
running the testcases. If nothing is passed, default timeout or
timeout set as per the testcase will be followed.
Signed-off-by: smit-gardhariya <gardhariya.smit@gmail.com>
Tests of rate-limiter and vfio are running on bare-metal, so we need to
ensure the custom kernel is updated and rebuilt for them.
Signed-off-by: Bo Chen <chen.bo@intel.com>
The tests require a fixed amount of huge page memory to run, rather than
encode that as fixed number of huge pages it should be calculated from
the size of memory required and the huge page size
Signed-off-by: Dom <peng6662001@163.com>
Significant API changes have occured, most significantly is the switch
to an approach which does not require vm-memory and can run no_std.
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
This latest stable Rust toolchain can help reduce the final binary size
around ~300k when comparing with building with Rust 1.62. Specifically,
the binary size now is 3.3M (previously 3.6M) with default feature, and
4.0M (previously 4.2M) with "kvm+mshv" feature (e.g. our release action
build) .
Signed-off-by: Bo Chen <chen.bo@intel.com>
Now cloud hypervisor will start signal thread to catch
SIGWINCH signal, cloud hypervisor then will resize the
guest console via vconsole.
This patch skip starting signal thread when there is no
need to resize guest console, such as console is not
configured.
Signed-off-by: Yong He <alexyonghe@tencent.com>
The PR #2333 added I/O rate limiter on block device, with some options
in `DiskConfig`. And the PR #2401 added rate limiter on virtio-net
device with same options, but it still throws `Error::ParseDisk`.
This commit fixes it with correct values.
Fixes: #2401
Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com>
Unused import 'xml' is redefined at:
> xml = ET.ElementTree(element=svg)
Hence, remove unused xml import.
Signed-off-by: Ruslan Mstoi <ruslan.mstoi@intel.com>
Any error from the backend or from the protocol handling code will now
result in an error reply being sent. This is cleanly achieved by
splitting the command handling out into its own method and using the
Rust Result<> based error handling to trigger the generation of the
error reply.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
This allows the implementation of PCI devices in a different process
using the vfio-user protocol.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
Replace the use of an enum with a bitfield representation which means
that is now possible to logical OR flags together.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
Once error occur, vcpu thread may exit, this should
be critical event for the whole VM, we should fire
exit event and set vcpu state.
If we don't set vcpu state, the shutdown process
will hang at signal_thread, which is waiting the
vcpu state to change.
Signed-off-by: Yong He <alexyonghe@tencent.com>
This test (which relies on nesting) is failing on the VFIO worker. The tests that use the
dedicated hardware pass fine.
See: #5190
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
We need to provide valid FDs while creating 'NetConfig' instances even
for unit tests. Closing invalid FDs would cause random unit test
failures.
Also, two identical 'NetConfig' instances are not allowed any more,
because it would lead to close the same FD twice. This is consistent
with the fact that a clone of a "NetConfig" instance is no
longer *equal* to the instance itself.
Fixes: #5203
Signed-off-by: Bo Chen <chen.bo@intel.com>
These are owned by the config (and are duplicated before being used to
create the `Tap` for the virtio-net device.)
By implementing Drop on NetConfig we have issues with moving out of
members that don't implement the Copy trait. This requires a small
adjustment to the unit tests that use the Default::default() function.
Fixes: #5197
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
The custom version duplicates any FDs that have been provided so that
the validation logic used on hotplug, which takes a clone of the config,
can be safely carried out.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
This code is indentical to what is in this repository. When a release
gets made we can then switch to that.
Fixes: #5122
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
If swtpm becomes unresponsive, guest gets blocked at "recvmsg" on tpm's
data FD. This change adds a timeout to the data fd socket. If swtpm
becomes unresponsive guest waits for "timeout" (secs) and continues to
run after returning an I/O error to tpm commands.
Signed-off-by: Praveen K Paladugu <prapal@linux.microsoft.com>
The updated image is configured in a same way as the previously used
2019, it has same
- Credentials
- Services configured, like SAC, SSH, RDP
- Size
All the Windows updates are applied so the state is current to the date.
Also, the latest stable version 0.1.229 of the VirtIO Windows drivers
is installed.
Signed-off-by: Anatol Belski <anbelski@linux.microsoft.com>
We can ideally defer the address space allocation till we start the
vCPUs for the very first time. Because the VM will not access the memory
until the CPUs start running. Thus there is no need to allocate the
address space eagerly and wait till the time we are going to start the
vCPUs for the first time.
Signed-off-by: Jinank Jain <jinankjain@microsoft.com>
This hypervisor leaf includes details of the TSC frequency if that is
available from KVM. This can be used to efficiently calculate time
passed when there is an invariant TSC.
TEST=Run `cpuid` in the guest and observe the frequency populated.
Fixes: #5178
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
Updates include:
- Add references to 'TDX Tools'
- Expand instructions on buidling and using TDShim
- Add version information of guest/host kernel, TDVF, TDShim being tested
Signed-off-by: Bo Chen <chen.bo@intel.com>
This is required for booting Linux:
From: https://lore.kernel.org/all/20221028141220.29217-3-kirill.shutemov@linux.intel.com/
"""
Virtualization Exceptions (#VE) are delivered to TDX guests due to
specific guest actions such as using specific instructions or accessing
a specific MSR.
Notable reason for #VE is access to specific guest physical addresses.
It requires special security considerations as it is not fully in
control of the guest kernel. VMM can remove a page from EPT page table
and trigger #VE on access.
The primary use-case for #VE on a memory access is MMIO: VMM removes
page from EPT to trigger exception in the guest which allows guest to
emulate MMIO with hypercalls.
MMIO only happens on shared memory. All conventional kernel memory is
private. This includes everything from kernel stacks to kernel text.
Handling exceptions on arbitrary accesses to kernel memory is
essentially impossible as handling #VE may require access to memory
that also triggers the exception.
TDX module provides mechanism to disable #VE delivery on access to
private memory. If SEPT_VE_DISABLE TD attribute is set, private EPT
violation will not be reflected to the guest as #VE, but will trigger
exit to VMM.
Make sure the attribute is set by VMM. Panic otherwise.
There's small window during the boot before the check where kernel has
early #VE handler. But the handler is only for port I/O and panic as
soon as it sees any other #VE reason.
SEPT_VE_DISABLE makes SEPT violation unrecoverable and terminating the
TD is the only option.
Kernel has no legitimate use-cases for #VE on private memory. It is
either a guest kernel bug (like access of unaccepted memory) or
malicious/buggy VMM that removes guest page that is still in use.
In both cases terminating TD is the right thing to do.
"""
With this change Cloud Hypervisor can boot the current Linux guest
kernel.
Reported-By: Jiaqi Gao <jiaqi.gao@intel.com
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
Do the following:
1. Use from_be_bytes to drop mutable slices.
2. Check for the exact buffer size throughout.
3. Simplify ptm_to_request where possible.
4. Make error messages style consistent.
Fix a typo in code comment while at it.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
There is no guarantee that the write can send the whole buffer at once.
In those rare occasions, we should return a sensible error.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The largest possible PTM response is only 16 bytes. Size the output
buffer correctly.
In the socket read function, rely on the caller to provide a
sufficiently large buffer. That eliminates another large stack variable.
In total this saves almost 8KB stack space.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Change "thead" to "thread".
Also make sure the two messages are distinguishable by adding "vmm" and
"vm" prefix.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
The number of aligned operations can not be larger than the number of
descriptors. Initializing the capacity to 1 is good enough per the
observation that most of time there is only one data descriptor in a
given request.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Before Linux v6.0, AArch64 didn't support "socket" in "cpu-map"
(CPU topology) of FDT.
We found that clusters can be used in the same way of sockets. That is
the way we implemented the socket settings in Cloud Hypervisor. But in
fact it was a bug.
Linux commit 26a2b7 fixed the mistake. So the cluster nodes can no
longer act as sockets. And in a following commit dea8c0, sockets were
supported.
This patch fixed the way to configure sockets. In each socket, a default
cluster was added to contain all the cores, because cluster layer is
mandatory in CPU topology on AArch64.
This fix will break the socket settings on the guests where the kernel
version is lower than v6.0. In that case, if socket number is set to
more than 1, the kernel will treat that as FDT mistake and all the CPUs
will be put in single cluster of single socket.
The patch only impacts the case of using FDT, not ACPI.
Signed-off-by: Michael Zhao <michael.zhao@arm.com>
Fix lowercase label to avoid "mkfs.fat: Warning: lowercase labels
might not work properly on some systems".
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
This patch adds a global execution timeout to the Jenkinsfile to avoid
infinite pending Jenkins pipelines, such as when certain worker nodes
are not available. The global execution timeout is now set to 4 hours
which is derived from total timeout of our longest stage (e.g. the
`Worker build`).
Fixes: #5148
Signed-off-by: Bo Chen <chen.bo@intel.com>
As a first time user of cloud-hypervisor and Rust environment
you get build errors with out this.
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@intel.com>
Right now integration test fails during the test run if
/dev/mshv or /dev/kvm does not exist. We should not
progress and exit early if not present.
Signed-off-by: Muminul Islam <muislam@microsoft.com>
Make the code more idiomatic by wrapping the actual size configured in
the returning Result type. This further allows simplifying
get_buffer_size.
The debug message in startup_tpm is more useful if it prints out the
actual size than the wanted size.
No functional change.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
In order to comply with latest TDX version, we rely onto the branch
kvm-upstream-2022.08.07-v5.19-rc8 from https://github.com/intel/tdx
repository. Updates are based on changes that happened in
arch/x86/include/uapi/asm/kvm.h headers file.
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Dependabot will create a branch on the repo for it's updates this
triggers the release action (because it's the same event as a tag) which
will then fail leading to dependabot PRs not being automerged. Instead
only run the release check test on PRs or tag creation.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
KSM doesn't work with MSHV stack since guest memory is pinned
(`pin_user_pages`) and pinned pages cannot be merged.
So, don't run the test for mshv.
Signed-off-by: Anirudh Rayabharam <anrayabh@linux.microsoft.com>
This makes it much more convenient to build binaries for profiling
without having to modify the Cargo.toml file.
Signed-off-by: Rob Bradford <robert.bradford@intel.com>
A few breaking changes:
1. `-vvv` needs to be written as `-v -v -v`.
2. `--disk D1 D2` and others need to be written as `--disk D1 --disk D2`.
3. `--option=value` needs to be written as `--option value`
Change integration tests to adapt to the breaking changes.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
Since argh does not support `--option=value`, we need to change the
integration test code to become `--option value`.
Signed-off-by: Wei Liu <liuwe@microsoft.com>
You will need to `chmod +x .git/hooks/pre-commit` to have it run on every
@@ -36,7 +58,7 @@ commit you make.
## Certificate of Origin
In order to get a clear contribution chain of trust we use the [signed-off-by language](https://01.org/community/signed-process)
In order to get a clear contribution chain of trust we use the [signed-off-by language](https://web.archive.org/web/20230406041855/https://01.org/community/signed-process)
used by the Linux kernel project.
## Patch format
@@ -79,12 +101,12 @@ you want to merge your changes to `cloud-hypervisor`:
1. Fork the [cloud-hypervisor](https://github.com/cloud-hypervisor/cloud-hypervisor) project
into your github organization.
2. Within your fork, create a branch for your contribution.
3. [Create a pull request](https://help.github.com/articles/creating-a-pull-request-from-a-fork/)
1. Within your fork, create a branch for your contribution.
1. [Create a pull request](https://help.github.com/articles/creating-a-pull-request-from-a-fork/)
against the main branch of the Cloud Hypervisor repository.
4. To update your pull request amend existing commits whenever applicable and
1. To update your pull request amend existing commits whenever applicable and
then push the new changes to your pull request branch.
5. Once the pull request is approved it can be integrated.
1. Once the pull request is approved it can be integrated.
sh 'scripts/dev_cli.sh tests --metrics -- -- --report-file /root/workloads/metrics.json'
}
}
stage('Upload metrics report') {
steps {
sh 'curl -X PUT https://cloud-hypervisor-metrics.azurewebsites.net/api/publishmetrics -H "x-functions-key: $METRICS_PUBLISH_KEY" -T ~/workloads/metrics.json'
}
}
}
}
stage('Worker build - Rate Limiter') {
agent { node { label 'focal-metrics' } }
when {
branch 'main'
beforeAgent true
expression {
return runWorkers
}
}
stages {
stage('Checkout') {
steps {
checkout scm
}
}
stage('Run rate-limiter integration tests') {
options {
timeout(time: 10, unit: 'MINUTES')
}
steps {
sh 'scripts/dev_cli.sh tests --integration-rate-limiter'
}
}
}
}
stage('Worker build - SGX') {
agent { node { label 'jammy-sgx' } }
when {
beforeAgent true
allOf {
branch 'main'
expression {
return runWorkers
}
}
}
stages {
stage('Checkout') {
steps {
checkout scm
}
}
stage('Run SGX integration tests') {
options {
timeout(time: 1, unit: 'HOURS')
}
steps {
sh 'scripts/dev_cli.sh tests --integration-sgx'
}
}
stage('Run SGX integration tests for musl') {
options {
timeout(time: 1, unit: 'HOURS')
}
steps {
sh 'scripts/dev_cli.sh tests --integration-sgx --libc musl'
}
}
}
post {
always {
sh "sudo chown -R jenkins.jenkins ${WORKSPACE}"
deleteDir()
}
}
}
stage('Worker build - VFIO') {
agent { node { label 'jammy-vfio' } }
when {
beforeAgent true
allOf {
branch 'main'
expression {
return runWorkers
}
}
}
stages {
stage('Checkout') {
steps {
checkout scm
}
}
stage('Run VFIO integration tests') {
options {
timeout(time: 1, unit: 'HOURS')
}
steps {
sh 'scripts/dev_cli.sh tests --integration-vfio'
}
}
stage('Run VFIO integration tests for musl') {
options {
timeout(time: 1, unit: 'HOURS')
}
steps {
sh 'scripts/dev_cli.sh tests --integration-vfio --libc musl'
}
}
}
post {
always {
sh "sudo chown -R jenkins.jenkins ${WORKSPACE}"
deleteDir()
}
}
}
}
}
}
post {
regression {
script {
if (env.BRANCH_NAME == 'main') {
slackSend(color: '#ff0000', message: '"main" branch build is now failing', channel: '#jenkins-ci')
}
}
}
fixed {
script {
if (env.BRANCH_NAME == 'main') {
slackSend(color: '#00ff00', message: '"main" branch build is now fixed', channel: '#jenkins-ci')
}
}
}
}
}
def cancelPreviousBuilds() {
// Check for other instances of this particular build, cancel any that are older than the current one
The following scenarios are supported by Cloud Hypervisor to bootstrap a VM, i.e.,
to load a payload/bootitem(s):
- Provide firmware
- Provide kernel \[+ cmdline\]\ [+ initrd\]
Please note that our Cloud Hypervisor firmware (`hypervisor-fw`) has a Xen PVH
boot entry, therefore it can also be booted via the `--kernel` parameter, as
seen in some examples.
### Custom Kernel and Disk Image
#### Building your Kernel
Cloud Hypervisor also supports direct kernel boot. For x86-64, a `vmlinux` ELF kernel (compiled with PVH support) is needed. In order to support development there is a custom branch; however provided the required options are enabled any recent kernel will suffice.
Cloud Hypervisor also supports direct kernel boot. For x86-64, a `vmlinux` ELF kernel (compiled with PVH support) or a regular bzImage are supported. In order to support development there is a custom branch; however provided the required options are enabled any recent kernel will suffice.
Direct kernel boot to userspace should work with a rootfs from most
distributions although you may need to enable exotic filesystem types in the
@@ -372,8 +387,8 @@ are all equal and welcome means of contribution. See the
## Slack
Get an [invite to our Slack channel](https://join.slack.com/t/cloud-hypervisor/shared_invite/enQtNjY3MTE3MDkwNDQ4LWQ1MTA1ZDVmODkwMWQ1MTRhYzk4ZGNlN2UwNTI3ZmFlODU0OTcwOWZjMTkwZDExYWE3YjFmNzgzY2FmNDAyMjI)
and [join us on Slack](https://cloud-hypervisor.slack.com/).
Get an [invite to our Slack channel](https://join.slack.com/t/cloud-hypervisor/shared_invite/enQtNjY3MTE3MDkwNDQ4LWQ1MTA1ZDVmODkwMWQ1MTRhYzk4ZGNlN2UwNTI3ZmFlODU0OTcwOWZjMTkwZDExYWE3YjFmNzgzY2FmNDAyMjI),
[join us on Slack](https://cloud-hypervisor.slack.com/), and [participate in our community activities](https://cloud-hypervisor.slack.com/archives/C04R5DUQVBN).
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
