When pausing a block I/O device model as part of suspending a VM, wait
for all active block I/O requests to finish before saving snapshot
data. This avoids having to save information about in-flight requests
both in the block_if layer and in storage device models.
For the AHCI device model, the queues are now guaranteed to be idle
when taking a snapshot, so remove the code to save queue state and
rely on the initial state in a resumed VM having all queues already
idle.
This will also simplify adding NVMe snapshot support in the future.
Reviewed by: jhb
Sponsored by: vStack
Differential Revision: https://reviews.freebsd.org/D26267
Allow clients of blockif to register a resize callback handler. When
a callback is registered, register an EVFILT_VNODE kevent watching the
backing store for a change in the file's attributes. If the size has
changed when the kevent fires, invoke the clients' callback.
Currently resize detection is limited to backing stores that support
EVFILT_VNODE kevents such as regular files.
Reviewed by: grehan, markj
MFC after: 1 month
Differential Revision: https://reviews.freebsd.org/D30504
Replace the existing ad-hoc configuration via various global variables
with a small database of key-value pairs. The database supports
heirarchical keys using a MIB-like syntax to name the path to a given
key. Values are always stored as strings. The API used to manage
configuation values does include wrappers to handling boolean values.
Other values use non-string types require parsing by consumers.
The configuration values are stored in a tree using nvlists. Leaf
nodes hold string values. Configuration values are permitted to
reference other configuration values using '%(name)'. This permits
constructing template configurations.
All existing command line arguments now set configuration values. For
devices, the "-s" option parses its option argument to generate a list
of key-value pairs for the given device.
A new '-o' command line option permits setting an individual
configuration variable. The key name is always given as a full path
of dot-separated components.
A new '-k' command line option parses a simple configuration file.
This configuration file holds a flat list of 'key=value' lines where
the 'key' is the full path of a configuration variable. Lines
starting with a '#' are comments.
In general, bhyve starts by parsing command line options in sequence
and applying those settings to configuration values. Once this is
complete, bhyve then begins initializing its state based on the
configuration values. This means that subsequent configuration
options or files may override or supplement previously given settings.
A special 'config.dump' configuration value can be set to true to help
debug configuration issues. When this value is set, bhyve will print
out the configuration variables as a flat list of 'key=value' lines.
Most command line argments map to a single configuration variable,
e.g. '-w' sets the 'x86.strictmsr' value to false. A few command
line arguments have less obvious effects:
- Multiple '-p' options append their values (as a comma-seperated
list) to "vcpu.N.cpuset" values (where N is a decimal vcpu number).
- For '-s' options, a pci.<bus>.<slot>.<function> node is created.
The first argument to '-s' (the device type) is used as the value of
a "device" variable. Additional comma-separated arguments are then
parsed into 'key=value' pairs and used to set additional variables
under the device node. A PCI device emulation driver can provide
its own hook to override the parsing of the additonal '-s' arguments
after the device type.
After the configuration phase as completed, the init_pci hook
then walks the "pci.<bus>.<slot>.<func>" nodes. It uses the
"device" value to find the device model to use. The device
model's init routine is passed a reference to its nvlist node
in the configuration tree which it can query for specific
variables.
The result is that a lot of the string parsing is removed from
the device models and centralized. In addition, adding a new
variable just requires teaching the model to look for the new
variable.
- For '-l' options, a similar model is used where the string is
parsed into values that are later read during initialization.
One key note here is that the serial ports use the commonly
used lowercase names from existing documentation and examples
(e.g. "lpc.com1") instead of the uppercase names previously
used internally in bhyve.
Reviewed by: grehan
MFC after: 3 months
Differential Revision: https://reviews.freebsd.org/D26035
Save and restore (also known as suspend and resume) permits a snapshot
to be taken of a guest's state that can later be resumed. In the
current implementation, bhyve(8) creates a UNIX domain socket that is
used by bhyvectl(8) to send a request to save a snapshot (and
optionally exit after the snapshot has been taken). A snapshot
currently consists of two files: the first holds a copy of guest RAM,
and the second file holds other guest state such as vCPU register
values and device model state.
To resume a guest, bhyve(8) must be started with a matching pair of
command line arguments to instantiate the same set of device models as
well as a pointer to the saved snapshot.
While the current implementation is useful for several uses cases, it
has a few limitations. The file format for saving the guest state is
tied to the ABI of internal bhyve structures and is not
self-describing (in that it does not communicate the set of device
models present in the system). In addition, the state saved for some
device models closely matches the internal data structures which might
prove a challenge for compatibility of snapshot files across a range
of bhyve versions. The file format also does not currently support
versioning of individual chunks of state. As a result, the current
file format is not a fixed binary format and future revisions to save
and restore will break binary compatiblity of snapshot files. The
goal is to move to a more flexible format that adds versioning,
etc. and at that point to commit to providing a reasonable level of
compatibility. As a result, the current implementation is not enabled
by default. It can be enabled via the WITH_BHYVE_SNAPSHOT=yes option
for userland builds, and the kernel option BHYVE_SHAPSHOT.
Submitted by: Mihai Tiganus, Flavius Anton, Darius Mihai
Submitted by: Elena Mihailescu, Mihai Carabas, Sergiu Weisz
Relnotes: yes
Sponsored by: University Politehnica of Bucharest
Sponsored by: Matthew Grooms (student scholarships)
Sponsored by: iXsystems
Differential Revision: https://reviews.freebsd.org/D19495
The Windows virtio driver ignores the advertized seg_max field and
assumes the host can accept up to 67 segments in indirect descriptors,
triggering an assert in the bhyve process.
This brings back r282922 but with a couple of changes:
- It raises the block interface segment limit to 128 instead of 67.
- Linux's virtio driver assumes that the segment limit is no
larger than the ring size. To avoid breaking Linux guests,
raise the VirtIO ring size to 128, and cap the VirtIO segment
limit at ring size - 2 (effectively 126).
Reviewed by: rgrimes, Patrick Mooney <pmooney@pfmooney.com>
Obtained from: Joyent (Linux workaround)
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D18831
The initial work on bhyve NVMe device emulation was done by the GSoC student
Shunsuke Mie and was heavily modified in performan, functionality and
guest support by Leon Dang.
bhyve:
-s <n>,nvme,devpath,maxq=#,qsz=#,ioslots=#,sectsz=#,ser=A-Z
accepted devpath:
/dev/blockdev
/path/to/image
ram=size_in_MiB
Tested with guest OS: FreeBSD Head, Linux Fedora fc27, Ubuntu 18.04,
OpenSuse 15.0, Windows Server 2016 Datacenter.
Tested with all accepted device paths: Real nvme, zdev and also with ram.
Tested on: AMD Ryzen Threadripper 1950X 16-Core Processor and
Intel(R) Xeon(R) CPU E5-2609 v2 @ 2.50GHz.
Tests at: https://people.freebsd.org/~araujo/bhyve_nvme/nvme.txt
Submitted by: Shunsuke Mie <sux2mfgj_gmail.com>,
Leon Dang <leon_digitalmsx.com>
Reviewed by: chuck (early version), grehan
Relnotes: Yes
Sponsored by: iXsystems Inc.
Differential Revision: https://reviews.freebsd.org/D14022
Mainly focus on files that use BSD 2-Clause license, however the tool I
was using misidentified many licenses so this was mostly a manual - error
prone - task.
The Software Package Data Exchange (SPDX) group provides a specification
to make it easier for automated tools to detect and summarize well known
opensource licenses. We are gradually adopting the specification, noting
that the tags are considered only advisory and do not, in any way,
superceed or replace the license texts.
No functional change intended.
While there is no issued with the number of descriptors in
a virtio indirect descriptor, it's a guest's choice as to
whether indirect descriptors are used. For the case where
they aren't, the virtio block ring size is still 64 which
is less than the now reported max_segs of 67. This results
in an assertion in recent Linux guests even though it was
benign since they were using indirect descs.
The intertwined relationship between virtio ring size,
max seg size and blockif queue size will be addressed
in an upcoming commit, at which point the max descriptors
will again be bumped up to 67.
The Windows virtio driver ignores the advertized seg_max
field and assumes the host can accept up to 67 segments
in indirect descriptors, triggering an assert in the bhyve
process.
No objection from: mav
Reviewed by: neel
Reported and tested by: Leon Dang (ldang@nahannisys.com)
MFC after: 2 weeks
GEOM does not support scatter/gather lists in its I/Os. Such requests
are cut in pieces by physio(), that may be problematic, if those pieces
are not multiple of provider's sector size. If such case is detected,
move the data through temporary sequential buffer.
MFC after: 2 weeks
It works only for virtual disks backed by ZVOLs and raw devices supporting
BIO_DELETE. Virtual disks backed by files won't report this capability.
MFC after: 2 weeks
Relnotes: yes
