The branch main has been updated by markj: URL: https://cgit.FreeBSD.org/src/commit/?id=d26c8ae527bbf2cc3da98c0857f634f162622752
commit d26c8ae527bbf2cc3da98c0857f634f162622752 Author: Mark Johnston <[email protected]> AuthorDate: 2026-02-19 14:38:59 +0000 Commit: Mark Johnston <[email protected]> CommitDate: 2026-02-19 17:16:15 +0000 vmm.4: Add information on VM access control Add a section to vmm.4 explaining how vmm device file ownership works. MFC after: 2 months Sponsored by: The FreeBSD Foundation Sponsored by: Klara, Inc. Differential Revision: https://reviews.freebsd.org/D54742 --- share/man/man4/vmm.4 | 39 +++++++++++++++++++++++++++++++++++++++ usr.sbin/bhyve/bhyve.8 | 25 ++++++++++++++++++++----- 2 files changed, 59 insertions(+), 5 deletions(-) diff --git a/share/man/man4/vmm.4 b/share/man/man4/vmm.4 index 07c40541f404..9c16bff8398a 100644 --- a/share/man/man4/vmm.4 +++ b/share/man/man4/vmm.4 @@ -59,6 +59,29 @@ riscv: The CPUs must implement the H (hypervisor) RISC-V ISA extension. .Pp PCI device passthrough to a virtual machine requires hardware with VT-d support and is available only on amd64. +.Sh ACCESS CONTROL +Only the super-user and processes with write access to the +.Pa /dev/vmmctl +device file may create and destroy virtual machines. +By default, members of the +.Va vmm +group have such access. +Once created, a virtual machine may be destroyed only by that user or +the super-user. +.Pp +Unprivileged users must use +.Dq monitor mode +to run the virtual machine; in this mode, the virtual machine is automatically +destroyed when its device file is closed. +When running +.Xr bhyve 8 , +this mode can be selected by specifying the +.Fl M +flag. +.Pp +Virtual machines can be created in a jail if the jail has the +.Va allow.vmm +flag set. .Sh PCI PASSTHROUGH On amd64 where the hardware supports VT-d, PCI devices can be reserved for use by the hypervisor. @@ -99,6 +122,12 @@ The and .Va pptdevs3 variables can be used for additional entries. +.Pp +In general, PCI passthrough cannot be used when running +.Xr bhyve 8 +as an unprivileged user or in a jail, as this feature requires write +access to +.Pa /dev/pci . .Sh LOADER TUNABLES Tunables can be set at the .Xr loader 8 @@ -109,6 +138,14 @@ prompt before booting the kernel or stored in Maximum number of virtual CPUs. The default is the number of physical CPUs in the system. .El +.Sh FILES +.Bl -tag -width /dev/vmmctl -compact +.It Pa /dev/vmmctl +control interface for creating and destroying virtual machines. +.It Pa /dev/vmm/* +device interface for individual virtual machines. +.It Pa /dev/vmm.io/* +device interface for device memory mapped into virtual machines. .Sh EXAMPLES Reserve three PCI devices for use by the hypervisor: bus 10 slot 0 function 0, bus 6 slot 5 function 0, and bus 6 slot 5 function 1. @@ -143,8 +180,10 @@ back: .Xr bhyve 4 , .Xr loader.conf 5 , .Xr bhyve 8 , +.Xr bhyvectl 8 , .Xr bhyveload 8 , .Xr devctl 8 , +.Xr jail 8 , .Xr kldload 8 .Sh HISTORY .Nm vmm.ko diff --git a/usr.sbin/bhyve/bhyve.8 b/usr.sbin/bhyve/bhyve.8 index 4c73a119b1ed..be906419d74e 100644 --- a/usr.sbin/bhyve/bhyve.8 +++ b/usr.sbin/bhyve/bhyve.8 @@ -25,7 +25,7 @@ .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" -.Dd January 5, 2026 +.Dd January 23, 2026 .Dt BHYVE 8 .Os .Sh NAME @@ -117,8 +117,23 @@ on arm64 the package provides a U-Boot image that can be used to boot the guest. .Pp .Nm -runs until the guest operating system reboots or an unhandled hypervisor -exit is detected. +runs until the guest operating system reboots (if +.Ql monitor +mode is not enabled) or halts, or an unhandled hypervisor exit is +detected. +.Pp +Generally +.Nm +must be run by the super-user, but users belonging to the +.Va vmm +group can create and run virtual machines as well. +See +.Xr vmm 4 . +When run by an unprivileged user, +.Nm +must have access to any required resources such as disk images or +network devices. +PCI passthrough cannot be used by unprivileged users. .Sh OPTIONS .Bl -tag -width 10n .It Fl a @@ -605,8 +620,8 @@ addressing rules. The .Cm slirp backend can be used to provide a NATed network to the guest. -This backend has poor performance but does not require any network -configuration on the host system. +This backend has limited performance but does not require any network +configuration on the host system and can be used by unprivileged users. It depends on the .Pa net/libslirp port.
