--- name: "umvu" section: "1" platform: "linux" package: "vuos-umvu" version: "0.9.2-1" release: "Debian 13 (Trixie)" platforms: ["linux"] url: "https://new.carta.tech/umvu" --- # umvu(1) user-mode implementation of VUOS ## NAME **umvu** – user-mode implementation of VUOS ## SYNOPSIS **umvu** \[*options* ...\] *command* \[*args*\] ## DESCRIPTION VUOS: view based operating system. VUOS gives processes their own view of the system resources. VUOS is a different perspective on namespaces, anykernels and related concepts. The main idea behind VUOS is that it is possible to give processes their own *view* using partial virtual machines. A partial virtual machine intercepts the system call requests and operates like a filter: system call can be forwarded to the kernel of the hosting system or processed by the partial virtual machine hypervisor. In this way processes can see a mix of resources provided by the kernel (on which they have the same view of the other processes) and virtual resource. It is possible to mount filesystems, load networking stacks, change the structure of the file system tree, create virtual devices, etc. **umvu** is a user mode implementation of VUOS concepts It is a modular partial virtual machine. Before loading any module, **umvu** is completely transparent, a process behaves inside **umvu** as it would have behaved outside. [*vu\_insmod*(1)](/vu_insmod/1.md) is the tool to load modules, e.g. *vufuse* for file system virtualization, *vunet* for networking, *vudev* for virtual devices, etc. These are just some examples of modules provided by the VirtualSquare team: modules are dynamic library based plugins so can be designed and implemented independently and loaded/unloaded at run time. **umvu** (VUOS hypervisor) is just a user process so, while it gives new perspective for processes, it does not widen the attack surface of the kernel. ## OPTIONS - **-h**, **--help** Print a short help message and exit. - **-x**, **--nonesting** **umvu** provides nested virtualization. It means that the system call requests of the modules can refer to virtual resource. (for example it is possible to mount a file system stored on a virtual device). The **-x** or **--nonesting** option disables this feature. - **-S**, **--noseccomp** **umvu** uses a seccomp filter to speed-up the virtualization. The **-S** or **--noseccomp** option disables this feature. - **-f** *file*, **--rc** *file* Execute commands from file instead of the system wide initialization file */etc/vurc* and the standard personal initialization file *~/.vurc* - **-N**, **--norc** Do not run the system wide initialization file */etc/vurc* and the standard personal initialization file *~/.vurc* - **-V** *name*, **--vu\_name** *name* set the name of the view. This name can be read and set using [**vuname**(1)](/vuname/1.md) - **-d** *debugtags*, **--debugtags** *debugtags* set the debug tags enabled from the beginning. e.g. **-d s** enables log messages of system call requests. (see [**vudebug**(1)](/vudebug/1.md) for a detailed description) - **-D** *colorspec*, **--debugcols** *colorspec* set color and font effects for log messages. e.g. **-D p:r -D c:g#** means that path resolution messages will be displayed in red while module choice log will be in reverse green. Color definition can refer to enable and disabled categories (the option defines the colors for a successive activation). (see [**vudebug**(1)](/vudebug/1.md) for a detailed description) - **-l** *level*, **--loglevel** *level* set the debug level. Only messages having level less or equal of the parameter of this option will be displayed. (0 = emergency, 1 = alert, 2 = critical, 3 = error, 4 = warning, 5 = notice, 6 = info, 7 = debug). The default level is 4. - **-s** *level*, **--syslog** *level* use [**syslog**(2)](/syslog/2.md) for logging and log messages having level less or equal of the parameter of this option. (0 = emergency, 1 = alert, 2 = critical, 3 = error, 4 = warning, 5 = notice, 6 = info, 7 = debug). This option is disabled by default. ## EXAMPLES ### mount a file system image The following example mounts /tmp/linux.img on /mnt. start the hypervisor, and run a bash *inside* the partial virtual machine - ``` $ umvu bash ``` This is the prompt of the partial virtualized shell, let us change it to $$ to show the difference - ``` $ PS1='\$\$ '` ``` let us load vufuse: a user-mode implementation of FUSE (source compatible with FUSE modules) - ``` $$ vu_insmod vufuse ``` nothing is currently mounted on /mnt - ``` $$ ls /mnt ``` the following command mounts the filesystem image /tmp/linux.img - ``` $$ vumount -t vufuseext2 -o ro /tmp/linux.img /mnt ``` now the image has been mounted: - ``` $$ ls /mnt bin boot dev etc lib lost+found mnt proc sbin tmp usr $$ vuumount /mnt $$ ls /mnt $$ exit ``` We have left the partial virtual machine Comments: user can *mount* any filesystem they like, on any directory. The linux kernel is not involved for all the system calls related to files in the mounted filesystem. The effects of this *mount* is just *perceived* by the processes running in the partial virtual machine. **vumount** is just a wrapper to the **mount**(1) system call (the command **[mount(8)](/mount/8.md)** does much much more, it is setuid root and requires real uid to be root to permit filesystem mounting ([**mount**(8)](/mount/8.md) works in **umvu** adding a module of uid/gid virtualization). ### create a disk image, partition it, create a filesystem and mount it In this example an 1GiB empty file is seen as a virtual disk, this disk gets partitioned (GPT), an ext4 file system is created on its first partition and finally this file system is mounted on /mnt start the hypervisor, and run a bash *inside* the partial virtual machine - ``` $ umvu bash ``` This is the prompt of the partial virtualized shell, let us change it to $$ to show the difference - ``` $ PS1='\$\$ ' ``` let us load vudev and vufuse: vudev to virtualize devices and vufuse as in the previous example - ``` $$ vu_insmod vudev vufuse ``` create a 1 GiB large empty file - ``` $$ truncate -s 1G /tmp/disk $$ ls -l /tmp/disk -rw-r--r-- 1 renzo renzo 1073741824 Jun 3 11:55 /tmp/disk ``` let us mount the empty file as a partitioned virtual disk: - ``` $$ vumount -t vudevpartx /tmp/disk /dev/hda Bad MBR signature 0 0 ``` clearly if not a partitioned disk, yet. Let us add a partitioning scheme: - ``` $$ /sbin/gdisk /dev/hda GPT fdisk (gdisk) version 1.0.3 Partition table scan: MBR: not present BSD: not present APM: not present GPT: not present Creating new GPT entries. Command (? for help): n Partition number (1-128, default 1): First sector (34-2097118, default = 2048) or {+-}size{KMGTP}: Last sector (2048-2097118, default = 2097118) or {+-}size{KMGTP}: +200M Current type is 'Linux filesystem' Hex code or GUID (L to show codes, Enter = 8300): Changed type of partition to 'Linux filesystem' Command (? for help): n Partition number (2-128, default 2): First sector (34-2097118, default = 411648) or {+-}size{KMGTP}: Last sector (411648-2097118, default = 2097118) or {+-}size{KMGTP}: Current type is 'Linux filesystem' Hex code or GUID (L to show codes, Enter = 8300): Changed type of partition to 'Linux filesystem' Command (? for help): p Disk /dev/hda: 2097152 sectors, 1024.0 MiB Sector size (logical): 512 bytes Disk identifier (GUID): F2A76123-73ED-4052-BAFE-6B37473E6187 Partition table holds up to 128 entries Main partition table begins at sector 2 and ends at sector 33 First usable sector is 34, last usable sector is 2097118 Partitions will be aligned on 2048-sector boundaries Total free space is 2014 sectors (1007.0 KiB) Number Start (sector) End (sector) Size Code Name 1 2048 411647 200.0 MiB 8300 Linux filesystem 2 411648 2097118 823.0 MiB 8300 Linux filesystem Command (? for help): w Final checks complete. About to write GPT data. THIS WILL OVERWRITE EXISTING PARTITIONS!! Do you want to proceed? (Y/N): Y OK; writing new GUID partition table (GPT) to /dev/hda. The operation has completed successfully. The disk has been partitioned: $$ ls -l /dev/hda1 brw------- 0 renzo renzo 0, 1 Jan 1 1970 /dev/hda1 $$ ls -l /dev/hda2 brw------- 0 renzo renzo 0, 2 Jan 1 1970 /dev/hda2 ``` Now it is possible to create an ext4 partition on /dev/hda1 - ``` $$ /sbin/mkfs.ext4 /dev/hda1 mke2fs 1.45.1 (12-May-2019) warning: Unable to get device geometry for /dev/hda1 Creating filesystem with 204800 1k blocks and 51200 inodes Filesystem UUID: c96c6499-40cd-43df-addf-52e06d7e6842 Superblock backups stored on blocks: 8193, 24577, 40961, 57345, 73729 Allocating group tables: done Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done ``` now the file system on /dev/hda1 can be mounted on /mnt - ``` $$ vumount -t vufuseext2 -o rw+ /dev/hda1 /mnt ``` add a significative file on /mnt - ``` $$ echo ciao * /mnt/hello $$ ls -l /mnt total 13 -rw-r--r-- 1 renzo renzo 5 Jun 3 12:09 hello drwx------ 2 root root 12288 Jun 3 12:06 lost+found $$ vuumount /mnt $$ vuumount /dev/hda $$ exit $ ``` ### mount a user-level networking stack It is possible to provide network partial virtualization using the **vunet** module start the hypervisor, and run a bash *inside* the partial virtual machine - ``` $ umvu bash ``` This is the prompt of the partial virtualized shell, let us change it to $$ to show the difference - ``` $ PS1='\$\$ ' ``` let us load vunet - ``` $$ vu_insmod vunet ``` the following command #mounts# a vde network on /dev/net/myvde. (see https://github.com/rd235/vdeplug4) - ``` $$ vumount -t vunetvdestack vxvde:// /dev/net/myvde ``` vustack is the command to select the stack to use. - ``` $$ vustack /dev/net/myvde ip link 1: lo: *LOOPBACK* mtu 65536 qdisc noop state DOWN mode DEFAULT group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: vde0: *BROADCAST,MULTICAST* mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000 link/ether 7e:76:c0:d7:3b:37 brd ff:ff:ff:ff:ff:ff ``` without vustack I can still access the stack provided by the linux kernel - ``` $$ ip link 1: lo: *LOOPBACK,UP,LOWER_UP* mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth0: *BROADCAST,MULTICAST,UP,LOWER_UP* mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000 link/ether 80:aa:bb:cc:dd:ee brd ff:ff:ff:ff:ff:ff ``` let us start a bash using /dev/net/myvde as itsdfault net - ``` $$ vustack /dev/net/myvde bash $ PS1='\$N\$ ' ``` let us configure the net - ``` $N$ ip addr add 192.168.250.250/24 dev vde0 $N$ ip link set vde0 up $N$ ip route add default via 192.168.250.1 $N$ ip addr 1: lo: *LOOPBACK* mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: vde0: *BROADCAST,MULTICAST,UP,LOWER_UP* mtu 1500 qdisc pfifo_fast state UNKNOWN group default qlen 1000 link/ether 7e:76:c0:d7:3b:37 brd ff:ff:ff:ff:ff:ff inet 192.168.250.250/24 scope global vde0 valid_lft forever preferred_lft forever inet6 fe80::7c76:c0ff:fed7:3b37/64 scope link valid_lft forever preferred_lft forever $N$ ip route default via 192.168.250.1 dev vde0 192.168.250.0/24 dev vde0 proto kernel scope link src 192.168.250.250 $N$ ping 80.80.80.80 PING 80.80.80.80 (80.80.80.80) 56(84) bytes of data. 64 bytes from 80.80.80.80: icmp_seq=1 ttl=52 time=56.9 ms 64 bytes from 80.80.80.80: icmp_seq=2 ttl=52 time=57.9 ms ^C $N$ ``` ## SEE ALSO [vu\_insmod(1)](/vu_insmod/1.md), [vu\_lsmod(1)](/vu_lsmod/1.md), [vu\_rmmod(1)](/vu_rmmod/1.md), [vumount(1)](/vumount/1.md), [vuumount(1)](/vuumount/1.md), [vudebug(1)](/vudebug/1.md) ## AUTHOR VirtualSquare. Project leader: Renzo Davoli