Skip to content

kmod

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

kmod

README.md

vpnhide -- Kernel module

Kernel-probe module that hides VPN interfaces from selected apps. Part of vpnhide.

Using VPN Hide? Install the matching APK and unified Core ZIP from the main guide. This directory builds an internal kernel payload input; users do not install a separate kmod ZIP.

The module does not modify the target app's process: there are no userspace function patches, injected framework classes, or module-owned anonymous memory regions. Detection paths outside the hooks listed below remain out of scope.

What it hooks

Hook target What it filters Detection path covered
dev_ioctl SIOCGIFFLAGS, SIOCGIFNAME, and other per-interface ioctls: returns -ENODEV for VPN interfaces Direct ioctl() calls from native code (Flutter/Dart, JNI, C/C++)
sock_ioctl SIOCGIFCONF: compacts VPN entries out of the returned interface array Interface enumeration via ioctl(SIOCGIFCONF)
rtnl_fill_ifinfo Trims VPN entries from RTM_NEWLINK netlink dumps via skb_trim and returns 0 getifaddrs() (which uses netlink internally), any netlink-based interface enumeration
inet6_fill_ifaddr Trims VPN entries from RTM_GETADDR IPv6 responses via skb_trim IPv6 address enumeration over netlink
inet_fill_ifaddr Trims VPN entries from RTM_GETADDR IPv4 responses via skb_trim IPv4 address enumeration over netlink
fib_route_seq_show Forward-scans for VPN lines and compacts them out with memmove /proc/net/route reads
ipv6_route_seq_show Forward-scans for VPN lines and compacts them out with memmove /proc/net/ipv6_route reads
fib_dump_info Trims IPv4 VPN route entries and public physical-interface host-route hints from netlink route dumps via skb_trim RTM_GETROUTE route table dumps
rt6_fill_node Trims IPv6 VPN route entries from netlink route dump replies via skb_trim IPv6 RTM_GETROUTE dumps
fib_nl_fill_rule Trims target-UID policy rules and VPN interface rules from netlink rule dumps via skb_trim RTM_GETRULE policy routing dumps
sock_setsockopt / sk_setsockopt entry redirect Returns -ENODEV for hidden VPN names and indices before socket state changes Raw SO_BINDTODEVICE / SO_BINDTOIFINDEX calls
filename_lookup, do_filp_open, vfs_getattr, iterate_dir entry redirects (optional) Makes VPN interface nodes look absent under sysfs and /proc/sys/net/*/{conf,neigh} stat/access/readlink, open/fstat, and getdents64 probes

All filtering is per-UID: only processes whose UID is a target in the config written to /proc/vpnhide_ctl see the filtered view. Everyone else (system services, VPN client, NFC subsystem) sees the real data.

The VFS row is additionally reboot-gated and disabled by default. Enable Settings → Experimental protection → Hide VPN filesystem paths and reboot. When disabled, those four probes are not registered, so the default boot pays no VFS hot-path trampoline cost. The same canonical setting also controls the equivalent optional KPM hooks.

Optional filesystem hook loader contract

filesystem_hiding is a read-only boolean module parameter consumed only by insmod; it is not a control-v2 field. VPN Hide Core loads the module with finit_module(2) and reads canonical optional feature filesystem_iface_paths to pass filesystem_hiding=1 or filesystem_hiding=0 accordingly. Omitting the parameter has the same disabled behavior as 0.

When enabled, module init registers filename_lookup, do_filp_open, vfs_getattr, and iterate_dir as one optional group. A registration failure rolls the whole group back; telemetry then omits hook bit 27 and reports a partial install. The control-v2 target mask remains a separate per-UID gate: loading the group does not hide paths for targets whose mask omits filesystem_iface_paths. Because the parameter is init-only and the module is not unloadable, changing it requires a reboot.

Why kernel-level?

Some anti-tamper SDKs read /proc/self/maps via raw svc #0 syscalls (bypassing any libc hook) and check ELF relocation integrity. No userspace interposition can hide from them.

Kernel kretprobes modify kernel function behavior, not userspace code. The target app's process memory, ELF tables, and /proc/self/maps are completely untouched.

GKI compatibility

The exported symbols used by the module are part of the GKI KMI. A build targets one GKI generation and its matching headers/CRCs; the C source stays identical across generations.

CI builds are provided for all 7 GKI generations: android12-5.10 through android16-6.12.

For old/non-GKI kernels or devices where the .ko cannot load, see the KPM backend. It covers the same kernel-level Native role through KernelPatch inline hooks and is packaged as vpnhide-kpm.zip.

Build

See BUILDING.md for the full guide (DDK Docker build, kernel source preparation, toolchain setup, Module.symvers generation).

./kmod/build.py --kmi android14-6.1

Delivery inside Core

The .ko is included in the unified Core bundle. The kmod build ZIP is an internal producer input, not a public installation choice. Core verifies its manifest, applicability and ownership before loading the pinned payload with finit_module; then it projects the canonical package policy into the control-v2 wire.

Bundled KPM/Zygisk alternatives are not conflicts. A compatible builtin driver has priority; a conflicting live/unknown owner is a different condition and cannot be resolved by blindly loading another module. Unknown loads remain held. The loaded .ko stays resident until reboot; ordinary rmmod is unsupported. Legacy standalone module inventory remains for migration/recovery only.

Target management

VPN Hide app (recommended): open the VPN Hide app (the lsposed APK). It lists all installed apps with icons, search, and checkboxes. Saves through VPN Hide Core, which writes the canonical JSON config and delivers it to the selected native backend immediately. Works on KernelSU, Magisk, and APatch. Managing targets this way is all a normal setup needs; the shell equivalents are in the collapsible For developers / advanced usage section at the end of this file.

Core writes /data/system/vpnhide_config.json, the persistent package-keyed desired state, and derives the /proc/vpnhide_ctl snapshot from it; LSPosed independently reads the canonical JSON from system_server.

Combined use with system_server hooks

Covering both native and Java API detection paths requires two layers, without placing vpnhide hooks in the target app's process:

  • vpnhide-kmod (this module) covers the native side: ioctl, getifaddrs() (netlink), /proc/net/route, /proc/net/ipv6_route, netlink address/route/rule dumps, pre-mutation SO_BINDTODEVICE / SO_BINDTOIFINDEX denial, and optional sysfs/proc-sys interface-path concealment.
  • lsposed hooks writeToParcel() on NetworkCapabilities, NetworkInfo, LinkProperties inside system_server -- stripping VPN data before Binder serialization reaches the app.

Together they cover the detection paths documented in docs/detection-vectors.md. That document also lists known gaps and environment-dependent signals.

KPM is the other kernel-level Native backend for this same role. Do not run KPM and the .ko at the same time; choose one kernel backend, then pair it with LSPosed for Java APIs.

Setup

  1. Install vpnhide-kmod as a KSU module (this module).
  2. Install lsposed as an LSPosed/Vector module and add "System Framework" to its scope (no other apps in scope).
  3. Pick target apps in the VPN Hide app -- it manages targets for both the kernel module and the system_server hooks.

Architecture notes

Why kretprobes work here

kretprobes instrument kernel functions by replacing their return address on the stack. Unlike userspace inline hooks (which modify instruction bytes), kretprobes:

  • Don't modify the target function's code in a way visible to userspace -- /proc/self/maps and the function's ELF bytes are unchanged
  • Keep the instrumentation outside the target app's process; kernel-level observability still depends on the device's access controls and debug surface
  • Can target eligible non-inlined functions with available symbols, including static functions; kprobe blacklists and compiler inlining can still prevent registration or leave a symbol off the live path

dev_ioctl calling convention (GKI 6.1, arm64)

int dev_ioctl(struct net *net,       // x0
              unsigned int cmd,       // x1
              struct ifreq *ifr,      // x2 -- KERNEL pointer
              void __user *data,      // x3 -- userspace pointer
              bool *need_copyout)     // x4

Important: x2 is a kernel-space pointer (the caller already did copy_from_user). Using copy_from_user on it will EFAULT on ARM64 with PAN enabled. The return handler reads via direct pointer dereference.

Why sock_ioctl, not dev_ifconf, for SIOCGIFCONF

SIOCGIFCONF does NOT go through dev_ioctl(). The call path is sock_ioctl → dev_ifconf() -- a completely separate function from dev_ioctl, which handles SIOCGIFFLAGS, SIOCGIFNAME, etc.

The natural choice would be to hook dev_ifconf directly, but Clang LTO can inline it into sock_do_ioctl while leaving an unused dev_ifconf symbol in kallsyms. A kretprobe can then register successfully without observing the live path. On 6.1+, SIOCGIFCONF is dispatched directly from sock_ioctl, so sock_do_ioctl is not a cross-version hook point either.

sock_ioctl is the stable hook point because (1) it is the file_operations->unlocked_ioctl callback for socket fds and therefore remains address-taken; (2) the supported GKI paths dispatch socket ioctls through it; and (3) after it returns, the ifconf data (ifreq array + ifc_len) is already in userspace, so the module can filter it uniformly via copy_from_user/copy_to_user.

The entry handler stashes the userspace argp; the return handler reads back the buffer, compacts out VPN entries, and updates ifc_len via put_user. Cost is one cmd == SIOCGIFCONF compare per socket ioctl for non-target paths.

rtnl_fill_ifinfo / inet_fill_ifaddr / inet6_fill_ifaddr: skb_trim

All three netlink fill functions are skipped the same way: the entry handler saves skb->len before the fill writes anything; the return handler calls skb_trim(skb, saved_len) to undo whatever was written, then returns 0 (success). The dump iterator sees a successful entry of zero new bytes and advances to the next interface/address.

We do not return -EMSGSIZE to skip a VPN entry. On Android 14 / 6.1 GKI kernels, the dump iterator interprets -EMSGSIZE on an empty skb as "buffer too small for even one entry" and retries the same entry forever — observed in production as a hang of getifaddrs() (issue #38). The skb_trim-and-return-0 path avoids the retry loop on every netlink dump function uniformly.

fib_route_seq_show: seq_file buffer compaction

fib_route_seq_show(struct seq_file *seq, void *v) appends one or more tab-separated route lines to seq->buf. Each call can write multiple lines (one per fib_alias in the routing table entry).

The kretprobe entry handler saves the seq pointer and seq->count (current buffer position) in ri->data. The return handler scans the newly written region [saved_count, seq->count) line by line, extracts the first tab-delimited field (interface name), and compacts out VPN lines using memmove. Finally, seq->count is adjusted to reflect the reduced content.

Why we save seq in the entry handler: in a kretprobe return handler, regs->regs[0] (x0 on arm64) contains the function's return value, not the original first argument. The original code tried to read seq from x0 in the return handler, which was reading the return value (0) as a pointer -- a bug that would crash or silently fail. The fix is standard kretprobe practice: save arguments in ri->data during the entry handler.

For developers / advanced usage

You do not need these to use the module — the VPN Hide app saves the config through core, which delivers it for you. They are for development and debugging only; there is no activator binary in 2.0, and a hand-written snapshot is replaced by core's next delivery.

# Push a control-config snapshot straight to the kernel (docs/protocol.md):
# header + folded debug flag + grouped target UIDs + mandatory end count
# (a0003ff = shared kernel hooks plus the optional .ko filesystem hook;
# control v2 uses bare hex).
adb shell su -c 'printf "vpnhide 2 config\ndebug 0\ntargets a0003ff 28b7\nend 1\n" > /proc/vpnhide_ctl'

License

MIT. See LICENSE.

The compiled module declares MODULE_LICENSE("GPL") as required by the Linux kernel to resolve EXPORT_SYMBOL_GPL symbols (register_kretprobe, proc_create, etc.) at runtime.