This allows the atrace cmd to notify cameraserver (the host of
media.camera service) that the set of tracing-related system properties
have changed. This allows the cameraserver to notice that it might need
to enable its trace events.
The atrace cmd has the necessary permission when running as shell, but
not when it is running as the "atrace" domain (notably when exec'd by
perfetto's traced_probes).
We're adding cameraserver to the whitelist as it contains important
events for investigating the camera stack.
Example denial:
05-14 22:29:43.501 8648 8648 W atrace : type=1400 audit(0.0:389): avc: denied { call } for scontext=u:r:atrace:s0 tcontext=u:r:cameraserver:s0 tclass=binder permissive=0
Tested: flashed blueline-userdebug, captured a perfetto trace with "camera" atrace category, confirmed that userspace atrace events are included in the trace.
Bug: 130543265
Change-Id: Ifd3fd5fd3a737c7618960343b9f89d3bf7141c94
It doesn't make sense to write neverallow assertions where an attribute
negation exists allowing the operation. When such a negation exists,
domains can "opt-out" of the neverallow assertion by declaring their
use of the attribute. Such trivially bypassable assertions provide
no security nor architectural guarantees.
"netdomain" is such an attribute. This attribute is used by processes to
indicate that they communicate with the network, for example, using
TCP/UDP sockets. Vendor code is freely allowed to use network
communication by declaring their use of the attribute.
Because the attribute is usable to any vendor domain, the "no socket
connections to netd" restriction is pointless and provides a false sense
of security. Any process can opt-out of these restrictions by just
declaring their use of networking functionality. This also results in
ineffective policy bloat, making it difficult to reason about the policy
and make changes.
Delete the ineffective, misleading neverallow assertion.
Test: compiles
Change-Id: Ia72d9660a337ef811e56c9227af29b17d043b99f
Clatd is effectively an internal implementation detail of netd.
It exists as a separate daemon only because this gives us a better
security boundary. Netd is it's only launcher (via fork/exec) and
killer.
Generated via:
{ echo; cat public/clatd.te; echo; } >> private/clatd.te
rm -f public/clatd.te
plus a minor edit to put coredomain after clatd type declaration
and required changes to move netd's clatd use out of public into private.
Test: build and install on non-aosp test device, atest, check for selinux clat denials
Change-Id: I80f110b75828f3657986e64650ef9e0f9877a07c
These denials are intermittent and unnecessary. Hide them while we
investigate how to properly fix the issue.
Bug: 131096543
Bug: 132093726
Test: Build
Change-Id: I1950c10a93d183c19c510f869419fcfccd5006d2
(cherry picked from commit 654ceeb93f)
No longer needed, since this is now done by netd.
In a separate commit so it can potentially not be backported to Q
if we so desire.
Test: build/installed on crosshatch with netd/clatd changes,
and observed functioning ipv4 on ipv6 only network with no
avc denials
Bug: 65674744
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Change-Id: Id927ee73469d3e90f5111bd5e31ed760a58c8ebe
Kernel commit da69a5306ab92e07224da54aafee8b1dccf024f6
("selinux: support distinctions among all network address families")
modified the kernel to support fine grain differentiation of socket
families, if userspace enables it (which Android does).
Modify the mtp SELinux policy to allow the use of pppox_socket
(needed for kernels 4.14 or greater) and the generic "socket" family
(for kernels below 4.14).
Bug: 130852066
Test: compiles
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Change-Id: I8ac4c2f98f823120060e51438b39254898f4a27e
VTS tests are run after flashing a GSI image on the device.
The properties ro.boot.dynamic_partitions and ro.boot.dynamic_partitions_retrofit
are currently placed in product partition and will be overwritten by the GSI image.
We need to move these properties to vendor partition so that they will be available
even after the device is flashed with GSI.
Bug: 132197773
Test: build and flash, adb getprop ro.boot.dynamic_partitions
Change-Id: Ib04896ef744d8d2daa5cb3feee2cbf45aae2ba51
Kernel commit da69a5306ab92e07224da54aafee8b1dccf024f6
("selinux: support distinctions among all network address families")
modified the kernel to support fine grain differentiation of socket
families, if userspace enables it (which Android does).
Modify the ppp SELinux policy to allow the use of pppox_socket
(needed for kernels 4.14 or greater) and the generic "socket" family
(for kernels below 4.14).
Addresses the following denials:
04-19 20:25:34.059 16848 16848 I pppd : type=1400 audit(0.0:8703): avc: denied { read write } for dsm=HS_Q path="socket:[171178]" dev="sockfs" ino=171178 scontext=u:r:ppp:s0 tcontext=u:r:mtp:s0 tclass=pppox_socket permissive=1
04-19 20:25:34.075 16848 16848 I pppd : type=1400 audit(0.0:8704): avc: denied { ioctl } for dsm=HS_Q path="socket:[171179]" dev="sockfs" ino=171179 ioctlcmd=0x7437 scontext=u:r:ppp:s0 tcontext=u:r:mtp:s0 tclass=pppox_socket permissive=1
Bug: 130852066
Test: compiles
Change-Id: I00cc07108acaac5f2519ad0093d9db9572e325dc
bpf programs/maps are now loaded by the bpfloader, not netd
Test: built/installed on crosshatch which uses eBPF - no avc denials
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Change-Id: I1ebd82e6730d62d1966da3c4634ecd78ce703543
