In b/73062966, we add new AID ranges for each partition that doesn't
yet have them (system, system_ext, odm, product). We also add group
and passwd files to these partitions to be able to map these AIDs into
human readable user and group names, and vice versa.
All processes should be able to read all users and groups. We divide
the ranges into non-overlapping regions for each partition and we
namespace the names with the partition name as a prefix.
Allow domain r_file_perms to
/(system|product|system_ext)/etc/(group|passwd).
Vendor and odm passwd and group files already have this access, since
/(vendor|odm)/etc/* is already domain readable.
Example contents:
blueline:/ $ cat /system/etc/passwd
system_tom::6050:6050::/:/bin/sh
blueline:/ $ cat /product/etc/passwd
product_tom::7013:7013::/:/bin/sh
Bug: 73062966
Test: tree-hugger selinux denial during boot test
Change-Id: Ib4dc31778e95e952174e1365497feaf93dca7156
/metadata/ota will store critical bits necessary to reify
system and vendor partition state during an OTA. It will be accessed
primarily by first-stage init, recovery/fastbootd, and update_engine.
Bug: 136678799
Test: manual test
Change-Id: Ib78cb96ac60ca11bb27d2b2fe011482e64ba0cf8
Move wifi services out of system_server into a separate APK/process.
Changes:
a) Created sepolicy for the new wifi apk.
b) The new APK will run with network_stack uid (eventually will be moved
to the same process).
Used 'audit2allow' tool to gather list of permissions required.
Note: The existing wifi related permissions in system_server is left
behind to allow the module to be loaded into system_server or
network_stack process depending on device configuration.
Bug: 113174748
Test: Device boots up and able to make wifi connection.
Test: Tested hotspot functionality.
Test: Ran WifiManagerTest & WifiSoftApTest ACTS tests locally.
Test: Will send for wifi regression tests.
Change-Id: Id19643a235bf0c28238f2729926b893ac2025b97
(cherry-picked from c7aa90091e6bec70a31a643cc4519a9a86fb0b38)
There was some plan to use binder from init, but it was abandoned. As
ServiceManager was recently re-written in C++, and as part of a
continued effort to reduce large dependencies in init and make sure it
is reliable, I'm removing these here.
Bug: 135768100
Test: N/A
Change-Id: I12b57709399c87ee25f689b601572b19abf7fb00
There is no reason to deny dumpstate from reading /data/misc/logd on
user builds. Logpersist is disallowed from running on those builds,
so there is no harm in copying this directory.
Bug: 136978224
Test: build
Change-Id: Ia58bde10e1f45978975597cd2ea1951a784d3b49
Ueventd can't set properties currently, but this is an artificial
limitation, since ueventd communicates to init that it has finished
cold boot via a file, and init polls this file instead of returning to
the epoll loop, where properties are handled.
A related change replaces that file with a property and thus frees
ueventd to be able to set properties. This change creates the
cold_boot_done property type for this property and gives only ueventd
permissions to set it.
Bug: 62301678
Test: boot, check that properties are set
Change-Id: I40843b423b192ea841db6a82f648e5bab9738e0e
This property will be set by system_server (to indicate the currently
selected theme for device), and can be accessed by vendor init.rc.
avc: denied { read } for property=persist.sys.theme pid=0 uid=0 gid=0 scontext=u:r:vendor_init:s0 tcontext=u:object_r:theme_prop:s0 tclass=file
Bug: 113028175
Test: Set a vendor init trigger that waits on `persist.sys.theme`. Check
that the trigger fires without denial.
Change-Id: Ia85b1a8dfc118efdbb9337ca017c8fb7958dc386
Merged-In: Ibb4e392d5059b76059f36f7d11ba82cd65cbe970
(cherry picked from commit 75182a1ea6)
This CL fixes the following denials during libdm_test
that is part of VTS.
avc: denied { read } for comm="loop1" path=2F6D656D66643A66696C655F32202864656C6574656429
dev="tmpfs" ino=97742 scontext=u:r:kernel:s0 tcontext=u:object_r:appdomain_tmpfs:s0
tclass=file permissive=0
W loop1 : type=1400 audit(0.0:371): avc: denied { read } for
path=2F6D656D66643A66696C655F32202864656C6574656429 dev="tmpfs" ino=97742 scontext=u:r:kernel:s0
tcontext=u:object_r:appdomain_tmpfs:s0 tclass=file permissive=0
Bug: 135004816
Test: adb shell libdm_test
Change-Id: Ifb6d58ee6f032cdf3952a05667aa8696d6e2a2fa
Merged-Id: Ifb6d58ee6f032cdf3952a05667aa8696d6e2a2fa
Bug: http://b/135139675
Coverage files are written to /data/misc/trace (governed by the
method_trace_data_file selinux type). Allow all domains to access
(create directories, access files) this directory when native coverage
is enabled (by setting NATIVE_COVERAGE to true) in an userdebug or eng
build.
Also relax neverallow constraints to allow access to
method_trace_data_file for native coverage builds.
Test: Build 32-bit cuttlefish with coverage:
m NATIVE_COVERAGE=true COVERAGE_PATHS="*"
and verify that there are no selinux denials in kernel log and
logcat.
Change-Id: I3fe7c77612854b9de7de7a0ddd5cbf44a2f5c21e
apexd stops itself when it finds that it is running on a device with
flattened APEXes (i.e. ro.apex.updatable = false).
Bug: 133907211
Test: launch sdk_phone_x86_64
adb logcat -d | grep apexd | wc -l
returns 3
Change-Id: I7fa161b069aa34adb028194b55f367fe740a0cfc
Although this may appear very permissive, it ok since the current allow
rule already gives every domain access to /dev/ashmem.
Change-Id: I1f121a3c6a911819b2c3e0605a0544a039cb5503
Bug: 134161662
Test: Check logcat for Sepolicy denials (logcat -d | grep shmem)
/system/apex/com.android.runtime is labeled as runtime_apex_dir
and init is allowed to mount on it.
When TARGET_FLATTEN_APEX is true (= ro.apex.updatable is unset or set to
false), apexd is not used to activate the built-in flattened APEXes.
Init simply bind-mounts /system/apex to /apex.
However, there is a special case here. The runtime APEX is installed as
either /system/apex/com.android.runtime.debug or
/system/apex/com.android.runtime.release, whereas it should be activated
on /apex/com.android.runtime - without the .debug or .release suffix.
To handle that case, the build system creates an empty directory
/system/apex/com.android.runtime and the .debug or .release directory
is bind-mounted to the empty directory by init at runtime.
Bug: 132413565
Test: marlin is bootable
Merged-In: I3fde5ff831429723fecd1fa5c10e44f636a63f09
Change-Id: I3fde5ff831429723fecd1fa5c10e44f636a63f09
(cherry picked from commit 99902a175b)
The execmem capability indicates that the processes creates anonymous
executable memory, which is most commonly used for JITing functionality.
All of the healthd executable code comes from the filesystem, and
healthd does not rely on JITing or loading code from non-file based
sources, so this permission is unnecessary.
Bug: 32659667
Test: compiles and boots
Change-Id: Ifb2b68625b191cb002dbb134cace6ddd215236e8
These lines are copied from update_engine.te, and are needed to update
dynamic partitions in recovery.
Bug: 132943965
Test: sideload OTA on cuttlefish
Change-Id: Id03a658aac69b8d20fa7bb758530a4469c75cf9c
The space between 2K and 16K in /misc is currently reserved for vendor's
use (as claimed in bootloader_message.h), but we don't allow vendor
module to access misc_block_device other than vendor_init.
The change in the topic adds a `misc_writer` tool as a vendor module,
which allows writing data to the vendor space to bridge the gap in the
short term. This CL adds matching labels to grant access.
Long term goal is to move /misc as vendor owned, then to provide HAL
access from core domain (b/132906936).
Bug: 132906936
Test: Build crosshatch that includes misc_writer module. Invoke
/vendor/bin/misc_writer to write data to /misc.
Change-Id: I4c18d78171a839ae5497b3a61800193ef9e51b3b
