Commit Graph

2020 Commits

Author SHA1 Message Date
Tom Cherry 821cb5e16c Merge "init: run vendor commands in a separate SELinux context"
am: 8e09b0b953

Change-Id: I7e0272f29bd8bab029a9f9c07aa413c9172f6f5a
2017-10-02 20:40:03 +00:00
Tom Cherry 8e09b0b953 Merge "init: run vendor commands in a separate SELinux context" 2017-10-02 19:48:58 +00:00
Tom Cherry cb0f9bbc85 init: run vendor commands in a separate SELinux context
One of the major aspects of treble is the compartmentalization of system
and vendor components, however init leaves a huge gap here, as vendor
init scripts run in the same context as system init scripts and thus can
access and modify the same properties, files, etc as the system can.

This change is meant to close that gap.  It forks a separate 'subcontext'
init that runs in a different SELinux context with permissions that match
what vendors should have access to.  Commands get sent over a socket to
this 'subcontext' init that then runs them in this SELinux context and
returns the result.

Note that not all commands run in the subcontext; some commands such as
those dealing with services only make sense in the context of the main
init process.

Bug: 62875318
Test: init unit tests, boot bullhead, boot sailfish

Change-Id: Idf4a4ebf98842d27b8627f901f961ab9eb412aee
2017-09-29 13:06:26 -07:00
Tom Cherry aceda86059 Merge "init: remove non-protobuf single persistent property file"
am: 61578ab50b

Change-Id: Id4b622ebde1db431d5dcb8cc3edccc6843d4c72b
2017-09-27 20:38:29 +00:00
Tom Cherry 61578ab50b Merge "init: remove non-protobuf single persistent property file" 2017-09-27 20:35:09 +00:00
Mark Salyzyn 387dbf9ab0 Merge "init: report shutdown,container for sigterm of init"
am: d2b5a2417f

Change-Id: Ifb4ab5fa166ee055838c5ad52d69e0d0c07c7e59
2017-09-27 19:40:16 +00:00
Mark Salyzyn d2b5a2417f Merge "init: report shutdown,container for sigterm of init" 2017-09-27 19:32:31 +00:00
Tom Cherry 9614e4d4d9 init: remove non-protobuf single persistent property file
This has moved to being serialized with libprotobuf.

Test: persistent properties work on bullhead
Test: init unit tests
Change-Id: I26ebe135e37d352f9c53612301bde703144853e7
2017-09-27 18:22:35 +00:00
Tom Cherry 531cd0f6d3 Merge "init: cleanup exit() uses"
am: 9c568d0fc0

Change-Id: I52ee55a8717f60bbf47b0cbb2a3115c63a824f84
2017-09-27 17:14:17 +00:00
Tom Cherry 9c568d0fc0 Merge "init: cleanup exit() uses" 2017-09-27 17:10:31 +00:00
Mark Salyzyn 161b8626bd init: report shutdown,container for sigterm of init
Change HandleSigtermSignal() handler to report shutdown,container. Add
the new reason to bootstat.  Remove log stutter as
HandlPowerctlMessage will also do a LOG(INFO) reporting
shutdown,container as reason.

Sending SIGTERM to init is to allow a host OS to ask an Android
Container instance to shutdown.  The temptation is to report
shutdown,sigterm but that does not accurately describe the usage
scenario.

Test: compile
Bug: 63736262
Change-Id: I3c5798921bdbef5d2689ad22a2e8103741b570b4
2017-09-27 09:10:00 -07:00
Mark Salyzyn a27a48590a Merge "init: shutdown,thermal shut off display immediately"
am: 82ee4060ed

Change-Id: I7c81548a3bbfefa8fadbe27d168708b2cb1d125a
2017-09-27 15:00:00 +00:00
Tom Cherry 4a679454d7 init: cleanup exit() uses
Primarily, this fixes a bug where a forked child of property service
uses exit() instead of _exit, which has the unintended consequences of
running the global destructors of init proper, which leads to
unintended cleanup.

Secondly, this replaces the remaining calls of exit() that really
should be LOG(FATAL).

Test: boot sailfish
Change-Id: I779228e7d44a73186bc7685bb723c4b9278a0a2d
2017-09-26 16:30:03 -07:00
Mark Salyzyn bfd05b69a5 init: shutdown,thermal shut off display immediately
Remove a source of heat.

Test: manual
Change-Id: I9a7391ddb121b6748a4f49f4b5bfb77453ca1514
2017-09-26 14:06:51 -07:00
Tom Cherry 00a15fc9bf Merge "init: fix typo"
am: 15a14d1ad9

Change-Id: I3805f03671f0ed09320df88b96b5cd958a9f742d
2017-09-20 00:45:07 +00:00
Tom Cherry a141907ecf init: fix typo
Test: build
Change-Id: I2f6ff60356087f9fd6c6f24e3923d70b9e93dcf8
2017-09-19 13:13:17 -07:00
Tom Cherry 61b81a0341 Merge "init: use protobuf for serialization of persistent properties"
am: f5dba11085

Change-Id: I6177b5b86290a50884f6212d45a09604b69cffd6
2017-09-19 17:16:59 +00:00
Tom Cherry f5dba11085 Merge "init: use protobuf for serialization of persistent properties" 2017-09-19 17:07:00 +00:00
Tom Cherry a97faba653 init: use protobuf for serialization of persistent properties
I probably should have done this from the start...  There's a shim to
convert my manually serialized format to protobuf, and since that has
not yet shipped, it'll be reverted in a short period of time.

Test: init unit tests
Test: upgrade from legacy and intermediate property formats successfully
Change-Id: Iad25f6c30d0b44d294230a53dd6876222d1c785b
2017-09-19 09:56:31 -07:00
Tom Cherry a6d75117e2 Merge "init: add exec_background command"
am: 424ed42fb4

Change-Id: I9529146ea9b29b2fb2e2c92f3533efc4e3106943
2017-09-18 21:37:38 +00:00
Tom Cherry 424ed42fb4 Merge "init: add exec_background command" 2017-09-18 21:33:26 +00:00
Tom Cherry 3631c545b8 init: add exec_background command
This command functions similarly to `exec` except that it does not
cause init to halt executing commands until the process has
terminated.  It is useful for launching simple one time background
tasks.

Bug: 65736247
Test: create an exec_background service and see it function properly
Change-Id: I719c8b85479b65201770aedc0a13191303007c11
2017-09-18 12:19:47 -07:00
Tom Cherry 171587d10e Merge "init: fix crash when reboot is triggered by a builtin"
am: b92415cee2

Change-Id: I4137679d88e58c706f14adeeaf84b41426079de4
2017-09-15 21:14:08 +00:00
Tom Cherry b92415cee2 Merge "init: fix crash when reboot is triggered by a builtin" 2017-09-15 21:08:52 +00:00
Tom Cherry 3633a4014a init: fix crash when reboot is triggered by a builtin
Builtin commands may set the sys.powerctl property, which causes
reboot to be immediately processed.  Unfortunately, part of the reboot
processing involves clearing the action queue, so when this scenario
happens, ActionManager::ExecuteOneCommand() can abort due to its state
being unexpectedly changed.

Longer term, the real fix here is to split init and property service.
In this case, the property sets will be sent to property service and
the reboot will only be processed once property service responds back
to init that the property has been set.  Since that will not happen
within the action queue, there will be no risk of failure.

Short term, this change sets a flag in init to shutdown the device
before the next action is run, which defers the shutdown enough to fix
the crash, but continues to prevent any further commands from running.

Bug: 65374456
Test: force bullhead into the repro case and observe that it no longer
      repros

Change-Id: I89c73dad8d7912a845d694b095cab061b8dcc05e
2017-09-15 21:07:41 +00:00
Tom Cherry 6213ce0dc1 Merge "init: fix hiding of move constructors of Result<T>"
am: db7b8f5e28

Change-Id: I933df61a6d5a523b7872ae0d8af091722f7c0e90
2017-09-15 20:29:05 +00:00
Tom Cherry d1c9cd0499 init: fix hiding of move constructors of Result<T>
This is needed to have Result<Result<T>> work correctly.

Test: init unit tests
Change-Id: If7d23d1ea13f3727b567d3baf0eee1d8d0e5a196
2017-09-15 20:17:38 +00:00
Tom Cherry 698981df84 Merge "init: fix variable scope issue with ExpandArgs()"
am: 1973110355

Change-Id: I6a257ae81ee2b4a3c128f5d4f4724cf1534670c7
2017-09-12 16:37:34 +00:00
Tom Cherry 1973110355 Merge "init: fix variable scope issue with ExpandArgs()" 2017-09-12 16:31:13 +00:00
Tom Cherry 5e405cacb1 init: fix variable scope issue with ExpandArgs()
ExpandArgs() was factored out of Service::Start() to clean up init,
however this introduced a bug: the scope of expanded_args ends when
ExpandArgs() returns, yet pointers to the c strings contained within
those std::strings are returned from the function.  These pointers are
invalid and have been seen to cause failures on real devices.

This change moves the execv() into ExpandArgs() and renames it
ExpandArgsAndExecv() to keep the clean separation of Service::Start()
but fix the variable scope issue.

Bug: 65303004
Test: boot fugu
Change-Id: I612128631f5b58d040bffcbc2220593ad16cd450
2017-09-11 16:08:54 -07:00
Luis Hector Chavez 721b2d6035 Merge "init: Allow clean system shutdown upon SIGTERM"
am: 61cb88add2

Change-Id: If0a495fa067440f0b59969ac0dbc010b4687b52f
2017-09-07 20:06:58 +00:00
Treehugger Robot 61cb88add2 Merge "init: Allow clean system shutdown upon SIGTERM" 2017-09-07 19:59:29 +00:00
Luis Hector Chavez 9f97f47940 init: Allow clean system shutdown upon SIGTERM
This allows Android to cleanly shutdown when running in a PID namespace
in a way that does not rely on adbd running. This is useful to allow
Android to be running in a container and its lifetime managed by an
OCI-compliant tool.

Bug: 65415372
Test: `kill -TERM 1` as root is correctly dropped.
Test: `kill -TERM 1` from the init PID namespace causes init to cleanly shutdown.
Change-Id: Ia66ebdb436221919081bc4723337c0c7f1e53b09
2017-09-07 10:47:04 -07:00
Tom Cherry c1968fb0c9 Merge "Log pid for writes to sys.powerctl"
am: 2827106d7d

Change-Id: I2fb780c5809d4c18950c114c07a7363723aa121a
2017-09-06 17:09:08 +00:00
Tom Cherry a84e14da1e Log pid for writes to sys.powerctl
Unless a process logs that it is requesting a device to reboot, there
are no logs to show which process triggered a reboot.  This change
introduces such a log in property service such that system initiated
reboots can be clearly blamed back to a calling process.

Bug: 64214361
Test: reboot and check kernel log for reboot string
Change-Id: I18de33d2a0933d20bdb581025b78020c88c5c6eb
2017-09-05 12:47:08 -07:00
kaichieh 1aecf73b43 Merge "Add odm sepolicy support to selinux.cpp"
am: f899548cdb

Change-Id: I33f7ded656fa2e61e130f158a2d8915f68524b91
2017-09-04 04:55:58 +00:00
Treehugger Robot f899548cdb Merge "Add odm sepolicy support to selinux.cpp" 2017-09-04 04:45:33 +00:00
Wei Wang df9ac65c75 Merge "init: add option to read file fully on readahead"
am: d97a1710b8

Change-Id: I2a49bd384b1fe621314916c1f325c03b7fa43162
2017-09-02 03:01:14 +00:00
Wei Wang 02628f3b23 init: add option to read file fully on readahead
Bug: 62413151
Test: boottime, dumpcache
Change-Id: I1a7b69f0619428e4db31c5a7639c5d895c89ecdb
2017-09-01 15:15:51 -07:00
kaichieh eef4cd7d08 Add odm sepolicy support to selinux.cpp
init: support loading odm sepolicy

Currently init merges two sepolicy cil files:
    - /system/etc/selinux/plat_sepolicy.cil
    - /vendor/etc/selinux/nonplat_sepolicy.cil

This change replaces nonplat_sepolicy.cil with the following two files:
    - /vendor/etc/selinux/declaration/nonplat_declaration.cil
    - /vendor/etc/selinux/vender_sepolicy.cil

And support merging another default (but optional):
    - /odm/etc/selinux/odm_sepolicy.cil.

Bug: 64240127
Test: boot sailfish normally without odm.cil
Test: boot another device having odm.cil
Change-Id: I0b7f8c656c73ddb0fd46f2af3c625d7c81566f2f
2017-09-01 18:13:29 +08:00
Tom Cherry 39088d35ee Merge "use a single file for storing persistent properties"
am: e1f9a58c86

Change-Id: Iea9d022eee79e3a4ccf9f5ebc6739d565b6e51b1
2017-08-30 19:42:08 +00:00
Tom Cherry e1f9a58c86 Merge "use a single file for storing persistent properties" 2017-08-30 19:30:18 +00:00
Tom Cherry 16fad42007 use a single file for storing persistent properties
We have seen that storing persistent properties in separate files
causes increased boot latency compared to if they were stored in a
single contiguous file.

This change creates a simple format for a contiguously stored property
file, and adds the support for arbitrary characters in the names of
persistent properties, which previously had been restricted.  It has a
mechanism for converting older devices to the new format as well.

Bug: 64392887
Test: boot bullhead with new properties
Test: boot bullhead and verify old properties are converted to the new
      property file
Test: corrupt property file and ensure that it gets recovered from memory
Test: new unit tests
Change-Id: I60d8201d655ce5c97b33faae81d5ca8dbbb21a14
2017-08-29 17:45:06 -07:00
Tom Cherry 43e92299e9 Merge "init: fix signal handling and LOG(FATAL) in child processes"
am: 30bf4b7500

Change-Id: I9147d02a9e710f9e8068a9b3b061768daf2648e9
2017-08-28 21:55:45 +00:00
Tom Cherry 30bf4b7500 Merge "init: fix signal handling and LOG(FATAL) in child processes" 2017-08-28 21:45:38 +00:00
Tom Cherry c39a5082dc Merge "init: support setting rlimits per service"
am: 459aa1cac6

Change-Id: I4ef5abc3371ce52783bfb14669bcffe33febb73e
2017-08-28 19:39:08 +00:00
Treehugger Robot 459aa1cac6 Merge "init: support setting rlimits per service" 2017-08-28 19:27:08 +00:00
Tom Cherry 849b481786 Merge "init: log all failures of Service::Start()"
am: 9bde0dc769

Change-Id: Ibb27338bb694482108ffc558b2823c583a9701d3
2017-08-28 17:30:28 +00:00
Tom Cherry 9bde0dc769 Merge "init: log all failures of Service::Start()" 2017-08-28 17:22:11 +00:00
Tom Cherry 7ac013de7e init: support setting rlimits per service
Add a new service option, `rlimit` that allows a given rlimit to be
set for a specific service instead of globally.

Use the same parsing, now allowing text such as 'cpu' or 'rtprio'
instead of relying on the enum value for the `setrlimit` builtin
command as well.

Bug: 63882119
Bug: 64894637

Test: boot bullhead, run a test app that attempts to set its rtprio to
      95, see that the priority set fails normally but passes when
      `rlimit rtprio 99 99` is used as its service option.
      See that this fails when `rlimit rtprio 50 50` is used as well.
Test: new unit tests

Change-Id: I4a13ca20e8529937d8b4bc11718ffaaf77523a52
2017-08-28 10:19:50 -07:00