This change adds an explicit cleanup for the subcontext processes and
avoids them from respawning, which causes a bunch of LOG(FATAL)s when
the system is going down.
Bug: 80425914
Test: kill -TERM $INIT_PID, no crashes for subcontext inits
Change-Id: I135191d959c1dd921b102af316b24d2bc161d6c9
This CL is largely and adaptation of
Ie996def20e25dc1afe0c74af2096af844934b2dc
Bug: 80741439
Test: Successfully built product-services.img with one module in it, and
flashed on device. Also successfully built image with
/system/product-services and no dedicated /product-services partition.
Change-Id: I1046dfe19bf874da929c9131b76a1f85422fbb80
Merged-In: I1046dfe19bf874da929c9131b76a1f85422fbb80
For some platforms it is not known at build time what devices
will be attached at runtime. Building into the kernel or pre-loading
at init all the modules that could be needed would unnecessary bloat
the kernel. The solution is dynamic kernel module loading.
The kernel will generate uevents when devices are added, userspace
should monitor for these events and load the compatible modules.
The init process already monitors for uevents, add here the ability
to respond to modalias events and preform the correct action.
Adding this to init is preferred over an external program as we
can read and process the module alias and dependency files once,
instead of for each module needing to be loaded.
Test: Run on Beagle-X15, check all needed modules are loaded (lsmod)
Change-Id: I1b57d9aeb0a9770f309207183dc4bc2b7b905f14
Signed-off-by: Andrew F. Davis <afd@ti.com>
Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
Executables should be in /system/bin
rather than sbin. Change lookup paths
in a few places to reflect this.
Test: recovery is in system/bin
Bug: 78793464
Change-Id: Ibcdff7abf1f21e2566e02cad8bdb3c26ee500534
In preparation for splitting first stage init from the rest of the
init executable, rename these files which are specifically involved in
first stage mount operations to a more appropriate name.
Test: builds
Change-Id: I8a2d4e8c7e1deea1bab45cc8e738727bc2ceb3e5
Clean up a few mistakes in logging initialization
1) Only init needs to clear stdout/stderr/stdin, so remove this from
ueventd, watchdogd, and subcontext init
2) Only init should reboot due to FATAL errors. This was true even
before this change due to getpid() checks, but there's no reason to
not just use the DefaultAborter for other processes.
3) It's probably a mistake for FATAL logs in init to try to gracefully
shutdown the system, so simply call RebootSystem() here.
4) Lastly, remove log.cpp since it's not actually shared code anymore
Test: build
Change-Id: Ic8c323393dc7ee98ed6bb9691361b51d0d915267
This early code is no longer needed now that logical partitions can be
created and flashed normally.
Bug: 79173901
Test: N/A
Change-Id: I393ef23b3d3ce1cd9c80833358051838d0e9f333
When using AVB, init's first stage regenerates uevent to find the paths
of block devices for verified partitions. If it can't find them all, it
will panic, causing the device to boot to recovery.
This does not work with logical partitions, since devices for those
partitions are created later in the first stage. In fact, they cannot be
created until uevent regeneration completes, since logical partitions are
are created by finding the "super" partition.
To address this we exclude logical partitions (as marked in fstab) from
the device finding process. Note that partitions moved from GPT to liblp
will no longer appear in by_name_symlink_map_.
Bug: 79173901
Test: AVB 2.0 device boots after deleting the vendor partition,
creating a super partition, and flashing it with a dynamic
vendor partition.
Change-Id: I19371b05912240480dc50864a2c04131258a7103
Since some vendors will have firmware in mount points in
/mnt/vendor/..., we extend the ueventd script language to allow
specifying the firmware directories.
Also, move the existing 4 directories to ueventd.rc as a primary user
of this mechanism.
Bug: 111337229
Test: boot sailfish; firmwares load
Change-Id: I0854b0b786ad761e40d2332312c637610432fce2
1. init creates /mnt/product used to mount product-specific rw partitions.
2. If a device tree fstab entry for early mount specifies a mount point
under /mnt/product e.g. /mnt/product/foo, init will create
/mnt/product/foo mount point.
Bug: 110808288
Test: change dt fstab entry to mount /mnt/product/foo; mount point is
created correctly, and partition is mounted in early mount.
Change-Id: I321e314992abe1084fd67a382c205f5c0c92bf3d
After dealing with some specical escape characters('\n','\r','\t','\\',"\r\n")
it doesn't goto the next position in the next loop, so it process the current
character twice.
For example, when parsing the string "test\ntoken" we expect the
"test'\n'token" but actually we got the "test'\n'ntoken"
Test: have espace characters in init .rc files
Change-Id: I015c087a5c6e5ee9c490f29a83b15b89443f7f81
Signed-off-by: liwugang <liwugang@xiaomi.com>
This functionality is useful for improving boottimes on the ARC++
project. Without this change, ro.serialno would be set to the empty
string when androidboot.serialno was unset in the kernel commandline.
Bug: 62039211
Test: boot with androidboot.serialno unset and ensure ro.serialno is
unset
Change-Id: Iaee339dfa3f0c871e5e9c1fc0534347f2b3e8a07
Let's start by fixing the build, then see how possible this will be to do fully.
Bug: 110477913
Test: n/a
Change-Id: If52e15bd070781a405c5494631ede033c1067903
Also get rid of the copy in parser. There's no incentive to switch to
a tokenizer that doesn't modify the input, nor is there a reason to
waste cycles making a copy of every init script as its processed.
Bug: 36970783
Test: boot
Change-Id: I8aca9c9d6f1961e1ab35dee50691a6791fc6ec66
Currently, init can create logical partitions by hardcoding them in
fs_mgr or by specifying them in device-tree. This change allows init to
also create logical partitions by using liblp, which stores partition
tables in a physical partition. The current name for this partition is
"android".
Two aspects of this code will change long-term. One, the prototype code
using device-tree will be deleted once fastboot supports logical
partitions. Two, libdm will obsolete most of the code in
fs_mgr_dm_linear.cpp. For now however we preserve how the prototype code
functions and we layer liblp on top of the existing dm_linear logic.
Bug: 79173901
Test: N/A
Change-Id: If014a109da78fa12269bf0df0dda39028ac2d1aa
init is now dynamically linked to libselinux. This gives us about 100KB
extra space, which is valuable in the recovery partition.
Bug: 63673171
Test: m -j
Test: adb reboot recovery
Change-Id: I72c4f886cbbb9ce54f8221f05547f5b9f0e1adb0
It's not going to be possible to safely assume $OUT has the right init
scripts to be parsed at a given point, so instead we fall back to
parsing init scripts individually.
This isn't a full revert of the previous commits. We retain parsing
correctness of the 'import' statements and we retain using the new
host side property functionality.
Also, fix a bug where main was not actually returning -1 on failure
Bug: 36970783
Test: testing individual files still works correctly
Change-Id: I4ae5620f234caa08993deb2c30825904a75f6654
Treehugger Robot