Test: Boot on bullhead.
Test: Ran the libbacktrace tests on bullhead.
Test: Added a temporary log message in the signal handler, and ran the
Test: backtrace tests.
Change-Id: I0a6888c9f311af2c8cc7fbb4929315911bd2bb3c
Replace a hard-coded 3 second sleep with logic to wait until we've
scanned USB devices once and they've all come online.
Before:
adb shell true 0.00s user 0.00s system 0% cpu 3.047 total
After:
adb shell true 0.00s user 0.00s system 9% cpu 0.041 total
Bug: http://b/37869663
Test: `time adb shell true` after adb kill-server
Change-Id: I251d42afb885908ed9d03167287594ea16650d3f
Use fdevent_run_on_main_thread to initialize mDNS in a thread and
register an fdevent from the main thread upon success.
This reduces the startup time of `adb server` by ~3 seconds when mDNS
can't be successfully started. With an already running adb server,
`time adb server nodaemon` goes from:
adb server nodaemon 0.00s user 0.16s system 4% cpu 3.817 total
to:
adb server nodaemon 0.00s user 0.01s system 1% cpu 0.665 total
Bug: http://b/37869663
Test: `adb server nodaemon` with an existing adb server
Change-Id: Ia5a1a2a138610f3bf6792400050ca68f95ae3734
Add a function to run a function on the main thread, to allow fdevents
that depend on a blocking function to be registered.
Bug: http://b/37869663
Test: adb_test on linux
Change-Id: I84a0b372360420b7647057297b8f437e8afa874e
First kill the process group before killing the cgroup to catch
the hopeful case that killing the cgroup becomes a no-op as all of its
processes have already been killed.
Do not report an error if kill fails due to ESRCH, as this happens
often when reaping processes due to the order in which we call
waitpid() and kill().
Do not call killProcessGroup in libprocessgroup if we have already
successfully killed and removed a process group.
Bug: 36661364
Bug: 36701253
Bug: 37540956
Test: Reboot bullhead
Test: Start and stop services
Test: Init unit tests
Change-Id: I172acf0f8e00189f910f865f4635a7b1782fc7e3
Private interface to permit testing only added to fs_config to
expose android_files and android_dirs.
Make sure that both paths to a partition are specified in fs_config
internal tables.
Test: gTest libcutils-unit-test --gtest_filter=fs_config.*
Bug: 37703469
Change-Id: Ida5fccdb786dc6d67325005d4fdd1fa1ffaef396
The exec_service documentation was difficult to read, clarify it.
Tests:
Run grip.py to verify that the markdown still works correctly.
Run aspell to verify spelling.
Change-Id: I29bdd456f3d3ea2a91c9d4772bd09a5a195f97a9
Signed-off-by: William Roberts <william.c.roberts@intel.com>
Also, add a link to the .clang-format-2 for this directory and clang
format the files that changed.
Bug: 31919199
Test: Boot bullhead.
Test: Run unit tests on bullhead. There are a few that fail, but they
Test: failed before and are not a result of this change.
Change-Id: I3d3b2111f6f6bf8a0d7039295d34d5168c191651
Files in the ramdisk by default have the rootfs label and must be
manually restoreconed.
Bug: 35219933
Change-Id: I2a749f128dc3a609907101ce703747f8990b4386
Invent keyutils.h to supply capability to set session keyring.
The keyring will hold things like the FBE encryption keys.
Test: gTest logd-unit-tests --gtest_filter=logd.statistics
Bug: 37751120
Bug: 36645158
Change-Id: Ieb44fa8f53dda6cf506a6243498c72d7f7f3cde7
For legacy reason, /data/data is a real dir and /data/user/0 is a
symbolic link to it. Overhead for linux kernel to walk through
symbolic link is not negligible. This is unnessary overhead to
carry over. This patch is to make /data/user/0 a a real dir and
make legacy folder /data/data a symbolic link. OTAed system does
not get impacted.
Test: Manual test
Change-Id: I419564a75f6ebf3154badb8725ba9831164592b6
Signed-off-by: cjbao <cathy.bao@intel.com>
Refine DAC security surrounding logd.daemon worker thread and add a
positive test for logd failure to access /data/system/packages.list.
- Add AID_PACKAGE_INFO to groups of worker thread.
- Move AID_SYSTEM to groups, setgid to AID_LOGD.
- Do not drop capabilities until after setting the uid and gids.
- Add a test that is part of logd.statistics test to check when
packagelistparser appears broken.
- If /data/system/packages.list is encrypted, ensure we do not pick
up the existing inode to ensure strong positive when finding access
problems.
- Replace all occurrences of NULL with nullptr in gTest code for
compliance with best practices.
Test: gTest logd-unit-tests --gtest_filter=logd.statistics
(expect consistent failure, later CLs fix)
Bug: 37751120
Bug: 36645158
Change-Id: I01b26fe5e25203246ae432d272c8daa9c07cab54
Delete the sysdeps/mutex tests that -Wthread-safety complains about.
We're using the standard library's std::mutex on all platforms now,
anyway.
Test: mma
Change-Id: I3bf958c72604b29dfb1d9c898d3c9aa34aed2685
Similar to what installkey used to do, init_user0 forks and
synchronously waits for vdc to return. This is dangerous to do in
init however as init also processes properties from a single thread.
I'm not aware of any specific issues that this is currently causing,
but it's a good preventative measure to match what installkey does and
use do_exec().
Test: Boot bullhead, see that init_user0 still happens
Change-Id: I853c61594fe3d97e91bbb2319ebddf2bbe80d457
Previously, adb was assuming a fixed maximum packet size of 1024 bytes
(the value for an endpoint connected via USB 3.0). When connected to an
endpoint that has an actual maximum packet size of 512 bytes (i.e.
every single device over USB 2.0), the following could occur:
device sends amessage with 512 byte payload
client reads amessage
client tries to read payload with a length of 1024
In this scenario, the kernel will block, waiting for an additional
packet which won't arrive until something else gets sent across the
wire, which will result in the previous read failing, and the new
packet being dropped.
Bug: http://b/37783561
Test: python test_device.py on linux/darwin, with native/libusb
Change-Id: I556f5344945e22dd1533b076f662a97eea24628e
Move the name of the "private/libc_logging.h" header to <async_safe/log.h>.
For use of libc_malloc_debug_backtrace, remove the libc_logging library.
The library now includes the async safe log functions.
Remove the references to libc_logging.cpp in liblog, it isn't needed because
the code is already protected by a check of the __ANDROID__ define.
Test: Compiled and boot bullhead device.
Test: Run debuggerd unit tests.
Test: Run liblog unit tests on target and host.
Test: Run libmemunreachable unit tests (these tests are flaky though).
Change-Id: Ie79d7274febc31f210b610a2c4da958b5304e402
Rather than quietly fall through to the legacy code with no explanation,
say why the first connection failed, and that the second attempt is only
going to be relevant for folks trying to sideload to a pre-KitKit device.
Before:
$ adb sideload mysid-ota-424425.zip
opening 'mysid-ota-424425.zip'...
connecting...
falling back to older sideload method...
error: closed
After:
$ adb sideload mysid-ota-424425.zip
adb: sideload connection failed: no devices/emulators found
adb: trying pre-KitKat sideload method...
adb: pre-KitKat sideload connection failed: no devices/emulators found
Also switch the legacy code to just read+write blocks rather than trying to
load the whole OTA package into RAM (which is in short supply on Win32).
This is probably of no practical use other than that it lets us report a
better error message (better than "terminate called after throwing an
instance of 'std::bad_alloc'").
Bug: https://issuetracker.google.com/37139736
Test: "adb sideload" with no device/a mysid/an angler
Change-Id: Ia568044aac61bd054301ebfd7fcd5ad5d6f3147a
The previous fix was taking account just the progress reported by dumpstate,
not progress/percentage. As such, it was not detecting the cases where the
percentage decreased but the progress didn't.
Bug: 37878670
Test: m -j32 adb_test && ./out/host/linux-x86/nativetest64/adb_test/adb_test --gtest_filter=BugreportTest.*
Change-Id: I5830028f3191a9b17f63aeed5c049b29fa7d1179
It has been reported that fs_mgr failed to open /dev/device-mapper
during the first stage mount. It's because other uevent (e.g., i2c
charger device) happens to be sent at the same time we're triggering
the device-mapper uevent to be sent. Current implementation returns
COLDBOOT_STOP unconditionally so it will only process the first received
uevent, leaving device-mapper uevent unhandled when the race happens.
Fix this by only returning COLDBOOT_STOP when the received uevent->path
matches that of device mapper.
Bug: 37745254
Test: first stage mount /vendor with vboot 2.0 (avb) on bullhead
Test: first stage mount /vendor with vboot 1.0 on sailfish
Change-Id: I4a77093ec8f90a5ca981a088f34d082d0270533b
Christopher Ferris