Update the entries only when the list is modified by the runtime.
Check that the list wasn't concurrently modified when being read.
Bug: 124287208
Test: libunwindstack_test
Test: art/test.py -b --host -r -t 137-cfi
Change-Id: I87ba70322053a01b3d5be1fdf6310e1dc21bb084
Occasionally we see cases when 40ms polling is still too conservative.
Change to 10ms polling period. Since the polling happens only after PSI
signal and continues for 1sec this should not affect system performance.
Test: lmkd_unit_test
Bug: 129358844
Change-Id: Ib759b865b2104be23741fc0eacaa541e22d50dde
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Monitor for FUNCTIONFS_UNBIND as well, so that in the case where we get
FUNCTIONFS_BIND, FUNCTIONFS_UNBIND, FUNCTIONFS_BIND, we don't trigger an
assertion failure from seeing two FUNCTIONFS_BINDs.
Bug: http://b/129134256
Test: manual
Change-Id: I80af5f4b833513e932262638b9f8d76bbcb35504
* changes:
libbase: realpath is wrapped with TEMP_FAILURE_RETRY
liblp: Replace open with GetControlFileOrOpen
init: expand prop in 'file'
libcutils: android_get_control_file uses realpath.
If we fail to submit writes for some reason (e.g. the USB cable was
unplugged), another thread that's waiting on the write mutex can enter
SubmitWrites and attempt to resubmit the writes that we already failed
to submit, leading to a failed assertion of !IoBlock::pending.
Increment writes_submitted_ before actually calling io_submit, so we
skip over these writes and fall through to exit.
Bug: http://b/129134256
Test: manually unplugged a blueline
Change-Id: I2fd0034e45db22c8f637c81039ce686b7aa6a03b
The function do_for_partitions() takes a const std::string
reference, so it's inefficient to pass a std::string::c_str().
Test: TreeHugger
Change-Id: Ia84ed9ec691ee2f524c61dd25a81b2995bb0bb33
When a function argument takes a std::string, it's inefficient to
pass std::string::c_str(), since that creates an additional copy.
So we change these calling sites.
Test: TreeHugger
Change-Id: I1c0fde7275eb3ebd3baf05ec62581b0243655608
Update debuggerd to print BuildId information by default.
Bug: 120975492
Test: New unit tests pass.
Test: debuggerd -b <PID> shows build id information.
Test: tombstones include build id information.
Change-Id: I019b031113d0b77385516223c63455b868924440
There are a bunch of branches that check "id >= LOG_ID_MAX", but because
C++ hates you, this does a promotion to signed int despite the
fact that both sides of the comparison are the same enum with an
underlying type of unsigned int. (C++17 §7.6.3)
Return LOG_ID_MAX instead of a value that gets promoted to signed -1, to
avoid this.
Bug: http://b/129272512
Test: /data/nativetest64/logcat-unit-tests/logcat-unit-tests
Change-Id: I4b3ee662d76d5cc80d9a9625d17f7e5b5980de41
Previous change If154dc364711bf7c86f32e24ddcd10be359386de called
"lmkd: Do not downgrade/ignore events when swap is full" added SwapTotal
into meminfo structure without adding the field into events.logtag file.
This results in logs which missing field and all fields starting with
"SwapFree" get reordered as a result. Fix this by adding the missing field
into events.logtag.
Bug: 129274901
Test: Confirm correct information in the logcat
Change-Id: Ia4de3790a7e9d49a0e4cba8b3161a715eaf6532e
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
SetupCgroups is called by init process during early-init stage and is not
supposed to be called again by anyone else. Ensure that the caller is the
init process, make sure cgroup.rc file is written only one time, keep the
file descriptor to cgroup.rc file open by the init process to ensure all
its further mappings stay valid even if the file is deleted.
Bug: 124774415
Test: build, run, verify no errors or warning in the logcat
Change-Id: I70ccec551fc07c380333566f618b969667dcf783
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Some of the legacy devices may not have vndk namespace. To make this
work, libnativeloader should not fail when it fails to search for vndk
namespace.
Bug: 129011845
Test: m -j
Change-Id: I6f68aeb6ae84a00985cc20561b2fcdb461be9724
Merged-In: Id9147d4aeee9ba2b301b6893dea5884dce9c14fa
Although man page for realpath doesn't say so, the bionic
implementation of realpath may exit with error code
EINTR. In such cases, retry.
Test: boots (sanity)
Change-Id: Ic5feb7152374a81bd2f475ad74c4bc8c3afb3a20
On taimen-eng build, the test reports:
[ WARNING ] user fastboot missing required to invalidate, ignoring a failure
ERROR: expected "cat: /vendor/hello: No such file or directory"
got "Hello World! Fri Mar 22 08:56:32 PDT 2019"
[ FAILED ] vendor content after flash vendor
Which is a result of a corner case problem on devices that need to
use overlayfs to support adb remount, but do not have fastbootd as
required by Dynamic Android Partitions (DAP). These legacy non-DAP
devices we consider this a Known Issue.
The message however is too alarming when reporting this Known Issue
that the test has notified you it accepts. Use standard notification
output format, and change the result to a series of WARNING instead.
The output would then look like:
[ WARNING ] user fastboot missing required to invalidate, ignoring a failure
[ WARNING ] expected "cat: /vendor/hello: No such file or directory"
[ WARNING ] got "Hello World! Fri Mar 22 08:56:32 PDT 2019"
[ WARNING ] vendor content after flash vendor
Which still tells us there is an issue, but does not cause a state
of extreme alarm when looking at the test output.
Test: adb-remount-test.sh
Bug: 120448575
Bug: 128876541
Bug: 123079041
Change-Id: I1d8d60f23f7670ade7eae1be629264f7507e0cfd
Allow having properties in 'file' option of a service.
Test: boots (sanity)
Test: lpdumpd
Bug: 126233777
Change-Id: I55158b81e3829b393a9725fd8f09200690d0230f
If the path to android_get_control_file is a symlink, the final sanity
check will fail that the fd does not have the same path as the given
file. We can't expect callers to readlink() because this would change
the environment key. Instead, try to call realpath on both paths.
Bug: 126233777
Test: lpdump
Change-Id: I0df10d7dbe3e572b8335faad812e5cd80bff1733
