AVB 1.0 support in fs_mgr has been removed in commit
Ibfb46aa6c2f761dbb3a9b5f0b16336e510417620.
Adding a VTS test case to ensure the legacy 'verify' fs_mgr
flag isn't used anymore.
Also converting GetFstabPath() to a public API in the fstab.h,
so the test case can read then parse the fstab file.
Bug: 204948957
Test: atest vts_fs_test
Change-Id: Ib6ed7cb8b6ad719b19cd876acf10f4312ecb51b6
On ctl.apex_load <apex_name> control message, init invokes linkerconfig
to update linker configuration for the updated apex.
Bug: 232173613
Test: CtsInitTestCases
Test: VendorApexHostTestCases
Change-Id: I01d975849c3f4efe74205a7b04ebbd6864ba1121
mmap_rnd_compat_bits is for address space randomization of 32-bit
applications on 64-bit system. Configuring it is not only unnecessary
for 64-bit "only" builds, but also can cause a boot failure if the
kernel is built without CONFIG_COMPAT which is the case for Microdroid.
Use ro.product.abilist32 to determine whether 32-bit applications are
supported and if not, don't configure it, but mmap_rnd_bits.
Bug: 237950549
Test: run Microdroid with the kernel built with aosp/2153639
Change-Id: Ifca6fa02f14ad4c7d8f9b2ab8852494c12945c3a
Was "check_most_at_once", should be "check_at_most_once".
Also straighten the if-then-else logic from:
if (cond) {
// then
} else {
// error...
return;
}
to:
if (!cond) {
// error...
return;
}
// fallthrough is "then"
Test: Presubmit
Change-Id: I85c1e94c47a727089374d5d05ecd40f4187b66a1
Currently, for an API symbol to be accessible to APEXes, it needs to be
marked as either # systemapi or # apex. It was originally just # apex,
but we added # systemapi to clearly identify the origin of the APIs.
The intended use is
* #apex is for APEX-visible symbols that are defined in an APEX
* #systemapi is for APEX-visible symbols that are defined in the
platform (the non-updatable part)
This intention is documented build/soong/docs/map_files.md, but isn't
enforced at all.
With b/239274367, this is now enforced and therefore the #apex tags in
the platform library are replaced with `#systemapi`
This change does not alter any functionality.
Bug: 239274367
Test: m
Change-Id: Ibdb9122c9969749e055404078bc2280edaa7346d
For consistent error message. This is a preparation step for the
follow-up change: to generate linker configuration in DoLoadApex()
Bug: 232173613
Test: CtsInitTestCases
Change-Id: I27954ae2429e82da1bde1925ef62b277c24129d3
1.In ANP, the path of the executable file recovery is sbin/recovery, and in ANR it is system/bin/recovery
2.Add a judgment on whether sbin/recovery exists in fs_mgr_fstab.cpp,
Make it possible to correctly load etc/recovery.fstab in recovery mode
Bug: 237368169
Change-Id: I8df76a003bc2edbaa01d9dd173a0cba1401ef3ac
This reverts commit da94c7f650.
Reason for revert: It appears this change slows down boot on normal devices.
Technically, this change is not necessary, but it prevents starting the secondary and having it throw an error in the only run 64 bit zygote config. But it's easier to throw the error than slow down boot up.
Bug: 238971179
Test: Verified that on a 64 with 32 config, the secondary zygote
Test: starts but exits.
Change-Id: I7ab0496a402db83e70168d52e5d5911b82a3b06a
servicemanager/hwservicemanager are pre-apexd services but still wants
to see VINTF fragments from APEXes, especially from /data.
Like ueventd, these services need to be started in "default" mount
namespace.
Bug: 237672865
Test: m && boot
Change-Id: I0266c5be5530a1a07f8ffa23a26186d45a55613f
Add "power_on_key" and "power_on_cable" to match
heuristics in system/core/bootstat/bootstat.cpp
Change-Id: I947026d658c8a385aab7a3ade9efacffb9538d56
Signed-off-by: jiajia tang <tangjiajia@xiaomi.com>
Right now, it just prints "no directories to process". There is also no
indication anywhere about the availability of the -f and -d flags.
Also took the opportunity to perform small quality-of-life improvements,
like updating error messages, hiding internal functions.
Test: (cd system/core/mkbootfs && mma) && ./out/host/linux-x86/bin/mkbootfs
Change-Id: Ia81eff0308e6e2686dde979d47c0bf938afbaf82
Path of "frameworks/base/cmds/statsd/src/atoms.proto"
is out of date. Latest path is
"frameworks/proto_logging/stats/atoms.proto".
Signed-off-by: jiajia tang <tangjiajia@xiaomi.com>
Change-Id: Ib9471e812466238d79e836b026123f167aaad701
Currently OTA merge process prints enum values in logs in numerical
form. This patch changes the log messages from numerical enums to
more friendly names.
Test: Manual OTA
===========Logs==========
CheckTargetMergeState for vendor_dlkm_b returned: Merging
ProcessUpdateState handling state: Merging
CheckTargetMergeState for product_b returned: Merging
CheckTargetMergeState for vendor_dlkm_b returned: MergeCompleted
ProcessUpdateState handling state: Merging
CheckTargetMergeState for product_b returned: MergeCompleted
ProcessUpdateState handling state: MergeCompleted
Bug: 234518211
Change-Id: If690d3ea86e6fe37e8b3e4bd52e87b70ae92495c
We don't initialize subcontext in microdroid. subcontext->pid() will
crash in such case.
Test: atest MicrodroidTests
Change-Id: Ifca65acededbf7f113d8bf16afbb8e2c1aa91b41
Thew new client will use AIDL if available, and fallback to HIDL
otherwise.
Test: th
Bug: 227536004
Change-Id: I9af21037a76a4a6db00144f5b2774ea23f3a5cc2
