/acct is not necessary anymore for process group creation (it migrated
to /sys/fs/cgroup and is automatically mounted by libprocessgroup)
Bug: 168907513
Test: booted in remote mode and tested adb sideload
Change-Id: Iaa455df65f393ac1f67ae81c6e78d8f4934d88a2
cc_test_host doesn't need TEST_MAPPING config anymore to run in
presubmit. (go/a-unit-tests)
Change-Id: Ia07ea311cf6febc2f937ab14966ad21dfd623c69
Test: presubmit
Bug: 178646865
New() should be used instead of NewForFirstStageMount().
Bug: 168258606
Test: data wipe with VABC merge in progress
Change-Id: Idf2b01a504b577766da303091721764242e99a69
Libadf is being removed since there are no android common kernels
that support adf. Remove support from miniui.
Test: Compiles and boots
Bug: 150467766
Change-Id: Ic66ea7a012b0466810ddeb8f884150b50c8ae769
Now we added a libz variant without the offending optimizations,
re-enable tests.
Test: treehugger
Bug: 177076632
Change-Id: I6969090b2cb4c059d952df7cc034d0ed1ac366b2
libz contain platform dependent optimization flags, and sometimes that
cause reconstruction of blobs to fail. Use libz_stable instead
Bug: 177076632
Test: treehugger
Change-Id: I3a8c1591672537d1c754b2bc5b26f939dd80ed47
For a proper solution, add a variant of libz which doesn't have platform
dependent optimizations, and make imgdiff use that version.
Test: treehugger
Bug: 177076632
Change-Id: Ia9e926c1adf22d351315eeec5ad1fabc3d48efd5
Use free to release GRSurface object will not call data_'s
DataDeleter, then the buf data_ points to will leakage.
Use delete release GRSurface object.
Test: manual
Change-Id: I53a9ea5a4ad67de1662c707a5ea59d738bd9c2d8
Signed-off-by: lijiazi <lijiazi@xiaomi.com>
Wait for socket to close before exit(), so that last "DONEDONE" message
is correctly sent.
Test: enter recovery, sideload an OTA, verify that no read error occured
on client side. Repeat 5 times.
Bug: 155158483
Change-Id: Ibb02548d8c03fb77f8f60cb69cff84c33903adc0
This is a pretty simplistic approach, it just shoves random data at the
verifier. The OTA format isn't too complicated so this should hopefully
be sufficient to let the fuzzer exercise the potentially interesting
parsing code.
Test: Let the fuzzer run on device for awhile:
1) FUZZ=libinstall_verify_package_fuzzer
2) SANITIZE_TARGET=hwaddress make ${FUZZ}
3) cd ${ANDROID_PRODUCT_OUT} && adb root && adb sync data
4) adb shell /data/fuzz/arm64/${FUZZ}/${FUZZ}
Change-Id: Icac6bde017b497d9f92c06191eb29e107ba9c0a7
Make it possible to build a custom UI which uses EthernetRecoveryUI.
Since EthernetRecoveryUI must be used with EthernetDevice, but the
make_device() function is the only place a custom UI can be specified,
we need to move the common EthernetDevice to the recovery_ui core and
make a new "default_ethernet_device" which looks like the other
default device library stubs, for backwards compatibility.
This makes it possible to use EthernetRecoveryUI on cuttlefish.
Bug: 172693524
Change-Id: I6a66bb949ef97d06f532adb286d2db4c002d2924
Bump the timeout as we see some flakiness in b/170178152. The other
part of the config is copied from the auto-generated config in
out/host/linux-x86/testcases/recovery_host_test/recovery_host_test.config
Bug: 170178152
Test: treehugger
Change-Id: Ia84c90ba6a686c47ecc7d8331c7e8c7cb4b78292
Revert submission 1433573-vab-libsnapshot-linkage
Reason for revert: b/169981170, update crash for droidfooders.
Reverted Changes:
Ie75bba98c:Link to libsnapshot_cow where libsnapshot is linke...
Ieedfadc55:libsnapshot: Partially implement OpenSnapshotWrite...
I28a5d4a88:Link to libsnapshot_cow everywhere libsnapshot is ...
Change-Id: I8c774ca4a8dec21dd308694bb8205861a19c3e12
