This was expected to return an error in the HIDL version, so we
explicitly say which it should return on failure.
Brought up in ANAPIC review in ag/21108335
Test: none
Bug: none
Change-Id: I6f183a490f91aca5e08b5ae83017e707b63268ea
Modify the semantics of transmit() to return a service
specific error with code CHANNEL_NOT_AVAILABLE if there
was an error in communicating with the secure element.
This can happen if the SE is put in low power mode when
no logical or basic channel is opened, e.g.
Bug: 270091254
Test: m VtsHalSecureElementTargetTest
Change-Id: I7df3ec6d9b6d5eeb2272971c44fc078a8558d2e6
Add missing checks for the proper use of init() interface.
Test correct channel management.
Test: atest VtsHalSecureElementTargetTest
Bug: 123254068
Change-Id: I411b070aa1a3e1d4fb4ac38d004a1cfc81e86023
Bug: 205762050
Test: atest VtsHalSecureElementTargetTest (8/8 passed)
Test: atest CtsOmapiTestCases
- all tests pass except testStatusWordTransmit,
testTransmitApdu, and testSelectNext. This is because
these tests don't increase coverage of the HAL API
except they call 'transmit' several times. Doing a full
mock OMAPI/APDU implementation for cuttlefish/AOSP is
tracked in b/123254068
Change-Id: I0ec0eb08fbcc4207d8c07427ccdf7ff106995627
