* Even though camera.device@3.4-external-impl still will look for the
file, this is seemingly no longer required in Q due to changes
elsewhere in the stack.
This reverts commit cf207232cb.
Change-Id: Ibd19f4d7abeafb7f4d8af1ca6226d6065058538c
* 'lineage-16.0' of http://www.github.com/LineageOS/android_device_samsung_msm8974-common:
msm8974-common: Build common legacy gnss hal
msm8974-common: Switch to GNSS HIDL HAL
msm8974-common: gps: Fix for buffer overrun crash at copying nmea string
msm8974-common: gps: Replacing copy headers with header libraries
msm8974-common: gps: Don't rely on transitively included headers
msm8974-common: gps: Remove nmea log
msm8974-common: gps: Return the correct length of nmea sentence
msm8974-common: gps: Default apn ip type to ipv4
msm8974-common: gps: Suppress unused-parameter warnings
msm8974-common: gps: Remove owner of gps hal modules
Change-Id: I910806396d54c33727ef3b807404d71786aa74e6
Add zero clearing of allocated nmea buffer to ensure
the nmea string is null terminated.
Change-Id: Ie36010a7d3eca16dabb3067ae891a94e4b63b10c
CRs-Fixed: 2041933
One must explicitly include what it need.
time.h for struct tm.
Bug: 37629934
Test: build
Merged-In: Iacd5317d958f61df062fca40fbf2e8992c7fbb44
Change-Id: I19e1043f23dbe85836b73714a4f1440159ebe419
Return the correct length of nmea sentence in the
function loc_eng_nmea_put_checksum(). The correct
length is length of NMEA string including ‘$’
character alongwith length of checksum.
CRs-Fixed: 1020044
Change-Id: I50d80b9c4492de4664bb3ccb7295b0eb4fee15a9
If a apn ip type is not found for a agps
data connection, default to using ipv4
instead of invalid.
CRs-fixed: 1007417
Bug: 28986221
Change-Id: Icc71de1fab3fd5967901c39a5c11eaea17b7766c
Signed-off-by: D. Andrei Măceș <Andrei.Maces@alumni.nd.edu>
Test: None
Bug: 122292884
[haggertk: This is only the libril side of this change. We quite
obviously do not carry our own librilutils]
Change-Id: Idb0d56749e41aa97a760dfed5246330041c3da3a
Include MCC/MNC provided by legacy (versions 14 and older) vendor RIL
implementations as part of data registration state, in data registration
state reported to clients of the radio HAL service.
Bug: 119904357
Test: local build and did the local test on Marlin device,
the mcc/mnc value are correct (not -1 and empty string value).
[haggertk: This purposely omits updating the numStrings check in
getDataRegistrationStateResponse(), given that we purposely changed
that check in our libril]
Change-Id: I5a26939402b86d889133e16f3508ab76b8bedddc
RIL_SignalStrength_v10 not support gw.timingAdvance. But default 0 will
be taken as valid value. Set to INT_MAX as invalid value.
Bug: 123096279
Test: Build pass. Correct value for GW timingAdvance in radio log.
01-22 16:22:41.150 1779 1910 V RILJ : [UNSL]< UNSOL_SIGNAL_STRENGTH
SignalStrength:{ mGsm=CellSignalStrengthGsm: rssi=2147483647 ber=99 mTa=2147483647
[haggertk: Upstream commit updated to patch both
convertRilSignalStrengthToHalV8 and convertRilSignalStrengthToHalV10,
as the upstream libril only has a single convertRilSignalStrengthToHal]
Change-Id: I37cc2c246d045a07ffad863fb0cc852d8184c3ca
The index value for GSM/CDMA/IMS application shall be -1 if there is no
relevant application according to the comment written for CardStatus
structure in radio/types.hal, so it shall be initialized to -1.
Bug: 63967442
Test: Confirm that the index values are correctly initialized.
Change-Id: I692e9049145d0f0c3c57879c25d0697879c76b39
Use strlcpy instead of strncpy when copying strings to make sure
the copy is always null-terminated.
Bug:73436938
[haggertk: Our original CAF base had the actual "replace strncpy
with strlcpy" part of this change already. This just adds the
expected sendErrorResponse()]
Change-Id: I12d4883c22a180e2136dc8c85bc0394ddcdcb706
Legacy RIL uses an integer to encode the number of
MNC digits. Because the size is not fixed, leading
zeroes result in ambiguity in the length of the mnc.
This change adds support for passing the number of
encoded digits in the most-significant nibble of the
mnc integer (which is only 10 bits). Thus, on any
implementation that is 16-bits or wider, the mnc info
will be properly encoded and decoded with the
correctly-sized string.
Bug: 111971808
Test: ril::util::mnc::test
Change-Id: I24aeba5328a63f80b0d6b25b068bd19160191dff
* 'lineage-16.0' of http://www.github.com/LineageOS/android_device_samsung_msm8974-common:
msm8974-common: gps: Correct week rollover timestamps
msm8974-common: gps: Fix race condition when GPS app is closed around an SSR event
msm8974-common: gps: Adding fix for KW warning
msm8974-common: gps: Assure event mask is set properly to get positions
msm8974-common: gps: Fix unable to clear existing event bitmasks
msm8974-common: gps: Add tech_mask in UlpLocation to support report
msm8974-common: gps: NMEA report indication to ULP
msm8974-common: gps: Making SUPL ES run time configurable
msm8974-common: gps: Fixes SUPL_MODE
msm8974-common: gps: Consider nsec while ranking the timer objects
msm8974-common: gps: Adding a new NMEA sentence : $GNGSA.
msm8974-common: gps: Set correct target for msm baseband
msm8974-common: gps: Add Cleanstep.mk
Change-Id: Ibfdb99e86d3bc3829b565ee8732a7a399ff27915
This adjustment will help with applications checking the time coming from gps.
Some applictions will consider times in the far past as an error.
Change-Id: I69d6971249c4898ac3895f725e46f268cdb2b00d
Fix race condition to get avoid stuck in modem if GPS application
is closed right around a SSR event.This fix will prevent resume the
session once HAL loc_stop has been called.
Change-Id: I4e5c3f09c097b835d8a685a8b465eb86dd06cd7a
CRs-Fixed: 1058559
Fix for buffer overflow possibility for the thread name passed in to
pthread_setname_np() method.
Change-Id: Id323dd058eddcf50d6fd9ec8908e9997b30c561b
CRs-Fixed: 999457
The event mask can be retrieved in the context of
client thread as zero and then queued up to go to
msg task thread. By the time the msg is actually
handled in msg task thread, the actual event
mask at LOC API layer may have already changed, but
this mask would then be overridden by zero. This
can cause no modem events to ever come, including
position reports.
The fix is to not retrieve the event mask in the
client thread, but instead wait for msg to be
handled in msg task thread before retrieving it.
Change-Id: I48562d028bbfa187732686c060b5cdd62c6d5a89
CRs-fixed: 2219519
Fix for an issue updateEvtMask call chain is
unable to clear existing event bitmasks stored
in LocApiV02.
Change-Id: Id34e2d92a506e7300f3d9b3ec74a80a1f6822787
CRs-Fixed: 2085379
SUPL ES gets updated as part of the sim specific
AGPS config hot swap update. We keep that lastest
SUPL ES and send that as part of the NI request,
to let NI handler to know the config status.
Change-Id: I566725d5db9b3abdbd9a423b86c10b819217f4f6
CRs-Fixed: 1024879
Carrier specific SUPL_MODE is not updated upon sim card insert.
gpc_conf must be updated with the new value for config to be
updated.
Change-Id: I2a5f10fa7d27be9253e9ca6f34c0c886285d1b11
CRs-Fixed: 953942
Currently only $GPGSA was being generated for GPS satellites.
This change adds $GNGSA for GLONASS satellites as well.
Change-Id: I7ab0e92c7ed4ce73be11909eefe0d570a1a0fc1e
CRs-Fixed: 978745
* This served a purpose for a period of time, but we decided to embed
the relevant symbols directly into libril.so with 8f6082f
This reverts commit 61a18ade33.
Change-Id: Ifb92a32be007fd0f842af38cf652edffde0604e5
