GitLab

Found when using reader mode APIs: control is accidentally
granted to the HAL, causing us to leak command buffers.

Also:
- Prevent multiple calls to NFC_Deactivate()
- Proper handling of congestion in P2P

Bug: 10360259
Change-Id: Ia52a948f46f8668cde5b135bd073f175cdd434f5

New API's include:
NFA_EnableListening
NFA_DisableListening
NFA_PauseP2p
NFA_ResumeP2p

Bug: 10360259
Change-Id: I7759c17fe1f83c2af251431e1dc56116a9fc78fd

Bug: 10360259
Change-Id: I57160e7073b9d625c470feff311f305dec16cb5e

Until we upgrade kernel driver on Mako/Manta, just
call the old address change ioctl.

Change-Id: I98cee0126b3f1d8f03eb63773f92ad95e7fe623e

Until we figure out how to deal with HAL binary
compatibility.

Also, do init the EE subsystem even if no SEs
are present; we still want to use the EE subsystem
on the host.

Change-Id: I17aad42e75508e2d04f53107527bce5bfe099ffe

Change-Id: I513ff28608bfbdc3725af9a6e705669c3e693cc7

Added 'get_max_ee' to the HAL interface to allow the NFC stack to
inquire the HAL layer of the maximum number of EE on the platform.

Added the necessary HAL interface to allow the stack to get the maximum
number of EE on the system.  This allows the system to dynamically set
this value at run-time in order to support chip-specific differences
in the number of EE on the platform.

Change-Id: I3ff00d117d715b1f46636501932ff92accc39483

Upgrade stack to allow JNI to control whether
to enable stack's secure element features.

Change-Id: Id4162d0a97a031fb1a0625f64faef90091282b3f

Sometimes presence-check times out when reading the tag.  Therefore, use
a longer timeout.

Change-Id: I905b36b64fc076adaf94418fc0ce121991c21866

After upgrading to NFA 1.3.62 we took the change to set the max
number of EE to be 4 (to support 20795).  But the value must
remain at 3 for the current set of chips.  We'll need to add
dynamic support later (or .conf file entry can be used).

Change-Id: If706e345b92a8e7630ad1cc4fdbfe908bceabae8

Improve reliability when creating NCI connection for HCP.
Improve firmware download during error condition.

Change-Id: I5655d4a66d04debd4a2a91c706393e8a45d6dc59

Updated the HAL device init configuration to include and entry
for 20795 to set the proper XTAL frequency/index.
By Paul Chaisson

Change-Id: I00c156822ccc9705a8a2ea670d65c204b05e5290

By Kevin Park. GKI_getbuf is executed in critical section.
Sometimes, when GKI getbuf fails due to low memory,
code logic in GKI_getbuf just returns NULL without releasing lock.
This introduces deadlock situation.
So fixed the code to call GKI_enable before returning.

Increasing GKI buf pool size for the
specific condition like system very busy on first boot.

Change-Id: I8c7de8d0f472f5dfa41a43c08c30025de4e073c4

Stack adds ability to adjust RF discovery frequencies.

Change-Id: I3c9a843c83c3a86181e7c8726b95c9f6517d8706

The stack modified the HAL to only compare the chip generation
(e.g."43341") instead of including the chip revision, when trying
to determine if the XTAL frequency needs to be updated.

Change-Id: Iab31706f67d8bab7927ae3180aec8fc36a745b47

Remove unused values from config.h:
-ENABLE_BRCM_EXTRAS_API
-LOG_TO_FILE
-LOGCAT_FILTER
-NFA_DM_ENABLE_SLEEP

Change-Id: If850c16a3bdac22b9535b27d9a4f3e103442303f

.conf variable SPD_IGNORE_VERSION is removed.  Stack always download
firmware if a difference is detected.

Change-Id: I2a5e21254e6c0df7829506c1c3c2f6ffa44c0048

Change-Id: Ifc970a5b68cc962e5fdfa9af11e54e569d25ef1b

Change-Id: Ifd54edaa0da38a637c2c010ed59f99974c0fc901

This logic is no longer needed.

Change-Id: I490e20ec599213f73aeddbaa9f99958854c08e8b

Fix potential problems discovered by MPS team's Coverity static analyzer.
CID 14304, 10630, 10626, 23660, 23655, 23654.

Change-Id: Iec34fd793b44710fa971175cd058e87ab17c76cc

The TLV with T=0xC2 can be sent from one or two locations in the code.
Add ability to dynamically adjust the TLV's value at both locations.
This adjustment is needed to disable one or both secure elements.

Change-Id: I404c4afa350daf571542b83cf0a1b80fa1645073

When non-volatile storage files are preserved during cold boot,
JNI must detect corruption in their contents by verifying their
checksums.

Change-Id: I191ac013f12dabda43e590449a80523bc42c0c90

Change-Id: I4be1b5f970300bdf23400be678d802beba65a5d7

Take more changes from main_prasadk_2163714.  Fix bug in the operation
that handles a NACK response from type-2 tag.

Change-Id: Iaea8fde94acf82c0f2d63f09378d9ee2818b159c

It's always 0, moved change client address logic to USERIAL.

Change-Id: If5e46c35a8dcb9c2007611043912815ca436137a

A tag from Korea is Mifare Ultralight C.  When the stack performs
presence-check, the tag responds with NACK and enters HALT state.
This patch properly performs presence-check for this tag.

Change-Id: Ia19e6216628bd74e8367f21d90ac806751eacfda

Change-Id: Ia9c970bdf761c776df99392e65c6ff4607746524

When encountering a corrupted firmware file, power-cycle the controller
and tell stack not to download it during next initialization.
Re-initialize HAL.

Change-Id: I5d330da4f3666a53ac6336b2d6643e5bf6383ffd

Store 2-byte checksum in the first two bytes of the non-volatile store.
If corruption is detected, delete all content in non-volatile store.

Change-Id: I79ca57796da1950a430064e50752c4a9fae6d05e

The NAME_SNOOZE_MODE_CFG variable shall be defined in the
chip-specific .conf.

Change-Id: I17a86ede4b8bb365f811138d8e37219b52ae8cb7

* commit '3f0440ad':
  Fix Apache license on some files.

Change-Id: I0fe5480bd8ef816c01a34147c85acdc1d3fd6612

The NFCC may have transitioned to a power-off
state when the timer is expired, causing a command
timeout. We'll control the routing table strictly
from the higher levels.

Change-Id: Ic5fa100d2a8ba16c5c911118f38ffef0268dff8f

Change-Id: I16c6082222e8eda5c5dde54cb9421233284e7115

When removing active listen modes, the original listen
mode configuration is stored. If there is no new LLCP
activation within a configured time interval (currently
5s), the listen mode configuration is restored to the
old value.

The configuration that is restored is by then not
necessarily the correct configuration anymore. In
particular, this may disturb eSE access which disables
discovery completely.

Disable this at least temporarily to verify that this
is causing crashes on Manta.

Bug: 8775164
Change-Id: I926535a73466c857f7a0f2cac912b5ab1f8c38e1

Bug: 8776949
Change-Id: Idbfb9c77304ce0df8521b167f25004bfd18b59d8

Bug: 8580410
Change-Id: Ib685275875629bc313ba63c53ec0326e329357cd