| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
The packet was always sized to MIU instead of actual buffer size.
Change-Id: Iaf2fbd1cddf07a95e95a8dd69323d62dca8e6a2c
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
If the message is to large to fit in a single packet (its size is
defined by MIU), then it need to be splited in multiple packets.
Change-Id: I69eafac0b250b0f32901795b2ab80bad173ae35a
|
|\ \ |
|
| |/
| |
| |
| | |
Change-Id: Ic5c321c3dbf3ca1f7f4e0697791522dd53dccf3e
|
| |
| |
| |
| | |
Change-Id: Icaeb2ed76d87da64c88b496ba76295aadeb12a7e
|
|/
|
|
| |
Change-Id: Ia543f55135701154d2c639639b2183bfa07c9fd9
|
|
|
|
| |
Change-Id: Ia1c006c52aaf312408ef2fba96715b1b6c382ea9
|
|
|
|
|
|
|
| |
One file got left out of my last commit,
and fix some logging around My Tag.
Change-Id: Iafb47dd5a3f6ed8e337b1f951875d010fc328002
|
|
|
|
| |
Change-Id: I8f52bf20f493581aef80201b4b112af28cf3b5fa
|
|
|
|
| |
Change-Id: Ie13294a978b46d5e75128b2d0fdd4c2f7a70f2e2
|
|
|
|
| |
This reverts commit ae66f8e306a46a24e4121b72b30d8beb5e1c08f9.
|
|
|
|
|
|
| |
This reverts commit 825f01522a1d68cadb634c88101e96f842478926.
Change-Id: I75e027c066ba7bddaab87f910083707d8eb380b0
|
|
|
|
|
|
|
|
| |
NativeNfcTag contains the NDEF
read/write and NativeNdefTag is
not used anywhere.
Change-Id: Ic3f7c3c792419468ea9c02f3d08df2761627137e
|
|
|
|
| |
This reverts commit 1fc06b8d4dfd7e92c43cbd4c5aacf8fa66d181d1.
|
|
|
|
|
|
| |
Bug: 3124915
Change-Id: If5a3e6b21fe43d9ce2ff82ee65751baaa1c0e0a2
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
| |
NFC service can be restarted with screen off, and we will be put in
polling loop, or worse constant RF, if this happens with a card in field.
Change-Id: I0b8d2ba04f3af47908586cccf7210d9a7abb5d54
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
| |
Change-Id: Ie288273526d9e947545dde134ac1d23399a716c4
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
| |
We're about to also disconnect from tags during SCREEN_OFF, which can take
some time, so we need to make sure to hold the CPU on.
Change-Id: Ie551c028cecfb49faddc95a45ecd8eac07671968
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: I4f839d20d2be8c816fffedc9f64292008751a7c1
|
|
|
|
|
| |
Change-Id: I92776225eafd39e0c4dad15b46d5e0ce63f7a991
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: Ic8f962879dc8c226b0ac10d2a1d4e41ede5d9326
|
|
|
|
| |
Change-Id: I613e5755dea46e6920ab98621d0f8ffb2feeb358
|
|
|
|
|
|
|
| |
The LTO was previously set to 250 as a workaround for performance
issues. This is not needed any more with the current RF speed.
Change-Id: Id236b142d3150203781d00434b93387d925abda5
|
|
|
|
| |
Change-Id: I3ca5e6f90d85fdef262874bd9223ca5f49468af4
|
|
|
|
|
|
|
| |
This allows services to accept multiple times in
a row.
Change-Id: Ifef474cd753d0481f3a9befce91d2fec869d34ad
|
|
|
|
| |
Change-Id: I597e4be9051cc30c4432f90bd8399ae4bdd84b83
|
|
|
|
| |
Change-Id: Ie56036441ea41d9290d33915e3616e9379f6ea98
|
|
|
|
| |
Change-Id: I5cddb343bf0cad13768fb4d6d3d00f82f4205afa
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Record the socket parameters for registered
LLCP service sockets since they're used when
accept() is called before a connection is
active.
- Add a missing break when handling mock NDEF
messages to avoid a crash when a My Tag is
scanned.
- Fix up the accept and read loops in My Tag
server.
Change-Id: Ia994c9dacb5e6770ac5c0f1b5b9e0156eea6e5cb
|
|
|
|
|
|
|
|
|
| |
sem_timedwait() was unbalanced.
Also remove more dead code, and reduce watchdog timeout from 10 to 5 seconds.
Change-Id: Ica776e4a27905071c1453b2db02f9ffadf6f9461
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: I909b2623f66ca0f3f0370617e2df4b1075614c44
|
|
|
|
|
|
|
| |
Will now send an intent with an empty NDEF message array attached
Change-Id: I58d86c9b5cfff127581824ca6af410a503059cbb
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
|
| |
Rename functions that need to be called with lock held to _locked().
Remove dead code.
Replace asyncDisconnect() with disconnect() - its broken.
Change-Id: I04feb449394750954e202223366a0fa5ea425924
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: Ibd553a01f8790d1f38673bd08f90e3c1a847f727
|
|
|
|
|
| |
Change-Id: Ibd50e30c7f340eed7c84785d2060a8a237f9cd5b
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: Ib6e8c642a97abab6592200023c1bb0993d0bfdc0
|
|
|
| |
Change-Id: I6d7756739489b240e2e27c91a1afd8c5bd534342
|
|
|
|
| |
Change-Id: I9f46908d99d05804cd325024fc60da762f5beb1d
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
emergency_recovery() was not able to kill libnfc threads blocked in mutex,
and was just wedging everything. So instead kill the entire NFC service
process. The framework will immediately restart it, and the NFC API's will
rebind to the new service handle.
Also I cleaned up the code to handle boot complete. This is necessary to clean
up because during recovery restart we do not receive a boot complete event.
It turns out we can use FLAG_RECEIVER_REGISTERED_ONLY_BEFORE_BOOT on intent
broadcasts, and avoid having to track boot complete at all.
Change-Id: I41c885b441128fd447d6b8c63bb50a13a1d4cdfd
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
| |
Do not even take a mutex in SCREEN_ON/OFF receiver. Just schedule the thread.
Bug: 3131975
Change-Id: Ic8763315c815aff63299f77590e61d668d13031b
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
| |
Change-Id: Ie7340a34ecd1e0a81d9589f8c8e130b761a221b5
|
|
|
|
| |
Change-Id: Iba652af8efdec58d04d5e8e5ecc6af8281f57279
|
|
|
|
|
|
|
| |
Added missing native tag registration in NFC service.
Change-Id: Id470fc93fc83a4b9bde6d0bd78500f50635efb48
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
| |
Use a boxed boolean, not a primitive boolean.
Use the correct array index.
Change-Id: Id3cd16c8d21b9dacd6d167b0458b6583d7fba89f
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
|
| |
mWatchdog is only assigned during a successful connect(), but other NativeTag
methods were calling methods on it without a null check.
Bug: 3133259
Change-Id: Ia28a6cad7bbf8e029e876f902474f6b4af380e02
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
|
|
|
| |
These should be safe on the main thread, because they involve configuration of
the local NFC adapter, so it should be very quick. However the NFC stack is
currently wedging frequently, and the resulting ANR's during screen on/screen
off are very disruptive.
Bug: 3131975
Change-Id: I06d3ee17a991a95676666300942054797c6adf00
Signed-off-by: Nick Pelly <npelly@google.com>
|
|
|
|
|
|
|
|
| |
The tag server has some problems when being
stopped and the client never gets notified
about the LLCP connection.
Change-Id: Ifdfc60d48ed85a9916f8abb085afcf01dd5a0155
|
|
|
|
| |
Change-Id: Ia54094291726fa4fac152a49b1b182d73ff9958f
|
|
|
|
|
| |
Change-Id: Ib122c9a952f7f41798f7e6b2b8e9884b53e67723
Signed-off-by: Nick Pelly <npelly@google.com>
|