BT-841 #done
TEST=1. manually check that Bluetooth cast setup works
     2. fx run-test bluetooth-tests -t bt-host-unittests

Change-Id: Ic5889243cbad7577598af2b16d7977576c45aa1e

BT-841 #progress
TEST=fx syslog no longer shows error log messages indicating firmware
     could not be loaded.

Change-Id: Ie292fd4e811f649b350724084003f81f359f2578

cp writeback.cpp
 -> buffer.cpp
 -> write-txn.cpp
 -> writeback-queue.cpp
 -> writeback-work.cpp

cp journal.cpp
 -> journal-entry.cpp

cp writeback.h
 -> buffer.h
 -> write-txn.h
 -> writeback-queue.h
 -> writeback-work.h

Change-Id: Ic0d594810f226d7a7760cf3d4ae91b79e971abce

Disabling test to unblock CQ.

Bug: FLK-189 #comment disable
Change-Id: I8cb709ab1d6d921da3ed69b98fe7a01c4c318761

This patch makes a few corrections of references to the original
garnet/docs/gen/build_arguments.md. The new location is
docs/gen/build_arguments.md . This change also add reference
to docs/development/tracing/README.md .

Bugs: BLD-454
Test: CQ
Change-Id: If0cb49f57445484fc9e8270aeaa86b3aeadc9f8d

Cleanup while reading the code.

TEST: No behavior change. fx run-test audio_mixer_tests
Change-Id: Ic63d36dafe4e8bd122d34a05a8af09ebee943abb

These tests have been accidentally disabled for some time. Reenabling in advance
of more changes within eapol.

Test: Existing tests run and pass
Change-Id: I816e85db9e7abc58251d93863b4556f49b3d82e7

While we don't want participants to be too picky about buffer_count, we
should be able to support cases where a participant must constrain both
the min and max buffer count.  This adds min_buffer_count.

This also moves some constraint checks to make sure we don't have any
constraint checks in the Accumulate* methods, which potentially don't
run if there's only one participant or only one participant specifying
buffer constraints or image constraints.

Tested: CQ, test_sysmem_min_buffer_count added
Change-Id: I1920dc26f92f463c692415e0e1f424fc4e0a16db

Also add the new //bundles/buildbot:bringup group.  The new model is
that for each product whose build is tested by bots, there should be
a //bundles/buildbot:$product group.  The bot configs will use that,
so `fx set PRODUCT.BOARD --with-base //bundles/buildbot:PRODUCT` will
approximate in a local build what the bots will do.

Bug: INTK-1015 #comment Define //bundles/buildbot:{bringup,core} groups
Change-Id: Iacc19043d34e948bde2215cb5441e91d97c67417

Change-Id: I09ad0fb80c7af8eaa94cec3d2f4fce1cd008d03b

This reverts commit 2dcfdba2.

Reason for revert: Failed boot on vim2

Original change's description:
> [zircon] Reformat zxcrypt and minfs partitions when the partition is
> missing. Added a unit test that failed without the changes.
>
> Change-Id: Idaf9913bf9006384e021f39047df1e8e8a855369

TBR=smklein@google.com,aarongreen@google.com,abarth@google.com,zarvox@google.com,jseibert@google.com

Change-Id: I0e755d84dc4839e8968308ca753aa1d9c9c5038b

Insert ops into streams, fetch and issue. Convert SchedulerOp from
struct to class.
Scheduler can balance the ordering by streams (currently dummy
round-robin).

Change-Id: I4b8f3e64eff1360f2334afa67036cbc09db105fc

... and add link to tracing-boot.md.

Change-Id: I0818ba4b0de23940b88b9b14b61f77d435077f39

Blast the reader with all the potential data that is available.

Also, re-indent, improves cut-n-pastability.

Change-Id: I4944aecc1616c41fc60ed7be13f55586b26b11b6

The old - names were only ever used because of limitations of the
old Zircon build system, which is no longer an issue.  The target
name in banjo_library() must always match the library name.  For
public Banjo libraries, the directory name matches that name too.

Change-Id: Ifff61436ec7d15eb53dfd21349c4367047e691d4

Migration of mediaplayer to ViewToken is complete.

Tested: CQ; no behavior changed
SCN-1018 #comment
SCN-1291 #comment

Change-Id: Iae3446ec6c7ea4eb280c985dbec9d73bfe3e5204

Test: no effect on the build
Change-Id: I819338ff6309f7487f0ebe72b63f11031a8c631b

SL4F calls can now be made on the IPv4 and IPv6 interfaces

Bug: CONN-56
Test: Verifed SL4F could be connected on both IPv4 and IPv6
      interfaces.
Change-Id: Ia6a68abc3c82a1484a4718f2aec3f8a88b399c77

LowEnergyConnectionManager now responds to local identity information
requests during the feature exchange and key distribution phases of SMP
pairing.

Bug: BT-243 #done
Test: 1. Run bt-pairing-tool, bt-le-battery-peripheral, and
         bt-le-peripheral -c -n <NAME> on Fuchsia. This will put Fuchsia
	 into LE peripheral mode with one GATT service that requires
	 security.
      2. Connect to Fuchsia from iOS using the nRF Connect app. This
         should trigger LE pairing. Respond to the authentication challenge
	 using bt-pairing-tool (local passkey entry on ToT). The battery
	 service characteristic should be readable and notifiable
	 without any errors.
      3. Pipe the output of bt-snoop-cli to Wireshark and confirm that:
         a. Fuchsia configured advertising using a random address.
	 b. Local IRK and identity address were distributed over SMP.
	 Check that the IRK matches what has been persisted in the
	 stash. Check that the identity address matches the address of
	 the active adapter (confirm via bt-cli).
      4. Reboot Fuchsia, repeat step #1. Connect again using nRF
         Connect. The battery service should be accessible without
	 additional pairing errors.
      5. Check the snoop logs to confirm that Fuchsia used a different
         RPA when advertising.
Change-Id: Ibb78b2de3049f9f351215ed4d878acfcca5569df

While several images already contain these packages, some images don't
and the packages need to be available in the universe set.

Bug: DX-1427 #done
Change-Id: If10223ec017312c797defca113ab21e30cf159b9

Clearer than using unique_ptr, and makes copying easier, which we need
for templates.

Change-Id: Ie851ab3c461d8b4c3b58c08c7837685e2865f96c

There has been no active development in about a year.
We can resurrect this from the git history if we need it in the future.
In the mean time, lets eliminate code that is not being used or tested.

TEST: builds without imx8 support
Change-Id: Ie9b04b476e777e3311bfaec3465804a1e81c350e

TEST: builds, bringup build boots normally on cleo.
Change-Id: I94227a565ce40b8e93db79aa4457862666466bec

Bug: BLD-461
Change-Id: I60372456d27eebdc5ebcb3e1b1d097b190603d1b

Now that Chromium is handling OnOutputConstraints() and
OnOutputFormat(), we can remove OnOutputConfig().

Tested: use_h264_decoder_test uses the new events on CQ
Change-Id: I57bf213ba8fa5d225b5bfe961c5f5fc9772ce601

TEST: n/a
Change-Id: I5e827e029e61dbc36392f1063dedf64808ff4dea

A name is scoped (by a library), and either has a source location, or an
anonymous name. Modeling this as a std::variant for clarity, and to
simplify owning semantics. We can now add a copy constructor with ease.

Change-Id: I7e6602e0abdc5c08535452253e324b13b8b9dade

This new function supports reading an inspect::Location into an
inspect::Source asynchronously.

CF-714: #progress

Change-Id: Ia7bf90b7b9ba174ebedcabf9303ed9a29b085676

This CL changes the behavior of the media player when the audio
renderer holds on to packets too long, which generally happens due to
scheduling issues. Because the player has allocated a limited amount
of space for transferring packets between itself and the renderer,
that space can be exhausted if the renderer doesn't retire packets
quickly enough.

The fix is to drop packets instead of panicking. There are a few
paths that had to be fixed to enable this. Sometimes the memory
allocation failures occur in a callback from the audio decoder.
Sometimes audio has to be interleaved after the decoder, and the
allocation failure is detected there. Sometimes (when no decoder
is needed) audio needs to be copied from system memory into vmo
memory, and the failure happens prior to the copy.

Two tests have been added to replicate these failure modes. Those
tests produced panics before the fixes were introduced.

TEST: new CQ tests
Change-Id: I26582d19481c6716f6a4141e6717b68670ea44f1

AUTH-196 #comment

Change-Id: I213f7e4440f104ce27535b685a4da100f0526cbd

Tested: CQ; no behavior change
SCN-580 #comment

Change-Id: I496893099aef041de0c721b4840bd104adc10781

Change-Id: I3df4cbd22b97e362c8c80bcb86e618ba0336cfe0

AUTH-202 #comment

Change-Id: I52c5ed31524d5c50cde393f1c1d56f1a0e97bbe5

This places an upper bound on the size of the cache.
Evicts expired tokens on get/put operations, and evicts a random token
when needed to make room for another token.

Change-Id: I8a9eb2926ab134c3e756b1f6b152c56e5bd1a199

Bug: BLD-457 #done
Change-Id: I615d2fb526ff8811f2c94219dbfe7fef2443f480

To make fx run -N even more friendly, I added "&&" (continue with
the next command if the first command succeeds) and "\" (continue
command on the next line) so users can simply copy and paste the
helpful commands in a single copy-and-paste action.

I find I have to re-enable tun/tap after each reboot of my Linux
machine. Of course, I don't remember until I try the "fx run" command,
but "fx run" helpfully prints the suggested commands to do that:

  To use qemu with networking on Linux, configure tun/tap:
  sudo ip tuntap add dev qemu mode tap user richkadel
  sudo ifconfig qemu up

Without thinking about it, it might seem like you should be able to
copy and paste both lines, but this fails because the sudo in the first
line tries to use the second line as the password.

The change simplifies running the commands, and may be less confusing.

Test: Verified results locally (very minor change)
Change-Id: I45ebfda917c96ba5704f108119b65a06aaba0ae4

- Sandbox was using fuchsia::sys::TerminationReason for its termination
codes, refactored termination process to return a netemul-specific
status code.
- Unit tests now check for expected status codes.

Change-Id: I7bc48a34da311d8de1f5a5d6caab27f600e3c5ba

- Sandbox was using a default destructor, which could cause race with
the FIDL bindings error callbacks kept in Sandbox::procs_.

BUG: FLK-178 #done
Change-Id: I46189bde74fa4d83faec27f503f9ec032c605906

The old build command is no longer working. I believe this change
addresses the original intent.

Test: tested locally to confirm it works

Change-Id: Ib097f5e713b4794a66daaf9332da2db2c0158cc1

primitives.

With the introduction of coding tables for boring types inside FIDL tables
to support linearization, array coding tables may have a null element,
when the array is in a FIDL table but the element is a boring type.

TEST: fx run-host-tests fidl_cpp_unittests
Change-Id: I56380c36369e4b501f5b82aa9cfc673c564f82a1