We suspect that something may be causing occasional deadlock around the
time that we attempt to bind the zxcrypt driver.  Add more logging, and
increase verbosity in a couple places to see if we can narrow down where
we're going awry.

DNO-492 #comment
TEST: no behavioral change, just more logging

Change-Id: Ib03af33da1d67533599c80e4453ef27dc160950e

The parser grammar accepted enum declarations without an explicit value,
but the AST processing does not.  We intend to enforce FIDL-like
behavior of requiring all such declarations to have a value, so this
change modifies the grammar to reflect this.

Change-Id: Ieaa5e0f5dbd40f5b78341be3c626a3e073709328

Change-Id: Id31ae5ae9c19d2f37f2d34b3e4af701b7af8e05e

This makes mocking a fidl service slightly less boilerplatey.

Change-Id: Ie98a5c5798bc32082763c1d63c33ef650cef356b

dhcp_test_client is removed; it was apparently only used for manual
testing, and we now have an automated test.

Change-Id: I686ebd5d92970d0c809e461e4b6149074eb469b2

Change-Id: Ia60b1ce5509059f20932a6d2676e9fc6d7690330

This change does not have new behaviors. It changes code
structure to prepare for channel_write_etc implementation

- New overload of RemoveHandleLocked(Handle*) and use of
  it instead pf RemoveHandleLocked(zx_handle_t)
- Centralize handle operations related to channel in
  user_handles.

TEST: existing channel-tests

Change-Id: I45ae6b4915264e3a2d428a80dace28af7d54786b

Test: CQ

Change-Id: I91823a0d0f12f2a26093ba5b0602a92afb4c7141

Add unit tests for the virtio touchscreen
functionality. This verifies that X and
Y values are sent in the correct range,
and that multiple fingers are tracked
correctly.

TEST: runtests -t virtio-test

Change-Id: I5bd575dcd6155a18f6440ec6426d724f6187581f

This test is inherently flaky, since it has
a dependency on sleeping. This is not a useful
end-to-end test, so remove it.

FLK-205 #done

Change-Id: Ic216e3b814b9bfecff028851c31edf9da7658c12

Add support for a new version id that QEMU
can pass through for a multitouch device.

DX-1364 #done

TEST: Ran in AEMU and manually verified that the
      new virtio driver registered as an HID
      touchscreen and was sending valid HID
      reports.

Change-Id: Ia5e57b07ff5278358abc6815426c5cee7b3a80be

When a BR/EDR connection completes, we may search for services on the
remote peer if any searches are registered.  If no searches are
required, we should not connect to do nothing.

Added a failure to FakeDomain when an expected outgoing L2CAP connection is
added but never connected to.

Also fx format-code

BT-861 #done
BT-857

Test: bt-host-unittests
      (updated GAP_BrEdrConnectionManagerTest.ServiceSearch,
       GAP_BrEdrConnectionManagerTest.AddServiceSearchAll,
       SDP_ServiceDiscovererTest.*)
Change-Id: Ib5730fa738718eba10ad29bd26695103cdfe01ee

This works around an issue where the first transaction after enabling
the VGP1 regulator gets a NACK error.

Also add a test to verify this behavior.

Bug: ZX-3486
Test: Rebooted 30 times, didn't see any I2C sensor failure messages
Test: runtests -t mt8167-i2c-test
Test: runtests -t mt8167s_ref-test
Change-Id: Ife7555f8037c520d0e3c6b4aa050ed5bf2620612

This changes the test to perform a system OTA upgrade after the
downgrade, assuming the `--amber-files` argument or `AMBER_FILES`
environment variable is specified.

Change-Id: I02aabaec293c4e8da568bf0ec03ac6ea29add577

Change-Id: I828c47a7444896da759437fc69bd4f4931a02fbc

This rewrites the OTA tests to be more object oriented and flexible.
Major changes:

* "amber" package is renamed to "packages" since amber is going away.
* "packages" grows a bunch of methods to interact with the package
  repository. It's now possible to open packages and blobs and extract
  out package files.
* "build" package is renamed to "artifacts".
* "artifacts" grows a struct to represent a build artifact, which has
  the ability to provide the package repository and system snapshot.
* Added a method to download the build-archive.tgz file (which is
  currently unused, but may be used later to support paving).

Change-Id: I53c0fe1b3f0b3a2659828c6172be175f9b38c125

BUG: NET-2241 #done

Conditonally send ICMP Parameter Problem repsponses when failing to
parse an IP Packet with a parameter problem error.

Change-Id: Icc4424da5f47ac2c0113cf89cb14bb7125f44278

This allows low-level C++ bindings for Zircon FIDL libraries to be used
in the Fuchsia build.

TEST: Verify that the low-level C++ binding
targets can be included in the fuchsia build.

Change-Id: I4ca5b09a563b1eec8001c5d11d4a87f014d4f079

Documentation has since been moved into the "docs/" subdirectory.

Change-Id: I87fdadbd4eac87412efe48a80301370ad5a9c209

Putting dependencies in public_deps allows their headers to be exposed
correctly.

TEST: 1. Observe that the correct public_deps are emitted in
out/default.zircon/legacy_targets-x64.json; 2. In a subsequent patch,
  verify that the low-level C++ binding targets can be included in the
  fuchsia build.

Change-Id: Icb09077c33ef1d95922b51d112ebae1f4cbc5401

Change-Id: I8154dac79e9e569fe54f705d41ebaa868f7958d1

Change-Id: Ia0a9a19a920c8b2ece9f220f28c2f1bb08dbdb6e

It's looking like 5s is too short -- sometimes,
appmgr takes upwards of 2 seconds (sometimes even
longer) to launch basemgr. Let run_component_test
handle timeouts instead.

Change-Id: I3f6a07ede5cdd2ac104165a7e61eac3684518ed2

This reverts commit f7a77a9a.

Reason for revert: broke GI

Original change's description:
> [modular] Fix basemgr to pass config override to sessionmgr
>
> * Re-factor ModularConfigReader to parse config file during construction, and
> expose some getters instead. Have it accept a root directory to work off
> of, which makes it more testable. (i.e., fake out a directory and pass it to
> ModuleConfigReader). Provide a factory for using the incoming namespace.
>
> * Have basemgr always pass config data to sessionmgr in its
> /config_override namespace, since appmgr enforces that /config is not
> overridable. sessionmgr uses ModularConfigReader, which prefers
> /config_override over /config when available.
>
> * Fix modular_config_xdr to parse SessionShellMapEntry -- it wasn't
> parsing session shell before.
>
> Test: modular_config_unittest tests that ModularConfigReader APIs
> parse and return basemgr config given a config directory. Tests that session
> shell URL is being parsed.
>
> Change-Id: I31c2f352a0ff5bb607e1b0a0628aaefd23affd8a

TBR=jamesr@google.com,vardhan@google.com,thatguy@google.com,lindkvist@google.com,vickiecheng@google.com

Change-Id: I63efe2134d73077573f5e40ccbc456f26e6a5317
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

Bug: ZX-2834 #done
Test process:
fx set smart_display_dev.frank \
--with //bundles:tools \
--with //garnet/packages/examples:media \

fx build

fx scp -r
~/fuchsia/garnet/test_data/media/third_party/chromium_media_test_data/bear.h264 \
fuchsia@[$(~/fuchsia/tools/devshell/netaddr --fuchsia)]:/tmp

fx shell "present_view
fuchsia-pkg://fuchsia.com/use_media_decoder#meta/use_media_decoder.cmx
--h264 --imagepipe --fps=24 /tmp/bear.h264"

Change-Id: I757a0929d8d24b5ac7fc2085c63e802342531ba8

This allows the imagination driver to determine the highest address
anything can be mapped to.

Change-Id: I990f5995c732e2fa3760ff4f5a9740c17995468d

Added FIDL change, ComponentContextImpl, and partial implementation in
agent_runner.cc.

Test: fx run-test agent_runner_unittests

Change-Id: I404bb846fca10b86e5b05ab9709fa68eaed60e87

Also updated banjo headers to specify requirements for status.

ZX-3960

Change-Id: Icdff35a13be15303068894ad692b09714a610e54

Added a few casts in byteswap to make it compatible with C++'s stricter
implicit conversion rules. Can now be included from C++ files.

Tests: built clang and gcc, ran QEMU.
Change-Id: If85bf150c8afc3f58ef12a961f5e55ccef6010dc

Change-Id: I02ebd7adcd0d9126a7e792325f2f3e7c5fccbb5f

Change-Id: I92b1cbbb0703e3890471e919ad4ecb6731ae4332

This CL adds a unit test. There's no non-test logic changes.

Change-Id: Ia77f76db029c61ae13cc90a0acdb3d3799aaa27f

Implemented FromStr trait.
Also changed from array to Vec to support version with less than 4 numbers.

Bug: PKG-678 #comment implemented parsing versions
Test: fx run-test omaha_client_lib_tests
Change-Id: Ibe9d8b79cf174808c7d64561845642b0f453c5ac

non-const references.

To align with the Google style guide.

FIDL-607 #done

TEST: No behavior change.

Change-Id: Ib75106f3eebc44a466ef438db87b01c890caa4a5

Initial wlan client api surface to support toggling wlan operating
state, saved networks, connection attempts and provides an update
mechanism for system components to receive wlan state information.

Bug: WLAN-948
Test: builds (this CL just adds the API, no implementation)

Change-Id: I4e2fd3d81e7d3d8440ac4b7a2efe0c53e81ec790

Basemgr and base shells don't need access to system/data/sysui now that we're
not reading base_shell_config.json.

MF-340 #comment Remove "data/sysui" from basemgr.cmx and dev_base_shell.cmx

Change-Id: Ia3a49ffa8b4cab61d430cfb0a63a0f0624ccf8fb

The TextFieldMultiplexer connects multiple TextFieldRequestStreams to a
single TextFieldProxy. Since we'd like to simplify TextField
implementations by allowing them to assume there's only a single client,
we need this code to de-interlace transactions.

This technique does have a latency penalty — it means a transaction has
to reach CommitEdit before it gets forwarded. However, I think that this
could be improved in the future without changing the overall structure,
by forwarding the first transaction that comes in message-by-message,
and making just the second transaction queue. This could be combined
with timeouts that call AbortEdit on the TextFieldProxy and switch to
the second transaction queue to avoid the first holding a lock for too
long.

This code isn't actually in use yet; it will get used in a future CL.

Change-Id: I86eeedbe22a175d23e1382c0dbf40028a3577d4e

Added a drawing example that works only with software framebuffer and
exercises the new Canvas features.

Added very simplistic implementations of fill_circle() and fill_roundrect(). These should
be good enough for recovery-mode UI and for working up the non-drawing portions of
Carnelian until Spinel is available.

Added a simple Canvas test to break the ice.

Change-Id: I557c60a3e7ea4098240f622017f367183c8fda76

Change-Id: I9a1ebe44bfa862bd99e98afd2949ad9b63d518dd

The handle passed to CreateDevice is now always a channel, but for the
platform-bus it will be a channel to fuchsia.boot.Items.

Change-Id: I44ccb0dd7e50ef3063fcaf5e47c42ed30ffcc988