Change-Id: Ic050b27bd8ef017a40414e48503afe1376164cfa

Other kinds of directories will be added in the future.  The first one
will be a lazy directory.  They all will need to deal with the FIDL
connections.  So, extracting common functionality into a separate file.

Simple directories order their entries alphabetically, but a lazy
directory may use a different kind of ordering.  Making
DirectoryConnection polymorphic on the traversal position type will
allow it to be used with different kinds of directories.

Change-Id: I5571c847aa467f9f08cf4691cfffa22d93a56c4d

This change breaks up the "model" module in multiple submodules:
- error.rs: ModelError
- instance.rs: Instance, Execution
- model.rs: Model
- routing.rs: capability routing (formerly part of Model)
- namespace.rs: IncomingNamespace (new struct, contains most namespace
construction and capability discovery logic formerly residing in
Execution)
- tests.rs: Tests for the overall model

CF-678 #comment broke up model/mod.rs

Change-Id: If27dead6e41dc9dea6f226ef5b55c7a3360afa09

We need the ability to recognize information frames (as opposed to
supervisory frames). And stylistically,
|!designates_supervisory_frame()| is an oblique was of expressing our
intent. Add |designates_information_frame()|, to more directly express
our intent when trying to recognize an information frame.

Bug: BT-442
Test: bt-host-unittests

Change-Id: I5c25d3162a7854b70770fe4dc77db22e839f38a7

This reapplies 3dec4ecb

> [virtio-magma] build libvulkan_intel for linux
> This change adds a linux_x64 target for libvulkan_intel, which will be
> used by VM guests in a future change. It also adds this library as a
> resource to the virtio_wl package, and exposes it to the guest using a
> new temporary ioctl.
>
> Test: manual run of magma_test in biscotti
> Change-Id: I1cc151d421cdd5e7c2144cb1b87985bf5a050ede

Test: fx build
Change-Id: I7a98c636a1d9c2adea8bf9b724db42b0524af47a

Change-Id: I2f6d016256930bd6d4b066eb501892ea41957502

Objects are referenced counted appropriately;
command buffers retain references to the context,
and contexts retain references to the address space.

Change-Id: Id2da395f5f3a387b30062d7b4085d85c8a476532

ZX-3612

Test: `k ut handle` and `runtests -t event-pair-test`
Change-Id: I93cd7f1f3714c25e078978a415ae0ad2fcd85385

ZX-3612

Test: runtests -t fifo-test
Change-Id: I4746c3a5eb60c93ecaf321153e580fe2a4db7c47

VK_EXTERNAL_SEMAPHORE_HANDLE_TYPE_TEMP_ZIRCON_EVENT_BIT_FUCHSIA is
the correct enum after switching to VK_FUCHSIA_external_semaphore.

Change-Id: Iddcb2a1825876eb3c929d24bebec51fcb6e5a5ee

Previously, the journal entries were only replayed
while mounting blobfs. This patch causes fsck to replay
the journal first before mounting.

Change-Id: I9d9801685c0604add663f5b567c3aa1ae6c56555

To look up type names we will be expanding on the current variable
finding code. This patch is a first step to rename the associated
classes and functions to "Name" instead of "Variable".

There is no functionality change.

Change-Id: I2e695b7b0101f5f1a40553c353595840aa93bae3

Refactor fidl echo server examples to use fuchsia_component rule.

Add dependency to echo_server_rust component from
component_manager_tests, as the tests rely on the echo server being
available.

Change-Id: I627af135cd3a937e9d9de50b3a8226d5d9673db8

When I changed the way message types work I missed this case.

Added a script to run all the samples so it will be easier to check that
they are all still working.

Change-Id: I8d6aa5bb9d1cfbd964ff36db2016b085a0a396cf

Change-Id: Ia4a4bac565745dfea9fcf1d7617560b6b786bb9a

This enables line buffering, which folks were relying on for debugging.

MA-573 #done

Change-Id: I03c522eeab2cb5ceec8ff3459dc418d82b0a8c03

To bridge between function arguments/returns and encoding/decoding, the
Go bindings generate request/response structs holdings the various
argument/return values. However, developers should not interact which
these, instead using method calls and receiving returns.

With the introduction of the error syntax, the fidlc compiler creates a
name for the struct holding the value, i.e. the first variant of the
result union where the second is the error. The name for this struct is
suffixed by "Response" which used to clash with the response struct
mentioned above.

By keeping the request/response struct private (unexported), we avoid
this clash.

(This does not solve the more general problem of name clashing -- nor
does it try to.)

FIDL-519 #comment

Change-Id: I50767de8d4b3e2d2cfd1b4f7fd4585ead9a455fd

Change-Id: I2287ecb6f1a5092863d5e6ab4f66b1b2dc85de65

Test appears flaky, but only in CQ.  Disabling for now until I can get
to the bottom of things.

FLK-129  #comment test being taken offline pending debugging.

Change-Id: Ifc703993f2c30bfe97c7f1b1f2fc2dc102e836cf

eTempZirconEventFUCHSIA is the correct enum after switching to
VK_FUCHSIA_external_semaphore.

Change-Id: I918e96cfeb223b16e2a409dd12a60af48859f460

DX-387 #comment

TEST=CQ

Change-Id: I36ad246c5a8ec9e59000419492b4e05e154c874f

'Using the error syntax' should be sub-section of 'How should I
represent errors?'

Change-Id: Ifeb567a2502e60788e0418e0712735f04a9e1821

This just moves the files, merging them comes later.

Change-Id: I1e98c3758a0840f711a9b829b74f1e847814797e

This switches the system updater to use package resolver
for caching packages, instead of talking directly to amber,
which will eventually make it easier to remove amber without
disruption.

Test: Manually ran system_ota_tests and confirmed update worked
Bug: PKG-607 #done
Change-Id: I5111f8011b5cf49e913208096529a9c7cc98dd7e

Change-Id: Ia8a8b7a27d53040f87e58cb99b9a15826f76501b

File name change only. This CL doesn't migrate the code.

ZX-3715

Test: qemu. clock-test passes.
Change-Id: I892d9dfffff1be96cc280afdcd67749b7712490c

This reverts commit 8c615366.

Reason for revert: Could be causing tests to fail

Original change's description:
> [pty] switch apps/drivers to use FIDL instead of ioctls
> 
> Change-Id: Ie00688161e1d4313f603bf3ef2f29c5440b0e7c1

TBR=voydanoff@google.com,smklein@google.com,teisenbe@google.com,cjn@google.com

Change-Id: Ifa33598f68aa8e5ecd072e7f7ca8b6d75e2144d1
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

Now that output to debuglog is fixed, this adds printfs to bootsvc,
fixes other printfs in bootsvc, and removes a superfluous printf in
devcoordintor.

Change-Id: Ic22db1e626d07b862c62ad4d52867e7e3028f8aa

People sometimes try to name their packages one of these names, which we
should avoid because they're too generic.

Change-Id: I6ebaf9e26b8d88c53c494fb98cfbbc05f4a0f56b

Change-Id: If8aa4243a71e479f84b9947392853669c9bd5a81

Iterate over all HID Reports in the Report
Descriptor and create a Device for each
Report ID that requires its own device.

This adds support for composite devices
that have multiple Reports. E.g some
touchscreens have seperate reports
for the touchscreen and the stylus.

Only fail if we are unable to find
a single HID Report that we can
parse correctly.

SCN-1251 #comment

Change-Id: Ia366232fc243ec634365615e16e3f48c68c12d94

The touchpad and touchscreen drivers depend
on the touch driver. Call the touch driver's
report Id function so the ReportId() call
works correctly.

Change-Id: I974f8650d39b76b85124ca13cb23fad7ec30ad4f

Add an InputDevice struct to InputInterpreter that
will hold all of a given device's information.
This will allow us to have an array of devices
from a single Hid endpoint, and it will allow
us to support composite devices.

It also allows us to add more device types by
only editing a single location in InputInterpreter

SCN-1251 #comment

Change-Id: I2bef76256b46771db4b1c19db9ffae7b82dec13c

The callback for RegisterDevice in the mock framework
would std::move the callback function. This prevented
the mocking framework from registering more than one
device successfully.

Change-Id: I07560239035b169202a985b0d45e6f972a86e1b8

PortDispatcher is retained by its PortObservers. Prior to this change,
PortDispatchers were not freed until the source Dispatcher was
canceled or closed. This made it easy to "leak" PortDispatcher and
PortObserver objects.

Now, when PortDispatcher's handle count drops to zero, it cancels the
observed source dispatchers and frees all PortObservers that might be
retaining it.

Testing:

1. Manually verified that port dispatchers are freed by running a test
(as described in ZX-3379) that continually creates a port, does an
object_wait_async, then closes the port. Watched kcounters and
observed the number of port dispatchers did not grow without bound.

2. Ran port-test's cancel_stress on x64/qemu+kvm and arm64/qemu+kvm
with a high kStressCount.

3. Ran port-test's new signal_close_stress on x64/qemu+kvm for several
minutes.

4. Ran fuchsia tests under lockdep and verified no new lockdep
issues (build with "ENABLE_LOCK_DEP := true" in local.mk, "fx run -k",
then "runtests").

Performance:

Ran port related zircon_benchmarks on Eve (--filer=Port). While
results varied from run to run both at HEAD and with this patch, no
consistent performance regression were observed (i.e. sometimes HEAD
was faster, sometimes this patch was faster).

Bug: ZX-3379 #done
Test: various tests described above
Change-Id: I9aa80844415c932f755d6818d3b38e24ccec89cd

This CL extends fuchsia-hyper to add support for communicating
with https web services by using rustls to provide for encrypted
communication.

In addition, it modifies HyperConnector to default to port 80/443
if a port was not specified in the url.

Bug: PKG-538 #done
Change-Id: Ie020bc0c3f8fd4a8f0128ace0e744b16fa3c6977

This switches amberctl to use package resolver for caching packages,
instead of talking directly to amber, which will eventually make it
easier to remove amber without disruption.

Note that this also removes `amberctl get_up -nowait ...`, because
that functionality cannot be replicated with the PackageResolver API
and golang's synchronous implementation of FIDL.

Test: Manually confirmed it can cache packages.
Bug: PKG-520 #done
Change-Id: I4e0b878bf3b31edfe7d0fe23afc080e26c0e7e04

Change-Id: I60a493d9d20807af0f2d7f686f2aec311d27b4d7

The shader while(true) loop was being removed by the Intel compiler.
Instead, read and write a memory location forever.

MA-530 #done

Test:
nuc:go/magma-tps#L0
vim2:go/magma-tps#L0

Change-Id: I6462151e9141d14fe46b15b8000bc38af0274539

TEST=fx build

Change-Id: Ieffee5e4e62d77b9168fea4f8e3ff315d851dd9c