Test: no functional change
Change-Id: I7917af4cdc34d1a432e993c11d82fb67a298f276

This reverts commit adc760b5.

Reason for revert: <INSERT REASONING HERE>

Original change's description:
> [fidl][cpp] Table support: fixes
> 
> See https://fuchsia-review.googlesource.com/c/garnet/+/223232
> 
> Test: compiles
> Change-Id: I29449947ebb6a114275f46973bc9d461ed7d747c

TBR=ctiller@google.com,pascallouis@google.com

Change-Id: I2067819fadd52cd5c8431177a9d01039ef3234c6
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

We were already disabling the virtual timer when exiting the guest, so
no need to repeat it.

MAC-180

Test: Ran "zircon_guest" on a VIM2.
Change-Id: I30600ddcb0a4ea5cb856f108f9620fcfc457aff9

Quiet unbinding doesn't currently work with
zx_task_resume_from_exception() (when exception port is checked) and it
is somewhat involved to make it work properly.

As it's not currently being used, remove
ZX_EXCEPTION_PORT_UNBIND_QUIETLY and related plumbing for now, to be
readded in a modified form in the future if necessary.

At the same time, switch utest/exception to stop using zx_task_resume()
in favour of zx_task_resume_from_exception(), as zx_task_resume() is
deprecated.

ZX-988 #comment
ZX-2720 #comment

Test: CQ, /boot/test/sys/exception-test, unbind_rebind_while_stopped_test removed.
Change-Id: If7f4706f7599c417cdc9447011441e42e892f0c2

zx_task_resume() is deprecated and being removed.

ZX-2720 #comment

Test: CQ
Change-Id: I87a33baec56708cd5cfd84fca093a8c28804373b

If the system was booting slowly (e.g. because ASAN + emulation), the
serial console might not be started.

Test: Booted in qemu-arm64 with ASAN and some prints enabled to record
how long it took for /dev/misc/console to come up.  Observed that
console connected even if it took over 3s for the console device to come
up.

Change-Id: Ibec19f4db78c7641cd31e0b817024d2cb3185252

zx_task_resume() is deprecated and being removed.

ZX-2720 #comment

Test: CQ
Change-Id: Iec99809728bc7d528b6026d8848739f0de9fe0d0

The change in minfs format requires that the zedboot key contains a driver
new enough to read the paved filesystems.

Test: manual
Bug: ZX-3042 #done
Change-Id: Id9691fc8e97e741fda5e0364d9e51a3063d61ab0

All clients have moved to the zxio vmofile backend.

Test: No behavior change
Change-Id: I7c7b4ca6721ca4901ae57c6db6ea544d498f43aa

This CL creates the ability to make a file descriptor from an arbitrary
zxio object. This CL then uses this ability in chromeos-disk-setup to
replace its use of fdio_vmo_fd.

Test: chromeos-disk-setup
Change-Id: I51eca7ba9f326d44df97c8993c082b09a113634f

zx_task_resume() is deprecated and being removed.

ZX-2720 #comment

Test: /boot/test/sys/debugger-test and CQ
Change-Id: I2bbc36d575bf938ce187ba67c8b07bdb19d0d97a

Also, fix formatting of doc comments for consistency.

Test: No functional changes, CQ
Change-Id: I31584af482ef763911a822c5ae6f4fb73e4d6596

See https://fuchsia-review.googlesource.com/c/garnet/+/223232

Test: compiles
Change-Id: I29449947ebb6a114275f46973bc9d461ed7d747c

We should not be using goto statements ever. Refactoring is usually
cleaner and leads to better readability.

TEST=fs-tests

Change-Id: I6608c2925a277d40038ebec4767ffd2e821e9f20

- Remove mentions of READ/WRITE
- Refactor event to point to rights.md to create section on
ZX_RIGHTS_BASIC

ZX-2399 #comment

Test: CQ
Change-Id: Id245980d6074372dd41e1b05f845f2917d33b72c

Test: no functional change
Change-Id: I4c524fe56d8392f7ef1947e1de5cb15479f4f6c3

These were moved out of the header in 36242efc but these structures are
used in garnet; we need access to these definitions to avoid duplicating
the structures there.

Tested: CQ

Change-Id: Iba658c7c8936d28564dc78d5eef3caa581999abf

zx_task_resume() is deprecated and being removed.

ZX-2720 #comment

Test: /boot/test/sys/policy-test and CQ
Change-Id: Ib9a787c69cbf511169e884a9e052a64c440bd144

This patch adds support for Hisilicon kirin970 board

Test: Booted zircon on kirin970 based Honor Play smartphone
Change-Id: I15129749213bcd9aa248e1be76d00fc98d45b160

TEST=guest_integration_tests

Change-Id: I2efb75703f243431ef6fb82ada2f399adbbf8486

The new name does a better job conveying the behavior of the executor
and is less ambiguous particularly in the presence of the newly
introduced fit::sequencer which really does apply sequencing.

Reland of abf79d33

Test: fit-test
Change-Id: I50c58c055620d39209ff39088865e71431776588

Test: fit-test
Change-Id: I9b20208689f55a4835f02093ee49abd4666782cc

Reland of 7baf23bc with build fix

Test: fit-test
Change-Id: I02296cd441cd999268c1f686d334ef98f7e02eee

After this CL, all the FDIO objects that use a ZXIO backend use the same
memory layout. This CL prepares us for unifying the ZXIO-based FDIO
objects.

Test: No behavior change
Change-Id: I2568eb3034bcea06804c37e10e40e0b7de704b7a

All clients have been migrated to zx_socket_shutdown.

ZX-2994 #done
Test: Removing dead code
Change-Id: I6240cd9c446aa49c5bbc36b42f0fbafd8ac7f707

This reverts commit 7baf23bc.

Reason for revert: blocking Zircon roll
...
sequencer.h:47:14: error: use of class template 'fit::bridge' requires template arguments
...

Original change's description:
> [fit] Add fit::sequencer for ensuring sequential execution order.
> 
> Test: fit-test
> Change-Id: I2f44bdba931f8a7bb9a0917549c1eb7c43858b70

TBR=jeffbrown@google.com,thatguy@google.com

Change-Id: I30bc48691e32dca236a0164f002135d46f8d4690
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

Test: This adds a test.
Change-Id: I4c380678301bf724c95693c538d2c97ad48567df

Revert "[fit] Rename sequential_executor to single_threaded_executor."
Revert "[fit] Add fit::sequencer for ensuring sequential execution order."

This reverts commit abf79d33.
This reverts commit 7baf23bc.

This code does not build, at least as C++14:

In file included from headers/headers.cc:98:
.../pkg/fit/_virtual_includes/fit/lib/fit/sequencer.h:47:14: error: use
of class template 'fit::bridge' requires template arguments
        fit::bridge bridge;

Test: Failed SDK builders
Change-Id: I95ff20ee448faea541ce7526ad139bbaf9449809

In order to reduce verbosity in the output of k counters, instead of
always printing both the sum and the per-core count, only output the sum
unless:

1) the user specifies a new -v option as in e.g.
   `k counters view -v kernel.channel.messages`

2) there's a outlier in the per-core counts that might be indicate that
   one core is working unusually hard. This outlier calculation used is
   https://en.wikipedia.org/wiki/Outlier#Tukey's_fences with k=1.5,
   which is not perfect. It is fairly conservative and might still
   output in not-so-interesting cases.

ZX-3008 #comment

Test: New unittest counters_tests added, quartile calc manually compared to https://docs.google.com/spreadsheets/d/1D58chwOpO-3_c41NMGJkpmFuOpGSYH-W50bD6MdOAjo/edit?usp=sharing
Change-Id: I2ec5fcdb0d5320abf70fc5cea7785b380d90c29b

Update symbolizer to revision 7140d924a314402c36bae66f34d0e77587ffb825

Test: CQ
Change-Id: I323ad5af628ccfbce59d3cd4c4861f3b8f45a66f

The new name does a better job conveying the behavior of the executor
and is less ambiguous particularly in the presence of the newly
introduced fit::sequencer which really does apply sequencing.

Test: fit-test
Change-Id: I7d4375fd340c929658652f1b7e8bfc1167676b52

Test: fit-test
Change-Id: I2f44bdba931f8a7bb9a0917549c1eb7c43858b70

TEST: install-disk-image-test passes, paving mt8167s starting to work

Change-Id: I804614df6c771fec739f735b3e331da9c5b64a78

Test: No behavior change
Change-Id: If6d8c2c215208765d2687712ac39b00a08d14560

This bootstraps the required devices to enable communication with the
bcm43458 wifi/bluetooth transceiver chip.

Test: wlan scan+connect; ping w/eth0 disconnected
Change-Id: I948bf41df1058dd93e77dea7036e9b3b4f6fb668

The pci bus driver will now obtain platform information from
pciroot and map an ecam into its address space if one exists.

Test: Checked config dump of 00:00.0, and verified kpci behavior
has not regressed;

Change-Id: I50c393c543de0323671b623c4deebe445305b727

These symbols are unused.

Test: No behavior change
Change-Id: I643a6b7633a1ca5dbf824403ee7773947ee1d6a2

As far as I can tell these haven't been used in well over a year.

Test: Booted
Change-Id: I7c713a1d565011617d193647fb9704201a614ac1

We no longer need this header.

Test: No behavior change
Change-Id: I03679671db7c478709c0f78f815c310cb238b3c5

Test: build/ran kpci with graphics under qemu

Change-Id: I87a256ccc299f11cd1fd4aa4667a9acf89d364b7