Test: clang lock annotation checking
Change-Id: I4f708e1449db34f0694ce1635eecd8d272f460f6

Test: no functional change
Change-Id: Id68768b6e3ca29a3a58c95737cc9b14cd468e0c6

Test: no functional change
Change-Id: I17dc9b7cfc9ddd6eba80165e777310ce2639a334

Add "DEBUG" and "INFO" levels to FS_TRACE macros. Currently compiled
to no-ops; these will eventually invoke the unified logging API for
Zircon.

ZX-3060 #comment In Progress
Test: CQ (No runtime visible changes)

Change-Id: I23c8877f9b0478766b77f3e2ad8be8e827a75b65

ZX-3012 #done

Test: hid uapp in sherlock.
Change-Id: I4ddec9b597e1dacd48abe98ba8971a736d2e9501

Edge interrupts do not require the state machine handling required by
level interrupts in the kernel and work as well in hid-buttons.
Issues with level interrupt handling has been seen at least on one device
that is being tracked by ZX-3077.  Since we have support for edge interrupts
in all HW that uses buttons today, switched to edge.

Test: hid uapp in astro and mt8167s.

Change-Id: I36725ee87baa7b007b1c523c17d2daca5efda65b

Add PROCESS management to jobs by default, add THREAD management to
processes by default. The logic here is that job handles need to be able
to create processes inside themselves, and that process need to be able
to create threads inside themselves.

Additionally, accept MANAGE_PROCESS on a job as sufficient to
process_create() with a fallback to WRITE for all the exisiting callers
that are calling with WRITE.

ZX-2967 #comment [rights] Add MANAGE_PROCESS to DEFAULT_JOB, MANAGE_THREAD to DEFAULT_PROCESS

Test: CQ including garnet and topaz
Change-Id: Icf767a369299596c2d5974225fb9e31fa59a923b

This has been replaced by fidl

Test: vkcube on vim2
ZX-2829 #done

Change-Id: I5ac6d00600a9a831ccfb27be22fe4b0ebaebadf9

It's different than the rest of kernel/syscalls, doesn't make anything
shorter, and confuses me as I try to remove WRITE usage.

No intended behaviour change.

ZX-2967 #comment [rights] Remove get_process() helper in kernel/syscalls/task.cpp

Test: CQ
Change-Id: I07e85d4686b494153a2007571795ec968e5eb975

usb_request_pool is a pool of usb_requests that a driver keeps track
of, to avoid reallocation of requests. With the recent changes to usb stack,
each driver in the usb stack has its own list_node_t* so as to avoid stamping
on other driver's lists. The library is unaware of the private portions of the
usb_request_t where the node pointer is stored for each driver. To queue
the usb_requests to the pool, this changeset informs the
usb-request library, the offset of list_node_t* from the
usb_request_t.

Test: vim2: fx serve. lsusb, iochk on mass storage device.
Astro: lsusb.
Pixelbook: fx serve. iochk

Change-Id: I78ea612472742ad93b6e5c61ffe09da7819276fd

Tested: Compiles
Change-Id: Ibde34775fbf199d1a9f13d09e01694c6569b8064

Test: eMMC driver probe fails if MSDC0 clock is
      disabled.
Change-Id: Ief62959162f650a9d8beb352da6a89a024ce4a97

Tested: Compiles
Change-Id: I834805865509a03446172e29d8ad7c0348e8751e

Tested: Compiles
Change-Id: Ib313b0dc682be403bc3b68cca8ceb7af206d6ce9

Tested: Ran qemu with network.
Change-Id: Ib76a6e2473627fe5f7319c4cfb6a62223bf74a16

ParseNumericLiteral() uses strtoull() to parse unsigned numeric types.
strtoull() parses negative numbers with unsigned integer wraparound
rules instead of failing. For all <64-bit unsigned types, this still
failed properly, since the wraparound would overrun the upper limit of
the type, but for uint64, this would incorrectly succeed. This change
addresses this error case.

Test: Run new test cases.
Change-Id: I4b419865bcb3059a362685ee1ac8229ee8034844

The purpose of this change is to allow wlan drivers to reserve a small
amount of headroom (64 bytes) in every tx buffer so that it can prepend
its own headers, such as DataFrameHeader and LlcHeader.
This could eliminate the need to copy the data in the ethernet frame.

64 is the closest power of 2 that is larger than 44, which is
DataFrameHeader::max_size (36) + LlcHeader::max_size(8)

Test: Build and smoke test with wlan and ethernet. Value not used yet.

NET-1911 #comment

Change-Id: Ic88c8963d7ce7d8be15a4e26736df8cd116fcce8

usb_request_t now has public and private portions for each driver in usb
stack. The complete_cb is different for each layer. At the moment we are
using the same req->complete_cb and modifying it when the request passes
down the stack. By making a private field for complete_cb in each
driver, we make the code less error prone. This changeset modifies the
request_queue hci protocol op to take complete_cb as a parameter. The
hci layer saves the callback in its private context and uses the one in
the private context when the request is complete. After all the usage
of complete_cb in usb_request_t are removed, I wll remove the
unnecessary fields.

Test: Vim2: fx serve. iochk on usb mass storage device.
Astro: netboot. lsusb. Verified usb network adapter showed up
Pixelbook: fx serve using usb network adapter. iochk on usb mass storage
device. audio -d 1 tone on usb audio device.

Change-Id: Ia5bb083325a9ec2cb7547a378b58a9cbb1e60824

Tested: Booted astro and pixelbook.
Change-Id: I600b706b0229d37f70b4e2636484e91efff87881

Tested: Compiles
Change-Id: I92808a817e688626f71a97b0286c2926a68b6c20

This ioctl is only used to set the clock frequency source on mali on
amlogic. I'll remove the ioctl version once thermd in garnet is
converted.

Test: vkcube on vim2
ZX-2829

Change-Id: I0121c0fdb4a84712cc82ac23ab9dac810fdb3464

Test: Tested on VIM2 and ran display-test (the osd and vd version)

ZX-2667

Change-Id: I8a08a877dc41d9fa3a1f82f9f3d69131d1844abf

In refactoring rights, I broke a bunch of these, but things mostly keep
working quietly in a not-quite-right way, because launchpad happily
takes ZX_HANDLE_INVALID and fires new things into the default root job,
rather than where they were supposed to be going.

No intended change during normal behavior.

ZX-2967 #comment [rights] Add some status checking to devmgr

Test: CQ
Change-Id: Ic0226e5c128665e425f137bb0472250d2fba6048

A channel with id DEVMGR_LAUNCHER_DEVFS_ROOT_HND will have the devfs
server attached to it.  This will let tests interact with devfs.

Test: Booted and ran a test program that successfully reported
the unprivileged devmgr's devfs being populated.

Change-Id: Ic65dafd6b5197e82e7b7f345766b6aee281e00d8

This removes the inner ecam structure that used to be part of
pci_platform_info_t and moves end_bus_num to top level. Since
ecam base was only used for debugging it isn't carried over into
the bus driver anymore. This means the bus driver can now determine
if it has an ecam available by simply checking if a valid handle was
passed to it by the pciroot protocol upon calling GetPciPlatformInfo.

Test: Confirmed config output with upci, and checked booting fuchsia
with a gui for kpci.

Change-Id: I24760bf3faed4a468c910b147297b20f60adf9fa

Test: boot on mt8167s

Change-Id: I6101cfe2ce680a2687e0298e9526cf57fea15a41

This is just a cleanup for consistency.

Test: Compiled
Change-Id: I98fda634ae47f00d84c83ecdf0a9e83ce0097867

This will support launching isolated instances of devmgr for testing.

Test: Will be consumed by tests in the future
Change-Id: I4c44f3fad350ac878fa6b95c44d68d4acc3e03ef

Test: no functional change

Change-Id: I04afee763cc9d607ebe8ffdc67b6aa49590f4dd7

One step towards class-ifying this

Test: booted under asan
Change-Id: Ic54403d27bb84d19a00736256c98d98bc5402b9d

Test: no functional change
Change-Id: I391472fa5e8310cab192b75af6eca09bbb195e2f

The docs talk about Zedboot as a bootloader, which is an accident of
incremental changes, as the docs originally only discussed Gigaboot. The docs
also reference scripts that have been deleted.

Test: n/a docs.
Change-Id: Ib61bec7167a3636534e8c3f2b4354a5e0c50d713

There should be no clients using this.

Test: builds.

ZX-2677 #comment cleanup

Change-Id: Ie18e9f3a28de92b9309c529ca7cd339205a89cd0

It is gated by HandleWrappers attribute until all protocols are
converted.

Tested: CQ. Also tested on pixelbook, vim2, and astro with all protocols
migrated.
Change-Id: I2fed84266bbab7ff7fbec4120112d88bc095b48f

Change-Id: I7fd4ff5c3d726d862d78ee8dcfa1580752555e6c

Previously, we got this value from zx_socket_read by passing NULL and 0
as the buffer and size arguments.

Test: socket-test
Change-Id: I33ecd117172467473a21be376a27837d95f41c01

Added bulk transfers tests for now. The test can choose to use
scatter gather for the OUT and/or IN EP, and we unscatter
the transferred sets of data to compare.

ZX-936 #done

TEST= with fx3 device, usb-test-fwloader and runtests -t usb-test

Change-Id: Iafc4693ad4cede907d5bb5be46cc3a2610e28325

Useful for DDK case.

Tested: `runtests -t libzx-test`
Change-Id: I74b74569ea9aa6dd7f4f99b9bcf7035deafb4514

Tested: `runtests -t libzx-test`
Change-Id: Ice742fb6bcbe4cdbb4c2f3de3fe6eae6c7f1076e

We currently report the total amount of data that clients can ever read
out of the socket, but we need to know the amount of data that can be
read out of the socket in one zx_socket_read call in order to replace
the NULL/0 behavior from zx_socket_read.

This CL adds rx_buf_available to zx_info_socket to contain that
information.

Test: socket_buffer_test
Change-Id: Icb8f2bc314eed5d4afce6190fd11cc08ef6b9d00

--driver-search-path specifies what path to search for driver DSOs
(defaulting to "/boot/driver").

--sys-device-driver specifies what path to open for the system
device driver (defaulting to "/boot/driver/bus-acpi.so" on x64 and
"/boot/driver/platform-bus.so" on arm64)

These will be used in the future for launching unprivileged devmgr
instances for the purposes of testing.

Test: Booted in qemu with and without each argument specified.

Change-Id: I8a2ddc299c7ddcd61b1d436bc3cc675488a927b1