Bug: ZX-1751 #done
Test: docs-only change
Change-Id: I07d59f9b5179b2faa357dba306d3e13c80c8e14e

Bug: ZX-1751 #comment Full libc++ use now available in Zircon.
Change-Id: I3f2f7e2e9a8f81129d27374e6e396a75586f6774

This CL adds environments derived from the standard environments that
instrument code with '-fsanitize=fuzzer' and link against
libclang_rt.fuzzer.a.

Bug: SEC-251
Test: noop-test; see zircon/public/gn/toolchain/environment.gni
Change-Id: I7d25134b78eb2fbdead551599e15218b68ecd5ba

Test: CQ

Change-Id: I47754d1308745f71fffe2197150e4e66c64b850d

The kernel supplies its own C++ ABI functions directly.

Bug: ZX-1751 #comment Removed zxcpp entirely
Change-Id: I3707353210d26aa98682d2b29e22dbbf3b44078e

Bug: ZX-1751 #comment Removed zxcpp from userland
Change-Id: Ic4fedc8e0c2362c1f26864629078cd86c6a59cac

... and add VLOG message to perfmon event processing

Change-Id: I3029a9c97a8c2710025d5e435fccf508e168f156

Change-Id: Ib244d3afefbad49c5534bf71882c7f21ceff3b9b

Remove the ::common:: namespace for sapphire-wide objects.

fx format-code

Test: bt-host-unittests, manual test of supported platforms (no
    behavior change)
BT-870

Change-Id: I810f29852e0da0d3d5bec25d57b2706d39be4e61

This allows for a zero copy conversion of a package uri into
a repo uri.

Change-Id: I6281b1a323ac7fe8b65c0556f1d638cd3cd8d114

DX-1013 #comment Add XUnion support

Change-Id: I778175f975f9783824531018dab47fab8cadba7f

This new variant enables all the kinds of sancov instrumentation,
not just the basic one that the standard runtime handles.  This
exercises more of the instrumentation features that e.g. fuzzing
also uses.  The special magic in the dynamic linker to cope with
the instrumentation is extended to handle all the cases.

Bug: ZX-3986 #comment ubsan-sancov-full variant
Test: /boot/test/sys/noop-test.ubsan-sancov-full
Change-Id: I2ca2bbaca732526ee10e0d60822410f48d8eb1ef

This makes it possible for Fuchsia GN targets to depend
on //zircon/public/lib/trace-handler or on a static library
like //zircon/public/lib/trace-test-utils that uses it.

Bug: BLD-473 #done
Test: no immediate effect on the build
Change-Id: If29ed149c178c2aae0d03316d518de6e2d3c22a0

Change-Id: I572c24dbcaf13e0197037499e379f94956cc5aa7

Several customers have started copying these libraries into their tree.
We're now at a point where we can add these libraries to the SDK.

DX-387 #done

Test: Captured by the "api" files.
Change-Id: I0d2ecb6a259a145a9563ed5d3c39e28b5b467c89

This reverts commit 96f91478.

Reason for revert: Breaks tests of SDK consumers.

Original change's description:
> [amberctl] Mirror source edits to pkg_resolver
> 
> To ease the migration from using amber source configs to using
> pkg_resolver repository configs, this change modifies the source
> management commands in amberctl to perform roughly equivalent operations
> on the RepositoryManager and rewrite.Engine.
> 
> Test: modified amberctl test to verify pkg_resolver state
> 
> PKG-704 #done
> 
> Change-Id: I92853a61b59a60693014853862a9494c9d227caa

TBR=raggi@google.com,etryzelaar@google.com,kevinwells@google.com,computerdruid@google.com,galbanum@google.com

Change-Id: Ie0621d15882b5f71bc737fa2a9558b6d965cc361
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

Change-Id: Ie2919f76b09db6f7419fc3be598af72f6f10ef24

Test: No behavior change
Change-Id: Ibd3183ee8337a77d30701a4e75cf635458dc10f4

the default name (from when the lookup fails) changes from "unknown-package" to "", which is fine and package was a bit of a misnomer anyway

TESTED=`fx run-test crashpad_agent_tests`

Change-Id: I51e4b86e2ccc74bc3e2d8751480211139f68aab2

for now, there are no data provider processes, but in the future there
will be one per connection to fuchsia.feedback.DataProvider. This CL is
just to introduce the test utilities.

DX-1497 #comment

TESTED=`fx run-test feedback_agent_tests -t feedback_agent_integration_test`

Change-Id: Ia5cb07334d3d27874318cd6b54fcf4afaa336da0

This is a reland of e52f9b4c

Original change's description:
> [build-id] pass extended ids.txt to dump_breakpad_symbols
>
> This change creates a concatenation of the ids.txt produced by the build
> that that produced by the .build-id archives downloaded from CIPD - and
> passes that to dump_breakpad_symbols.
>
> Tested locally.
>
> Bug: INTK-884 #comment
> Bug: INTK-969 #comment
> Change-Id: I9d96f8098239ca6a1c828be0a44dbe36a0c3b14a

Bug: INTK-884 #comment, INTK-969 #comment
Change-Id: I90adfc83c16bc00e80ff97f8bfe7190ae4607e21

DeviceId only referred to peers and is renamed to be more consistent
with the idea that it is an identifier for a bluetooth peer.

fx format-code because "Peer" is 2 characters shorter

Test: bt-host-unittests (no behavior change)
BT-870

Change-Id: Ib97223e106bf9a9efb506de343ce218ba0b50c62

When library A depends on B and B depends on C, declarations from C are
not emitted as part of the JSON IR for library A. This would not work in
backend generation if A references declarations from C.

TEST:
./scripts/build-zircon-x64 -A
./scripts/ninja -C ./build-asan system/utest:host
build-asan/host-x64-linux-clang/obj/system/utest/fidl-compiler/fidl-compiler-test

Change-Id: Ic48ec8b2e55280ac6bedbd78853c2bd1303df288

Begin in remoteio.cpp, by rewriting calls to the C binding as llcpp.

Public APIs converted:
- fdio_open_at
- fdio_service_connect_at
- fdio_service_clone
- fdio_service_clone_to

As of this patch, some implementations are less than ideal: most
helper functions take zx_handle_t as arguments as opposed to zx::channel
or zx::unowned_channel, which encodes ownership in the type. As a
result, there are frequent casting from zx_handle_t to zx::channel etc.
and vice versa. As greater proportion of the library is converted to
C++, these conversions may be removed.

Change-Id: I9a8bc4df228610daafa64eec7ebc99ba194a3489

Instantiate the noop-test target in each variant of each environment()
defined by standard_environments().  The top-level build-tests target
collects all of these and is in the default set.  This ensures that
each supported variant gets exercised and can successfully build an
executable.

Bug: BLD-430
Change-Id: Iccd823f7711c9cff04e0c6bf891f7d4541023b27

Bug: NONE
Test: fx build-push sl4f
Change-Id: If2b2f90cd059f82122c98b1ca73f99ad743b1512

Change-Id: Iad9571687b97d6f41b40d5ee17c3e2c356295c8a

Change-Id: I9ac02eced65672a7fbedad4dd8337875846dd530

Bug: BLD-478 #done
Test: observed built bringup.x64 system has no /boot/test
Change-Id: Ib31ecd17306448f7a2600d3213316d8ead843536

Test code needs to use the mock mmio library header while driver
code needs to use the real mmio library header.  Hence, a library
like ddktl or hwreg that is used in both contexts cannot rely on
this header directly.  Instead, make the APIs agnostic to mmio
implementation by using forward declarations and requiring users of
the mmio-related APIs to #include what they use.

The renaming of the mock library to aaa-mock-mmio-reg was a previous
attempt to work around this issue, which is actually wholly
inadequate in the current build system.  So the library is renamed
back to mock-mmio-reg.

In the hwreg library, this was achieved by using a template common
to the kernel and non-kernel cases, rather than an explicit forward
declaration.  To make that work, the argument order for
RegisterIo::Write was swapped to match MmioBuffer::Write so that
they are interchangeable in template instantiation.

Both implementations of the <lib/mmio/mmio.h> were using templates
where they should have been using simple forward declarations.  Fix
that so that separate forward declarations become tenable.

Some cases of non-driver code still using the real mmio library
remain and need to be untangled.

Bug: ZX-4062 #comment hwreg, ddktl libraries made mmio-header-agnostic
Change-Id: Ie20c246049b37569346790439b6b315fef69234a

This attempts both netruncmd and ssh.

Change-Id: I7fb6d2a16c1ae436f4a113207de1c536d789cfbc

This adds support for reading notes from sections before reading from segmnets.
This allows split debug files to be used.

Bug: TC-450 #done

Change-Id: Ib2f66b9ea08f1767fb992dbf41955ff64abbc929

On powervr the alignment requirements are large (around 4096), causing
the bind offset to be large enough so that the resulting buffers OOM in
ManyReadback.

Change-Id: Ib6e046a16fe491742f971004bb3088eadec72721

Previously the two auth providers were packaged together. There wasn't a
good reason for this and its likely they will eventually be used
separately so we give them their own package.

AUTH-196 #comment

Change-Id: I5dd60fdd18a10261636e46fde90d30587e76f0a9

as part of the effort to use consistent vocabulary, rename RemoteDevice
to Peer, including PeerCache and related variables.

fx format-code for reflows.

Test: bt-host-unittests (no behavior change)
BT-870

Change-Id: Ib96e929f93e2fac23ccf098e6df9921678cd2e11

A subtle case in the move constructor of llcpp unions was exposed in the
fdio refactor, when the code uses more than two std::moves, and compiled
by GCC for ARM64. The move constructor delegates the move of the union
payload to the corresponding move assignment operator of the payload.
However, the payload's memory is not initilized when its move assignment
operator is invoked, causing potential undefined behaivor.

GCC perhaps took advantage of this and generated unexpected code.

TEST: https://fuchsia-review.googlesource.com/c/fuchsia/+/274654 passes
all the tests, when rebased over this patch. I'm not exactly sure how to
write a test for this exact case.

Change-Id: I05b1eb4be60d921a7ab12f167b45189ab1c4de94

Test: CQ

Change-Id: Ib88d643a5fe64fce99fc33e68588453dddf66521

This makes it easier for other tests to use the infrastructure
created in the system ota tests for other host/target-style testing.

Change-Id: Ib704b713d2889512656ebdbc0a58fd3063649414

This reverts commit f50bcc0c.

Reason for revert: Build is failing

Original change's description:
> [wlan-inspect] Be explicit about logging Option
> 
> Previously, we added a blanket WriteInspect implementation to Option
> types that would log a value only if an Option doesn't exist. When
> looking at an inspect_log call on an optional value, it's non-obvious
> that in some cases, no value would be logged, and this blanket
> implementation was only added because inspect_log didn't have a way to
> specify optional logging.
> 
> With this commit, optional logging can be specified with a question
> mark on the key, so the blanket WriteInspect implementation is also
> removed.
> 
> Test: fx run-test wlan-inspect-tests
> Change-Id: I926c20fe2e64fc2e1221c2b84b61d5d9b062e817

TBR=qsr@google.com,hahnr@google.com,eyw@google.com,kiettran@google.com

Change-Id: Iac6d853aa9561dd04089408b071909da66dcb63e
No-Presubmit: true
No-Tree-Checks: true
No-Try: true

This was a band-aid put in place during the NUC shortage that allowed
NUC environments by default only for debug builds.

Change-Id: I55939864854b1f8ea640976d4cfc891becdc675a