llvm-nm doesn't seem to accept shorten form of --format arguments,
see below an error excerpt:
arm-rdk-linux-gnueabi-llvm-nm: for the --format option: Cannot find option named 'p'!
So, let's pass the full name of the option (which also works for binutils version of nm).
Signed-off-by: Damian Wrobel <dwrobel@ertelnet.rybnik.pl>
Mirrors the change from 8e82c6ab1b%5E%21/
Also adds a new variable needed to build ToT ANGLE. This is because the toolchain change is unfortunately linked to rolling ANGLE at the same time, since the current ANGLE pull has build errors under VS 2017, but ToT hits internal compiler errors with the old toolchain, so they need to be updated in lockstep.
Part of fix for https://github.com/flutter/flutter/issues/48831
This new flag will be used by Skia GN rules to determine if it should be built
using the Fuchsia SDK. This is to support standalone builds of Skia using the
Fuchsia SDK (it could previously only be built with the Fuchsia SDK as part
of Flutter).
Unblocks https://github.com/flutter/engine/pull/13123
* We should only enable static thread safety checks for Flutter translation units. Not all third party code is correctly annotated for static checks. This will be enabled seprately in //flutter.
* Allow third party users of libcxx to use RTTI (used by ICU).
By default, address sanitizer enabled builds can detect leaks if the `ASAN_OPTIONS=detect_leaks=1` environment option is specified. If only leaks need to detected, this standalong option may be used.
These warnings were present for a short duration on LLVM/Clang. The latest rolls
no longer have the same. This is causing toolchain autorollers to fail on
the specification of the unknown flag.
We should figure out if it makes sense to slow down the cadence of these rolls.
Note that as of C++11, the standard specifies that such files shall be
treated as though they include a newline at end-of-file. We add this
warning solely to match Fuchsia's clang warnings and ensure we don't
commit files without newline at EOF that will later break when rolled to
Fuchsia.
To detect and rebuild all TUs on toolchain updates, we inherited a rule from
Chromium whereby the Git SHA of the //buildtools repo would be used to insert a
preprocessor define in each TUs build invocation. But we don't use a
//buildtools as a repo anymore. Instead buildtools comes from CIPD. So the
script now picks the //buildroot revision instead. But we don't specify the
toolchain revision in that repo either though. It is instead specified in the
engine. So if the engine updates the toolchain, only the next (unrelated)
//buildroot rev will ensure all old TUs are rebuilt unless the `out` directory
is cleared on each invocation.
This patch replaces the old buggy technique with one where the system headers
(that include headers in buildtools) are tracked in addition to user headers.
When the toolchain updates, these "system" header timestamps will be invalidated
and the old artifacts invalidate.
The simple change of just tracking the //flutter (where the toolchain SHA is
present) would make it so that each engine change (frequent) would rebuild all
TUs instead of just each buildtools change (rarer). Also, this technique would
leave us susceptible to the kind of issue that led us in this situation in the
first place.
I also just searched and replaced all calls to track user dependencies with ones
to track system + user dependencies. Not all toolchain are actually being used.
We should clean these up in a separate patch.
Separates the min SDK version from the min deployment version on macOS
(as it already is on iOS), and lowers it to 10.11.
Using 10.11 as a minimum allows a wider range of hardware support, and
there are currently no issues compiling the engine with a target of
10.11.
Part of https://github.com/flutter/flutter/issues/33200
This adds minimal (in some cases dummy) files sufficient to allow
building ANGLE in this buildroot using the build_with_chromium
option.
This relanded version uses the new swiftshader_ prefixed versions of the
DLL names in the swiftshader build, which are loaded preferentially if
available. This ensures that the swiftshader tests will use those,
rather than the potentially incompatible ANGLE versions. In local
testing, this resolved the ShellTest failures that led to the previous
revert.
Reland of fe3b82877e (#276)
This adds minimal (in some cases dummy) files sufficient to allow
building ANGLE in this buildroot using the build_with_chromium option.
