Since we don't support using the component build to produce mojo apps,
we can simplify the build targets in a few ways:
*) every mojo_native_application must depend on the c system thunks,
so just make that part of the template instead of requiring the dep
*) there's no such thing as depending on gles2 headers from a component,
so delete the forwarding group.
Most targets that want to use the gles2 headers in a mojo context
want to depend on an implementation through the thunks, so
//mojo/public/c/gles2 does just that. A smaller number of targets (such
as the implementation of the thunks) want to just depend on the headers
but not an impl, so they can depend on //mojo/public/c/gles2:headers.
The //mojo/public/gles target isn't that useful since the only thing we
expose is a set of C entry points.
We can probably also simplify the c system targets, but that's trickier
due to more extensive use from the chromium side.
BUG=438701
R=viettrungluu@chromium.org
Review URL: https://codereview.chromium.org/780733002
This adds a tracing service that can aggregate tracing data from
multiple sources and write a json file out to disk that trace-viewer can
understand. This also teaches the shell, sky_viewer, and various other
services how to register themselves with the tracing service and provide
tracing data on demand. Finally, this teaches the skydb prompt to tell
the tracing service to start/stop tracing when the 'trace' command is
issued.
The tracing service exposes two APIs, a collector interface and a
coordinator interface. The collector interface allows different entities
to register themselves as being capable of producing tracing data. The
coordinator interface allows asking the service to collect data from
all registered sources and flushing the collected data to disk.
The service keeps track of all open connections to registered sources
and broadcasts a request for data whenever the coordinator's Start
method is called, then aggregates all data send back into a single
trace file. In this patch, the tracing service simply gives all sources
1 second to return data then flushes to disk. Ideally it would keep
track of how many requests it sent out and give each source a certain
amount of time to respond but this is simple and works for most cases.
The tracing service can talk to any source that is capable of producing
data that the trace-viewer can handle, which is a broad set, but in
practice many programs will want to use //base/debug's tracing to
produce trace data. This adds code at //mojo/common:tracing_impl that
registers a collector hooked up to //base/debug's tracing system. This
can be dropped in to the mojo::ApplicationDelegate::Initialize()
implementation for most services and applications to easily enable
base tracing. Programs that don't use //base, or that want to register
additional data sources that can talk to trace viewer (perhaps providing
data that's more easily available from another thread, say) may want
to create additional connections to the tracing service.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/769963004
1. Print out the sky_server commandline as well so it's easy
to run the server without running test_sky.
2. Quote the --args-for arguments, so that they copy-paste
correctly.
R=abarth@chromium.org
Review URL: https://codereview.chromium.org/762703005
This moves the tracing code and mojoms from sky/viewer/services/ to
mojo/common/ in anticipation of refactoring these to be more widely
usable. This doesn't actually change behaviors yet.
BUG=436639
R=abarth@chromium.org
Review URL: https://codereview.chromium.org/751303003
Previously, skydb could be used only for Debug builds. As part of working on
graphics performance, I'd like to use it for release builds as well. This CL
teaches it to use the same configuration arguments as sky_server, test_sky, and
test_perf.
The downside of this CL is that you'll need to supply --debug in the common
case of working with a debug build.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/746473003
Flakiness dashboard assumed each buildbot step corresponds to one
test suite. So, it assumes the test type matches the buildbot
step name. This fix is somewhat hacky in that we won't be
able to make other test suites in mojob test upload to the
flakiness dashboard, but that's technical debt we already have
from smashing multiple test suites into one buildbot step.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/725993003
Lets see if this removes the flake on the bots seen in sky tests. In
local testing it looks like sky_server's CherryPy becomes overloaded and
then the requests take a very long time to complete which makes the tests
timeout. I increased the timeout locally and it made the tests start
passing consistently when using high --repeat values.
R=ojan@chromium.org
Review URL: https://codereview.chromium.org/741603002
Since Sky is checked directly into mojo and does not contain
Blink history, I knew of no easy way to resurrect Blink files.
In making the V8 debugger work, I had to bring back a ton
of files and thus wrote a script to help me do so.
Maybe others will find this useful.
R=ojan@chromium.org
Review URL: https://codereview.chromium.org/732193002
When we removed Aura from the build, we broke event targeting. This CL hacks
around the issue by blindly forwarding events from the root view to the content
view in skydb. Once we have proper event targeting, we should rip out this code.
R=erg@chromium.org
Review URL: https://codereview.chromium.org/714213003
[Changes: Remove additional deps which Linux/Windows was fine with, but
which broke android because it implied an aura dependency. window_manager
should now be properly deauraed.]
This patch is only half done, however the main driver (removing aura) is
needed by other teams for them to make progress.
Input handling is regressed in this patch; while FocusController is
rebuilt to the point where it should theoretically work, all of the
details about ui::EventTargeter don't work correctly. There is an empty
subclass ViewTargeter which will need to be filled out before
Thankfully, this should only regress people who are using the wm_flow
demo, and this should unblock the android guys who don't have aura
support on their platform.
BUG=431047
TBR=eseidel@chromium.org, sky@chromium.org
Review URL: https://codereview.chromium.org/710203006
This patch is only half done, however the main driver (removing aura) is needed by other teams for them to make progress.
Input handling is regressed in this patch; while FocusController is rebuilt to the point where it should theoretically work, all of the details about ui::EventTargeter don't work correctly. There is an empty subclass ViewTargeter which will need to be filled out before
Thankfully, this should only regress people who are using the wm_flow demo, and this should unblock the android guys who don't have aura support on their platform.
BUG=431047
R=eseidel@chromium.org, sky@chromium.org
Review URL: https://codereview.chromium.org/698543005
This CL teaches test_perf how to walk the benchmarks directory to find the
benchmarks and how to actually run them. We're still not uploading results to
the server yet, however.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/710133002
This teaches the SkyDebugger prompt how to tell
the sky debugger (server) to shut down instead of
just calling exit(0).
This also teaches the WindowManagerApp (server) how
to tear down all of its connections itself instead
of depending on the pipes to do so (which would
crash when youd delete the WindowManagerApp as the
pipes could outlive it with WindowManagerImpl objects
containing raw pointers back to the WindowManagerApp).
Shutdown is not yet clean. It errors out trying to
talk to the X11 server, but it's closer to clean
than it was prior to this change. I may add back
and exit(0) to side-step shutdown until it can be
made fully clean.
R=jamesr@chromium.org, sky@chromium.org
BUG=430291, 430242
Review URL: https://codereview.chromium.org/695183003
This script isn't quite working, but it's valuable to checkpoint the work at
this point. I'm coordinating with the chromeperf.appspot.com folks to get the
server to accept the data.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/702603005
Previously we were using string concatenation to build up the directory
listing. Now we fetch some JSON from the server and use data binding to inflate
a directory listing.
This exmaple doesn't work yet because we're missing support for
template@repeat, but hopefully we'll get that soon.
Also, added support for setRequestHeader to XMLHttpRequest.
R=rafaelw@chromium.org
Review URL: https://codereview.chromium.org/688413005
Now the SkyDebugger implements NavigatorHost and actually navigates the
mojo::View. This CL pulled a big refactor of sky/tools/debugger to separate out
MojoMain from debugger.cc.
R=eseidel@chromium.org
Review URL: https://codereview.chromium.org/690363002
So, instead of:
import "bar.mojom";
module foo {
...
}
You'll do:
module foo;
import "bar.mojom";
...
(Note that the grammar in the parser doesn't capture the ordering
constraint. It's hard to do so while keeping the grammar LR(1) AFAICT.
So instead we enforce the ordering "manually".)
Also delete mojo/spy/PRESUBMIT.py since it's broken. We'll delete
mojo/spy separately.
R=jamesr@chromium.org
Review URL: https://codereview.chromium.org/687793004
Turns out that the /sky mapping to the gen
directory was interfering with our new pattern
of mapping the actual /sky directory.
So now we use a more restrictive mapping pattern
of only mapping /sky/services, similar for
/mojo/services and /mojo/public.
Fixed an exception of not finding _sky_server
when a custom sky_server was already running.
R=abarth@chromium.org
Review URL: https://codereview.chromium.org/692163002
Automatically shut-down the sky_server when quitting.
Always start the skyserver at the mojo-root
instead of the parent directory to the path.
Log when starting a sky_server outside the mojo root.
Re-use already running sky_servers when possible.
R=esprehn@chromium.org, abarth@chromium.org
Review URL: https://codereview.chromium.org/679333005
This makes it possible to change the viewport size
which mojo is using. The sky debugger uses this new
API to set the viewport to something closer to mobile
instead of 800x600 ancient desktop size. :)
Still the (800, 600) constant is prevelant throughout
mojo (just search for it) and mojo does not seem to
fully update when the viewport size changes, so we actually
only see 320x600 pixels instead of 320x640 in
sky debugger after this change.
Mostly this change is about validating that this is the
correct way to plumb this information.
R=sky@chromium.org
Review URL: https://codereview.chromium.org/692693002
This makes the output much more readable and you can actually see
stacktraces. Long-term, we need to stick this output into the result
of the previous test instead of having them be printed to stderr.
R=esprehn@chromium.org
Review URL: https://codereview.chromium.org/686723004
