| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It does not make a difference in functionality, but after engineAdded()
the server won't wait on a mutex anymore. Before this change, if you
managed to send a message to the V4 debugger after the server had
called aboutToBeAdded(), but before it had stopped waiting, you could
produce a deadlock by scheduling an event for the GUI thread that was
never delivered.
This is a cherry-pick of 18c4295e25503ae637a715858de5c94a3d105a92
from 5.7 as apparently the problem also occurs in 5.6.
Change-Id: Ie2343e6da4bd0b8956d41ff8ebd4d7594616ebd1
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
|
|
|
|
|
|
|
|
|
| |
All setters now store tags, so no-one can play loosy-goosy with the
boxed values (and accidentally forget to "tag" a value, resulting in
random garbage).
Change-Id: Ia0b78aa038d3ff46d5292b14bd593de310da16a0
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
Some of the adapters immediately return dataReady() when reportData()
is invoked. This means that there is only one adapter in the start
times list then, which in turn causes all the data from that adapter
to be sent at once, without caring for the other adapters' timestamps.
Change-Id: Ic1e12fdcefb0a691067518fba100368f13c927f7
Task-number: QTBUG-53590
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
|
|
|
|
|
|
|
|
|
|
| |
qmake is clever enough to include the plugins without this.
Furthermore, the explicit imports duplicate the plugins in static
builds and they lead to cross linking from QtQml to QtQuick.
Task-number: QTBUG-50306
Change-Id: I822e000481f583f513b863f54f65eb8ff772c0c3
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
| |
Change-Id: I677ab9fce527b5220317d4d30f48d55b7f081e6a
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
this fixes static builds by ensuring that all dependencies are exported.
Task-number: QTBUG-51071
Change-Id: I5928f853a1d13b6a73533f9c6f6eae9da0e52a26
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
| |
We don't need to pass debuggers and engines around as the data
collector already has all the necessary information. Also, the
exception collect job is only used in the test case, so we don't
need to define it in the collector.
Change-Id: I3197dd5c2d99e95465aa787097c6f8bf8aee794e
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
To avoid interaction with the engine from the debugger thread we
move the value lookup functionality into the data collector, and
drop the RefHolder. Also, we define some more debugger jobs to
move the work the request handlers do into the GUI thread.
Task-number: QTBUG-50481
Change-Id: Ia73ebd01e715f13110ef010a766e9b0a1e202df9
Reviewed-by: Nils Jeisecke <jeisecke@saltation.de>
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: I408dcb81ba654c929f25ffff1427560b6e7962ad
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: Id3295937bdbd2372a77f273ce5c048807d43f51e
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Saves iterating the whole container and unneeded allocation of a list
Change-Id: Iae1f8e0cf3a17b163cf930c43a27b2ebb4315e5c
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
| |
qWarnings() usually are not translated.
Change-Id: I0375e0e21054d33349d3f56ee9ac6f580e366736
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This service provides information about the state of the
most recently used QmlEngine in a way that can be integrated
with "native" debuggers (GDB, LLDB, potentially CDB).
Unlike the existing QV4 debugger service, the data extraction
is triggered by direct calls from the native debuggers, i.e.
has the examined QmlEngine stopped at the time of inspection.
Since roundtrips through the native debuggers are expensive,
an additional goal is to minimize roundtrips.
Change-Id: Ic420129bbc15162197e35b52d92e882c027c9f6d
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
| |
This debug connector offers a few simple C functions a native debugger
can hook into to communicate with QML debug services.
It is the base of the 'native mixed' debugging mode in Qt Creator.
Change-Id: I32ce1f21be8e3522dccf1210d4de3b131b819bc5
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Otherwise different debug services could steal each other's debugers.
Change-Id: Ic0a50333d21c7d20a7124240ea598f8446400ae3
Reviewed-by: hjk <hjk@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
| |
... into a pure interface and a QV4::Debugging::V4Debugger implementation.
This is in preparation of a second implementation of this interface
to be used with 'native mixed' debugging.
Change-Id: I3078dcfe4bdee392a2d13ef43a55ca993e7b88d8
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Make the distinction between connectors and services a bit clearer.
Change-Id: I5a8de587beaaed4b4b2138d9d50c5b92fea38df9
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Use character literals where applicable.
Change-Id: I294fc4cb5cbbd23df9735ba2b398118f37cbe08a
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
| |
In some cases the connection may relay a message to the server directly
when calling one of the connect methods. This leads to a crash if the
server, due to not knowing the connection yet, cannot reply.
Change-Id: Iac2da08c621b70c2e5a1922efad0cd6ae4edc4e1
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
| |
Reduce usage of ScopedContext.
Change-Id: I84a6a7478065de3398fd0b21596ca1308e78ceb3
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
Now that the other method is gone, let's use
the shorter currentContext
Change-Id: I2a6fb3b77f83a1ffdf314ad29081e303d17030ed
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This saves one pointer per allocated execution context.
Now every execution context that is pushed, allocates two
Values on the js stack. One contains the context itself, the
other one the offset to the parent context.
Things are a bit tricky for with and catch scopes, as those
are called from the generated code, and can't open a Scope
anymore. In addition, all methods iterating over the js
stack frames need to work with ExecutionContext pointers,
not ScopedContext's.
Change-Id: I6f3013749d4e73d2fac37973b976ba6029686b82
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The compilation of .js resources was missed and the compile times were
underestimated because the dependency resolution and error checking
wasn't factored in. In particular we only profiled the done() step for
QQmlTypeData. However, JavaScript compilation also occurs in the
dataReceived() step of QQmlScriptBlob and QQmlTypeData will already
parse the program into an AST at that step.
Compile steps can be nested now, but considering the fact that
significant time may be spent before and after compiling dependencies
for a parent module, this seems to be the best way to model them.
Furthermore, in order to not needlessly convert QUrl to QString at
runtime, the compilation profiler saves the files now as QUrl.
Change-Id: I215a87787f9117c069ecd77b2d913cc0b0ff3c89
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Fix CLANG build error:
qqmldebugserver.cpp:115:10: error: private field 'm_block' is not used
[-Werror,-Wunused-private-field]
bool m_block;
Change-Id: I3be87c1ec0347b46e90e40c8769aeb507cf22c26
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This changes the type announced for functions to the actual JavaScript
type "function".
The type for null is also wrong: it should be "object". However, older
QtCreators cannot distinguish between null and {} if null gets the
correct type, unless you explicitly compare x === null in an expression
evaluator. For this reason the fake "null" type is kept for now.
Also, the value field of undefined is now set as QJsonValue::Undefined
which causes it to be omitted when sent over the wire. This is the
logical thing to do.
In addition we add type and value fields for all data members mentioned
in a response, not only the ones specifically asked for. The value
field is the actual value for any primitives (including strings), or the
number of properties for composite types: objects, arrays, functions. In
turn, the "ref" members are omitted for primitive types, so that we
don't have to hold references to them in the debug service anymore. Even
old QtCreators can deal with verbatim data members without "ref".
Task-number: QTBUG-47746
Task-number: QTBUG-47747
Change-Id: I773e6418c39cd9814aadb5fb5ef7e109f9a4e618
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
We need to collect the refs in the debugService's list in order for
them to show up on addRefs() and we need to generate proper error
responses if either the debugger is not stopped or the evaluation
throws an exception.
Task-number: QTBUG-47797
Task-number: QTBUG-47816
Change-Id: I98f17c1f3976859ee50b9bfac41091276ff60982
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The "values" of "special" refs are always JavaScript "undefined"s.
Thus, when trying to deduplicate an actual "undefined" we'd
accidentally hit the special refs. Avoid this by checking if a found
ref is special.
In an ideal world we'd get rid of the special refs altogether and
save QV4::FunctionObject in the values array, but that is not quite
trivial as long as the QV4::ExecutionEngine::stackTrace() doesn't
give us FunctionObjects.
Task-number: QTBUG-47788
Change-Id: Idf358c285f40930220fad2207c6ab5c9101573b2
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
| |
We don't want to load the debugger when profiling and vice versa. This
makes it easier to prevent unwanted services from getting loaded.
Task-number: QTBUG-47623
Change-Id: I28893b6218110274a6d30b27805d89dbb443add3
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
| |
We will need them to determine which services to load. Also, drop all
the complicated indirection. With all relevant code in one place we
don't need the actions anymore. Setting the server's blocking mode only
makes sense before the thread starts. Afterwards the result is
undefined, so we don't have to specially protect it.
Change-Id: I58135cd598c31d3a02ccd3f67b238da01a176934
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
The data collector and all the jobs it uses to interact with the engine
are only used from the debugger plugin. We can as well move them there.
Change-Id: Ia48251f08b48c7e1e607b8ae2a3d1de29f80f742
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
| |
The debugger is the only thing that actually needs it. Note that for
this to work we need to make QV4::Debugging::Debugger a QObject and
add some signals. The net effect is still a reduction in binary size
of about 1kb.
Change-Id: Ibecb8cfa140fc26fc13c8cbefb3d027ebdcd28a4
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch changes the variable collection to store QV4::Value values
into a JS array, which is retained by the collector. This prevents any
GC issues, and gives a nice mapping from handle (used in the debugging
protocol) to JS value. It also allows for easy "shallow" object
serialization: any lookup can start with the QV4::Value, and add any
values it encounters to the array.
Testing is changed to use this collector directly, thereby testing the
class that is actually used to generate protocol data.
Task-number: QTBUG-47061
Change-Id: Iec75c4f74c08495e2a8af0fedf304f76f8385fd7
Reviewed-by: Erik Verbruggen <erik.verbruggen@theqtcompany.com>
|
|
|
|
|
|
|
| |
It's only used by plugins now.
Change-Id: Ia73a2a22ba6bccbd85bbca6eda2fb9f5a8d6dd3b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: Icd4e6a6c57bc3ac65cb43d2329d236012b988678
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: I12627a07ceedea4aceafa6f0e630c0cab69d156d
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
| |
As we look up services by name we should also add and remove them by
name. As the thread doesn't run during adding and removing of services
we don't have to check the client plugins for the initial state.
It's also a good idea to eventually disconnect any signals that we
connect on addService().
Change-Id: I9acd17d2caafe15831f32b7b959dc2dea9cab08c
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: Ibae6602bd1725d98ddaa751bfc00391a20a1460e
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Nobody needs to access it anymore.
Change-Id: Ib782037ac0a8721997f6d36995d03f42c060602b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
We are not loading any secondary inspector plugins anymore, so the
logic to select them is unnecessary now.
Change-Id: Ic44c49e41c6bff4b19ce527df4657c6d73c0c82b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
This way we don't have to look up the implementation in order to send
messages.
Change-Id: I70cb122785875cf8b4ba6f7f2afd62cca77c8abb
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
| |
The inspector service doesn't do anything useful without the QtQuick2
plugin and vice versa. This way we can also use the
QQmlDebugPluginManager.
Change-Id: I78f154dcc9103ec9ec3d2eda216bfb293231583e
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
Now that they are self-contained plugins there is no reason for the
indirection anymore.
Change-Id: Ic2e2fe2075796c758057235e12981c8d40ce97c2
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
Also remove all the hacks that deal with static builds, QT_NO_LIBRARY
and friends. QQmlDebugPluginManager handles those cases.
Change-Id: I62f13b787292108fa25d09fabc775332394989be
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The packet protocol can be part of the server, now that the server is
not part of QtQml anymore. This enables us to remove some duplicated
code from the connections.
As an added benefit, with more control over the sending process,
QQmlDebugServer can now efficiently send single packets, without
creating a QList<QByteArray> first.
Change-Id: I13cc831e254c02b737e64816d6d3ab051d760995
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
| |
Plugins in qmltooling won't get built if that is set.
Change-Id: Ideced675064e74ccb78b9ac8c8fff2536150810c
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
| |
Change-Id: I8dc95f64c6df7303e8f580f191ee35da2284718b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
| |
This will allow us to remove the instance() methods and create the
services from factories in plugins. Also, it allows us to remove the
isDebugging member from QQmlEnginePrivate.
Change-Id: Id9d9820a910902ecfdb1e8175e215093ce3d0965
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
| |
We will access the services' functionality through those interfaces
once they live in their own plugins.
Change-Id: I0a0d7e73c07cb874b3b507cc4a9d304588c87bca
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
| |
Defining things outside the Qt namespace and then accessing Qt
internals is just asking for trouble. The redundant QtQuick2
namespace can be dropped, though.
Change-Id: Ifa83733a3f98704b7b51697dbca9486d0f504f65
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This results in much cleaner code than the previous implementation
using QMetaObject::invokeMethod().
We have to use read locks now for adding and removing engines, as we
should have done already before. If a condition is waiting on a write
lock you cannot acquire a read lock from another thread. So, if we kept
the write locks we wouldn't be able to receive messages while the engines
are waiting.
Change-Id: Icfe641601dec2f8d7181ae579146ed603d57a4c2
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
|