This effectively fixes the LLDB dumper for std::deque and std::stack by
avoiding a code path in which LLDB returns wrong data (the type and
offset for the _Deque_iterator::_M_first member - surprisingly the
other three members there get reported correctly).
Change-Id: Iae1ba8b81987442ed2d486b383a145a77ce49cd7
Reviewed-by: David Schulz <david.schulz@qt.io>
Reviewed-by: hjk <hjk@qt.io>
The Math. function like Math.max() are quite useful
to define more complex layouts. Therefore we allow them.
Change-Id: Ia95dcbcc1b8e96c117650dc8643da4a9de0ecdba
Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
Reviewed-by: Marco Benelli <marco.benelli@qt.io>
The disk cache time stamp is only accurate up to a second
and there might be .qmlc files from the actual application.
Change-Id: I6ec9cfdda8eb7aba5a630ac577d2cef999a8132c
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
This still does not solve the QFile dumper test fail with LLDB/Linux,
but its one obstactle less. The remaining one is LLDB not reporting
the QFile base object size at all.
Change-Id: I3144cf9469c4456d21b8c220c9df1cd890348491
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
More compact display and arbitrary inner types.
Change-Id: I68bc1f5cc88f51e993e6e8a91556d9d745a2574a
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Both LLDB 3.8/Linux and 360.x/Mac cannot display the contents
of a virtual base class when accessed to a 'secondary' path.
The problem is with LLDB itself, so don't expect Creator tests
to pass in those cases.
Change-Id: I25b005d66fa9f64766a4cc0aaaa2c865b6df8c5f
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
There have been cases observed where 'p QArrayData::shared_null' finds
valid symbols that are not found using gdb.lookup_symbols. The cause
for that is unknown.
Apply an expensive workaround by checking for (the equivalent of)
a working 'p QArrayData::shared_null' but execute it only when
a libQt5Core was found. This keeps the overhead for non-Qt setups
at a bearable (unsuccessful) iteration over known shared object
names.
Change-Id: Id398673b938d3c3a72c24317abdbefbe793e54df
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Include <memory> for std::shared_ptr.
Change-Id: I549eef8e8dc42b58b07e3528d55193ed8c54dd7e
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
(cherry picked from commit 6614b7a276)
It has a function directly doing that, use it instead of
guessing (badly, in some cases)
Change-Id: I1ce317c28d60f820f8e868e3c0c5a665f8de6cb7
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
... to allow more fine-grained decision making for kit choosers.
Change-Id: I9a4e97f7b4f4b074e1c764f6b134c15e2896dc79
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
We need DYLD_IMAGE_SUFFIX=_debug for the captures()
cache warming call to succeed even if the actual access
is by offset only.
Change-Id: I24a90b4c4187459904f14a664b992a5bba9a20eb
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Needs to be implemented in Qt Quick itself nowadays.
Task-number: QTCREATORBUG-14194
Change-Id: I0456dc5d86c03672d8b6b8f9c45799851db1b059
Reviewed-by: Thomas Hartmann <Thomas.Hartmann@theqtcompany.com>
For example when toolbars showed a menu for actions that no longer fit
(e.g. open diff viewer, reduce window size until the synchronize and
switch view buttons vanish).
Use the application flag to disable icons in menus on macOS globally
instead of doing that individually on the actions.
Change-Id: I3b682f48ee0e866f0f0d5bb0834690937f427d7f
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Adding API for that to the parallel map/reduce functions
Change-Id: Ic7987d899c124dcb04c1ca110b09fc97576873fe
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
We already ship with a cherry-picked change from version 3.9.
Change-Id: Iad09ac8db470506c5de0321f3a2346c549b11b31
Reviewed-by: Marco Bubke <marco.bubke@qt.io>
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Emit the signal only if we are not canceled.
Task-number: QTCREATORBUG-17249
Change-Id: I5082a5fe7554eecdc7ec5d148150c1b169002f6d
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
The TODO plugin provides default values for color-coding backgrounds
based on the severity of each item. These colors are rather light, and
there is just one default with no theme awareness.
It is a bug to use a theme-aware text color on a fixed-color background
-- just switch to a dark theme such as flat-dark and observe the
unreadability.
This patch simply ensures that the default settings have reasonable
readability. These defaults are still not theme-aware, but at least they
are consistent.
Change-Id: Ibf35a241c70e0f1ea001b55f84fd2e1dd1a0a2ea
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Some code path did not resolve typedefs when looking up
struct members by name, making e.g. the BoostList dumper
fail for LLDB.
Change-Id: I7adf235cff3941574cab9f03d6f15ec5f3e2f0bd
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
-arch is not allowed in compiler flags as it's automatically handled by
the qbs.architecture property, and is an error in current versions of
Qbs. If the architecture was successfully detected, remove the flags.
Change-Id: I85cce7b7f4ef5a92f857ec624a912861bcb267f5
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
We might end up with undefined boost version otherwise.
Change-Id: Ibcdc70cbce8aad42947ce4760df0208d5ebe83f7
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
The plugin is in use since many years and can be considered stable these
days.
Change-Id: Ia41b24849abd16e15313ab97f82d277673b3904b
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Only MinGW gcc defines WIN32, MSVC compiler does not. It's
also defined by qmake (msvc-desktop.conf), but not by qbs ...
Let's just use _WIN32, that's defined everywhere.
Change-Id: I8342a70498be54a965dcf7fae63eaf406aaa3c04
Reviewed-by: Tobias Hunger <tobias.hunger@qt.io>
Reviewed-by: Tim Jenssen <tim.jenssen@qt.io>
Avoid unintentional re-enabling of the code parser.
Handling of the enabled state broke several times before,
therefore separate it from other states of the parser to
avoid breaking it again when not taking enough care while
refactoring or adding features related to states.
Change-Id: If1eb0dd649225f10bfc3bf06f09851649da75983
Reviewed-by: David Schulz <david.schulz@qt.io>
"Save as" sent it before even the new name was asked.
Change-Id: I8f791689267e57e63d51990af96373f653e54943
Reviewed-by: David Schulz <david.schulz@qt.io>
Keep strings in lambdas instead of WatchItem pointers that
might get deleted while the menu is open.
Change-Id: Iec65924696da9754ffbbbb833ae0db990598c8e3
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
* Run a debugger (of any kind)
* Close Qt Creator window
* When prompted to shutdown choose Yes
* An error message box appears with "Unexpected GDB exit"
This is not a real solution, but a quick hack to make the error message go.
2 soft assertions are also hit on this case:
SOFT ASSERT: "state() == EngineShutdownRequested" in file debuggerengine.cpp, line 1130
Debugger::Internal::GdbTermEngine(0x55ef651cea30, name = "GdbEngine") "InferiorShutdownRequested"
*** UNEXPECTED STATE TRANSITION: Debugger::Internal::GdbTermEngine(0x55ef651cea30, name = "GdbEngine")
"State changed from InferiorShutdownRequested(16) to EngineShutdownOk(21) [master]"
SOFT ASSERT: "state() == InferiorShutdownRequested" in file debuggerengine.cpp, line 1068
Debugger::Internal::GdbTermEngine(0x55ef651cea30, name = "GdbEngine") "DebuggerFinished"
UNEXPECTED STATE: 22 WANTED: 16 IN gdbengine.cpp:1841
They are *not* addressed in this patch.
Task-number: QTCREATORBUG-16770
Change-Id: I419f134e527a154dfbe1c85a3fd6629597a363f0
Reviewed-by: hjk <hjk@qt.io>
UI has changed again, so adapt the test to reflect this.
Change-Id: I5d6bcf2b48b8b625f6086f5a1101041a73d7e329
Reviewed-by: Robert Loehning <robert.loehning@qt.io>
