- add new logic to ContainerTypeProvider to use a JobGroup with
maximum 1 thread so jobs from this class are never run
at the same time
- change ContainerTypeProvider.getStatus() to wait until there are no
scheduled, active, or cancelled jobs in the JobGroup
- change ContainerLaunchConfigurationDelegate.finalLaunchCheck() to
create a fake launch target and call LaunchTargetManager.getStatus()
which will eventually call the new ContainerTypeProvider.getStatus()
- fix NLS warnings in ContainerLaunchConfigurationDelegate
Change-Id: Icad11c5814dd8ee7fac99a41d74329ce65907a38
Note - these bundles aren't actually published, so they are not
actually part of compare and replace checks. But if we are going
to increment version numbers, they should be incremented following
the policy as if they were published.
[1] https://wiki.eclipse.org/Version_Numbering
Change-Id: I01c815b0ae075d704b4bae9aef666cc13d98161c
Because of a bug in the platform implementation of equals/hashCode under
some circumstances, we can workaround the problem by returning
the same instance of CElementImageDescriptor when the objects are equal().
Change-Id: I6ff15cffa4b48080dcc6c9a023622edf4ae2d497
Using the mouse and not the focus listeners leads to lots
of weird, unintended consequences. For example:
- If mouse does not exit the control (because, e.g. it was not
in the control) then captureKeyEvents(false) is never issued
- If mouse does leave control, then control stops capturing
events. This means that although the cursor and focus is still
on the control, keyboard shortcuts no longer go to the control,
so Ctrl-C becomes copy again. This means, depending on where you
move the mouse changes the behaviour of the control.
This reverts commit 19351cbc2b.
Change-Id: I4f57c659e21d823df049b095159a34e5c110ef29
(cherry picked from commit 56d741a50c)
As part of the rebuild of the infra images to get new Java 11, we
also got updated compilers that generate some different bytecodes
for the native libraries. This commit brings things up to date.
Change-Id: I1039a5e19172092353dc4e1ea7de1381c031d280
When a user does an upgrade with Arduino installed and an upgrade is done
now that Arduino is removed from master they can get into a broken state
because Arduino plug-ins had not traditionally defined upper ranges.
This commit adds such limits so users will get a fail to upgrade error to
CDT 10 if they have Arduino installed.
Change-Id: I05e7f40e21dd1a2f511d3d0ced13dac49ec0bec1
CDT has two representations for a pointer type wrapped in a
qualifier type: it can be an IPointerType with some qualifiers
set on it directly, or an IPointerType wrapped in an
IQualifierType. (We prefer the first representation to avoid
creating too many wrappers, but sometimes the second one arises
during template instantiation.) This patch makes sure that two
such types can compare equal even if they use different
representations.
Change-Id: Ia8c7d227c74378aae74f04545b9a69103c14e74b
The position of namespaces didn't take into account because the
declaration doesn't exist yet when we use this kind of refactoring,
so the find method of MethodDefinitionInsertLocationFinder didn't
look for namespaces.
Change-Id: I839194879c41f86653c837ca83a306ea1840c1d0
(cherry picked from commit d36ed7cfd5)
When using -ffunction-section, each function is placed in it's own
section name. For example the main() function would be placed in
.text.main, so the pattern looking for undefined references needs to
allow the extra dot there.
Since C++ will use other symbols than what's included in \w, lets allow
everything except a + sign and whitespace.
Change-Id: I66d9ee7d8cbc0a405e60d7cdeb43c38b30107245
Signed-off-by: Torbjörn Svensson <azoff@svenskalinuxforeningen.se>
(cherry picked from commit 36de92cc77)
In commit f69a613d72, a check for null and
0-length array was introduced, but the checks is not reliable unless the
checked value is later used.
Change-Id: Ifc32b369514ee03eeea21281a7af80616531af87
Signed-off-by: Torbjörn Svensson <azoff@svenskalinuxforeningen.se>
This makes it possible for a user to subclass CBreakpointPropertyPage
and use their own adapter to decide the breakpoint label for
watchpoints.
Change-Id: I086179c6b5561b77410fd2a6f7d77bb6084d94d4
Signed-off-by: Anton Backstrom <anton.backstrom@gaisler.com>
Straightforward conversion. Still usages left that need deeper
investigation how to be done proper.
Some long time commented code removed as java formatter breaks trailing
whitespaces on save.
Change-Id: If74259bed5735b0d4cc98fc2cfa609c9c53c80c9
Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
Default ARCH = x86_64 as x86 is no longer built nor needed.
Change-Id: Id1d0dc6b0b29438cc200d614cb3add55c8eb40f1
Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
