This is a mechanical change which was performed by a script based on the
contents of direct_call.py, and then manually checked over to fix
various rewriting bugs caused by my glorified sed script. See the
previous part for more context on the change.
Differential Revision: https://phabricator.services.mozilla.com/D137227
These attributes replace the previous direct_call.py table which
specified how to locate the concrete implementation of a protocol and
whether it should use a virtual implementation or not.
They work by specifying the concrete type for an actor, and disabling
the automatic inclusion of the implementation's header file, which can
be included explicitly with an `include "";` statement. This allows
customizing both the name and include path of the concrete
implementation of an interface.
Differential Revision: https://phabricator.services.mozilla.com/D137226
This simplifies the logic around descriptors significantly, which is
especially useful considering how few places use the type. There is a
small change required on Windows to create the NamedPipe directly and
transfer around each end's handle, rather than connecting between
processes after the fact.
A named pipe has to be used, rather than an anonymous pipe, as
bidirectional communication is required.
Differential Revision: https://phabricator.services.mozilla.com/D130381
`base::KillProcess`, with the `wait` parameter set to true, does a
bounded blocking wait for the process to exit by polling and sleeping in
a loop, with ad-hoc parameters. The only user of that case is the Gecko
Media Plugin code, which doesn't actually need it as discussed in bug
(comments 4-6); also, currently it's blocking the IPC I/O thread in the
parent process, which is not good for browser responsiveness.
Accordingly, this patch deletes that code and removes the parameter.
Differential Revision: https://phabricator.services.mozilla.com/D136662
This simplifies the logic around descriptors significantly, which is
especially useful considering how few places use the type. There is a
small change required on Windows to create the NamedPipe directly and
transfer around each end's handle, rather than connecting between
processes after the fact.
A named pipe has to be used, rather than an anonymous pipe, as
bidirectional communication is required.
Differential Revision: https://phabricator.services.mozilla.com/D130381
This patch adds a library that contains an interposer function for
pthread_create(). The interposer will setup an alternate signal stack to
handle crashes - thus enabling us to catch stack overflows - and then call the
real pthread_create() function. Since the interposer needs to appear in the
linker's search order before libpthread we manually link it into firefox,
plugin-container and xpcshell's executables ASAP.
Differential Revision: https://phabricator.services.mozilla.com/D132736
All uses of the intr message type have been removed from the tree, and the only
remaining uses are in IPDL tests, which currently do not test the IPDL runtime.
This test fully removes support for intr messages from the MessageChannel
interface.
Differential Revision: https://phabricator.services.mozilla.com/D136500
This is no longer necessary as the Quantum DOM project is no longer
happening, and removing support simplifies various components inside of
IPDL.
As some code used the support to get a `nsISerialEventTarget` for an
actor's worker thread, that method was replaced with a method which
instead pulls the nsISerialEventTarget from the MessageChannel and
should work on all actors.
Differential Revision: https://phabricator.services.mozilla.com/D135411
This can happen when some code is probing the UtilityProcessManager but
does not actually launch a process, as exposed by the pending patch to
enable FOG.
Differential Revision: https://phabricator.services.mozilla.com/D135941
This is possible if the socket process bridge ends up being restarted,
so needs to be handled as an edge-case. The approach taken here is to
tightly couple the task queue, pid, and BackgroundStarterChild instances
so that they can be swapped out at runtime whenever InitStarter is
called.
There is no test in the current patch as it is apparently still
unsupported to restart the socket process after a crash.
Differential Revision: https://phabricator.services.mozilla.com/D135746
