summaryrefslogtreecommitdiffstats
path: root/base/message_loop.cc
Commit message (Collapse)AuthorAgeFilesLines
...
* ThreadLocalStorage::Set should take care of validating its inputs. No need todarin@google.com2008-08-191-1/+0
| | | | | | | | | do that in the MessageLoop constructor. TBR=evanm git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1058 0039d316-1c4b-4281-b951-d872f2087c98
* Define MessagePumpDefault and use it to implement MessageLoop on non-Windowsdarin@google.com2008-08-191-2/+6
| | | | | | | | | | | | | | | | | platforms. This is actually just a first-step toward the real fix which is to use MessagePumpDefault on all platforms on non-UI and non-IO threads. This CL also fixes some GCC compilation errors. I renamed MessageLoopOwnable to TaskBase, which seems more appropriate since a MessageLoopOwnable has a next Task pointer and clearly is only meaningful in the context of Task. (I wonder why it is even a separate class, but that is another issue.) I had to make the next_task / set_next_task methods public since they are used by an inner class of MessageLoop. Perhaps those inner classes should be made into top-level classes, but that seemed like too much to change at this time. R=jar,mmentovai git-svn-id: svn://svn.chromium.org/chrome/trunk/src@1045 0039d316-1c4b-4281-b951-d872f2087c98
* Take 2 at the new MessageLoop implementation.darin@google.com2008-08-161-571/+120
| | | | | | | | R=jar git-svn-id: svn://svn.chromium.org/chrome/trunk/src@973 0039d316-1c4b-4281-b951-d872f2087c98
* rollback portions of r928 to test to see if it impacts perfdarin@google.com2008-08-151-133/+562
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@936 0039d316-1c4b-4281-b951-d872f2087c98
* Try only re-scheduling work when HandleWorkMessage is called.darin@google.com2008-08-151-0/+3
| | | | | | | TBR=jar git-svn-id: svn://svn.chromium.org/chrome/trunk/src@930 0039d316-1c4b-4281-b951-d872f2087c98
* Introduce MessagePump to represent the native message pump used to drive adarin@google.com2008-08-151-562/+130
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | MessageLoop. A MessageLoop now has a MessagePump. This will make it possible to port the MessagePump interface to other platforms as well as to use an IO completion port for our worker threads on Windows. Currently, there is only MessagePumpWin, which attempts to preserve the pre-existing behavior of the MessageLoop. API changes to MessageLoop: 1. MessageLoop::Quit means return from Run when the MessageLoop would otherwise wait for more work. 2. MessageLoop::Quit can no longer be called outside the context of an active Run call. So, things like this: MessageLoop::current()->Quit(); MessageLoop::current()->Run(); are now: MessageLoop::current()->RunAllPending(); 3. MessageLoop::Quit can no longer be called from other threads. This means that PostTask(..., new MessageLoop::QuitTask()) must be used explicitly to Quit across thread boundaries. 4. No protection is made to deal with nested MessageLoops involving watched objects or APCs. In fact, an assertion is added to flag such cases. This is a temporary measure until object watching and APC facilities are removed in favor of a MessagePump designed around an IO completion port. As part of this CL, I also changed the automation system to use an IPC::ChannelProxy instead of an IPC::Channel. This moves the automation IPC onto Chrome's IO thread where it belongs. I also fixed some abuses of RefCounted in the AutomationProvider class. It was deleting itself in some cases! This led to having to fix the ownership model for AutomationProvider, which explains the changes to AutomationProviderList and so on. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@928 0039d316-1c4b-4281-b951-d872f2087c98
* git-svn-id: svn://svn.chromium.org/chrome/trunk/src@760 ↵darin@google.com2008-08-121-15/+63
| | | | 0039d316-1c4b-4281-b951-d872f2087c98
* git-svn-id: svn://svn.chromium.org/chrome/trunk/src@739 ↵darin@google.com2008-08-121-63/+15
| | | | 0039d316-1c4b-4281-b951-d872f2087c98
* Make timer.cc portable by factoring its Windows bits into MessageLoop.darin@google.com2008-08-121-15/+63
| | | | | | Please note that the goal of this CL is merely to move the Windowisms out of timer.cc and into message_loop.cc. Next up will be to refactor message_loop.cc so that the Windowisms are further isolated. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@734 0039d316-1c4b-4281-b951-d872f2087c98
* Fix bug 1317458: Busy wait in IPC::SyncChannel::Send in some casesmpcomplete@google.com2008-08-091-6/+7
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@610 0039d316-1c4b-4281-b951-d872f2087c98
* Apparently, our WndProc can get called for other HWNDs :(darin@google.com2008-08-061-15/+17
| | | | | | TBR=brettw git-svn-id: svn://svn.chromium.org/chrome/trunk/src@461 0039d316-1c4b-4281-b951-d872f2087c98
* just some hopefully non-contentious stuff to get out of the way before doing ↵darin@google.com2008-08-061-22/+12
| | | | | | the real MessageLoop changes. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@459 0039d316-1c4b-4281-b951-d872f2087c98
* ObjectWatcher needs to know when the current thread's MessageLoop is being ↵darin@google.com2008-08-051-0/+18
| | | | | | destroyed. This might also be generically useful, so I added a new API on ML to observe when the ML is being destroyed. The notification is sent to observers just prior to ML::current() being modified to return NULL. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@407 0039d316-1c4b-4281-b951-d872f2087c98
* Rollback message loop testjar@google.com2008-08-031-13/+11
| | | | | | TBR git-svn-id: svn://svn.chromium.org/chrome/trunk/src@299 0039d316-1c4b-4281-b951-d872f2087c98
* This is a test, and will rollback asap after a build/test cycle starts.jar@google.com2008-08-031-11/+13
| | | | | | | | | | I use a counter of MessageLoop::Quit() calls so that I can re-post "extra" calls when MessageLoop::Run() terminates. This assures that no quits are lost... but I'm not sure if some code abused the Quit() calls, and already posts many-too-many quits :-/. I want to see if the tree tests go red, and how the distributed reliability test responds. The test should already pass the basic unit_test and base_unittest, but I'm not sure what the rest of the UI tests etc will do. If I'm lucky, they'll be "less" flakey. ...but they may just go red all over :-( TBR git-svn-id: svn://svn.chromium.org/chrome/trunk/src@298 0039d316-1c4b-4281-b951-d872f2087c98
* Call ResetBirthTime before dispatching the user's Task. Also, remove ↵darin@google.com2008-08-021-9/+0
| | | | | | | | PostSignaledTask since it won't be used. TBR=jar git-svn-id: svn://svn.chromium.org/chrome/trunk/src@289 0039d316-1c4b-4281-b951-d872f2087c98
* A new (private) interface is supplied that allows the object watcher (coding injar@google.com2008-08-011-0/+10
| | | | | | | | | | | | | | | | | | | | | proress by Darin) to Post the task (when the object is signaled) into a message loop. I also cleaned up the time-of-birth for tasks that sleep for a while before running (such as those held by the timer, or by passed to this new PostSignaledTask() interface. r=darin M base/tracked.cc M base/message_loop.h M base/message_loop.cc M base/tracked.h git-svn-id: svn://svn.chromium.org/chrome/trunk/src@262 0039d316-1c4b-4281-b951-d872f2087c98
* Support RanAllPending() rather than RunOnce(), and integrated into ↵jar@google.com2008-07-311-11/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | ipc_sync_channel. RunOnce() semantics were not sufficient to support the effort to pump messages in ipc_sync_channel, as ipc_sync_channel is not able to detect pending Tasks. This change list switches to RunAllPending() tasks, rather than just Running the items that avail themselves in one run of the loop. This is a very small mod of the existing code. I'm still trying to stay focused on the bug at hand, and minimize chances of unrelated regressions (we only modify IPC channel semantics). The slight semantic change (from original attempts to call Quit() externally and then Run()) is that we now terminate the message loop after also servicing the high-resolution timer tasks. bug=1291034 r=darin M base/message_loop.h M base/message_loop.cc M chrome/common/ipc_sync_channel.cc git-svn-id: svn://svn.chromium.org/chrome/trunk/src@176 0039d316-1c4b-4281-b951-d872f2087c98
* (Re-landing of) Support RunOnce() in message loop.jar@google.com2008-07-301-36/+39
| | | | | | | | | | | | | | | | | | | | | | | | | | | Deeply nested calls to MessageLoop::Run() were made, and a multitude of Quit() messages were handled at the most nested level. This left outer invocations hung, waiting for "their" kQuitMsg to arrive (but the the nested loop had already discarded them in its processing of pending messages before exit :-/ ). We now use a more controlled run of the message loop, which does not rely on kQuitMsg. This re-landing doesn't have the anti-hang assertion, which was breaking a lot of tests. bug=1291034 r=darin,mpcomplete M base/message_loop.h M base/message_loop.cc M chrome/common/ipc_sync_channel.cc git-svn-id: svn://svn.chromium.org/chrome/trunk/src@130 0039d316-1c4b-4281-b951-d872f2087c98
* Rollback 109jar@google.com2008-07-301-36/+36
| | | | | | | | | | M base/message_loop.h M base/message_loop.cc M chrome/common/ipc_sync_channel.cc git-svn-id: svn://svn.chromium.org/chrome/trunk/src@110 0039d316-1c4b-4281-b951-d872f2087c98
* M base/message_loop.hjar@google.com2008-07-301-36/+36
| | | | | | | | M base/message_loop.cc M chrome/common/ipc_sync_channel.cc git-svn-id: svn://svn.chromium.org/chrome/trunk/src@109 0039d316-1c4b-4281-b951-d872f2087c98
* Add base to the repository.initial.commit2008-07-261-0/+969
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@8 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-03-03 04:22:01 +0000