summaryrefslogtreecommitdiff
path: root/vcl/inc/osx
AgeCommit message (Collapse)Author
2019-09-23loplugin:fragiledestructor (macOS)Stephan Bergmann
Change-Id: I85542ed9e631ad8589d3bc3469d171ab1d5cb4f9 Reviewed-on: https://gerrit.libreoffice.org/79396 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2019-07-31Fix typosAndrea Gelmini
Change-Id: I2b9361433df39f66ec8328262a3128cf7294b6e1 Reviewed-on: https://gerrit.libreoffice.org/76720 Tested-by: Jenkins Reviewed-by: Julien Nabet <serval2412@yahoo.fr>
2019-06-20Demote from std::unique_lock to std::scoped_lock where applicableStephan Bergmann
Change-Id: I53a019f05978bab62ad0da3d0eb08f37f8ec1e18 Reviewed-on: https://gerrit.libreoffice.org/74414 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2019-03-17loplugin:typedefparam (macOS)Stephan Bergmann
Change-Id: I50cacafa8bbd90fef15603f0bde3f653f78393ea Reviewed-on: https://gerrit.libreoffice.org/69305 Tested-by: Jenkins Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2019-02-27convert more Start timer methods from sal_uLong->sal_uInt64Noel Grandin
Change-Id: I1ed43bab00a5bc456032410ccf32b3fd64cc970c Reviewed-on: https://gerrit.libreoffice.org/68419 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2019-02-26sal_uIntPtr->sal_uInt32 inSalClipRegion::BeginSetClipRegionNoel Grandin
sal_uInt32 seems reasonable given that this is the number of rectanges in an image. And then convert all of the other BeginSetClipRegion methods to use sal_uInt32 too. Change-Id: I85a712ec823662ac30f3859051e2b974fb99775e Reviewed-on: https://gerrit.libreoffice.org/68343 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2019-01-11OSX Re-Introduce NSApplicationMain usageJan-Marek Glogowski
This restores the nested NSApplicationMain and default run loop usage. Without it the Java AWT integration will start its own event loop, effectively blocking any non-system event processing. Reproducible via "Tools - Macros - Organize Macros - BeanShell... - LibreOffice Macros - HelloWorld - helloworld.bsh - Edit". The blocking can be prevented by overriding NSApplication::run and running our own event loop using Application::Execute. But this still doesn't show the Java AWT editor window and I couldn't find any information how to fix this. Since OSX now is a VCL plugin, this can't restore the old hook mechanism, but instead adds a new function to SalInstance. SalInstance initialization happens at InitVCL() start just a little bit later in the call stack. Somehow NSApplicationMain manages to run the Java VM in an extra thread, so it doesn't block the main loop. Probably this could also be handled by LO starting the JVM as a thread. Further information for an implementation eventually can be found in the "Technical Note TN2147" "JNI Development on Mac OS X." Change-Id: I04a0c2bf7949571f1b678ada9ab3592e0fe30c1f Regression-from: 925e2edb6f3f8fffcff9eddb31ed18bc77e2a690 Reviewed-on: https://gerrit.libreoffice.org/65836 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2019-01-07tdf#42949 Fix IWYU warnings in include/vcl/[v-x]*Gabor Kelemen
Found with bin/find-unneeded-includes Only removal proposals are dealt with here. Change-Id: I98f49765c6b74808dcbd692e0f375dd2848fcfd4 Reviewed-on: https://gerrit.libreoffice.org/65614 Tested-by: Jenkins Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
2018-12-11OSX implement bitmap-only rendering modeJan-Marek Glogowski
Eventually OSX should also depend on SVP and SVP should just export the platform agnostic code, so we can get rid of all the IOS and ANDROID ifdefs. Since the SVP mode now also uses a default DPI value of 96, we can get rid of those "broken" adjustments to the tests in commit a4ab0c08e747 ("Row height tolerance level increase for unittest") and commit 1e55a47e89a9 ("tdf#62268: allow row height recalculation on document load"). The DPI handling on OSX seems really broken. Additionally DPI should probably a float value, as rounding results in various of-by-one errors in tests. Change-Id: I47881683a9c3d3e1b4f8e7aba1d03842da4d6652 Reviewed-on: https://gerrit.libreoffice.org/64736 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2018-11-27tdf#114839: Drop ancient AquaBlinker crackTor Lillqvist
Most probably that was related to "pulsating" buttons that haven't been used since 10.9. Whether the code even worked (made buttons pulsate) on 10.9 anyway is unclear. Much of the code in this file uses various HITheme (Carbon?) API that as far as I know is long since deprecated and undocumented. It's weird that it still compiles, and that it doesn't get caught by the private API use checking when submitting to the the Mac App Store, in the LibreOffice Vanilla case. This change reduces the busy looping in Dialog::Execute() even when nothing happens (no user input, no mouse movement) when displaying a dialog. Change-Id: I9c62db608f637aa5d8493176feb2873f6426fd00
2018-11-01loplugin:useuniqueptr in SalFrame::PostEventNoel Grandin
Change-Id: Ib066b1d6df90f330f2f93ec639bd7bc59a08c024 Reviewed-on: https://gerrit.libreoffice.org/62507 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-10-31loplugin:useuniqueptr in DeletePrinterQueueInfoNoel Grandin
Change-Id: Ia124a4af642e449dc05f5bae2d5ca766bd67bd68 Reviewed-on: https://gerrit.libreoffice.org/62388 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-10-02Unify sal plugin loadersJan-Marek Glogowski
Change-Id: Ic099761eaff80349e985ccf62e3f4aa6b2e98022 Reviewed-on: https://gerrit.libreoffice.org/61103 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2018-09-19std::shared_ptr in AquaSalInfoPrinter can be std::unique_ptrNoel Grandin
also simplify the the call to reset Change-Id: Ie29b5ef754a8caf905e2dddbb56d8f1ffc7a43a0 Reviewed-on: https://gerrit.libreoffice.org/60698 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-09-18Drop NSApplicationMain usageJan-Marek Glogowski
Allows to drop all the special SVMain handling introduced for MacOSX. This way LO can also be build via SSH, because gengal won't abort / fail because of the missing window system. Currently this just implements the global menu. I don't know what else is missing. Eventually we want to reimplement the Cocoa debug option forwarding (NSAccessibilityDebugLogLevel). Change-Id: I359c46fd03e2436a8a37fd211e59e4e305f8aba0 Reviewed-on: https://gerrit.libreoffice.org/60571 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2018-09-18Move yield mutex handling into SalInstanceJan-Marek Glogowski
After the refectoring in commit 4c93de2c921b ("merge GenericSolarMutex and SolarMutex"), there is no more need to prevent instantiation of comphelper::SolarMutex objects. Since every VCL backend implements the yield mutex management in the same way, we can move the general implementation into the SalInstance. While at it use std::unique_ptr for the yield mutex on Mac and Windows platforms. Change-Id: Ibe0610bd92b4623152ee14e7a35b52465d403720 Reviewed-on: https://gerrit.libreoffice.org/60570 Tested-by: Jenkins Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2018-06-21merge GenericSolarMutex and SolarMutexNoel Grandin
Since nothing else is implementing the SolarMutex abstract class. Change-Id: I2a41254af3e9c7534033cdd0bece9dd8e0258b9d Reviewed-on: https://gerrit.libreoffice.org/56153 Tested-by: Jenkins Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-06-14return SalBitmap using std::shared_ptrNoel Grandin
since we hold it like that in Bitmap anyway Change-Id: I6264dfaaae6210cb008df5db8a421fc80c508f5b Reviewed-on: https://gerrit.libreoffice.org/55458 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-06-09hold and return SalMenu by std::unique_ptrNoel Grandin
and drop DestroyMenu, all the implementations just delete the object Change-Id: I673997ea1a9c5216b100cafdc70a8f697732769b Reviewed-on: https://gerrit.libreoffice.org/55501 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-06-09return and hold SalMenuItem by std::unique_ptrNoel Grandin
and drop DestroyMenuItem, all the implementations were just deleting the pointer Change-Id: I058817b963988b1e399279f61f45716a2cdba023 Reviewed-on: https://gerrit.libreoffice.org/55500 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-06-09hold and return SalPrinter with std::unique_ptrNoel Grandin
and remove DestroyPrinter, doesn't not anything beyond delete'ing the object Change-Id: I25e14b962e65a0e131fae3ff5771c82920a4e375 Reviewed-on: https://gerrit.libreoffice.org/55498 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-06-09return SalSession with std::unique_ptrNoel Grandin
Change-Id: I85527407c0cd2d3b94565fd7e24b0616948c9b0f Reviewed-on: https://gerrit.libreoffice.org/55497 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-05-30tdf#117872: Revert "tdf#115284: Unify LibreOffice and system full-screen..."Tor Lillqvist
Instead, never participate in the macOS system full-screen mode. There is just too much complexity involved, and the way LibreOffice works really isn't prepared for the concept of windows having the option from a system point of view to being full-screenable or not. This means that the green bubble in window title bars changes from being a (system) full-screen toggle to being a maximize/restore toggle. Sure, the "maximize" concept also probably can be confused with LibreOffice's own full-screen concept. For instance, the Start Centre window is not expecting to be made full-screen. Still, when you from the Start Centre open a Writer document, it is the *same* window that is re-used as the Writer window, and then suddenly should be prepared to handle going full-screen. Also, it is up to each separate kind of document window whether it can be made full-screen (from the LibreOffice point of view) or not. Writer windows can, but Impress windows can't, for example. The View>Full Screen menu entry is added separately each case. Maybe I will come back to this mess later, or not. Anybody else is welcome to have a go, too, of course. This reverts commit 4b42fd7e9516fbbd8a92d97680524f32dd260fb2. Change-Id: I6983481cbd30c0e5190c450483b1246006c80632 Reviewed-on: https://gerrit.libreoffice.org/55049 Reviewed-by: Tor Lillqvist <tml@collabora.com> Tested-by: Tor Lillqvist <tml@collabora.com>
2018-05-24Bin ancient AQUA_NATIVE_MENUS=FALSE crackTor Lillqvist
Using the system ("native") menubar works fine since a long time surely. No reason to keep a fallback possibility. Change-Id: I0d9ed86c28b0d832c8123b18980740dbf895ec1c Reviewed-on: https://gerrit.libreoffice.org/54775 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Tor Lillqvist <tml@collabora.com>
2018-05-24tdf#115284: Unify LibreOffice and system full-screen concepts on macOSTor Lillqvist
Also tdf#76476, and probably more. Make it so that when a window is in full-screen mode from LibreOffice's point of view, it is also full-screen from the system's point of view, and vice versa. All three ways to enter and leave full-screen mode can now be used with the same end result: The Ctrl-Cmd-F shortcut, the "View > Full Screen" menu entry, and the green bubble on the title bar. Don't disable/deactivate/etc menus while in full-screen mode. The menu auto-hides so there is no harm in having it function normally. Don't display the floating toolbar with a single "Full Screen" button in it as the way to leave full-screen mode. Instead, the same three ways that can be used to enter full-screen mode work to leave it, too. Sadly I could not figure out a way to set a window properly to full-screen at the point where a document window is created and set to be the same size as that kind of document window was the previous time it was open in LibreOffice. Thus don't save state for full-screen windows as we can't properly restore them. At least not for macOS. It is not good to just restore them as non-full-screened but still at the size they had when full-screen. One irritating glitch remains, and I was unable to fix that properly: I now prevent closing the document window that is in full-screen mode. Otherwise, if it is closed, the full-screen mode remains even if no window is open there; the desktop is completely black. Moving the cursor to the top edge, the LibreOffice menu is there, though. I tried to fix that but with no fully satisfying result. (Some attempts even lead to crashes, so just disabling closing is better than crashing at least.) Change-Id: Id909077ef9de9f19d48c8b9ad10d748a65b2417f Reviewed-on: https://gerrit.libreoffice.org/54760 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Tor Lillqvist <tml@collabora.com>
2018-05-22tdf#77444: Call TriggerUserEventProcessing() in a few key placesTor Lillqvist
Causes the updated aAppEventList to be handled right away, and not only when the next ten-second timed fires. Change-Id: Idbcc67f13c8ea69cc08b6c5f832d4461579d5dc6 Reviewed-on: https://gerrit.libreoffice.org/54643 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Tor Lillqvist <tml@collabora.com>
2018-04-25loplugin:useuniqueptr in VirtualDeviceNoel Grandin
Change-Id: I7c70f9c590dd63cca1dce16ce184fc7e14922de2 Reviewed-on: https://gerrit.libreoffice.org/53353 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-04-23weld SfxPrintOptionsDialogCaolán McNamara
and SwMMResultPrintDialog Change-Id: Icded6a26a3a151293bea0c9173334cf634283e89 Reviewed-on: https://gerrit.libreoffice.org/53299 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Caolán McNamara <caolanm@redhat.com> Tested-by: Caolán McNamara <caolanm@redhat.com>
2018-04-07Revert "long->sal_Int32 in tools/gen.hxx"Stephan Bergmann
This reverts commit 8bc951daf79decbd8a599a409c6d33c5456710e0. As discussed at <https://lists.freedesktop.org/archives/libreoffice/2018-April/079955.html> "long->sal_Int32 in tools/gen.hxx", that commit caused lots of problems with signed integer overflow, and the original plan was to redo it to consistently use sal_Int64 instead of sal_Int32. <https://gerrit.libreoffice.org/#/c/52471/> "sal_Int32->sal_Int64 in tools/gen.hxx" tried that. However, it failed miserably on Windows, causing odd failures like not writing out Pictures/*.svm streams out into .odp during CppunitTest_sd_export_ooxml2. So the next best approach is to just revert the original commit, at least for now. Includes revert of follow-up 8c50aff2175e85c54957d98ce32af40a3a87e168 "Fix Library_vclplug_qt5". Change-Id: Ia8bf34272d1ed38aac00e5d07a9d13fb03f439ae Reviewed-on: https://gerrit.libreoffice.org/52532 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2018-04-03long->sal_Int32 in tools/gen.hxxNoel Grandin
which triggered a lot of changes in sw/ Change-Id: Ia2aa22ea3f76463a85ea077a411246fcfed00bf6 Reviewed-on: https://gerrit.libreoffice.org/48806 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-03-02drop old tools/gen methods in vclNoel Grandin
Change-Id: I863ce5ae46ae90f06780261fa17b087a7153c807 Reviewed-on: https://gerrit.libreoffice.org/50445 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2018-02-23tdf#103571: Avoid spurious heavy SalEvent::DisplayChanged callbacksTor Lillqvist
It seems that on some Macs that the NSApplicationDidChangeScreenParametersNotification is sent for unknown reasons quite often. I can reproduce the problem by changing the Dock size in System Preferences while LibreOffice is running, but others seem to get it without resorting to such trickery. The code used to invoke the SalEvent::DisplayChanged callback in all cases, which can be extremely heavy, as it involves re-measuring text layouts all over the place in all open document windows. Avoid that if the geometry in fact has not changed. Sure, there still is the problem that LibreOffice can become unresponsive for several seconds when the display geometry *does* change, like when you attach or detach a display. Change-Id: I659881e5e392bd599f6be190835e32a77d9f4725 Reviewed-on: https://gerrit.libreoffice.org/50249 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Tor Lillqvist <tml@collabora.com>
2018-01-20More loplugin:cstylecast on macOSStephan Bergmann
Automatic rewrite (of loplugin:cstylecast and loplugin:unnecessaryparen) after cab0427cadddb3aaf1349c66f2fa13a4234ba4b2 "Enable loplugin:cstylecast for some more cases" and a409d32e7f6fc09e041079d6dbc3c927497adfed "More loplugin:cstylecast" Change-Id: Iff4877e8a42804c952c48c13332caf0a83c92870 Reviewed-on: https://gerrit.libreoffice.org/48216 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2017-12-16sal_uLong->SalPrinterError in SalPrinterNoel Grandin
and define a default implementation in the base class that returns SalPrinterError::NONE, since only one of the subclasses wants to override and return something useful Change-Id: Id41ff90693527e8624cc2993b7481114af85fcee Reviewed-on: https://gerrit.libreoffice.org/46509 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2017-12-15loplugin:salcall (macOS)Stephan Bergmann
Change-Id: I297ac09358ce948acae9b73e8ed605964520c73b Reviewed-on: https://gerrit.libreoffice.org/46437 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2017-11-09Avoid races when using OSX_RUNINMAIN_MEMBERSStephan Bergmann
...so that e.g. main thread in SalYieldMutex::doAcquire could reset m_aInMainTHreadCondition and then block waiting on it only after another thread had set it. (Saw such a deadlock in some 'make check' CppunitTest.) Change-Id: I5c676956a2bec6bf8f94d7dbeee64f100db39bd3 Reviewed-on: https://gerrit.libreoffice.org/44501 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
2017-10-31ReleaseYieldMutex is always called with trueNoel Grandin
so drop param and rename to ReleaseYieldMutexAll Change-Id: Ic4fcee24d46405659e54363c87f21d88696b0ce1 Reviewed-on: https://gerrit.libreoffice.org/44057 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
2017-10-24loplugin:includeform: vcl (macOS)Stephan Bergmann
Change-Id: I4382a45dcffb32c7c001ee722ac1deccb2b01c2d
2017-10-20WIN message handling cleanup and refactoringJan-Marek Glogowski
Replace a lot of duplicated case code with macros. Some minor constifications of function parameters. I restrained from shorten the SAL_MSG_* via preprocessor concat, so a grep will still find the whole names. Change-Id: If1f2477fc8817b4ae7816e807154e35004bb4da9 Reviewed-on: https://gerrit.libreoffice.org/43531 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-10-20OSX implement direct timeout processingJan-Marek Glogowski
Basically works in the same way then in the Windows backend. Change-Id: If48e2b55f037563839a5e716830ce75ed78381e4 Reviewed-on: https://gerrit.libreoffice.org/43530 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-10-16tdf#112076 OSX post timeout after shouldCloseWindowJan-Marek Glogowski
It seems that something filters application defined events when closing a window... DoYield expects a timeout event, which never arrives, stopping all redraw and generating a busy loop. Change-Id: Ic83468cceac7b967acaecf18b321178fc7565b94 Reviewed-on: https://gerrit.libreoffice.org/43381 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-10-04Convert tick-based timer events to versioned onesJan-Marek Glogowski
Instead of storing the system ticks in the timer event message simply store a version. Moves the version handling code into a VersionedEvent class, inherited by WinSalTimer and AquaSalTimer. Change-Id: I5add85031d36b3424a26a9ef798294cbfb00b2e4 Reviewed-on: https://gerrit.libreoffice.org/42959 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-29OSX rename mpFirstInstance to mpInstanceJan-Marek Glogowski
In the same spirit as the Windows commit 7c52d86f7b05fe7e0178f6d98a12a531b88a32ff. Change-Id: Ic45803c0715723b6f57c9f6a0c731edd559aa92c Reviewed-on: https://gerrit.libreoffice.org/42932 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-29tdf#112605 OSX refactor SalAquaTimerJan-Marek Glogowski
Drops all the static calls and variables from SalAquaTimer and moves all timer handling logic inside the class. This also fixes the leak in timerElapsed. Change-Id: Ie7a1a8fffb70b8579ec6876eed10a275d2f06d27 Reviewed-on: https://gerrit.libreoffice.org/42913 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-28Unify SalUserEvent handlingJan-Marek Glogowski
Merges the various SalUserEvent structs and their handling into a single class. This includes a common SalFrame* hash map, as all backends use such a map to verify alive SalFrames. It also reverts the "FIXME: lousy workaround" for i#90083, which was part of commit d6f7c94e5c27ba02ff5c3229760c9808cc9b5bea. At least on my current OSX box application based window switching "just works" "out of the box", even without the code. Change-Id: I188b567e44fd79c162b2d9cabbd771d1f66c7dc4 Reviewed-on: https://gerrit.libreoffice.org/42845 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-27Don't wait-yield non-main threads in the main threadJan-Marek Glogowski
This prevents blocking the main thread by a yielding non-main thread. The current solution is to wait on a condition, which is set by the main thread on wakeup. Change-Id: I8d680bb51a36ce1e0d3d4713d47d8e2ef93d7297 Reviewed-on: https://gerrit.libreoffice.org/42808 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-26WIN run main thread redirects ignoring SolarMutexJan-Marek Glogowski
This way we can drop all the special nReleased handling. Instead we use the same mechanism as on Mac, where we keep the lock, but disable it for the main thread. As a security measure we assert on duplicate redirects, which should not happen. As a result we can't use SendMessage on the main thread itself, which would normally just call the WinProc directly. This could be accomplished by converting the redirect bool into a counter, which should be safe, as no other thread could acquire the SolarMutex, as we don't release it. Change-Id: Icd87b3da37a2489f3cad2bc80215bf93fc41d388 Reviewed-on: https://gerrit.libreoffice.org/42583 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-21WaE: -Wunused-variableTor Lillqvist
Change-Id: I6bcc39cd749f8cdbe550b94346c5cb8fd9f67097
2017-09-21tdf#99784 OSX run GUI stuff in the main threadJan-Marek Glogowski
The extension manager starts dialogs from its own thread. But the OSX backend currently doesn't defer these calls to the main thread. This implements the deference by running the called function in the main thread, using a code ^Block, and returning the result via a SalYieldMutex member. Change-Id: Id8977991e3eda91da27c23d8021e028d4f4cefe5 Reviewed-on: https://gerrit.libreoffice.org/42448 Tested-by: Jenkins <ci@libreoffice.org> Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
2017-09-19loplugin:unnecessaryoverrideStephan Bergmann
Change-Id: I37c338fd08df3dc102963ccffc65c1f0684653be
generated by cgit (git 2.34.1) at 2025-01-01 11:22:18 +0000