There is no need to explicitly check in with Launch Services or call RegisterApplication in the WebContent process, since this is handled when calling [NSApplication _accessibilityInitialize].
<rdar://problem/61210496>
Created attachment 395271 [details] Patch
Comment on attachment 395271 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=395271&action=review r=me assuming it builds and passes tests > Source/WebCore/PAL/pal/spi/cocoa/CoreServicesSPI.h:28 > +extern "C" void _CSCheckFixDisable(); I would have expected this earlier in the file: #if USE(APPLE_INTERNAL_SDK) #include <CarbonCore/RuntimePriv.h> #endif Unless there is some reason we can’t or shouldn’t do that? > Source/WebCore/PAL/pal/spi/cocoa/LaunchServicesSPI.h:108 > +typedef bool (^LSServerConnectionAllowedBlock) (CFDictionaryRef optionsRef); > +extern "C" void _LSSetApplicationLaunchServicesServerConnectionStatus(uint64_t flags, LSServerConnectionAllowedBlock block); > +extern "C" CFDictionaryRef _LSApplicationCheckIn(LSSessionID sessionID, CFDictionaryRef applicationInfo); Similarly, kind of expected, within the #if USE(APPLE_INTERNAL_SDK) section: #import <LaunchServices/LSApplicationServicesPriv.h> > Source/WebKit/WebProcess/cocoa/WebProcessCocoa.mm:542 > + // _CSCheckFixDisable() needs to be called before checking in with Launch Services. This comment leaves out critical context. It’s not obvious how this is the right place to do this. How is this place "before checking in with Launch Services"? The comment should say.
Created attachment 395278 [details] Patch
(In reply to Darin Adler from comment #3) > Comment on attachment 395271 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=395271&action=review > > r=me assuming it builds and passes tests > > > Source/WebCore/PAL/pal/spi/cocoa/CoreServicesSPI.h:28 > > +extern "C" void _CSCheckFixDisable(); > > I would have expected this earlier in the file: > > #if USE(APPLE_INTERNAL_SDK) > > #include <CarbonCore/RuntimePriv.h> > > #endif > > Unless there is some reason we can’t or shouldn’t do that? > Fixed. > > Source/WebCore/PAL/pal/spi/cocoa/LaunchServicesSPI.h:108 > > +typedef bool (^LSServerConnectionAllowedBlock) (CFDictionaryRef optionsRef); > > +extern "C" void _LSSetApplicationLaunchServicesServerConnectionStatus(uint64_t flags, LSServerConnectionAllowedBlock block); > > +extern "C" CFDictionaryRef _LSApplicationCheckIn(LSSessionID sessionID, CFDictionaryRef applicationInfo); > > Similarly, kind of expected, within the #if USE(APPLE_INTERNAL_SDK) section: > > #import <LaunchServices/LSApplicationServicesPriv.h> > Added. > > Source/WebKit/WebProcess/cocoa/WebProcessCocoa.mm:542 > > + // _CSCheckFixDisable() needs to be called before checking in with Launch Services. > > This comment leaves out critical context. It’s not obvious how this is the > right place to do this. How is this place "before checking in with Launch > Services"? The comment should say. Fixed. Thanks for reviewing!
Committed r259396: <https://trac.webkit.org/changeset/259396> All reviewed patches have been landed. Closing bug and clearing flags on attachment 395278 [details].