Add MOZ_MUST_USE to prevent potential bugs.

Ongoing try build: https://treeherder.mozilla.org/#/jobs?repo=try&revision=29e70283d74e

Thank you for doing this! You have a few places where you've added MOZ_MUST_USE to an NS_IMETHOD function. I'm in the process of including MOZ_MUST_USE in the NS_IMETHOD macro itself, in bug 1292440, so those changes won't be necessary.

Try: https://treeherder.mozilla.org/#/jobs?repo=try&revision=a51020d5d193 (In reply to Nicholas Nethercote [:njn] from comment #2) > Thank you for doing this! > > You have a few places where you've added MOZ_MUST_USE to an NS_IMETHOD > function. I'm in the process of including MOZ_MUST_USE in the NS_IMETHOD > macro itself, in bug 1292440, so those changes won't be necessary. Thanks, I've updated the patch.

Comment on attachment 8779991 [details] Bug 1293212 - Add MOZ_MUST_USE to prevent potential bugs. https://reviewboard.mozilla.org/r/70822/#review68288 those fixed ::: uriloader/exthandler/ExternalHelperAppParent.h:53 (Diff revision 1) > NS_DECL_NSIMULTIPARTCHANNEL > NS_DECL_NSIRESUMABLECHANNEL > NS_DECL_NSISTREAMLISTENER > NS_DECL_NSIREQUESTOBSERVER > > - bool RecvOnStartRequest(const nsCString& entityID) override; > + MOZ_MUST_USE bool RecvOnStartRequest(const nsCString& entityID) override; Why the changes to Recv*. IPC machinery will kill the child process if false is returned, and do we really want to annotate all the IPC stuff with MOZ_MUST_USE? So I wouldn't add the annotation to any Recv* ::: uriloader/exthandler/HandlerServiceParent.h:20 (Diff revision 1) > private: > virtual ~HandlerServiceParent(); > virtual void ActorDestroy(ActorDestroyReason aWhy) override; > > > - virtual bool RecvFillHandlerInfo(const HandlerInfo& aHandlerInfoData, > + virtual MOZ_MUST_USE bool RecvFillHandlerInfo(const HandlerInfo& aHandlerInfoData, also these recv ::: uriloader/exthandler/mac/nsOSHelperAppService.h:41 (Diff revision 1) > // platformAppPath --> a platform specific path to an application that we got out of the > // rdf data source. This can be a mac file spec, a unix path or a windows path depending on the platform > // aFile --> an nsIFile representation of that platform application path. > - virtual nsresult GetFileTokenForPath(const char16_t * platformAppPath, nsIFile ** aFile); > + virtual MOZ_MUST_USE nsresult GetFileTokenForPath(const char16_t * platformAppPath, nsIFile ** aFile); > > - nsresult OSProtocolHandlerExists(const char * aScheme, > + MOZ_MUST_USE nsresult OSProtocolHandlerExists(const char * aScheme, now the params don't align anymore. Please fix

Ongoing try: https://treeherder.mozilla.org/#/jobs?repo=try&revision=71ff13f86047

Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/autoland/rev/26a015091dbc Add MOZ_MUST_USE to prevent potential bugs. r=smaug

Backed out in https://hg.mozilla.org/integration/autoland/rev/09c45bad0ba5 for OS X failures in the web-platform-test /html/webappapis/system-state-and-capabilities/the-navigator-object/protocol.html like https://treeherder.mozilla.org/logviewer.html#?job_id=25620789&repo=try

> You have a few places where you've added MOZ_MUST_USE to an NS_IMETHOD > function. I'm in the process of including MOZ_MUST_USE in the NS_IMETHOD > macro itself, in bug 1292440, so those changes won't be necessary. I ended up abandoning bug 1292440 because it wasn't working well. The alternative is to use the new [must_use] property in IDL, which I added in bug 1296164.

Fixed the failed test: https://treeherder.mozilla.org/#/jobs?repo=try&revision=f84db78823b5 Only changed one line.

Pushed by wpan@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/18c67e7044be Add MOZ_MUST_USE to prevent potential bugs. r=smaug