Fix frontend crash related to forward declaration of NSNumber

[AlexDenisov]

Summary

Fixes this bug: http://llvm.org/bugs/show_bug.cgi?id=22561

[AlexDenisov]

Ping?

It’s a trivial one, but I don’t want to push it without review, though it fixes crash
-- 
AlexDenisov
Software Engineer, https://github.com/AlexDenisov

[hfinkel]

In D7643#124219, @AlexDenisov wrote:

Ping?

It’s a trivial one, but I don’t want to push it without review, though it fixes crash

Not to give you a hard time, but, why are you asking for pre-commit review? It sounds like you consider this to be a fairly-obvious one-line change to fix a crash-on-invalid case (and that is indeed what it looks like). If so, you don't need pre-commit review, so please feel free to commit. If you'd like a pre-commit review anyway, it is helpful to explain your doubts so potential reviews have some idea about what you're unsure, and are aware of any potential complications you're contemplating.

See: http://llvm.org/docs/DeveloperPolicy.html#obtaining-commit-access

Also, for the future, please upload full-context patches to Phabricator, see: http://llvm.org/docs/Phabricator.html#requesting-a-review-via-the-web-interface

-- 
AlexDenisov
Software Engineer, https://github.com/AlexDenisov

[AlexDenisov]

Not to give you a hard time, but, why are you asking for pre-commit review?
Well, due to first link (obtaining commit access) I got 'commit-after-approval’ access, so just trying to follow the rules.

Anyway, thank you for feedback and useful links.

-- 

AlexDenisov
Software Engineer, https://github.com/AlexDenisov

[hfinkel]

In D7643#124251, @AlexDenisov wrote:

Not to give you a hard time, but, why are you asking for pre-commit review?
Well, due to first link (obtaining commit access) I got 'commit-after-approval’ access, so just trying to follow the rules.

Understood, but we do have this rule:

2. You are allowed to commit patches without approval which you think are obvious. This is clearly a subjective decision — we simply expect you to use good judgement. Examples include: fixing build breakage, reverting obviously broken patches, documentation/comment changes, any other minor changes.

and you're welcome to follow that one too ;)

Anyway, thank you for feedback and useful links.

-- 

AlexDenisov
Software Engineer, https://github.com/AlexDenisov

lib/Sema/SemaExprObjC.cpp

       S.Diag(Loc, diag::err_undeclared_nsnumber);
       return nullptr;
     }
-    
+  }
+
+  if (S.NSNumberPointer.isNull()) {
     // generate the pointer to NSNumber type.
     QualType NSNumberObject = CX.getObjCInterfaceType(S.NSNumberDecl);
     S.NSNumberPointer = CX.getObjCObjectPointerType(NSNumberObject);

test/SemaObjC/objc-literal-nsnumber.m

 typedef int NSInteger;
 #endif
 
+void checkNSNumberUnavailableDiagnostic() {
+  id num = @1000; // expected-error {{NSNumber must be available to use Objective-C literals}}
+
+  int x = 1000;
+  id num1 = @(x); // expected-error {{NSNumber must be available to use Objective-C literals}}\
+                  // expected-error {{illegal type 'int' used in a boxed expression}}
+}
+
+@class NSNumber;
+
+void checkNSNumberFDDiagnostic() {
+  id num = @1000; // expected-error {{NSNumber must be available to use Objective-C literals}}
+
+  int x = 1000;
+  id num1 = @(x); // expected-error {{declaration of 'numberWithInt:' is missing in NSNumber class}}\
+                  // expected-error {{illegal type 'int' used in a boxed expression}}
+}
+
+
 @interface NSObject
 + (NSObject*)nsobject;
 @end