Hi Hans, This is relatively minor, but it's safe and fixes a rejects-valid regression in some configurations. Might be worth taking for Clang 4.
On 24 January 2017 at 15:18, Richard Smith via cfe-commits < cfe-commits@lists.llvm.org> wrote: > Author: rsmith > Date: Tue Jan 24 17:18:28 2017 > New Revision: 292991 > > URL: http://llvm.org/viewvc/llvm-project?rev=292991&view=rev > Log: > PR31742: Don't emit a bogus "zero size array" extwarn when initializing a > runtime-sized array from an empty list in an array new. > > Modified: > cfe/trunk/lib/Sema/SemaInit.cpp > cfe/trunk/test/SemaCXX/new-delete-cxx0x.cpp > > Modified: cfe/trunk/lib/Sema/SemaInit.cpp > URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/ > SemaInit.cpp?rev=292991&r1=292990&r2=292991&view=diff > ============================================================ > ================== > --- cfe/trunk/lib/Sema/SemaInit.cpp (original) > +++ cfe/trunk/lib/Sema/SemaInit.cpp Tue Jan 24 17:18:28 2017 > @@ -1684,7 +1684,7 @@ void InitListChecker::CheckArrayType(con > // If this is an incomplete array type, the actual type needs to > // be calculated here. > llvm::APSInt Zero(maxElements.getBitWidth(), > maxElements.isUnsigned()); > - if (maxElements == Zero) { > + if (maxElements == Zero && !Entity.isVariableLengthArrayNew()) { > // Sizing an array implicitly to zero is not allowed by ISO C, > // but is supported by GNU. > SemaRef.Diag(IList->getLocStart(), > > Modified: cfe/trunk/test/SemaCXX/new-delete-cxx0x.cpp > URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/ > SemaCXX/new-delete-cxx0x.cpp?rev=292991&r1=292990&r2=292991&view=diff > ============================================================ > ================== > --- cfe/trunk/test/SemaCXX/new-delete-cxx0x.cpp (original) > +++ cfe/trunk/test/SemaCXX/new-delete-cxx0x.cpp Tue Jan 24 17:18:28 2017 > @@ -1,4 +1,4 @@ > -// RUN: %clang_cc1 -fsyntax-only -verify %s -std=c++11 > -triple=i686-pc-linux-gnu > +// RUN: %clang_cc1 -fsyntax-only -verify %s -std=c++11 > -triple=i686-pc-linux-gnu -pedantic > > void ugly_news(int *ip) { > (void)new int[-1]; // expected-error {{array size is negative}} > @@ -29,6 +29,7 @@ void fn(int n) { > (void) new int[2] {1, 2}; > (void) new S[2] {1, 2}; > (void) new S[3] {1, 2}; > + (void) new S[n] {}; > // C++11 [expr.new]p19: > // If the new-expression creates an object or an array of objects of > class > // type, access and ambiguity control are done for the allocation > function, > @@ -44,6 +45,7 @@ void fn(int n) { > (void) new T[2] {1, 2}; // ok > (void) new T[3] {1, 2}; // expected-error {{no matching constructor}} > expected-note {{in implicit initialization of array element 2}} > (void) new T[n] {1, 2}; // expected-error {{no matching constructor}} > expected-note {{in implicit initialization of trailing array elements in > runtime-sized array new}} > + (void) new T[n] {}; // expected-error {{no matching constructor}} > expected-note {{in implicit initialization of trailing array elements in > runtime-sized array new}} > } > > struct U { > > > _______________________________________________ > cfe-commits mailing list > cfe-commits@lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits >
_______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits