<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Thu, May 1, 2014 at 4:01 PM, David Blaikie <span dir="ltr"><<a href="mailto:dblaikie@gmail.com" target="_blank">dblaikie@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Thu, May 1, 2014 at 3:54 PM, Kaelyn Takata <<a href="mailto:rikka@google.com">rikka@google.com</a>> wrote:<br>
><br>
><br>
><br>
> On Thu, May 1, 2014 at 3:26 PM, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> wrote:<br>
>><br>
>> On Thu, May 1, 2014 at 2:15 PM, Kaelyn Takata <<a href="mailto:rikka@google.com">rikka@google.com</a>> wrote:<br>
>> > Author: rikka<br>
>> > Date: Thu May 1 16:15:24 2014<br>
>> > New Revision: 207796<br>
>> ><br>
>> > URL: <a href="http://llvm.org/viewvc/llvm-project?rev=207796&view=rev" target="_blank">http://llvm.org/viewvc/llvm-project?rev=207796&view=rev</a><br>
>> > Log:<br>
>> > When sorting overload candidates, sort arity mismatches in ascending<br>
>> > order by the number of missing or extra parameters. This is useful if<br>
>> > there are more than a few overload candidates with arity mismatches,<br>
>> > particularly in the presence of -fshow-overloads=best.<br>
>> ><br>
>> > Modified:<br>
>> > cfe/trunk/lib/Sema/SemaOverload.cpp<br>
>> > cfe/trunk/test/Misc/error-limit-multiple-notes.cpp<br>
>> ><br>
>> > Modified: cfe/trunk/lib/Sema/SemaOverload.cpp<br>
>> > URL:<br>
>> > <a href="http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaOverload.cpp?rev=207796&r1=207795&r2=207796&view=diff" target="_blank">http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaOverload.cpp?rev=207796&r1=207795&r2=207796&view=diff</a><br>
>> ><br>
>> > ==============================================================================<br>
>> > --- cfe/trunk/lib/Sema/SemaOverload.cpp (original)<br>
>> > +++ cfe/trunk/lib/Sema/SemaOverload.cpp Thu May 1 16:15:24 2014<br>
>> > @@ -9229,7 +9229,10 @@ static unsigned RankDeductionFailure(con<br>
>> ><br>
>> > struct CompareOverloadCandidatesForDisplay {<br>
>> > Sema &S;<br>
>> > - CompareOverloadCandidatesForDisplay(Sema &S) : S(S) {}<br>
>> > + size_t NumArgs;<br>
>> > +<br>
>> > + CompareOverloadCandidatesForDisplay(Sema &S, size_t nArgs)<br>
>> > + : S(S), NumArgs(nArgs) {}<br>
>> ><br>
>> > bool operator()(const OverloadCandidate *L,<br>
>> > const OverloadCandidate *R) {<br>
>> > @@ -9254,8 +9257,18 @@ struct CompareOverloadCandidatesForDispl<br>
>> > if (!L->Viable) {<br>
>> > // 1. Arity mismatches come after other candidates.<br>
>> > if (L->FailureKind == ovl_fail_too_many_arguments ||<br>
>> > - L->FailureKind == ovl_fail_too_few_arguments)<br>
>> > + L->FailureKind == ovl_fail_too_few_arguments) {<br>
>> > + if (R->FailureKind == ovl_fail_too_many_arguments ||<br>
>> > + R->FailureKind == ovl_fail_too_few_arguments) {<br>
>> > + int LDist = abs(L->Function->getNumParams() - NumArgs);<br>
>> > + int RDist = abs(R->Function->getNumParams() - NumArgs);<br>
>> > + if (LDist == RDist)<br>
>> > + return L->FailureKind == ovl_fail_too_many_arguments &&<br>
>> > + R->FailureKind == ovl_fail_too_few_arguments;<br>
>> > + return LDist < RDist;<br>
>> > + }<br>
>> > return false;<br>
>> > + }<br>
>> > if (R->FailureKind == ovl_fail_too_many_arguments ||<br>
>> > R->FailureKind == ovl_fail_too_few_arguments)<br>
>> > return true;<br>
>> > @@ -9442,7 +9455,7 @@ void OverloadCandidateSet::NoteCandidate<br>
>> > }<br>
>> ><br>
>> > std::sort(Cands.begin(), Cands.end(),<br>
>> > - CompareOverloadCandidatesForDisplay(S));<br>
>> > + CompareOverloadCandidatesForDisplay(S, Args.size()));<br>
>> ><br>
>> > bool ReportedAmbiguousConversions = false;<br>
>> ><br>
>> ><br>
>> > Modified: cfe/trunk/test/Misc/error-limit-multiple-notes.cpp<br>
>> > URL:<br>
>> > <a href="http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Misc/error-limit-multiple-notes.cpp?rev=207796&r1=207795&r2=207796&view=diff" target="_blank">http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Misc/error-limit-multiple-notes.cpp?rev=207796&r1=207795&r2=207796&view=diff</a><br>
>> ><br>
>> > ==============================================================================<br>
>> > --- cfe/trunk/test/Misc/error-limit-multiple-notes.cpp (original)<br>
>> > +++ cfe/trunk/test/Misc/error-limit-multiple-notes.cpp Thu May 1<br>
>> > 16:15:24 2014<br>
>> > @@ -4,20 +4,22 @@<br>
>> > void foo(int);<br>
>> > void foo(double);<br>
>> > void foo(int, int);<br>
>> > +void foo(int, int, int, int);<br>
>> ><br>
>> > int main()<br>
>> > {<br>
>> > - foo();<br>
>> > + foo(1, 2, 3);<br>
>> > }<br>
>> ><br>
>> > // error and note suppressed by error-limit<br>
>> > struct s1{};<br>
>> > struct s1{};<br>
>> ><br>
>> > -// CHECK: 10:5: error: no matching function for call to 'foo'<br>
>> > -// CHECK: 6:6: note: candidate function not viable: requires 2<br>
>> > arguments, but 0 were provided<br>
>> > -// CHECK: 5:6: note: candidate function not viable: requires 1<br>
>> > argument, but 0 were provided<br>
>> > -// CHECK: 4:6: note: candidate function not viable: requires 1<br>
>> > argument, but 0 were provided<br>
>> > +// CHECK: 11:5: error: no matching function for call to 'foo'<br>
>> > +// CHECK: 6:6: note: candidate function not viable: requires 2<br>
>> > arguments, but 3 were provided<br>
>> > +// CHECK: 7:6: note: candidate function not viable: requires 4<br>
>> > arguments, but 3 were provided<br>
>> > +// CHECK: 5:6: note: candidate function not viable: requires 1<br>
>> > argument, but 3 were provided<br>
>> > +// CHECK: 4:6: note: candidate function not viable: requires 1<br>
>> > argument, but 3 were provided<br>
>><br>
>> Would it be appropriate/worthwhile adding some cases where there are<br>
>> the same number of arguments but they're of different types? Or in<br>
>> that case do we not show the other candidates at all?<br>
><br>
><br>
> What exactly do you mean by "same number of arguments but different types"?<br>
> If you meant that a candidate accepts the right number of arguments but the<br>
> types are wrong,<br>
<br>
</div></div>Yep, that's what I meant.<br>
<div class=""><br>
> they are already sorted before any arity mismatches. Also,<br>
> there are already two one-argument candidates that have different argument<br>
> types.<br>
<br>
</div>Hmm - should they be in the check line to demonstrate/test that they<br>
do come before arity mismatches? Or is that covered by other tests<br>
somewhere?<br></blockquote><div><br></div><div>Presumably they are covered by other tests, though it looks like most of this comparator was written about 3 or 4 years ago...</div><div><br></div><div>This test had been the only one that broke when I changed the sort order, which is why it is the one I modified to test the arity sorting a bit more thoroughly.</div>
<div> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="HOEnZb"><div class="h5"><span style="color:rgb(34,34,34)"> </span><br></div></div></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb">
<div class="h5">
><br>
>><br>
>> > // CHECK: fatal error: too many errors emitted, stopping now<br>
>> > -// CHECK-NOT: 15:8: error: redefinition of 's1'<br>
>> > -// CHECK-NOT: 14:8: note: previous definition is here<br>
>> > +// CHECK-NOT: 16:8: error: redefinition of 's1'<br>
>> > +// CHECK-NOT: 15:8: note: previous definition is here<br>
>> ><br>
>> ><br>
>> > _______________________________________________<br>
>> > cfe-commits mailing list<br>
>> > <a href="mailto:cfe-commits@cs.uiuc.edu">cfe-commits@cs.uiuc.edu</a><br>
>> > <a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits</a><br>
><br>
><br>
</div></div></blockquote></div><br></div></div>