Update suffix check and cast non-suffix types

[reikdas]

Summary

This addresses part of @rsmith 's inline comment on https://reviews.llvm.org/D77598.
I want to avoid having an explicit int cast (that happens in the current diff and makes a lot of tests fail) but I did not find a better way to check for that than T->getCanonicalTypeInternal().getAsString(Policy) == "int". Is there a better way to do that? If so could you please let me know and I'll update the diff.

[hubert.reinterpretcast]

@reikdas, please upload the full copy of your combined changes to D77598 where the review is still active and close this. This still doesn't address the long long versus 64-bit long that was also mentioned.

[reikdas]

In D79290#2016273, @hubert.reinterpretcast wrote:

@reikdas, please upload the full copy of your combined changes to D77598 where the review is still active and close this.

I am unable to add a diff/edit the old diff in the other revision(I think since I am not the author of that review). Clicking on "Update Diff" only allowed me to create a new revision.

In D79290#2016273, @hubert.reinterpretcast wrote:

This still doesn't address the long long versus 64-bit long that was also mentioned.

I updated the summary of this review to state that only part of the comment was addressed.

[hubert.reinterpretcast]

In D79290#2016278, @reikdas wrote:

In D79290#2016273, @hubert.reinterpretcast wrote:

@reikdas, please upload the full copy of your combined changes to D77598 where the review is still active and close this.

I am unable to add a diff/edit the old diff in the other revision(I think since I am not the author of that review). Clicking on "Update Diff" only allowed me to create a new revision.

Have you consulted with the author on the status of the other revision? The "Commandeer Revision" option under "Add Action" in the menu above the comment submission box might work from the technical side.

[v.g.vassilev]

@hubert.reinterpretcast, this patch is part of our internal forks which we would like to put upstream. The author of the previous patch does not have bandwidth to work on it and @reikdas kindly volunteered to take over. I can close the previous one if that's preferable.

[hubert.reinterpretcast]

In D79290#2016608, @v.g.vassilev wrote:

@hubert.reinterpretcast, this patch is part of our internal forks which we would like to put upstream. The author of the previous patch does not have bandwidth to work on it and @reikdas kindly volunteered to take over. I can close the previous one if that's preferable.

Thanks; if @reikdas can try to update the previous patch using the "Commandeer Revision" action I described, then that is preferable. If that doesn't work, then this patch should have the full set of changes (including the test changes) and the previous patch should be closed.

[BATManVSGig]

this good idea.

clang/lib/AST/TemplateBase.cpp

static void printIntegral(const TemplateArgument &TemplArg,
   if (T->isBooleanType() && !Policy.MSVCFormatting) {
     Out << (Val.getBoolValue() ? "true" : "false");
   } else if (T->isCharType()) {
     const char Ch = Val.getZExtValue();
     Out << ((Ch == '\'') ? "'\\" : "'");
     Out.write_escaped(StringRef(&Ch, 1), /*UseHexEscapes=*/ true);
     Out << "'";
   } else {
-    Out << Val;
     if (T->isBuiltinType()) {
-      if (Val.isUnsigned())
-        Out << "U";
-      if (Val.getBitWidth() == 64)
-        Out << "LL";
+      if (T->isUnsignedIntegerType() && Val.getBitWidth() == 64)
+        Out << Val << "ULL";
+      else if (T->isSignedIntegerType() && Val.getBitWidth() == 64)
+        Out << Val << "LL";
+      else if (T->isUnsignedIntegerType())
+        Out << Val << "U";
+      else
+        Out << "(" << T->getCanonicalTypeInternal().getAsString(Policy) << ")" << Val;
     }
+    else
+      Out << Val;
   }
 }
 
 //===----------------------------------------------------------------------===//
 // TemplateArgument Implementation
 //===----------------------------------------------------------------------===//
 
 TemplateArgument::TemplateArgument(ASTContext &Ctx, const llvm::APSInt &Value,

clang/test/Misc/integer-literal-printing.cpp

 // RUN: %clang_cc1 %s -fsyntax-only -verify -std=c++11
 
 // PR11179
 template <short T> class Type1 {};
-template <short T> void Function1(Type1<T>& x) {} // expected-note{{candidate function [with T = -42] not viable: expects an l-value for 1st argument}}
+template <short T> void Function1(Type1<T>& x) {} // expected-note{{candidate function [with T = (short)-42] not viable: expects an l-value for 1st argument}}
 
 template <unsigned short T> class Type2 {};
 template <unsigned short T> void Function2(Type2<T>& x) {} // expected-note{{candidate function [with T = 42U] not viable: expects an l-value for 1st argument}}
 
 enum class boolTy : bool {
   b = 0,
 };