[PATCH] D50901: [clang][ubsan] Split Implicit Integer Truncation Sanitizer into unsigned and signed checks

[Filipe Cabecinhas via Phabricator via cfe-commits]

filcab added inline comments.


================
Comment at: lib/CodeGen/CGExprScalar.cpp:305
   enum ImplicitConversionCheckKind : unsigned char {
-    ICCK_IntegerTruncation = 0,
+    ICCK_IntegerTruncation = 0, // Legacy, no longer used.
+    ICCK_UnsignedIntegerTruncation = 1,
----------------
vitalybuka wrote:
> lebedev.ri wrote:
> > vitalybuka wrote:
> > > why do you need to keep it?
> > *Here* - for consistency with the compiler-rt part.
> > 
> > There - what about mismatch in the used clang version
> > (which still only produced the `(ImplicitConversionCheckKind)0`), and compiler-rt version
> > (which has `(ImplicitConversionCheckKind)1` and `(ImplicitConversionCheckKind)2`)?
> > Is it 100.00% guaranteed not to happen? I'm not so sure.
> I don't think we try support mismatched versions of clang and compiler-rt
We don't make a big effort in open source. But we try to at least make it work (check previous work on type_mismatch handler, before versions were introduced), or error "loudly" when linking (versioned checks).

I think having keeping the older check number free is a good thing and allows us to have binary compatibility with older objects (and keep supporting old object files (we *did* release llvm 7.0 with the old-type check)).

If we hadn't had a release in between, I'd be all for reusing.


Repository:
  rC Clang

https://reviews.llvm.org/D50901