[PATCH] D18584: Complete support for C++ Core Guidelines Type.6: Always initialize a member variable.

Alexander Kornienko via cfe-commits cfe-commits at lists.llvm.org
Mon Apr 18 18:11:38 PDT 2016


Would be nice, if you could write a patch. I don't yet have a reduced test
case (it crashes without delayed template parsing), since my creduce is
still running after many hours ;)
On Apr 18, 2016 8:45 PM, "Michael Miller" <michael_miller at motu.com> wrote:

> michael_miller added a comment.
>
> In http://reviews.llvm.org/D18584#403872, @alexfh wrote:
>
> > FYI, the check has started crashing after this patch. I'll try to
> provide a minimal test case soon. The relevant part of the stack trace is:
> >
> >   @     0x7fc9c255efa0          8  clang::Stmt::getLocStart()
> >   @     0x7fc9c48fdac1         64
> clang::tidy::cppcoreguidelines::(anonymous
> namespace)::IntializerInsertion::getLocation()
> >   @     0x7fc9c49026d5       1696
> clang::tidy::cppcoreguidelines::ProTypeMemberInitCheck::checkMissingBaseClassInitializer()
> >   @     0x7fc9c490424f         96
> clang::tidy::cppcoreguidelines::ProTypeMemberInitCheck::check()
>
>
> Hmm... I can get this to crash in that location but only if I run with
> -fdelayed-template-parsing. I've got an easy fix and test case if that's
> the same issue you're running into. We should move the check at line 307
> into ProTypeMemberInitCheck::check before both checks. I've got a test case
> I can add to cppcoreguidelines-pro-type-member-init-delayed.cpp that
> reproduces it. Let me know if I should write up a patch for this.
>
>
> Repository:
>   rL LLVM
>
> http://reviews.llvm.org/D18584
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20160419/29960b90/attachment.html>


More information about the cfe-commits mailing list