Add 'remark' diagnostic type in clang

[Arthur O'Dwyer]

FWIW, I'm strongly strongly opposed to making up a second parallel-universe
diagnostic system (-W versus -R, -Werror= versus <nothing>, etc.) that will
have to be maintained in perpetuity alongside the old system, when it's
trivially possible to extend the old system to support varying severity
levels. Still, as I don't have time to write the patch, I'm exiting the
discussion at this point.

–Arthur


On Fri, Feb 28, 2014 at 3:32 AM, Chandler Carruth <chandlerc at google.com>wrote:

>
> On Fri, Feb 28, 2014 at 3:18 AM, Joerg Sonnenberger <
> joerg at britannica.bec.de> wrote:
>
>> On Fri, Feb 28, 2014 at 02:52:55AM -0800, Chandler Carruth wrote:
>> > On Fri, Feb 28, 2014 at 2:10 AM, Dmitri Gribenko <gribozavr at gmail.com
>> >wrote:
>> >
>> > > On Fri, Feb 28, 2014 at 9:56 AM, Chandler Carruth <
>> chandlerc at google.com>
>> > > wrote:
>> > > > I also think that it is ultimately a mistake to re-use the same
>> > > commandline
>> > > > syntax for remarks. While I'd rather re-use all the infrastructure,
>> I'd
>> > > > rather separate these into a different spelling eventually.
>> > >
>> > > +1.  Our current logic around -W* flags is already complex enough (to
>> > > understand -- both for users and developers, to extend and to
>> > > maintain), and overloading the meaning of -W* for remarks will only
>> > > make it even more complex, and possibly confusing as well (it is well
>> > > established that -W* flags enable warnings).
>> >
>> >
>> > Ya know, -R* flags don't appear to be in use in the GCC commandline at
>> all.
>> > I'm liking the idea of taking over that capital letter for
>> remarks-related
>> > flags more.
>>
>> Please don't. It is a linker flag that has historically often been
>> passed without -Wl to gcc.
>
>
> GCC has rejected this flag since at least GCC 4.5, the oldest version I
> have lying around to test with. Clearly this syntax has long since gone
> away in GCC, and I think it is entirely reasonable for it to not work with
> Clang.
>
> We don't even come *close* to supporting all of the flags GCC accepts
> today, much less those it started rejecting over four years ago.
>
> _______________________________________________
> cfe-commits mailing list
> cfe-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140228/5358550d/attachment.html>