[petsc-dev] C++11, compatibility versus simplicity

Tim Tautges tautges at mcs.anl.gov
Thu Oct 3 16:11:45 CDT 2013


Switching your default C++ compilation type in response to one (of very many) of your dependent libs changing theirs 
seems odd, even for an otherwise-great one.  C++11 has some great stuff, but just how critical is it?

- tim

On 10/03/2013 04:02 PM, Jed Brown wrote:
> Tim Tautges <tautges at mcs.anl.gov> writes:
>
>> Lots of missing spaces for XL.
>
> Fortunately, the LLVM backend is competitive with XL on performance and
> is standards-compliant.  With any luck, IBM will stop shipping XL.
>
>> What about Cray, can't remember the compiler flavor they use.
>
> Cray has their own compiler, but it's not the default on any of their
> machines.
>
>> But anyway, surely you're not expecting your users to install their
>> own GCC or LLVM, right?
>
> I've seen brand-new near-petascale machines installed with 5+ year old
> software stacks.  If the people running the facility are stuck in the
> stone age, then yes, users can install a contemporary compiler.  Or the
> facility can provide a recent version.
>
> (Yes, it's a serious impediment and lack of cross-vendor ABI
> compatibility often means that it's not even an option.  People jump
> through far greater hoops, so it depends whether you are targeting the
> average or the extreme.)
>

-- 
================================================================
"You will keep in perfect peace him whose mind is
   steadfast, because he trusts in you."               Isaiah 26:3

              Tim Tautges            Argonne National Laboratory
          (tautges at mcs.anl.gov)      (telecommuting from UW-Madison)
  phone (gvoice): (608) 354-1459      1500 Engineering Dr.
             fax: (608) 263-4499      Madison, WI 53706




More information about the petsc-dev mailing list