Mariah, can I assume that gcc-4.3.1 and gcc-4.3.2 are not available on varro. There seem to be files or something with those names in /usr/local/bin, but they don't seem to do anything.
BTW, we all *really* appreciate your efforts in getting gcc working on these machines for us to use. Just to be clear, when I said that I thought gcc was broken on varro, this was not in any way intended to be critical of your excellent work installing it. I only meant that I thought either gcc has an (until now, unnoticed) bug or there is some bug in the Apple compiler on the machine so that it miscompiles gcc. Now I know that the gcc folks intend to have things this way, even though oodles of documentation on the web says otherwise. Thanks again for your hard work!! Bill. 2009/3/10 Bill Hart <goodwillh...@googlemail.com>: > Got it!! __APPLE_CC__ is defined to be 1 by FSF GCC, but is defined to > be the APPLE compiler revision by a genuine APPLE GCC. > > So the fix is trivial. > > The number of websites and even books that have the wrong information > about this is just immense! So much damned misinformation about GCC on > the web. > > Bill. > > 2009/3/9 <ja...@njkfrudils.plus.com>: >> >> >> If FSF GCC builds on the apple system then I suppose it would need to know >> which compiler is being used to bootstrap it , so somewhere in FSF GCC >> code/or configure would be the recommended safe way of detecting which. >> >> >> On Monday 09 March 2009 22:37:58 Bill Hart wrote: >>> OK, I did some reading, and apparently FSF GCC defines __APPLE_CC__ on >>> Apple machines. So testing if this is defined does not tell you if the >>> compiler vendor is Apple or FSF. >>> >>> The reason is that apparently Apple's include files break if this is >>> not defined by GCC. >>> >>> I'm surprised we haven't seen this before. But apparently if you use >>> FSF GCC on an Apple machine, then that is a situation we have not >>> encountered before. >>> >>> We need a new way to distinguish compiler vendor. I prefer to do this >>> at compile time if possible. Anyone have any ideas? >>> >>> Bill. >>> >>> 2009/3/9 Bill Hart <goodwillh...@googlemail.com>: >>> > Hi Mariah, >>> > >>> > Perhaps I can answer your question with another. >>> > >>> > How would I write the program so that gcc does not identify itself as >>> > Apple? >>> > >>> > Are you saying that by including stdio.h that it defines __APPLE_CC__? >>> > >>> > Are the standard libraries set up to tell me which compiler it is? If >>> > so they are telling me the wrong thing. This is gnu cc and not the >>> > Apple version. No? >>> > >>> > Bill. >>> > >>> > On 09/03/2009, Mariah <mariah.le...@gmail.com> wrote: >>> >> Bill, >>> >> >>> >> You claim that there is a problem with gcc-4.3.3 on varro. >>> >> I am afraid that I do not understand why you think so. >>> >> >>> >> It looks like you added an include clause to a hello world >>> >> program, and then because the include clause was evaluated >>> >> you claim there is a problem with gcc. I am confused. >>> >> >>> >> As I am the person who built gcc-4.3.3 for varro, I am interested in >>> >> hearing if there is a problem (and fixing it). I did not >>> >> notice any problem during the build of gcc. >>> >> >>> >> Can you enlighten me why you believe that there is problem >>> >> with gcc-4.3.3 on varro? >>> >> >>> >> Mariah >>> >>> >> >> >> >> >> > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "mpir-devel" group. To post to this group, send email to mpir-devel@googlegroups.com To unsubscribe from this group, send email to mpir-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/mpir-devel?hl=en -~----------~----~----~----~------~----~------~--~---
