Re: [mpir-devel] detecting CPU type, CFLAGS, MPN_PATH
On 22 May 2012 10:10, Jeroen Demeyer jdeme...@cage.ugent.be wrote: Within the Sage project, we discovered several issues with the CPU/CFLAGS/MPN_PATH configuration code in MPIR-2.4.0. The code mostly works well, but there are a few things I would like to change. I would also like to add a configuration option to build a generic binary meaning one which would run on any CPU of a given architecture (e.g. x86_64). It could still use assembly code, but only instructions which exist on every CPU of a given architecture. This should be possible, I think. this gives a generic fsf build ./configure --build=$(./configfsf.guess) make make check make install I dont think this will work on mingw64 though , as I've hacked other things for it. --- I would also like to reorganize the code to separate this configuration code from the rest of MPIR. Within Sage, I would then use this for all packages, not just MPIR. I'm not sure if I understand, but if you can demonstrate it working, I'll certainly take a look. -- I dont know what you mean? --- Does MPIR itself have any plans with this code? Is it likely you would accept these changes that I'm proposing? Not immediately, but if it is needed for Sage, we'll definitely consider it. So long as there are no performance regressions, it is very likely to be accepted. Bill. Jeroen. -- You received this message because you are subscribed to the Google Groups mpir-devel group. To post to this group, send email to mpir-devel@googlegroups.com. To unsubscribe from this group, send email to mpir-devel+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/mpir-devel?hl=en. -- You received this message because you are subscribed to the Google Groups mpir-devel group. To post to this group, send email to mpir-devel@googlegroups.com. To unsubscribe from this group, send email to mpir-devel+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/mpir-devel?hl=en. -- You received this message because you are subscribed to the Google Groups mpir-devel group. To post to this group, send email to mpir-devel@googlegroups.com. To unsubscribe from this group, send email to mpir-devel+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/mpir-devel?hl=en.
Re: [mpir-devel] detecting CPU type, CFLAGS, MPN_PATH
This configure line will be a much better generic x86_64 build , due to the way the directory's are laid out ./configure --build=$(if [ $(./configfsf.guess | cut -f 1 -d -) = x86_64 ] ; then echo k8-$(./config.guess | cut -f 2- -d -) ; else echo $(./config.guess) ; fi) make make check make install you can replace the k8 with core2 if you think its more common Jason -- 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.
Re: [mpir-devel] detecting CPU type, CFLAGS, MPN_PATH
On 22 May 2012 10:10, Jeroen Demeyer jdeme...@cage.ugent.be wrote: Within the Sage project, we discovered several issues with the CPU/CFLAGS/MPN_PATH configuration code in MPIR-2.4.0. The code mostly works well, but there are a few things I would like to change. I would also like to add a configuration option to build a generic binary meaning one which would run on any CPU of a given architecture (e.g. x86_64). It could still use assembly code, but only instructions which exist on every CPU of a given architecture. This should be possible, I think. I would also like to reorganize the code to separate this configuration code from the rest of MPIR. Within Sage, I would then use this for all packages, not just MPIR. I'm not sure if I understand, but if you can demonstrate it working, I'll certainly take a look. Does MPIR itself have any plans with this code? Is it likely you would accept these changes that I'm proposing? Not immediately, but if it is needed for Sage, we'll definitely consider it. So long as there are no performance regressions, it is very likely to be accepted. Bill. Jeroen. -- 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. -- 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.