#HOW TO INSTALL GLIBC 2.19 PORTABLE#

For your safety, please do change your compiler back to your current default one after successfully compiling IRFinder. I'm not sure if it influences your future compilation or if it brings incompatibility for your current packages. Please be aware, you're at your own risk to use anaconda GCC and Glibc. By doing so, you also have to make sure the GCC in anaconda is your default GCC.

If it is other distros, I think your only option as a non-admin user is to install glibc in your anaconda by conda install -c asmeurer glibc. Which linux distro are you using? If it is RHEL 6.x, the glibc is located at /usr/lib/x86_64-redhat-linux5E/lib64/ and you have to edit the Makefile of IRFinder by changing LDFLAGS_static := -static -static-libgcc to LDFLAGS_static := -static -static-libgcc -L/usr/lib/x86_64-redhat-linux5E/lib64. Usually this should be setup by your system admin. Could you please do a which gcc or gcc -v and figure out which version is in use. if your gcc 6.1 is your default compiler.

#HOW TO INSTALL GLIBC 2.19 PORTABLE#

Add some compilers and install the FNP toolkit, and you're ready to build portable daemons.There are several things going on here. In my case I created a VM in VirtualBox and installed CentOS 6.5, which has glibc 2.12. My solution was to create an 'old' system in a virtual machine to build our daemons. The ray of hope is that executables built on an 'older' system seem to run well on 'new' systems. It's when you build your executable on a 'new' system (glibc > 2.14) and try to run it on an 'older' system (glibc < 2.14) that there is trouble. Likewise if the older/newer systems are later than 2.14. If your older/newer systems are all earlier than 2.14, everything runs. The dividing line appears to be glibc 2.14. Going in the other direction doesn't always work an executable build with a newer version of glibc will not necessarily run on a system that features an older version of glibc. It works very well going forward, that is, running an executable built with an older version of glibc runs on a system that features a newer version of glibc. The maintainers of glibc go to great pains to ensure backward compatibility. The answer is to build your vendor daemon on an older system. Perhaps that means linking against an older glibc. I would like to be able to build a vendor daemon on our development system that runs on the maximum possible number of target systems.

The daemon produced by our older glibc 2.14 system ran everywhere, while the same daemon built with glibc 2.19 needs glibc 2.14 or newer. Since the upgrade, our customers often see the following message when they start up lmadmin: '/lib/libc.so.6: version `GLIBC_2.14' not found (required by fnpdaemon)'Īs I understand (from posts like this: ), glibc is very backward compatible in that old apps will run under new glibc versions, but it's a little more complicated for new apps running under older versions of glibc. Previous to the upgrade, we could build a vendor daemon on our system and it would work on pretty much any Linux flavor and version we tried. In the process, we movied from glibc2.14 to glibc 2.19. We recently upgraded our Linux development system to openSUSE LEAF 42.1. How do you compile your vendor daemon for maximum portability to older (Linux) systems?

How to install glibc 2.19