Hi Ramu,
On Monday 22 February 2010 18:59:03 Ramu Denduluri wrote:
I was trying to compile Epics base 3.14.9 on 64 bit Fedora 11 operating
system and got compilation errors as follows:
We don't have access to any Fedora 11 systems here so what I say below might
not be quite right, but we suspect that you're missing an RPM and that the
error you're getting isn't an EPICS issue as such.
However I would advise you that Base R3.14.9 built for a 64-bit machine should
*only* be used for Channel Access clients (and be aware that some of the
standard clients like StripTool and ALH may not work on 64-bit at all yet);
the IOC code was not properly patched to work on 64-bit until R3.14.10.
/usr/bin/ld:
/usr/lib/gcc/x86_64-redhat-linux/4.4.1/libstdc++.a(stdexcept.o):
relocation R_X86_64_32S against `vtable for std::runtime_error' can not
be used when making a shared object; recompile with -fPIC
/usr/lib/gcc/x86_64-redhat-linux/4.4.1/libstdc++.a: could not read
symbols: Bad value
Does your /usr/lib/gcc/x86_64-redhat-linux/4.4.1/ directory contain any files
or soft links with names starting "libstdc++.so"? If soft-links, where do
they point to? You can run these commands (replacing my 4.1.1 with 4.4.1 and
adjust the number after the .so as necessary) and compare their output to
mine:
tux% file /usr/lib/gcc/x86_64-redhat-linux/4.1.1/libstdc++.so
/usr/lib/gcc/x86_64-redhat-linux/4.1.1/libstdc++.so: symbolic link to
`../../../../lib64/libstdc++.so.6.0.8'
tux% file /usr/lib64/libstdc++.so.6.0.8
/usr/lib64/libstdc++.so.6.0.8: ELF 64-bit LSB shared object, AMD x86-64,
version 1 (SYSV), stripped
My machine tux is a RHEL5 64-bit system and has two separate RPMs which are
relevant to this issue:
tux% rpm -qf /usr/lib/gcc/x86_64-redhat-linux/4.1.1/libstdc++.so
gcc-c++-4.1.2-46.el5
tux% rpm -qf /usr/lib64/libstdc++.so.6.0.8
libstdc++-4.1.2-46.el5
I suspect you are missing the .so file completely, in which case you'll need
to install the FC11 versions of one or more of the above packages.
- Andrew
