We're trying to debug (what we assume is) a nasty application bug in software deployed on some RHEL3 boxes, and it looks like we might have to resort to instrumenting some low-level funcs in libc to accomplish that.
Having paid RHAT for access to the "channel" that entitles us to grab RPMs from their servers, I used up2date to install all the latest packages available from RHAT, so I believe we have a canonical, up-to-date RHEL3-WS box. I then downloaded glibc-2.3.2-95.50.src.rpm from RHAT since "rpm -qa" reports that glibc-2.3.2-95.50 currently installed. I then installed that source package thus: rpm -ivh glibc-2.3.2-95.50.src.rpm ...and built it, thus: cd /usr/src/redhat/BUILD rpmbuild --verbose -bp ../SPECS/glibc.spec ...with no apparent build problems. However, it seems like the resultant objects and libraries bear no resemblance to any of the currently installed objs and libs on the box. For example, the resultant libc.so looks like this: -rwxr-xr-x 1 root root 8937512 Jul 12 09:55 libc.so* libc.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), not stripped ...while no installed version of anything is even half that size: -rwxr-xr-x 1 root root 1560188 Apr 19 12:19 /lib/i686/libc-2.3.2.so* /lib/i686/libc-2.3.2.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), not stripped -rwxr-xr-x 1 root root 1569204 Apr 19 12:19 /lib/libc-2.3.2.so* /lib/libc-2.3.2.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), not stripped -rwxr-xr-x 1 root root 1573180 Apr 19 12:19 /lib/tls/libc-2.3.2.so* /lib/tls/libc-2.3.2.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), not stripped ...and the lists of symbols extracted from the pre-built libs are surprisingly different from the locally-built libs, with even some fundamental things like errno handling apparently different. So, this Debian user is wondering (publicly, for posterity) what's the canonical way to reproduce a given RHAT package like this from (what I've been lead to believe is) the corresponding source RPM? _______________________________________________ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/