Description of problem: Four object files -- bitmap.o, find_bit.o, hweight.o and rbtree.o -- are installed as part of the python-perf rpm. $ rpm -qlp python-perf-4.5.0-0.rc5.git0.2.fc24.x86_64.rpm /usr/lib64/python2.7/site-packages /usr/lib64/python2.7/site-packages/bitmap.o /usr/lib64/python2.7/site-packages/find_bit.o /usr/lib64/python2.7/site-packages/hweight.o /usr/lib64/python2.7/site-packages/perf-0.1-py2.7.egg-info /usr/lib64/python2.7/site-packages/perf.so /usr/lib64/python2.7/site-packages/rbtree.o This seems to have been introduced with the 4.2.0 kernel with just hweight.o and rbtree.o: $ rpm -qlp python-perf-4.2.6-300.fc23.x86_64.rpm /usr/lib64/python2.7/site-packages /usr/lib64/python2.7/site-packages/hweight.o /usr/lib64/python2.7/site-packages/perf-0.1-py2.7.egg-info /usr/lib64/python2.7/site-packages/perf.so /usr/lib64/python2.7/site-packages/rbtree.o The 4.1 kernel was the last clean build with just the .so file: $ rpm -qlp python-perf-4.1.13-100.fc21.x86_64.rpm /usr/lib64/python2.7/site-packages /usr/lib64/python2.7/site-packages/perf-0.1-py2.7.egg-info /usr/lib64/python2.7/site-packages/perf.so Version-Release number of selected component (if applicable): python-perf-4.2 and newer How reproducible: always Steps to Reproduce: 1. rpm -ql python-perf Actual results: leftover object files from the kernel build are installed in the python site-packages directory Expected results: only perf.so and the egg-info file are installed Additional info:
Ew. Thanks for the report. I have an idea as to why this is, but I don't have a fix in hand yet.
I've been poking through the Makefiles to see why this might be happening, but I haven't figured it out yet. We could add an rm command to the rpm spec file to clean up the *.o files, but that's just masking the problem.
(In reply to Jeff Bastian from comment #2) > I've been poking through the Makefiles to see why this might be happening, > but I haven't figured it out yet. We could add an rm command to the rpm > spec file to clean up the *.o files, but that's just masking the problem. Right, exactly. I described the problem and the reason for it to upstream. Jiri replied with a patch that should fix things. I'll grab it Monday and give it a shot. http://thread.gmane.org/gmane.linux.kernel/2163216
I've added the patch in Rawhide. Should make the 4.5-rc6 build.