Description of problem: When reading the /proc/bus/pci/XX/YY file while hot-removing the device that relates to this file, kernel panics with the following call trace. ID: 14583 TASK: ffff81046be23860 CPU: 15 COMMAND: "appname" #0 [ffff81046fe2bbe0] crash_kexec at ffffffff800aeb6b #1 [ffff81046fe2bc68] pci_user_read_config_dword at ffffffff80056e0e #2 [ffff81046fe2bce0] do_page_fault at ffffffff80067dd7 #3 [ffff81046fe2bdd0] error_exit at ffffffff8005ede9 [exception RIP: pci_user_read_config_dword+83] RIP: ffffffff80056e0e RSP: ffff81046fe2be88 RFLAGS: 00010046 RAX: 0000000000000246 RBX: ffff81044f095000 RCX: 0000000000000004 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 RBP: 0000000000000000 R8: ffff81046fe2be94 R9: ffff81047e45eac0 R10: ffff81046fe2a000 R11: 0000000000000000 R12: 0000000000000087 R13: 0000000000000246 R14: ffff81046fe2bedc R15: ffff81046fe2bf50 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 #4 [ffff81046fe2be80] pci_user_read_config_dword at ffffffff80056def #5 [ffff81046fe2bec0] proc_bus_pci_read at ffffffff80160b60 #6 [ffff81046fe2bf10] vfs_read at ffffffff8000b6b0 #7 [ffff81046fe2bf40] sys_read at ffffffff80011c01 #8 [ffff81046fe2bf80] sysenter_do_call at ffffffff8006249d RIP: 00000000ffffe410 RSP: 00000000ef29c130 RFLAGS: 00000296 RAX: 0000000000000003 RBX: ffffffff8006249d RCX: 00000000f2e3b000 RDX: 0000000000001000 RSI: 00000000f264bff8 RDI: 00000000f264bff8 RBP: 0000000000000000 R8: 0000000000000000 R9: 0000000000000000 R10: 00000000ffffe410 R11: 0000000000000000 R12: 0000000000000000 R13: 00000000ef29c158 R14: 0000000000000030 R15: 0000000100000000 ORIG_RAX: 0000000000000003 CS: 0023 SS: 002b Version-Release number of selected component (if applicable): 2.6.18-194.el5 How reproducible: Every time Steps to Reproduce: Step to Reproduce: 1) Open a /proc/bus/pci/XX/YY file. 2) Hot-remove the related device. # echo 0 > /sys/bus/pci/slots/ZZ/power 3) Read the /proc/bus/pci/XX/YY file. Actual results: Panic Expected results: No panic. Additional info:
Going to continue working this in the original bug. *** This bug has been marked as a duplicate of bug 618114 ***