Description of problem: The Intel Ethernet (82579LM) driver (e1000e) crash randomly when enabling the device. Version-Release number of selected component (if applicable): OS: Fedora 18 driver:e1000e driver version: 2.5.4-NAPI firmware-verison:0.13-3 How reproducible: Not very often. I can reproduce the crash 2 out of 10 times Steps to Reproduce: 1. Disable the network. 2. Reboot the OS 3. After boot up, type "ifup em1" 4. If the interface is enabled without crash, go to step 2. Finally it will crash. Actual results: Most of the time, the interface was enabled without any problem. Expected results: 2 out of 10 times it will crash Additional info: From vmcore created by Kdump: PID: 0 TASK: ffffffff81c13420 CPU: 0 COMMAND: "swapper/0" #0 [ffff88011e203b70] machine_kexec at ffffffff8103d985 #1 [ffff88011e203be0] crash_kexec at ffffffff810c2218 #2 [ffff88011e203cb0] oops_end at ffffffff8160e4e8 #3 [ffff88011e203ce0] no_context at ffffffff81602b57 #4 [ffff88011e203d40] __bad_area_nosemaphore at ffffffff81602d41 #5 [ffff88011e203d90] bad_area_nosemaphore at ffffffff81602d73 #6 [ffff88011e203da0] do_page_fault at ffffffff8161106f #7 [ffff88011e203eb0] page_fault at ffffffff8160d925 [exception RIP: smp_irq_move_cleanup_interrupt+183] RIP: ffffffff8103a6f7 RSP: ffff88011e203f68 RFLAGS: 00010046 RAX: 0000000000001717 RBX: 00000000000000a3 RCX: 00000000fffffffa RDX: 0000000000000017 RSI: 0000000000000032 RDI: ffff8801160ee38c RBP: ffff88011e203fa8 R8: ffff8801160ee300 R9: ffff880119400000 R10: 0000000000000001 R11: 0000000000000001 R12: 0000000000000000 R13: 00000000000000a3 R14: 0000000000000001 R15: ffff8801160ee300 ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018 #8 [ffff88011e203fb0] irq_move_cleanup_interrupt at ffffffff81615d67 --- <IRQ stack> --- #9 [ffffffff81c01d68] irq_move_cleanup_interrupt at ffffffff81615d67 [exception RIP: intel_idle+240] RIP: ffffffff81328b60 RSP: ffffffff81c01e18 RFLAGS: 00000206 RAX: 0000000000000000 RBX: 0000000000000000 RCX: 20c49ba5e353f7cf RDX: 0000000000005100 RSI: 0000000000000000 RDI: 0000000000000000 RBP: ffffffff81c01e78 R8: 0000000000000004 R9: 0000000000016544 R10: 0000000000000001 R11: 0000000000000000 R12: ffffffff81082809 R13: ffffffff81c01da8 R14: ffff88011e20e740 R15: ffff88011e20e940 ORIG_RAX: ffffffffffffffdf CS: 0010 SS: 0018 #10 [ffffffff81c01e80] cpuidle_enter at ffffffff814bb829 #11 [ffffffff81c01e90] cpuidle_idle_call at ffffffff814bbeb9 #12 [ffffffff81c01ee0] cpu_idle at ffffffff8101c53f
By the way, my test machine is DELL LATITUDE E6430
Seems another re-incarnation of the old: https://bugzilla.redhat.com/show_bug.cgi?id=869341
Fei, what exact kernel version are you running when you see this? thanks, Michele
Created attachment 854065 [details] output of dmesg from the vmcore of kdump This is the output of dmesg from the vmcore of kdump
Hi Michele, I am using kernel 3.6 with minor changes to TCP/IP and USB. Thanks. Fei
Hi Fei, apologies, but we can't really chase issues in self-compiled kernels. We're already short on people as it is. If you can reproduce it with a current stock fedora-released kernel, please do file another bug. Thanks, Michele *** This bug has been marked as a duplicate of bug 126342 ***