Hi Andy, I got the below dmesg and the first bad commit is git://git.kernel.org/pub/scm/linux/kernel/git/luto/linux.git fast-return-to-userspace commit e94ab903e725eaf9d98e56f3043878bde7fc1b58 Author: Andy Lutomirski AuthorDate: Tue May 6 13:26:33 2014 -0700 Commit: Andy Lutomirski CommitDate: Tue May 6 13:26:33 2014 -0700 Fast interrupt return to userspace This could be even faster if it were written in assembler :) The only reason it's Signed-off-by is that I agree to the DCO. That should not be construed to mean that anyone should apply this patch. It's an abomination and it will do terrible, terrible things. It boots, though :) I haven't tested it beyond that. Signed-off-by: Andy Lutomirski +--------------------------------------------------------------------------+-----------+------------+ | | v3.15-rc4 | e94ab903e7 | +--------------------------------------------------------------------------+-----------+------------+ | boot_successes | 162 | 0 | | boot_failures | 5 | 20 | | BUG:kernel_early_hang_without_any_printk_output | 2 | | | Out_of_memory:Kill_process | 2 | | | backtrace:do_sys_open | 1 | | | backtrace:SyS_open | 1 | | | BUG:kernel_boot_hang | 1 | | | backtrace:vm_mmap_pgoff | 1 | | | backtrace:SyS_mmap_pgoff | 1 | | | backtrace:SyS_mmap | 1 | | | BUG:sleeping_function_called_from_invalid_context_at_arch/x86/mm/fault.c | 0 | 20 | | WARNING:CPU:PID:at_kernel/locking/lockdep.c:lockdep_trace_alloc() | 0 | 20 | | INFO:lockdep_is_turned_off | 0 | 20 | | WARNING:CPU:PID:at_kernel/smp.c:smp_call_function_many() | 0 | 16 | | WARNING:CPU:PID:at_kernel/smp.c:smp_call_function_single() | 0 | 16 | +--------------------------------------------------------------------------+-----------+------------+ [ 1.664070] Key type encrypted registered [ 1.669356] Freeing unused kernel memory: 2684K (ffffffff816f7000 - ffffffff81996000) [ 1.671079] BUG: sleeping function called from invalid context at arch/x86/mm/fault.c:1167 [ 1.672358] in_atomic(): 0, irqs_disabled(): 1, pid: 1, name: init [ 1.673103] 1 lock held by init/1: [ 1.673518] #0: (&mm->mmap_sem){++++++}, at: [] __do_page_fault+0xc9/0x4d0 [ 1.674401] irq event stamp: 1069840 [ 1.674401] hardirqs last enabled at (1069839): [] retint_swapgs+0x13/0x25 [ 1.674401] hardirqs last disabled at (1069840): [] error_sti+0x5/0x6 [ 1.674401] softirqs last enabled at (1065946): [] __do_softirq+0x1e2/0x2b0 [ 1.674401] softirqs last disabled at (1065935): [] irq_exit+0xd5/0xe0 [ 1.674401] CPU: 6 PID: 1 Comm: init Not tainted 3.15.0-rc4-00001-ge94ab90 #1 [ 1.674401] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 [ 1.674401] 00007fff4af8dff8 ffff880000211d60 ffffffff813d5a87 0000000000000029 [ 1.674401] ffff880000211d70 ffffffff81094dcd ffff880000211e70 ffffffff81038b54 [ 1.674401] ffff880000208908 0000000000000002 ffff88000fb200a8 0000000000000002 [ 1.674401] Call Trace: [ 1.674401] [] dump_stack+0x4d/0x66 [ 1.674401] [] __might_sleep+0xed/0x130 [ 1.674401] [] __do_page_fault+0xe4/0x4d0 [ 1.674401] [] ? local_clock+0x16/0x30 [ 1.674401] [] ? lock_release_holdtime.part.30+0xf/0x190 [ 1.674401] [] ? _raw_spin_unlock+0x22/0x30 [ 1.674401] [] ? vtime_account_user+0x8c/0xa0 [ 1.674401] [] ? context_tracking_user_exit+0xbb/0xd0 [ 1.674401] [] do_page_fault+0x31/0x70 [ 1.674401] [] do_async_page_fault+0x24/0xe0 [ 1.674401] [] async_page_fault+0x25/0x30 [ 1.674401] [] ? install_sysret_trampoline+0x36/0x70 [ 1.674401] [] retint_swapgs+0x18/0x25 Attached dmesg for the parent commit, too, to help confirm whether it is a noise error. git bisect start 0dcebb805ee5cb03202711908c828a76c60dc51e 89ca3b881987f5a4be4c5dbaa7f0df12bbdde2fd -- git bisect good a2149d48144498ee00c03e7cfef13841dd459e65 # 20:33 20+ 4 Merge 'cgroup/review-unified-updates' into devel-hourly-2014050720 git bisect bad 88f3059da6b1c67e559fba548dc1442c4b7e846d # 20:37 0- 20 Merge 'cgroup/for-next' into devel-hourly-2014050720 git bisect bad cc71f24555d15787e3bc3de4c654b44db127d535 # 20:37 0- 20 Merge 'kvm/queue' into devel-hourly-2014050720 git bisect bad ba96bd4b4a7445154b9dfd9b990402efed483948 # 20:37 0- 20 Merge 'arm-perf/aarch64' into devel-hourly-2014050720 git bisect bad 0d621996b9340f6043fc61bfde2a1c0694661f75 # 20:37 0- 20 Merge 'luto/fast-return-to-userspace' into devel-hourly-2014050720 git bisect bad e94ab903e725eaf9d98e56f3043878bde7fc1b58 # 20:37 0- 20 Fast interrupt return to userspace # first bad commit: [e94ab903e725eaf9d98e56f3043878bde7fc1b58] Fast interrupt return to userspace git bisect good 89ca3b881987f5a4be4c5dbaa7f0df12bbdde2fd # 20:39 69+ 5 Linux 3.15-rc4 git bisect bad 0dcebb805ee5cb03202711908c828a76c60dc51e # 20:39 0- 23 0day head guard for 'devel-hourly-2014050720' git bisect good 7e338c9991ecee9c2ac7a4cee2c2e11ecb563d02 # 21:07 69+ 0 Merge branch 'for-3.15' of git://linux-nfs.org/~bfields/linux git bisect good 9f072b9386bc35ca2ed534c0392f5036902ce41b # 23:54 69+ 0 Add linux-next specific files for 20140509 This script may reproduce the error. ----------------------------------------------------------------------------- #!/bin/bash kernel=$1 initrd=quantal-core-x86_64.cgz wget --no-clobber https://github.com/fengguang/reproduce-kernel-bug/blob/master/initrd/$initrd kvm=( qemu-system-x86_64 -cpu kvm64 -enable-kvm -kernel $kernel -initrd $initrd -smp 2 -m 256M -net nic,vlan=0,macaddr=00:00:00:00:00:00,model=virtio -net user,vlan=0 -net nic,vlan=1,model=e1000 -net user,vlan=1 -boot order=nc -no-reboot -watchdog i6300esb -serial stdio -display none -monitor null ) append=( debug sched_debug apic=debug ignore_loglevel sysrq_always_enabled panic=10 prompt_ramdisk=0 earlyprintk=ttyS0,115200 console=ttyS0,115200 console=tty0 vga=normal root=/dev/ram0 rw ) "${kvm[@]}" --append "${append[*]}" ----------------------------------------------------------------------------- Thanks, Jet