Re: [LKP] Re: [mm/debug] 87c4696d57: kernel_BUG_at_include/linux/mm.h

[Rong Chen <rong.a.chen@xxxxxxxxx>]

On 1/7/20 2:46 PM, Anshuman Khandual wrote:

On 12/27/2019 07:52 PM, kernel test robot wrote:

[    9.781974] kernel BUG at include/linux/mm.h:592!
[    9.782810] invalid opcode: 0000 [#1] PTI
[    9.783443] CPU: 0 PID: 1 Comm: swapper Not tainted 5.5.0-rc3-00001-g87c4696d57b5e #1
[    9.784528] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
[    9.785756] EIP: __free_pages+0x14/0x40
[    9.786442] Code: 0c 9c 5e fa 89 d8 e8 5b f3 ff ff 56 9d 5b 5e 5d c3 8d 74 26 00 90 8b 48 1c 55 89 e5 85 c9 75 16 ba b4 b6 84 d6 e8 ac 49 fe ff <0f> 0b 8d b4 26 00 00 00 00 8d 76 00 ff 48 1c 75 10 85 d2 75 07 e8
[    9.789697] EAX: d68761f7 EBX: ea52f000 ECX: ea4f8520 EDX: d684b6b4
[    9.790850] ESI: 00000000 EDI: ef45e000 EBP: ea501f08 ESP: ea501f08
[    9.791879] DS: 007b ES: 007b FS: 0000 GS: 0000 SS: 0068 EFLAGS: 00010286
[    9.792783] CR0: 80050033 CR2: ffffffff CR3: 16d00000 CR4: 000406b0
[    9.792783] Call Trace:
[    9.792783]  free_pages+0x3c/0x50
[    9.792783]  pgd_free+0x5a/0x170
[    9.792783]  __mmdrop+0x42/0xe0
[    9.792783]  debug_vm_pgtable+0x54f/0x567
[    9.792783]  kernel_init_freeable+0x90/0x1e3
[    9.792783]  ? rest_init+0xf0/0xf0
[    9.792783]  kernel_init+0x8/0xf0
[    9.792783]  ret_from_fork+0x19/0x24
[    9.792783] Modules linked in:
[    9.792803] ---[ end trace 91b7335adcf0b656 ]---


To reproduce:

        # build kernel
	cd linux
	cp config-5.5.0-rc3-00001-g87c4696d57b5e .config
	make HOSTCC=gcc-7 CC=gcc-7 ARCH=i386 olddefconfig prepare modules_prepare bzImage

        git clone https://github.com/intel/lkp-tests.git
        cd lkp-tests
        bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email

Hello,

As the failure might be happening during boot when the test executes,
do we really need to run these LKP based QEMU environment in order to
reproduce the problem ? Could not this be recreated on a standalone
system.

- Anshuman

Hi Anshuman,

Thanks for your advice, we provided another reproduce method which is in the end of dmesg.xz file attached in original report:
but the initrd file need to be supplied.

qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-0 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-1 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-2 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-3 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-4 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-5 256G
qemu-img create -f qcow2 disk-vm-snb-89942dcb126f-6 256G

kvm=(
›   qemu-system-x86_64
›   -enable-kvm
›   -cpu SandyBridge
›   -kernel $kernel
›   -initrd initrd-vm-snb-89942dcb126f
›   -m 8192
›   -smp 2
›   -device e1000,netdev=net0
›   -netdev user,id=net0,hostfwd=tcp::32032-:22
›   -boot order=nc
›   -no-reboot
›   -watchdog i6300esb
›   -watchdog-action debug
›   -rtc base=localtime
›   -drive file=disk-vm-snb-89942dcb126f-0,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-1,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-2,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-3,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-4,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-5,media=disk,if=virtio
›   -drive file=disk-vm-snb-89942dcb126f-6,media=disk,if=virtio
›   -serial stdio
›   -display none
›   -monitor null
)

append=(
›   ip=::::vm-snb-89942dcb126f::dhcp
›   root=/dev/ram0
›   user=lkp
›   job=/job-script
›   ARCH=i386                                                                                                                                                                                                                                                                                                                                                              
›   kconfig=i386-randconfig-h001-20191225
›   branch=linux-devel/devel-hourly-2019122518
›   commit=87c4696d57b5e380bc2720fdff0772b042462b7d
›   BOOT_IMAGE=/pkg/linux/i386-randconfig-h001-20191225/gcc-7/87c4696d57b5e380bc2720fdff0772b042462b7d/vmlinuz-5.5.0-rc3-00001-g87c4696d57b5e
›   max_uptime=1500
›   RESULT_ROOT=/result/trinity/300s/vm-snb/yocto-minimal-i386-2019-05-20.cgz/i386-randconfig-h001-20191225/gcc-7/87c4696d57b5e380bc2720fdff0772b042462b7d/8
›   result_service=tmpfs
›   debug
›   apic=debug
›   sysrq_always_enabled
›   rcupdate.rcu_cpu_stall_timeout=100
›   net.ifnames=0
›   printk.devkmsg=on
›   panic=-1
›   softlockup_panic=1
›   nmi_watchdog=panic
›   oops=panic
›   load_ramdisk=2
›   prompt_ramdisk=0
›   drbd.minor_count=8
›   systemd.log_level=err
›   ignore_loglevel
›   console=tty0
›   earlyprintk=ttyS0,115200
›   console=ttyS0,115200
›   vga=normal
›   rw
›   rcuperf.shutdown=0
›   watchdog_thresh=60
)

"${kvm[@]}" -append "${append[*]}"


We will definitely give serious thought to your suggestion.

Best Regards,
Rong Chen