Getting Started - process getting killed

[CharithYMendis]

I am trying to setup nanoBench and use the kernel space implementation on a Haswell machine.

I built the code for the driver and installed it. However, when I run the example the process gets killed.

sudo ./kernel-nanoBench.sh -asm "ADD RAX, RBX; add RBX, RAX" -config configs/cfg_Haswell_common.txt ./kernel-nanoBench.sh: line 122: 21457 Killed $taskset cat /proc/nanoBench

[andreas-abel]

Which OS/kernel are you using?

Can you post the output of dmesg?

[CharithYMendis]

OS/kernel

Linux 4.4.0-148-generic #174~14.04.1-Ubuntu SMP x86_64 x86_64 x86_64 GNU/Linux

dmesg

[  549.285212] BUG: unable to handle kernel NULL pointer dereference at           (null)
[  549.285378] IP: [<ffff882014800154>] 0xffff882014800154
[  549.285487] PGD 80000000359b3067 PUD 359b4067 PMD 0
[  549.285687] Oops: 0002 [#1] SMP
[  549.285837] Modules linked in: nb(OE) nfsv3 ipmi_devintf ip6table_filter ip6_tables iptable_filter ip_tables x_tables openafs(POE) autofs4 nvidia_uvm(OE) nvidia_drm(POE) nvidia_modeset(POE) nvidia(POE) ipmi_ssif intel_rapl x86_pkg_temp_thermal bnep intel_powerclamp coretemp rfcomm kvm_intel bluetooth kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel joydev input_leds cryptd ttm drm_kms_helper drm fb_sys_fops syscopyarea sysfillrect sysimgblt sb_edac edac_core nfsd mei_me ioatdma shpchp mei lpc_ich auth_rpcgss nfs_acl wmi nfs lockd grace binfmt_misc sunrpc fscache mlx5_ib mlx4_ib ipmi_si 8250_fintek ipmi_msghandler mlx4_en vxlan ip6_udp_tunnel udp_tunnel mlx4_core ib_srp scsi_transport_srp ib_ipoib acpi_pad acpi_power_meter mac_hid ib_umad rdma_ucm rdma_cm iw_cm ib_uverbs ib_cm ib_sa ib_mad ib_core ib_addr hid_generic mlx5_core igb usbhid i2c_algo_bit ahci dca hid libahci ptp pps_core fjes [last unloaded: msr]
[  549.290600] CPU: 24 PID: 2855 Comm: cat Tainted: P           OE   4.4.0-148-generic #174~14.04.1-Ubuntu
[  549.290671] Hardware name: Supermicro SYS-1028GR-TR/X10DRG-H, BIOS 1.0c 05/20/2015
[  549.290738] task: ffff880035c76200 ti: ffff8800360d0000 task.ti: ffff8800360d0000
[  549.290822] RIP: 0010:[<ffff882014800154>]  [<ffff882014800154>] 0xffff882014800154
[  549.290973] RSP: 0018:ffffc9000dd95000  EFLAGS: 00010082
[  549.291052] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
[  549.291135] RDX: 0000000000000000 RSI: ffffc9000dc94000 RDI: ffffc9000da7e000
[  549.291216] RBP: ffffc9000d97d000 R08: 0000000000000000 R09: 0000000000000000
[  549.291298] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
[  549.291380] R13: 0000000000000000 R14: ffffc9000d87c000 R15: 0000000000000000
[  549.291463] FS:  00007fcd8544b740(0000) GS:ffff88103fb00000(0000) knlGS:0000000000000000
[  549.291565] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  549.291644] CR2: 0000000000000000 CR3: 0000000074e4c000 CR4: 0000000000160670
[  549.291727] Stack:
[  549.291799]  2bfb0769a37930ed d9f2d2187f695cf4 6608ba1bcdfd8b9b c91c428ccc92e2d4
[  549.292087]  56e165874f8cdfe7 3aa3acbe97d732e6 e7b87febb7aedf5f ace53722bb6b9da8
[  549.292376]  bd10c7462c766aa9 d1c7497af6056ed5 ee88c304fa7e933c a24bd133228e3be2
[  549.292665] Call Trace:
[  549.292738]  <UNK>
[  549.292790] Code: c1 e2 20 48 09 c2 49 29 57 10 0f ae e8 41 5f 0f ae e8 5a 0f ae e8 59 0f ae e8 58 0f ae e8 3c 81 0f ae e8 9e 0f ae e8 58 0f ae e8 <00> 00 37 1b 20 88 00 00 37 1b 20 88 00 00 37 1b 20 88 00 00 37
[  549.296050] RIP  [<ffff882014800154>] 0xffff882014800154
[  549.296175]  RSP <ffffc9000dd95000>```

[andreas-abel]

Can you also post the dmesg output you get when loading the kernel module? Normally, it should be something like this:

[  107.703159] Initializing nanoBench kernel module...
[  107.703160] Vendor ID: GenuineIntel
[  107.703162] Brand: Intel(R) Xeon(R) CPU E3-1225 v3 @ 3.20GHz
[  107.703163] DisplayFamily_DisplayModel: 06_3CH
[  107.703163] Stepping ID: 3
[  107.703163] Performance monitoring version: 3
[  107.703164] Number of general-purpose performance counters: 8
[  107.703164] Bit widths of general-purpose performance counters: 48

[CharithYMendis]

[  444.369938] Initializing nanoBench kernel module...
[  444.369944] Vendor ID: GenuineIntel
[  444.369950] Brand: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
[  444.369953] DisplayFamily_DisplayModel: 06_3FH
[  444.369955] Stepping ID: 2
[  444.369958] Performance monitoring version: 3
[  444.369960] Number of general-purpose performance counters: 4
[  444.369963] Bit widths of general-purpose performance counters: 48
[  490.387103] NMI watchdog: enabled on all CPUs, permanently consumes one hw-PMU counter.
[  531.507980] NMI watchdog: enabled on all CPUs, permanently consumes one hw-PMU counter.
[  549.285212] BUG: unable to handle kernel NULL pointer dereference at           (null)
[  549.285378] IP: [<ffff882014800154>] 0xffff882014800154
[  549.285487] PGD 80000000359b3067 PUD 359b4067 PMD 0
[  549.285687] Oops: 0002 [#1] SMP

From here on it is what posted before.

[andreas-abel]

I cannot reproduce the problem. I just tested it with a 4.4.0 kernel and it's working fine for me. Maybe you are using some non-default kernel option, or there is an incompatibility with some other kernel module. I would suggest booting from a flash drive with Ubuntu 18.04 and trying again in that environment.