------- Comment From [email protected] 2018-02-27 00:25 EDT------- Hi , Unable to hit the above call trace , hence the issue is resolved.
Verified on kernel : root@ltc-wcwsp1:~/shriya/kernel# uname -a Linux ltc-wcwsp1 4.15.0-10-generic #11 SMP Fri Feb 23 01:59:38 EST 2018 ppc64le ppc64le ppc64le GNU/Linux Perf fuzzer : ======= + ./perf_fuzzer -r 1492143527 Using user-specified random seed of 1492143527 *** perf_fuzzer 0.32-rc0 *** by Vince Weaver Linux version 4.15.0-10-generic ppc64le Processor: ppc64le UNKNOWN Watchdog enabled with timeout 60s Will auto-exit if signal storm detected Seeding RNG with supplied seed 1492143527 To reproduce, try: echo 0 > /proc/sys/kernel/nmi_watchdog echo 2 > /proc/sys/kernel/perf_event_paranoid echo 100000 > /proc/sys/kernel/perf_event_max_sample_rate ./perf_fuzzer -r 1492143527 Fuzzing the following syscalls: mmap perf_event_open close read write ioctl fork prctl poll Also attempting the following: signal-handler-on-overflow busy-instruction-loop accessing-perf-proc-and-sys-files trashing-the-mmap-page Pid=4506, sleeping 1s ================================================== Starting fuzzing at 2018-02-26 21:36:45 ================================================== Cannot open /sys/kernel/tracing/kprobe_events Iteration 10000, 98728 syscalls in 5.32 s (18.554 k syscalls/s) Open attempts: 90726 Successful: 917 Currently open: 54 ENOENT : 446 E2BIG : 7816 EBADF : 8522 EINVAL : 72718 ENOSPC : 81 EOVERFLOW : 2 EOPNOTSUPP : 224 Trinity Type (Normal 345/22780)(Sampling 28/22630)(Global 488/22586)(Random 56/22730) Type (Hardware 246/12690)(software 362/11897)(tracepoint 58/11874)(Cache 55/11258)(cpu 136/11638)(breakpoint 33/11690)(nest_capp0_imc 1/357)(nest_capp1_imc 1/365)(core_imc 0/460)(nest_mba0_imc 1/378)(nest_mba1_imc 0/352)(nest_mba2_imc 0/362)(nest_mba3_imc 1/371)(nest_mba4_imc 1/357)(nest_mba5_imc 0/320)(nest_mba6_imc 1/358)(nest_mba7_imc 0/439)(nest_mcs01_imc 0/359)(nest_mcs23_imc 2/361)(nest_nvlink0_imc 19/14840) Close: 863/863 Successful Read: 824/912 Successful Write: 0/855 Successful Ioctl: 371/924 Successful: (ENABLE 91/91)(DISABLE 81/81)(REFRESH 9/78)(RESET 84/84)(PERIOD 2/75)(SET_OUTPUT 8/80)(SET_FILTER 0/92)(ID 87/87)(SET_BPF 0/91)(PAUSE_OUTPUT 9/79)(#10 0/0)(#11 0/0)(#12 0/0)(#13 0/0)(#14 0/0)(>14 0/86) Mmap: 662/1060 Successful: (MMAP 662/1060)(TRASH 144/156)(READ 133/143)(UNMAP 653/988)(AUX 0/169)(AUX_READ 0/0) Prctl: 929/929 Successful Fork: 471/471 Successful Poll: 895/921 Successful Access: 0/0 Successful Overflows: 0 Recursive: 0 SIGIOs due to RT signal queue full: 0 Memset of mmap#18 at (nil) caused segfault! MAP_SHARED MAP_DENYWRITE MAP_FIXED MAP_NONBLOCK MAP_STACK PROT_READ PROT_WRITE size 27661 fd 47 event: cpu=4 pid=-1 group_fd=-1 flags=9 memset(&pe[47],0,sizeof(struct perf_event_attr)); pe[47].type=PERF_TYPE_SOFTWARE; pe[47].size=64; pe[47].config=PERF_COUNT_SW_PAGE_FAULTS; pe[47].sample_type=0; /* 0 */ pe[47].read_format=PERF_FORMAT_TOTAL_TIME_RUNNING|PERF_FORMAT_GROUP; /* a */ pe[47].disabled=1; pe[47].inherit=1; pe[47].pinned=1; pe[47].precise_ip=0; /* arbitrary skid */ pe[47].wakeup_events=0; pe[47].bp_type=HW_BREAKPOINT_EMPTY; -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1746225 Title: [P9,Power NV][WSP][Ubuntu 1804] : "Kernel access of bad area " when grouping different pmu events using perf fuzzer . (perf:) Status in The Ubuntu-power-systems project: In Progress Status in linux package in Ubuntu: Fix Committed Status in linux source package in Artful: In Progress Status in linux source package in Bionic: Fix Committed Bug description: == SRU Justification == Due to this bug, perf fuzzer resulted in crash and system goes for a reboot and results in a call trace shown in the bug. It is due to grouping of different PMU events, which is fixed by mainline commit 5aa04b3eb6fca63d2e9827be656dcadc26d54e1 Commit 5aa04b3eb6fca63d2e9827be656dcadc26d54e11 is in mailine as of v4.15-rc5. == Fix == commit 5aa04b3eb6fca63d2e9827be656dcadc26d54e11 Author: Ravi Bangoria <[email protected]> Date: Thu Nov 30 14:03:22 2017 +0530 powerpc/perf: Fix oops when grouping different pmu events == Regression Potential == Low. This fix is specific to powerpc. == Test Case == A test kernel was built with this patch and tested by the original bug reporter. The bug reporter states the test kernel resolved the bug. ==Original Bug Description== == Comment: #0 - Shriya R. Kulkarni <[email protected]> - 2018-01-30 03:24:47 == Problem Description : ============== Perf fuzzer resulted in crash and system goes for reboot and the call trace is shown below . It is due to grouping of different PMU events. Machine details : ========== OS : Ubuntu 1804 uname -r : 4.13.0-25-generic system : Witherspoon + DD2.1 perf -v : perf version 4.13.13 ltc-wspoon12 login: [78592.995848] Unable to handle kernel paging request for instruction fetch [78592.995914] Faulting instruction address: 0x00000000 [78592.995950] Oops: Kernel access of bad area, sig: 11 [#1] [78592.995982] SMP NR_CPUS=2048 [78592.995985] NUMA [78592.996011] PowerNV [78592.996045] Modules linked in: vmx_crypto idt_89hpesx crct10dif_vpmsum at24 ofpart uio_pdrv_genirq uio cmdlinepart powernv_flash mtd ibmpowernv opal_prd ipmi_powernv ipmi_devintf ipmi_msghandler sch_fq_codel ip_tables x_tables autofs4 nouveau lpfc ast i2c_algo_bit crc32c_vpmsum ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm mlx5_core nvmet_fc nvmet tg3 nvme_fc nvme_fabrics ahci nvme_core libahci mlxfw devlink scsi_transport_fc [78592.996367] CPU: 69 PID: 6010 Comm: perf_fuzzer Tainted: G W 4.13.0-25-generic #29-Ubuntu [78592.996422] task: c000003f77b5b500 task.stack: c000003d0b0c8000 [78592.996462] NIP: 0000000000000000 LR: c0000000000e9b1c CTR: 0000000000000000 [78592.996509] REGS: c000003d0b0cb780 TRAP: 0400 Tainted: G W (4.13.0-25-generic) [78592.996562] MSR: 9000000040009033 <SF,HV,EE,ME,IR,DR,RI,LE> [78592.996588] CR: 48002874 XER: 00000000 [78592.996642] CFAR: c0000000000e9b18 SOFTE: 1 [78592.996642] GPR00: c0000000000eb128 c000003d0b0cba00 c0000000015f6200 0000000000000000 [78592.996642] GPR04: c000003d0b0cbba0 c000003d0b0cbc20 0000000000000002 c000000001596b10 [78592.996642] GPR08: 0000000000000002 0000000000000000 c000000001596b10 c000003fecad0028 [78592.996642] GPR12: 0000000000000000 c000000007a8d480 0000000000000000 0000000000000000 [78592.996642] GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 [78592.996642] GPR20: 0000000000000001 c000003d0b0cbc1c c000003d0b0cbc24 c000003d0b0cbb98 [78592.996642] GPR24: c000003d0b0cbba0 c000003d0b0cbc20 0000000000001555 c000003fefeb4ea0 [78592.996642] GPR28: c000003d0b0cbc20 0000000000000002 0000000000003000 c000003fefeb5190 [78592.997170] NIP [0000000000000000] (null) [78592.997208] LR [c0000000000e9b1c] power_check_constraints+0x13c/0x5a0 [78592.997247] Call Trace: [78592.997267] [c000003d0b0cba00] [c000003d0b0cbaa0] 0xc000003d0b0cbaa0 (unreliable) [78592.997321] [c000003d0b0cbb80] [c0000000000eb128] power_pmu_event_init+0x298/0x6a0 [78592.997373] [c000003d0b0cbc70] [c00000000029e6b4] perf_try_init_event+0xd4/0x120 [78592.997424] [c000003d0b0cbcb0] [c0000000002a1038] perf_event_alloc.part.23+0x7b8/0xb90 [78592.997475] [c000003d0b0cbd30] [c0000000002aa0dc] SyS_perf_event_open+0x69c/0xfa0 [78592.997527] [c000003d0b0cbe30] [c00000000000b184] system_call+0x58/0x6c [78592.997568] Instruction dump: [78592.997597] XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX [78592.997664] XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX [78592.997733] ---[ end trace 57fb7542c4083583 ]--- [78594.008780] [78594.008932] Sending IP[78773.335857584,5] OPAL: Switch to big-endian OS I to other CPUs [78594.01029 Steps to reproduce : ============ #! /bin/bash set -x git clone https://github.com/deater/perf_event_tests.git cd perf_event_tests/include mkdir asm cd asm wget http://9.114.13.132/repo/shriya/perf_regs.h cd ../../lib make sleep 10 cd ../fuzzer make sleep 10 echo 0 > /proc/sys/kernel/nmi_watchdog echo 2 > /proc/sys/kernel/perf_event_paranoid echo 100000 > /proc/sys/kernel/perf_event_max_sample_rate ./perf_fuzzer -r 1492143527 To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu-power-systems/+bug/1746225/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : [email protected] Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
