Re: perf: perf_fuzzer still triggers bts warning

[Vince Weaver]

Probably not related but it also triggered this just now:

static void x86_pmu_start(struct perf_event *event, int flags)
{
        struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
        int idx = event->hw.idx;

        if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
                return;


[ 7501.295830] ------------[ cut here ]------------
[ 7501.295844] WARNING: CPU: 2 PID: 7166 at arch/x86/events/core.c:1236 x86_pmu_start+0xae/0x100
[ 7501.295948] CPU: 2 PID: 7166 Comm: perf_fuzzer Tainted: G        W       4.8.0-rc6+ #5
[ 7501.295950] Hardware name: LENOVO 10FY0017US/SKYBAY, BIOS FWKT53A   06/06/2016
[ 7501.295952]  0000000000000086 000000006d6df8e9 ffff9045fdc83d70 ffffffff93b3b863
[ 7501.295958]  0000000000000000 0000000000000000 ffff9045fdc83db0 ffffffff9387f7d1
[ 7501.295963]  000004d4ee352100 ffff9045fdc8a480 ffff9045f0d18800 0000000000000001
[ 7501.295968] Call Trace:
[ 7501.295970]  <IRQ>  [<ffffffff93b3b863>] dump_stack+0x63/0x90
[ 7501.295985]  [<ffffffff9387f7d1>] __warn+0xd1/0xf0
[ 7501.295989]  [<ffffffff9387f8fd>] warn_slowpath_null+0x1d/0x20
[ 7501.295994]  [<ffffffff93806f9e>] x86_pmu_start+0xae/0x100
[ 7501.295999]  [<ffffffff93977050>] perf_event_task_tick+0x1f0/0x2c0
[ 7501.296004]  [<ffffffff938abbc8>] scheduler_tick+0x78/0xd0
[ 7501.296008]  [<ffffffff938fe8b0>] ? tick_sched_do_timer+0x30/0x30
[ 7501.296014]  [<ffffffff938ee8a7>] update_process_times+0x47/0x60
[ 7501.296017]  [<ffffffff938fe285>] tick_sched_handle.isra.13+0x25/0x60
[ 7501.296020]  [<ffffffff938fe8ed>] tick_sched_timer+0x3d/0x70
[ 7501.296025]  [<ffffffff938ef633>] __hrtimer_run_queues+0xf3/0x280
[ 7501.296030]  [<ffffffff938efb08>] hrtimer_interrupt+0xa8/0x1a0
[ 7501.296035]  [<ffffffff93852d88>] local_apic_timer_interrupt+0x38/0x60
[ 7501.296040]  [<ffffffff93e1de8d>] smp_apic_timer_interrupt+0x3d/0x50
[ 7501.296045]  [<ffffffff93e1d1a2>] apic_timer_interrupt+0x82/0x90
[ 7501.296046]  <EOI>  [<ffffffff939bc799>] ? vm_normal_page+0x9/0xa0
[ 7501.296057]  [<ffffffff939bd637>] ? unmap_page_range+0x557/0x930
[ 7501.296062]  [<ffffffff939bda8d>] unmap_single_vma+0x7d/0xe0
[ 7501.296066]  [<ffffffff939bddda>] unmap_vmas+0x4a/0xa0
[ 7501.296070]  [<ffffffff939c6687>] exit_mmap+0xa7/0x170
[ 7501.296075]  [<ffffffff9387c962>] mmput+0x62/0xf0
[ 7501.296078]  [<ffffffff93883739>] do_exit+0x339/0xb60
[ 7501.296082]  [<ffffffff93883fe3>] do_group_exit+0x43/0xc0
[ 7501.296085]  [<ffffffff9388f4f8>] get_signal+0x2b8/0x6b0
[ 7501.296090]  [<ffffffff9382d5d7>] do_signal+0x37/0x7c0
[ 7501.296094]  [<ffffffff939c0b2c>] ? handle_mm_fault+0xb0c/0x15d0
[ 7501.296099]  [<ffffffff938373d9>] ? sched_clock+0x9/0x10
[ 7501.296102]  [<ffffffff938373d9>] ? sched_clock+0x9/0x10
[ 7501.296107]  [<ffffffff93831761>] ? nmi_handle+0x71/0x120
[ 7501.296111]  [<ffffffff9380358c>] exit_to_usermode_loop+0x8c/0xd0
[ 7501.296115]  [<ffffffff93803a36>] prepare_exit_to_usermode+0x26/0x30
[ 7501.296118]  [<ffffffff93e1bee5>] retint_user+0x8/0x13
[ 7501.296121] ---[ end trace 6247e022342bacf8 ]---