15.1-RELEASE page fault on kldload i915kms (refreshed for 15.1)

[Jan Henrik Sylvester]

Updating from 15.0 to 15.1 on my laptop, I had to rebuild the drm and
gpu-firmware ports to avoid a crash due to i915kms. It works fine now.

On my desktop computer with Coffee Lake CPU, I was not so lucky. Even
after rebuilding all drm and gpu-firmware ports on 15.1-RELEASE, kldload
i915kms from drm-612-kmod (or drm-latest-kmod) produces a page fault.

Needless to say, everything was fine on 15.0.

Any idea?

Jun 19 09:35:44 h kernel: iic0: <I2C generic I/O> on iicbus0
Jun 19 09:35:45 h kernel: [drm] Got Intel graphics stolen memory base
0x7d800000, size 0x2000000
Jun 19 09:35:45 h kernel: drmn0: <drmn> on vgapci0
Jun 19 09:35:45 h kernel: vgapci0: child drmn0 requested pci_enable_io
Jun 19 09:35:45 h syslogd: last message repeated 1 times
Jun 19 09:35:45 h kernel: drmn0: [drm] Found COFFEELAKE (device ID 3e92)
display version 9.00 stepping N/A
Jun 19 09:35:45 h kernel: drmn0: successfully loaded firmware image
'i915/kbl_dmc_ver1_04.bin'
Jun 19 09:35:45 h kernel: drmn0: [drm] Finished loading DMC firmware
i915/kbl_dmc_ver1_04.bin (v1.4)
Jun 19 09:35:45 h kernel: lkpi_iic0: <LinuxKPI I2C> on drmn0
Jun 19 09:35:45 h kernel: iicbus1: <Philips I2C bus> on lkpi_iic0
Jun 19 09:35:45 h kernel: iic1: <I2C generic I/O> on iicbus1
Jun 19 09:35:45 h kernel: lkpi_iic1: <LinuxKPI I2C> on drmn0
Jun 19 09:35:45 h kernel: iicbus2: <Philips I2C bus> on lkpi_iic1
Jun 19 09:35:45 h kernel: iic2: <I2C generic I/O> on iicbus2
Jun 19 09:35:45 h kernel: lkpi_iic2: <LinuxKPI I2C> on drmn0
Jun 19 09:35:45 h kernel: iicbus3: <Philips I2C bus> on lkpi_iic2
Jun 19 09:35:45 h kernel: iic3: <I2C generic I/O> on iicbus3
Jun 19 09:35:45 h kernel: lkpi_iic3: <LinuxKPI I2C> on drmn0
Jun 19 09:35:45 h kernel: iicbus4: <Philips I2C bus> on lkpi_iic3
Jun 19 09:35:45 h kernel: iic4: <I2C generic I/O> on iicbus4
Jun 19 09:35:46 h kernel: drmn0: [drm] [ENCODER:98:DDI A/PHY A] failed
to retrieve link info, disabling eDP
Jun 19 09:35:46 h kernel: sysctl_add_oid: can't re-use a leaf
(hw.dri.debug)!
Jun 19 09:35:46 h kernel: lkpi_iic4: <LinuxKPI I2C> on drm1
Jun 19 09:35:46 h kernel: iicbus5: <Philips I2C bus> on lkpi_iic4
Jun 19 09:35:46 h kernel: iic5: <I2C generic I/O> on iicbus5
Jun 19 09:35:46 h kernel: lkpi_iic5: <LinuxKPI I2C> on drm3
Jun 19 09:35:46 h kernel: iicbus6: <Philips I2C bus> on lkpi_iic5
Jun 19 09:35:46 h kernel: iic6: <I2C generic I/O> on iicbus6
Jun 19 09:35:46 h kernel: lkpi_iic6: <LinuxKPI I2C> on drm5
Jun 19 09:35:46 h kernel: iicbus7: <Philips I2C bus> on lkpi_iic6
Jun 19 09:35:46 h kernel: iic7: <I2C generic I/O> on iicbus7
Jun 19 09:35:46 h kernel: [drm] Initialized i915 1.6.0 for drmn0 on minor 0
Jun 19 09:35:46 h kernel: drmn0: [drm] Cannot find any crtc or sizes
Jun 19 09:35:46 h syslogd: last message repeated 1 times
Jun 19 09:35:46 h kernel: hdacc1: Unexpected unsolicited response with
tag 63: ffffffff
Jun 19 09:36:33 h syslogd: kernel boot file is /boot/kernel/kernel
Jun 19 09:36:33 h kernel: lkpi_iic7: <LinuxKPI I2C> on drmn0
Jun 19 09:36:33 h kernel: iicbus8: <Philips I2C bus> on lkpi_iic7
Jun 19 09:36:33 h kernel: iic8: <I2C generic I/O> on iicbus8
Jun 19 09:36:33 h kernel:
Jun 19 09:36:33 h syslogd: last message repeated 1 times
Jun 19 09:36:33 h kernel: Fatal trap 12: page fault while in kernel mode
Jun 19 09:36:33 h kernel: cpuid = 4; apic id = 08
Jun 19 09:36:33 h kernel: fault virtual address = 0x8
Jun 19 09:36:33 h kernel: fault code = supervisor read data,
page not present
Jun 19 09:36:33 h kernel: instruction pointer = 0x20:0xffffffff856de87e
Jun 19 09:36:33 h kernel: stack pointer = 0x28:0xfffffe00d36d49b0
Jun 19 09:36:33 h kernel: frame pointer = 0x28:0xfffffe00d36d49e0
Jun 19 09:36:33 h kernel: code segment = base 0x0, limit
0xfffff, type 0x1b
Jun 19 09:36:33 h kernel: = DPL 0, pres 1, long
1, def32 0, gran 1
Jun 19 09:36:33 h kernel: processor eflags = interrupt enabled,
resume, IOPL = 0
Jun 19 09:36:33 h kernel: current process = 0
(linuxkpi_long_wq_3)
Jun 19 09:36:33 h kernel: rdi: 0000000000000000 rsi: fffff8015b381200
rdx: 0000000000000008
Jun 19 09:36:33 h kernel: rcx: ffffffff8549cf7f r8: ffffffff8549f3a6
r9: ffffffff854a3100
Jun 19 09:36:33 h kernel: rax: 0000000000000001 rbx: fffff80420243a90
rbp: fffffe00d36d49e0
Jun 19 09:36:33 h kernel: r10: 0000000000000001 r11: fffff80001269000
r12: fffff80179995030
Jun 19 09:36:33 h kernel: r13: 0000000000000000 r14: fffff80420243000
r15: fffff80420243a80
Jun 19 09:36:33 h kernel: trap number = 12
Jun 19 09:36:33 h kernel: panic: page fault
Jun 19 09:36:33 h kernel: cpuid = 4
Jun 19 09:36:33 h kernel: time = 1781854546
Jun 19 09:36:33 h kernel: KDB: stack backtrace:
Jun 19 09:36:33 h kernel: #0 0xffffffff80bd609d at kdb_backtrace+0x5d
Jun 19 09:36:33 h kernel: #1 0xffffffff80b86c06 at vpanic+0x136
Jun 19 09:36:33 h kernel: #2 0xffffffff80b86ac3 at panic+0x43
Jun 19 09:36:33 h kernel: #3 0xffffffff81097fcd at trap_pfault+0x37d
Jun 19 09:36:33 h kernel: #4 0xffffffff8106e568 at calltrap+0x8
Jun 19 09:36:33 h kernel: #5 0xffffffff856f0097 at
intel_dp_add_mst_connector+0xb7
Jun 19 09:36:33 h kernel: #6 0xffffffff85486ff5 at
drm_dp_mst_port_add_connector+0x125
Jun 19 09:36:33 h kernel: #7 0xffffffff8548661c at
drm_dp_send_link_address+0x9bc
Jun 19 09:36:33 h kernel: #8 0xffffffff85485bd2 at
drm_dp_check_and_send_link_address+0x22
Jun 19 09:36:33 h kernel: #9 0xffffffff8548335e at
drm_dp_mst_link_probe_work+0x1fe
Jun 19 09:36:33 h kernel: #10 0xffffffff80e25594 at linux_work_fn+0xe4
Jun 19 09:36:33 h kernel: #11 0xffffffff80becb92 at
taskqueue_run_locked+0x182
Jun 19 09:36:33 h kernel: #12 0xffffffff80bedd62 at
taskqueue_thread_loop+0xc2
Jun 19 09:36:33 h kernel: #13 0xffffffff80b3d41b at fork_exit+0x7b
Jun 19 09:36:33 h kernel: #14 0xffffffff8106f58e at fork_trampoline+0xe
Jun 19 09:36:33 h kernel: Uptime: 3d20h0m11s
Jun 19 09:36:33 h kernel: Automatic reboot in 15 seconds - press a key
on the console to abort

Best,
Jan Henrik

[Anton Shepelev]

Jan Henrik Sylvester:

> On my desktop computer with Coffee Lake CPU, I was not
> so lucky. Even after rebuilding all drm and
> gpu-firmware ports on 15.1-RELEASE, kldload i915kms from
> drm-612-kmod (or drm-latest-kmod) produces a page
> fault.
>
> Needless to say, everything was fine on 15.0.

In my experience, the issue tracker at

<https://bugs.freebsd.org>

is the place to report such issues as yours. There, the
actual maintainers of the respective packages and
components have a higher chance of seeing and acting on
bug reports.