4K screen doesn't work with 60Hz on Intel graphics

Message

robert7k · Post by **robert7k** » Tue Jul 05, 2016 8:10 pm

Hi there,

I've got a Thinkpad T460p (with Intel Graphics) and a LG 27MU67-B screen, which supports a resolution up to 3840x2160.

My problem is that I only can get the resolution 3840x2160 to work @ 30Hz. As soon as I select 60Hz the screen stays blank. 1920x1080 works @60Hz.

However, xrandr recognizes that the screen supports 60Hz:

Code: Select all

DP3-1 connected 3840x2160+2560+0 (normal left inverted right x axis y axis) 600mm x 340mm
   3840x2160     60.00 +  30.00* 
   1920x1080     60.00    59.94  
   1600x900      60.00  
   1280x1024     60.02  
   1280x800      59.81  
   1152x864      59.97  
   1280x720      60.00    59.94  
   1024x768      60.00  
   800x600       60.32  
   720x480       60.00    59.94  
   640x480       60.00    59.94

I tried different cables and ports:

miniDisplayPort on the side of the laptop
DisplayPort on the UltraDock docking station
HDMI port on the laptop
HDMI port on the UltraDock

I used cables which came with the screen, so this shouldn't be an issue.

I'm using Gentoo with kernel gentoo-sources-4.6.0 and xf86-video-intel-2.99.917_p20160621-r1

Any ideas or suggestions how I could get 60Hz to work?

Thanks!

chithanh · Post by **chithanh** » Tue Jul 05, 2016 9:27 pm

Which resolution do you use on the console?

For 3840x2160@60 Hz, the DisplayPort needs to negotiate HBR2 transfer mode. Maybe that failed for some reason. If dmesg doesn't contain any hints, you can use the drm.debug=... kernel parameter to increase debug verbosity.

robert7k · Post by **robert7k** » Tue Jul 05, 2016 9:50 pm

chithanh wrote:Which resolution do you use on the console?

Console is at 1280x1024, but that was selected automatically.

chithanh wrote: For 3840x2160@60 Hz, the DisplayPort needs to negotiate HBR2 transfer mode. Maybe that failed for some reason. If dmesg doesn't contain any hints, you can use the drm.debug=... kernel parameter to increase debug verbosity.

Good hint, I didn't check dmesg so far, but found some errors which might be related to my issue:

Code: Select all

[    2.303018] ------------[ cut here ]------------
[    2.303021] WARNING: CPU: 3 PID: 59 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120
[    2.303022] WARN_ON(!wm_changed)
[    2.303022] Modules linked in:
[    2.303023] CPU: 3 PID: 59 Comm: kworker/3:1 Not tainted 4.6.0-gentoo #1
[    2.303024] Hardware name: LENOVO 20FWCTO1WW/20FWCTO1WW, BIOS R07ET63W (2.03 ) 03/15/2016
[    2.303026] Workqueue: events_long drm_dp_mst_link_probe_work
[    2.303027]  0000000000000000 ffff88042faff8f8 ffffffff81299908 ffff88042faff948
[    2.303028]  0000000000000000 ffff88042faff938 ffffffff81052e0c 00000e0200000001
[    2.303029]  ffff88042fc22000 ffff88042fc23000 ffff88042fb09d6c ffff88042f8baba8
[    2.303029] Call Trace:
[    2.303032]  [<ffffffff81299908>] dump_stack+0x4d/0x65
[    2.303033]  [<ffffffff81052e0c>] __warn+0xcc/0xf0
[    2.303035]  [<ffffffff81052e7a>] warn_slowpath_fmt+0x4a/0x50
[    2.303036]  [<ffffffff813aed04>] skl_update_other_pipe_wm+0x114/0x120
[    2.303037]  [<ffffffff813aee49>] skl_update_wm+0x139/0x5d0
[    2.303039]  [<ffffffff814358de>] ? intel_ddi_enable_transcoder_func+0x16e/0x240
[    2.303040]  [<ffffffff813b0289>] intel_update_watermarks+0x19/0x20
[    2.303041]  [<ffffffff81419d6c>] haswell_crtc_enable+0x31c/0x890
[    2.303043]  [<ffffffff81406019>] ? intel_finish_crtc_commit+0x9/0x10
[    2.303044]  [<ffffffff813747d0>] ? drm_atomic_helper_commit_planes_on_crtc+0x150/0x270
[    2.303045]  [<ffffffff814159d8>] intel_atomic_commit+0x628/0x1570
[    2.303047]  [<ffffffff81397602>] ? drm_atomic_add_affected_connectors+0x22/0xf0
[    2.303048]  [<ffffffff81397122>] drm_atomic_commit+0x32/0x50
[    2.303048]  [<ffffffff813756b2>] restore_fbdev_mode+0x232/0x260
[    2.303049]  [<ffffffff81395ff5>] ? drm_modeset_lock_all_ctx+0x95/0xb0
[    2.303050]  [<ffffffff8137786e>] drm_fb_helper_restore_fbdev_mode_unlocked+0x2e/0x70
[    2.303051]  [<ffffffff813778d8>] drm_fb_helper_set_par+0x28/0x50
[    2.303052]  [<ffffffff813777ed>] drm_fb_helper_hotplug_event+0x9d/0xf0
[    2.303053]  [<ffffffff8142d169>] intel_fbdev_output_poll_changed+0x19/0x20
[    2.303054]  [<ffffffff8136b3f2>] drm_kms_helper_hotplug_event+0x22/0x30
[    2.303055]  [<ffffffff81437300>] intel_dp_mst_hotplug+0x10/0x20
[    2.303056]  [<ffffffff8137100d>] drm_dp_send_link_address+0x18d/0x210
[    2.303057]  [<ffffffff8108608b>] ? pick_next_task_fair+0x12b/0x470
[    2.303058]  [<ffffffff81371138>] drm_dp_check_and_send_link_address+0xa8/0xb0
[    2.303060]  [<ffffffff81371193>] drm_dp_mst_link_probe_work+0x53/0xb0
[    2.303061]  [<ffffffff8106a146>] process_one_work+0x146/0x3f0
[    2.303061]  [<ffffffff8106a8b6>] worker_thread+0x46/0x470
[    2.303062]  [<ffffffff8106a870>] ? max_active_store+0x60/0x60
[    2.303063]  [<ffffffff8106a870>] ? max_active_store+0x60/0x60
[    2.303064]  [<ffffffff8106f434>] kthread+0xc4/0xe0
[    2.303065]  [<ffffffff816c0752>] ret_from_fork+0x22/0x40
[    2.303066]  [<ffffffff8106f370>] ? kthread_park+0x50/0x50
[    2.303067] ---[ end trace d6237fc6e26cda1f ]---

...

[    3.246568] [drm:intel_mst_disable_dp] *ERROR* failed to update payload -22
[    3.263300] ------------[ cut here ]------------
[    3.263306] WARNING: CPU: 1 PID: 60 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120
[    3.263314] WARN_ON(!wm_changed)
[    3.263316] Modules linked in:
[    3.263318] CPU: 1 PID: 60 Comm: kworker/1:1 Tainted: G        W       4.6.0-gentoo #1
[    3.263319] Hardware name: LENOVO 20FWCTO1WW/20FWCTO1WW, BIOS R07ET63W (2.03 ) 03/15/2016
[    3.263322] Workqueue: events drm_dp_destroy_connector_work
[    3.263323]  0000000000000000 ffff88042fabba60 ffffffff81299908 ffff88042fabbab0
[    3.263324]  0000000000000000 ffff88042fabbaa0 ffffffff81052e0c 00000e0200000000
[    3.263325]  ffff88042fc22000 ffff88042fc23000 ffff88042fb09d6c ffff88042f8baba8
[    3.263325] Call Trace:
[    3.263328]  [<ffffffff81299908>] dump_stack+0x4d/0x65
[    3.263330]  [<ffffffff81052e0c>] __warn+0xcc/0xf0
[    3.263331]  [<ffffffff81052e7a>] warn_slowpath_fmt+0x4a/0x50
[    3.263333]  [<ffffffff813aed04>] skl_update_other_pipe_wm+0x114/0x120
[    3.263334]  [<ffffffff813aee49>] skl_update_wm+0x139/0x5d0
[    3.263335]  [<ffffffff813b806b>] ? intel_display_power_put+0xcb/0x130
[    3.263337]  [<ffffffff813b0289>] intel_update_watermarks+0x19/0x20
[    3.263338]  [<ffffffff81415822>] intel_atomic_commit+0x472/0x1570
[    3.263341]  [<ffffffff8139647a>] ? drm_atomic_set_crtc_for_connector+0x2a/0xe0
[    3.263342]  [<ffffffff81397122>] drm_atomic_commit+0x32/0x50
[    3.263343]  [<ffffffff81374fc6>] drm_atomic_helper_set_config+0x76/0xb0
[    3.263345]  [<ffffffff814373a1>] intel_dp_destroy_mst_connector+0x51/0xe0
[    3.263346]  [<ffffffff81370226>] drm_dp_destroy_connector_work+0xa6/0x150
[    3.263348]  [<ffffffff8106a146>] process_one_work+0x146/0x3f0
[    3.263348]  [<ffffffff8106a8b6>] worker_thread+0x46/0x470
[    3.263349]  [<ffffffff8106a870>] ? max_active_store+0x60/0x60
[    3.263350]  [<ffffffff8106f434>] kthread+0xc4/0xe0
[    3.263352]  [<ffffffff816c0752>] ret_from_fork+0x22/0x40
[    3.263353]  [<ffffffff8106f370>] ? kthread_park+0x50/0x50
[    3.263354] ---[ end trace d6237fc6e26cda20 ]---

...

[    3.929558] ------------[ cut here ]------------
[    3.929563] WARNING: CPU: 3 PID: 24 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120
[    3.929564] WARN_ON(!wm_changed)
[    3.929565] Modules linked in:
[    3.929566]  arc4 rtsx_pci_sdmmc mmc_core iwlmvm kvm_intel mac80211 kvm irqbypass input_leds e1000e iwlwifi mei_me mei cfg80211 rtsx_pci mfd_core thinkpad_acpi led_class
[    3.929574] CPU: 3 PID: 24 Comm: kworker/3:0 Tainted: G        W       4.6.0-gentoo #1
[    3.929575] Hardware name: LENOVO 20FWCTO1WW/20FWCTO1WW, BIOS R07ET63W (2.03 ) 03/15/2016
[    3.929578] Workqueue: events_long drm_dp_mst_link_probe_work
[    3.929579]  0000000000000000 ffff8804302ef8f8 ffffffff81299908 ffff8804302ef948
[    3.929580]  0000000000000000 ffff8804302ef938 ffffffff81052e0c 00000e0200000001
[    3.929582]  ffff88042fc22000 ffff88042fc23000 ffff88042fb09d6c ffff88042f8baba8
[    3.929583] Call Trace:
[    3.929586]  [<ffffffff81299908>] dump_stack+0x4d/0x65
[    3.929601]  [<ffffffff81052e0c>] __warn+0xcc/0xf0
[    3.929603]  [<ffffffff81052e7a>] warn_slowpath_fmt+0x4a/0x50
[    3.929604]  [<ffffffff813aed04>] skl_update_other_pipe_wm+0x114/0x120
[    3.929606]  [<ffffffff813aee49>] skl_update_wm+0x139/0x5d0
[    3.929608]  [<ffffffff814358de>] ? intel_ddi_enable_transcoder_func+0x16e/0x240
[    3.929610]  [<ffffffff813b0289>] intel_update_watermarks+0x19/0x20
[    3.929611]  [<ffffffff81419d6c>] haswell_crtc_enable+0x31c/0x890
[    3.929613]  [<ffffffff81406019>] ? intel_finish_crtc_commit+0x9/0x10
[    3.929615]  [<ffffffff813747d0>] ? drm_atomic_helper_commit_planes_on_crtc+0x150/0x270
[    3.929616]  [<ffffffff814159d8>] intel_atomic_commit+0x628/0x1570
[    3.929618]  [<ffffffff81397602>] ? drm_atomic_add_affected_connectors+0x22/0xf0
[    3.929620]  [<ffffffff81397122>] drm_atomic_commit+0x32/0x50
[    3.929621]  [<ffffffff813756b2>] restore_fbdev_mode+0x232/0x260
[    3.929622]  [<ffffffff81395ff5>] ? drm_modeset_lock_all_ctx+0x95/0xb0
[    3.929623]  [<ffffffff8137786e>] drm_fb_helper_restore_fbdev_mode_unlocked+0x2e/0x70
[    3.929624]  [<ffffffff813778d8>] drm_fb_helper_set_par+0x28/0x50
[    3.929625]  [<ffffffff813777ed>] drm_fb_helper_hotplug_event+0x9d/0xf0
[    3.929626]  [<ffffffff8142d169>] intel_fbdev_output_poll_changed+0x19/0x20
[    3.929628]  [<ffffffff8136b3f2>] drm_kms_helper_hotplug_event+0x22/0x30
[    3.929629]  [<ffffffff81437300>] intel_dp_mst_hotplug+0x10/0x20
[    3.929631]  [<ffffffff8137100d>] drm_dp_send_link_address+0x18d/0x210
[    3.929633]  [<ffffffff8108635b>] ? pick_next_task_fair+0x3fb/0x470
[    3.929634]  [<ffffffff81371138>] drm_dp_check_and_send_link_address+0xa8/0xb0
[    3.929636]  [<ffffffff81371193>] drm_dp_mst_link_probe_work+0x53/0xb0
[    3.929638]  [<ffffffff8106a146>] process_one_work+0x146/0x3f0
[    3.929639]  [<ffffffff8106a8b6>] worker_thread+0x46/0x470
[    3.929640]  [<ffffffff8106a870>] ? max_active_store+0x60/0x60
[    3.929641]  [<ffffffff8106f434>] kthread+0xc4/0xe0
[    3.929643]  [<ffffffff816c0752>] ret_from_fork+0x22/0x40
[    3.929644]  [<ffffffff8106f370>] ? kthread_park+0x50/0x50
[    3.929645] ---[ end trace d6237fc6e26cda21 ]---

Any idea how I can fix this?

chithanh · Post by **chithanh** » Tue Jul 05, 2016 10:07 pm

You do have the Skylake firmware installed and it is loaded properly during boot, yes?

These messages are possibly unrelated to your problem, see https://bugs.freedesktop.org/show_bug.cgi?id=95521

You could try to force the console to higher resolution with the video=3840x2160 kernel parameter, but that might fail too. You could try a newer kernel (4.7-rc6) and if that still doesn't work report a bug at https://bugs.freedesktop.org/

robert7k · Post by **robert7k** » Tue Jul 05, 2016 10:25 pm

chithanh wrote:You do have the Skylake firmware installed and it is loaded properly during boot, yes?

I've got sys-kernel/linux-firmware-20160331 installed. However, I found this message in dmesg:

Code: Select all

[    0.360307] i915 0000:00:02.0: enabling device (0006 -> 0007)
[    0.371619] i915 0000:00:02.0: Direct firmware load for i915/skl_dmc_ver1.bin failed with error -2
[    0.371623] i915 0000:00:02.0: Failed to load DMC firmware [https://01.org/linuxgraphics/intel-linux-graphics-firmwares], disabling runtime power management.
[    0.387539] [drm] Initialized i915 1.6.0 20160229 for 0000:00:02.0 on minor 0
[    2.303021] WARNING: CPU: 3 PID: 59 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120
[    2.634293] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
[    3.263306] WARNING: CPU: 1 PID: 60 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120
[    3.929563] WARNING: CPU: 3 PID: 24 at drivers/gpu/drm/i915/intel_pm.c:3586 skl_update_other_pipe_wm+0x114/0x120

Edit: I guess I haven't set the kernel parameters CONFIG_EXTRA_FIRMWARE and CONFIG_EXTRA_FIRMWARE_DIR. I just changed that and will retry with a newly built kernel.

chithanh · Post by **chithanh** » Tue Jul 05, 2016 10:39 pm

If you have the i915 driver built into your kernel, then you need the firmware built-in too.

You can set the following in your kernel config:

Code: Select all

CONFIG_EXTRA_FIRMWARE="i915/skl_dmc_ver1.bin i915/skl_guc_ver1.bin"
CONFIG_EXTRA_FIRMWARE_DIR="/lib/firmware"

(if dmesg still complains about missing firmware files, you need to add them too)

robert7k · Post by **robert7k** » Tue Jul 05, 2016 10:51 pm

Thanks for your help, chithanh!

I recompiled the kernel with the firmware, it now seems to load fine:

Code: Select all

[    0.377983] [drm] Finished loading i915/skl_dmc_ver1.bin (v1.26)
[    0.391256] [drm] Initialized i915 1.6.0 20160229 for 0000:00:02.0 on minor 0

However, I still cannot switch into 60Hz mode. This is the error which appeared when calling xrandr to set 60Hz:

Code: Select all

[  119.231697] ------------[ cut here ]------------
[  119.231712] WARNING: CPU: 0 PID: 1617 at drivers/gpu/drm/i915/intel_display.c:13929 skl_max_scale.part.100+0x6a/0x80
[  119.231716] WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)
[  119.231718] Modules linked in:
[  119.231722]  rfcomm acpi_call(O) bnep twofish_x86_64_3way lrw glue_helper twofish_x86_64 twofish_common dm_crypt dm_mod loop fuse btusb btrtl btbcm btintel bluetooth arc4 iwlmvm rtsx_pci_sdmmc mac80211 kvm_intel mmc_core kvm irqbypass iwlwifi input_leds cfg80211 rtsx_pci mfd_core mei_me mei thinkpad_acpi led_class
[  119.231768] CPU: 0 PID: 1617 Comm: X Tainted: G        W  O    4.6.0-gentoo #3
[  119.231771] Hardware name: LENOVO 20FWCTO1WW/20FWCTO1WW, BIOS R07ET63W (2.03 ) 03/15/2016
[  119.231775]  0000000000000000 ffff8800ad6e7a30 ffffffff81299908 ffff8800ad6e7a80
[  119.231781]  0000000000000000 ffff8800ad6e7a70 ffffffff81052e0c 0000366900000870
[  119.231787]  ffff88042bc41900 ffff88042fa83800 ffff880431c36000 ffff88042fa83800
[  119.231792] Call Trace:
[  119.231801]  [<ffffffff81299908>] dump_stack+0x4d/0x65
[  119.231809]  [<ffffffff81052e0c>] __warn+0xcc/0xf0
[  119.231816]  [<ffffffff81052e7a>] warn_slowpath_fmt+0x4a/0x50
[  119.231820]  [<ffffffff8140af7a>] skl_max_scale.part.100+0x6a/0x80
[  119.231824]  [<ffffffff8140b04b>] intel_check_primary_plane+0xbb/0xd0
[  119.231830]  [<ffffffff8143cbe8>] ? intel_dp_source_rates+0x48/0x80
[  119.231834]  [<ffffffff8143cc66>] ? intel_dp_common_rates+0x46/0xd0
[  119.231841]  [<ffffffff813fddd9>] intel_plane_atomic_check+0x129/0x1f0
[  119.231849]  [<ffffffff813737f3>] drm_atomic_helper_check_planes+0x43/0x1d0
[  119.231855]  [<ffffffff81417606>] intel_atomic_check+0x2f6/0x1180
[  119.231861]  [<ffffffff816dbd01>] ? __ww_mutex_lock_slowpath+0x111/0x1b0
[  119.231868]  [<ffffffff81396bee>] drm_atomic_check_only+0x13e/0x640
[  119.231873]  [<ffffffff813964a2>] ? drm_atomic_set_crtc_for_connector+0x52/0xe0
[  119.231878]  [<ffffffff81397102>] drm_atomic_commit+0x12/0x50
[  119.231885]  [<ffffffff81374fc6>] drm_atomic_helper_set_config+0x76/0xb0
[  119.231890]  [<ffffffff81387d1d>] drm_mode_set_config_internal+0x5d/0xf0
[  119.231896]  [<ffffffff8138b553>] drm_mode_setcrtc+0x183/0x4e0
[  119.231901]  [<ffffffff8137d67d>] drm_ioctl+0x13d/0x560
[  119.231907]  [<ffffffff8138b3d0>] ? drm_mode_setplane+0x1b0/0x1b0
[  119.231915]  [<ffffffff81185ba9>] do_vfs_ioctl+0x89/0x580
[  119.231920]  [<ffffffff8105f08d>] ? __set_task_blocked+0x2d/0x70
[  119.231926]  [<ffffffff81175a6c>] ? __sb_end_write+0x1c/0x20
[  119.231931]  [<ffffffff81061621>] ? __set_current_blocked+0x31/0x50
[  119.231937]  [<ffffffff811860dc>] SyS_ioctl+0x3c/0x70
[  119.231941]  [<ffffffff81061881>] ? SyS_rt_sigprocmask+0x81/0xa0
[  119.231947]  [<ffffffff816de09b>] entry_SYSCALL_64_fastpath+0x13/0x8f
[  119.231951] ---[ end trace 5cf240cb25d2429d ]---
[  119.304372] [drm:intel_cpu_fifo_underrun_irq_handler] *ERROR* CPU pipe B FIFO underrun
[  134.530287] [drm:intel_mst_disable_dp] *ERROR* failed to update payload -22

chithanh · Post by **chithanh** » Tue Jul 05, 2016 10:54 pm

I think this error is related to the dock as per the freedesktop bug above.

Can you try booting the system undocked and see if something changes?

robert7k · Post by **robert7k** » Tue Jul 05, 2016 11:03 pm

chithanh wrote:I think this error is related to the dock as per the freedesktop bug above.

Can you try booting the system undocked and see if something changes?

Nope. I rebooted without the dock. I then attached the screen using a MiniDP-Cable, the picture appeared @30Hz. No errors in dmesg until this point. But when I changed to 60Hz I got the exact same error message.

chithanh · Post by **chithanh** » Tue Jul 05, 2016 11:06 pm

Ok, then I am out of ideas. If the latest 4.7-rc kernel does not improve the situation, you can create a bug report at https://bugs.freedesktop.org/ so the Intel upstream can look into it.

robert7k · Post by **robert7k** » Tue Jul 05, 2016 11:14 pm

OK, I'll try a newer kernel in the next days and let you know if that helped.

Thanks for your help and support, much appreciated!

robert7k · Post by **robert7k** » Wed Jul 06, 2016 7:09 am

I opened this issue: https://bugs.freedesktop.org/show_bug.cgi?id=96828

chithanh · Post by **chithanh** » Wed Jul 06, 2016 8:37 am

Good. To give upstream some more information, boot with log_buf_len=1M drm.debug=14 kernel parameters, try to set 4K@60 Hz, and then capture the entire dmesg output and attach to the bug.

robert7k · Post by **robert7k** » Wed Jul 27, 2016 10:14 pm

I tried again today using the new kernel 4.7.0, but the problem remains.

As you suggested, I added the kernel parameters and attached the log output to the bug.