drm-i915-update-38, no output on high resolution display

[Alexander Mishurov]

The same problems I have after I ported some code from ValleyView.
Here's lines from my log which I thinks a problem:

Stolen size is zero:

Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_probe_helper] drmn0 desc: 
Intel Valleyview (mobile)
Feb  6 17:25:55 fbsd kernel: drmn0: <Intel Valleyview (mobile)> on vgapci0
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_get_pci_dev]
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_ctxbitmap_next] bit : 0
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_ctxbitmap_init] 
drm_ctxbitmap_init : 0
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_get_minor]
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_get_minor] new minor 
assigned 64
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_get_minor]
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_get_minor] new minor 
assigned 0
Feb  6 17:25:55 fbsd kernel: [drm:KMS:pid1131:i915_dump_device_info] 
i915 device info: gen=7, pciid=0x0f31 
flags=is_mobile,need_gfx_hws,is_valleyview,has_hotplug,has_bsd_ring,has_blt_ring,info: 
[drm] Memory usable by graphics device = 2048M
Feb  6 17:25:55 fbsd kernel: [drm:KMS:pid1131:i915_gem_gtt_init] GMADR 
size = 256M
Feb  6 17:25:55 fbsd kernel: [drm:KMS:pid1131:i915_gem_gtt_init] GTT 
stolen size = 0M
Feb  6 17:25:55 fbsd kernel: [drm:pid1131:drm_addmap_core] offset = 
0x90000000, size = 0x00200000, type = 1

Then problems with communications

Feb  7 01:27:46 fbsd kernel: Keep display ON for 
testing[drm:KMS:pid1133:intel_panel_get_backlight] get backlight PWM = 3125
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:intel_gmbus_force_bit] 
enabling bit-banging on Philips I2C bus. force bit now 258912793
Feb  7 01:27:46 fbsd kernel: random: harvesting attach, 8 bytes (4 bits) 
from intel_sdvo_ddc_proxy397632
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:gmbus_xfer] GMBUS [i915 
gmbus dpb] NAK for addr: 0038 r(1)
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:intel_sdvo_read_byte] i2c 
transfer returned -6
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:intel_sdvo_init] No SDVO 
device found on SDVOB
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:intel_gmbus_force_bit] 
disabling bit-banging on Philips I2C bus. force bit now 258912792
Feb  7 01:27:46 fbsd kernel: [drm:KMS:pid1133:intel_dp_i2c_init] 
i2c_init DPDDC-B
Feb  7 01:27:46 fbsd kernel: random: harvesting attach, 8 bytes (4 bits) 
from drm_iic_dp_aux1
Feb  7 01:27:48 fbsd kernel: [drm:KMS:pid1133:intel_dp_aux_ch] dp_aux_ch 
timeout status 0x51450064
Feb  7 01:27:48 fbsd kernel: [drm:KMS:pid1133:intel_dp_i2c_aux_ch] 
aux_ch failed -60
Feb  7 01:27:48 fbsd kernel: [drm:KMS:pid1133:intel_dp_aux_ch] dp_aux_ch 
timeout status 0x51450064
Feb  7 01:27:48 fbsd kernel: [drm:KMS:pid1133:intel_dp_i2c_aux_ch] 
aux_ch failed -60





On 07/02/16 11:49, Ruediger Gad wrote:
> Hi,
>
> thanks a lot for the updates.
> Unfortunately, this does not seem to solve the issue in my case.
>
> With 10fada6d35b00431585d549f00ab609d7a3652ef my laptop screen still
> turns off.
>
> However, I noticed that doing "kldunload i915kms" (typing blindfold due
> to black screen) at least re-enables the backlight.
> There is still no visible output but at least the backlight switches
> back on.
>
> Attached, I include the dmesg output from my attempt with
> 10fada6d35b00431585d549f00ab609d7a3652ef.
> The CPU is an Intel(R) Core(TM) i7-4710MQ CPU @ 2.50GHz.
>
> Do you have any further hints what I could try?
>
> Thanks a lot to all who are working on this!
>
>
>
> Cheers,
>
> Ruediger
>
>
>
>
> On 03.02.2016 20:54, Stefan Wendler wrote:
>> I can reproduce this. I have the same CPU and the last commit that works
>> for me is 10fada6d35b00431585d549f00ab609d7a3652ef
>>
>> So commit 72bb3b3f9dcaf573be532ced9043eb59dedf21df introduces code that
>> breaks i7-4910QM CPUs
>>
>> Interesting .. looks like I have started testing after this commit then ^^
>>
>> thanks to Stefan Grundmann for giving the hint :)
>>
>> Cheers,
>> Stefan
>>
>> On 01/19/2016 11:35, Stefan Grundmann via freebsd-x11 wrote:
>>> hi,
>>>
>>> during testing on a Lenovo Thinkpad W541 (20EF0020GE: Core i7-4910QM, NVIDIA Quadro K2100M optimus)
>>> i found that after commit 72bb3b3f9dcaf573be532ced9043eb59dedf21df "drm/i915: set TRANSCODER_EDP even earlier",
>>> kldload i915kms results in a blank internal display (eDP1, 2880x1620).
>>>
>>> reverting 72bb3b3f9dcaf573be532ced9043eb59dedf21df make the output work again on all tested revisions.
>>>
>>> best regards
>>>
>>> Stefan Grundmann
>>> _______________________________________________
>>> freebsd-x11 at freebsd.org  mailing list
>>> https://lists.freebsd.org/mailman/listinfo/freebsd-x11
>>> To unsubscribe, send any mail to"freebsd-x11-unsubscribe at freebsd.org"
>>>
>
>
> _______________________________________________
> freebsd-x11 at freebsd.org  mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-x11
> To unsubscribe, send any mail to"freebsd-x11-unsubscribe at freebsd.org"