[Intel-gfx] [PATCH] drm/i915: Add debugfs to read any DPCD register
He, Shuang
shuang.he at intel.com
Fri Apr 10 01:06:41 PDT 2015
> -----Original Message-----
> From: Daniel Vetter [mailto:daniel.vetter at ffwll.ch] On Behalf Of Daniel
> Vetter
> Sent: Friday, April 10, 2015 3:39 PM
> To: He, Shuang
> Cc: Gao, Ethan; intel-gfx at lists.freedesktop.org; R, Durgadoss
> Subject: Re: [Intel-gfx] [PATCH] drm/i915: Add debugfs to read any DPCD
> register
>
> On Thu, Apr 09, 2015 at 02:54:13PM -0700, shuang.he at intel.com wrote:
> > Tested-By: Intel Graphics QA PRTS (Patch Regression Test System Contact:
> shuang.he at intel.com)
> > Task id: 6145
> > -------------------------------------Summary-------------------------------------
> > Platform Delta drm-intel-nightly Series Applied
> > PNV 276/276 276/276
> > ILK 302/302 302/302
> > SNB -21 313/313 292/313
> > IVB 337/337 337/337
> > BYT 286/286 286/286
> > HSW 395/395 395/395
> > BDW 321/321 321/321
> > -------------------------------------Detailed-------------------------------------
> > Platform Test drm-intel-nightly Series Applied
> > SNB igt at kms_mmio_vs_cs_flip@setplane_vs_cs_flip
> DMESG_WARN(4)PASS(3) DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at kms_rotation_crc@primary-rotation DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@cursor DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@cursor-dpms DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@debugfs-forcewake-user DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@dpms-mode-unset-non-lpsp
> DMESG_WARN(4)PASS(3) DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@dpms-non-lpsp DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@drm-resources-equal DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@fences DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@fences-dpms DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@gem-execbuf DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@gem-idle DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@gem-mmap-cpu DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@gem-mmap-gtt DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@gem-pread DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@i2c DMESG_WARN(4)PASS(3) DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@modeset-non-lpsp DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@modeset-non-lpsp-stress-no-wait
> DMESG_WARN(4)PASS(3) DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@pci-d3-state DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@reg-read-ioctl DMESG_WARN(4)PASS(3)
> DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > SNB igt at pm_rpm@rte DMESG_WARN(4)PASS(3) DMESG_WARN(1)
> > (dmesg patch
> applied)drm:intel_dp_start_link_train[i915]]*ERROR*too_many_voltage_ret
> ries,give_up at too many voltage .* give up
> >
> drm:intel_dp_complete_link_train[i915]]*ERROR*failed_to_train_DP,aborti
> ng at failed to train .* aborting
> > Note: You need to pay more attention to line start with '*'
>
> Something seems wrong with your test infrastructure, I've seen a bunch of
> result (from completely unrelated patches) that DP link training is
> failing all over the place on snb (and sometimes ilk).
[He, Shuang] Yes, it seems to happen on same test machine. Is it possible to related to some kernel issue? I see those dmesg warnings happens for almost same set of tests in two rounds of testing on that test machine: igt at pm_rpm*
http://lists.freedesktop.org/archives/intel-gfx/2015-April/064274.html
http://lists.freedesktop.org/archives/intel-gfx/2015-April/064289.html
while other testing didn't use same hardware seems doesn't have this issue:
http://lists.freedesktop.org/archives/intel-gfx/2015-April/064166.html
Is it possible it's hardware specific issue?
>
> Has someone tripped over the cable?
[He, Shuang] I have just gone to lab and double checked, it's still well connected.
Thanks
--Shuang
> -Daniel
> --
> Daniel Vetter
> Software Engineer, Intel Corporation
> http://blog.ffwll.ch
More information about the Intel-gfx
mailing list