The CI Bug Log issue associated to this bug has been updated.

### New filters associated

* GLK: igt@i915_selftest@live_gtt - dmesg-warn - *ERROR* rcs0 workaround lost on load!
  - https://intel-gfx-ci.01.org/tree/drm-tip/IGT_5021/fi-glk-dsi/igt@i915_selftest@live_gtt.html

That's worryingly peculiar. Hopefully just the read back is missing something.

Not too frequent at least, it's been seen once since reported. Probably not a "highest" priority but let's keep watching for a while still.

Still seen once, dropping to "high"

I've seen this failure on trybot: https://intel-gfx-ci.01.org/tree/drm-tip/Trybot_4424/fi-bxt-j4205/igt@i915_selftest@live_gtt.html

I notice one symptom which also ties both occurrences and comes a bit earlier in the log:

  [drm:intel_modeset_init [i915]] 0 display pipe available.

This is definitely suspect and very strange.

Chris mentioned DMC and power context corruption as potential culprits in this area. It seems to be turning out into something which could use a 1:1 session with a BXT or GLK to hunt it down locally.

dmesg from BXT:

7> [400.777593] [drm:intel_csr_ucode_init [i915]] Loading i915/bxt_dmc_ver1_07.bin
<6> [400.778771] [drm] Finished loading DMC firmware i915/bxt_dmc_ver1_07.bin (v1.7)
<7> [400.779203] [drm:__bxt_hpd_detection_setup [i915]] Invert bit setting: hp_ctl:10001010 hp_port:38
<7> [400.779945] [drm:intel_fbc_init [i915]] Sanitized enable_fbc value: 1
<7> [400.780193] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM0 latency 7 (7.0 usec)
<7> [400.780339] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM1 latency 7 (7.0 usec)
<7> [400.780442] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM2 latency 8 (8.0 usec)
<7> [400.780524] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM3 latency 22 (22.0 usec)
<7> [400.780602] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM4 latency 22 (22.0 usec)
<7> [400.780682] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM5 latency 22 (22.0 usec)
<7> [400.780764] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM6 latency 22 (22.0 usec)
<7> [400.780844] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM7 latency 22 (22.0 usec)
<7> [400.780938] [drm:intel_modeset_init [i915]] 0 display pipe available.
<7> [400.781047] [drm:intel_dump_cdclk_state [i915]] Current CDCLK 144000 kHz, VCO 1152000 kHz, ref 19200 kHz, bypass 19200 kHz, voltage level 6
<7> [400.781143] [drm:intel_update_max_cdclk [i915]] Max CD clock rate: 624000 kHz
<7> [400.781237] [drm:intel_modeset_init [i915]] Max dotclock rate: 624000 kHz
<7> [400.781945] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL A hw state readout: crtc_mask 0x00000000, on 0
<7> [400.782065] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL B hw state readout: crtc_mask 0x00000000, on 0
<7> [400.782172] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL C hw state readout: crtc_mask 0x00000000, on 0
<7> [400.787581] [drm:i915_init_ggtt [i915]] clearing unused GTT space: [1000, 100000000]
<7> [400.788738] [drm:__intel_engine_init_ctx_wa [i915]] Initialized 13 context workarounds
<7> [400.790120] [drm:i915_gem_contexts_init [i915]] logical context support initialized
<7> [400.792692] [drm:intel_execlists_submission_init [i915]] Initialized 6 rcs0 workarounds
<7> [400.792815] [drm:intel_execlists_submission_init [i915]] Initialized 3 whitelist workarounds
<7> [400.796126] [drm:intel_init_gt_powersave [i915]] BIOS enabled RC states: HW_CTRL off HW_RC6 off SW_TARGET_STATE 0
<3> [400.801170] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (20d4=bbbbbbbb/0, expected 40004, mask=4)
<3> [400.801492] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (20e0=bbbbbbbb/0, expected 40004000, mask=4000)
<3> [400.801580] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (20e4=900/0, expected 4000400, mask=400)
<3> [400.801667] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b100=0/0, expected f84000, mask=ffc000)
<3> [400.801754] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b118=42062e58/0, expected 200000, mask=200000)
<3> [400.801841] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b11c=ffff8881/0, expected 4, mask=4)

From GLK:

<7> [327.729602] [drm:intel_csr_ucode_init [i915]] Loading i915/glk_dmc_ver1_04.bin
<6> [327.730821] [drm] Finished loading DMC firmware i915/glk_dmc_ver1_04.bin (v1.4)
<7> [327.731935] [drm:__bxt_hpd_detection_setup [i915]] Invert bit setting: hp_ctl:10001010 hp_port:38
<7> [327.732421] [drm:intel_fbc_init [i915]] Sanitized enable_fbc value: 0
<7> [327.732610] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM0 latency 7 (7.0 usec)
<7> [327.732674] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM1 latency 7 (7.0 usec)
<7> [327.732766] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM2 latency 8 (8.0 usec)
<7> [327.732831] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM3 latency 22 (22.0 usec)
<7> [327.732895] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM4 latency 22 (22.0 usec)
<7> [327.732958] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM5 latency 22 (22.0 usec)
<7> [327.733021] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM6 latency 22 (22.0 usec)
<7> [327.733085] [drm:intel_print_wm_latency [i915]] Gen9 Plane WM7 latency 22 (22.0 usec)
<7> [327.733160] [drm:intel_modeset_init [i915]] 0 display pipe available.
<7> [327.733241] [drm:intel_dump_cdclk_state [i915]] Current CDCLK 79200 kHz, VCO 633600 kHz, ref 19200 kHz, bypass 19200 kHz, voltage level 4
<6> [327.733462] mei_hdcp mei::b638ab7e-94e2-4ea2-a552-d1c54b627f04:01: bound 0000:00:02.0 (ops i915_hdcp_component_ops [i915])
<7> [327.733537] [drm:intel_update_max_cdclk [i915]] Max CD clock rate: 316800 kHz
<7> [327.733612] [drm:intel_modeset_init [i915]] Max dotclock rate: 633600 kHz
<7> [327.734141] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL A hw state readout: crtc_mask 0x00000000, on 0
<7> [327.734224] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL B hw state readout: crtc_mask 0x00000000, on 0
<7> [327.734305] [drm:intel_modeset_setup_hw_state [i915]] PORT PLL C hw state readout: crtc_mask 0x00000000, on 0
<7> [327.737964] [drm:i915_gem_init_ggtt [i915]] clearing unused GTT space: [1000, 100000000]
<7> [327.738759] [drm:__intel_engine_init_ctx_wa [i915]] Initialized 12 context workarounds
<7> [327.739245] [drm:i915_gem_contexts_init [i915]] logical context support initialized
<7> [327.741143] [drm:intel_execlists_submission_init [i915]] Initialized 5 rcs0 workarounds
<7> [327.741233] [drm:intel_execlists_submission_init [i915]] Initialized 4 whitelist workarounds
<7> [327.743626] [drm:intel_init_gt_powersave [i915]] BIOS enabled RC states: HW_CTRL off HW_RC6 off SW_TARGET_STATE 0
<3> [327.746673] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (20d4=e167a808/0, expected 40004, mask=4)
<3> [327.746826] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (20e0=0/0, expected 40004000, mask=4000)
<3> [327.746898] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b100=0/0, expected f84000, mask=ffc000)
<3> [327.746969] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b118=0/0, expected 200000, mask=200000)
<3> [327.747039] [drm:wa_verify [i915]] *ERROR* rcs0 workaround lost on load! (b11c=0/0, expected 4, mask=4)

Seen once in two months, although only out of 15 runs (GLK seems to have been neglected during June/July).

Keeping it open with the same priority since it's BAT, but would like to see it at least once more before asking people to spend tine on it.

*** Bug 111377 has been marked as a duplicate of this bug. ***

A CI Bug Log filter associated to this bug has been updated:

{- GLK: igt@i915_selftest@live_gtt - dmesg-warn - *ERROR* rcs0 workaround lost on load! -}
{+ BXT GLK: igt@i915_selftest@live_gtt - dmesg-warn - *ERROR* (engine|rcs0) workaround lost on load! +}

New failures caught by the filter:

  * https://intel-gfx-ci.01.org/tree/drm-tip/IGT_5133/fi-bxt-j4205/igt@i915_selftest@live_gtt.html

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/305.