Re: [Intel-gfx] [igt-dev] [PATCH i-g-t] igt/kms_frontbuffer_tracking: Restore modparams around test

[Paulo Zanoni <paulo.r.zanoni@intel.com>]

Em Seg, 2018-08-06 às 22:07 +0100, Chris Wilson escreveu:
> Quoting Paulo Zanoni (2018-08-06 21:56:15)
> > Em Seg, 2018-08-06 às 21:22 +0100, Chris Wilson escreveu:
> > > Save the module parameters from setup and restore them on
> > > teardown,
> > > so
> > > that we leave the system in the same state as we found it.
> > 
> > Currently kms_fbt uses igt_set_module_param_int() which uses
> > igt_save_module_param(), which installs an exit handler, which is
> > already supposed to accomplish what's described in the sentence
> > above.
> > What's missing from the current exit handler? Why is the new
> > strategy
> > better?
> 
> Hmm, it's clearly not working then as fbc remains enabled after
> kms_frontbuffer_tracking:
> 
> <7>[ 3305.252856] [IGT] kms_frontbuffer_tracking: exiting, ret=0
> <5>[ 3305.252940] Setting dangerous option enable_psr - tainting
> kernel
> <5>[ 3305.253033] Setting dangerous option enable_fbc - tainting
> kernel
> <7>[ 3305.279279] [drm:intel_atomic_check [i915]] [CONNECTOR:67:HDMI-
> A-1] checking for sink bpp constrains
> <7>[ 3305.279317] [drm:intel_hdmi_compute_config [i915]] picking bpc
> to 12 for HDMI output
> <7>[ 3305.279346] [drm:intel_hdmi_compute_config [i915]] forcing pipe
> bpp to 36 for HDMI
> <7>[ 3305.279376] [drm:intel_atomic_check [i915]] hw max bpp: 36,
> pipe bpp: 36, dithering: 0
> <7>[ 3305.279405] [drm:intel_dump_pipe_config [i915]] [CRTC:51:pipe
> B][modeset]
> <7>[ 3305.279434] [drm:intel_dump_pipe_config [i915]] output_types:
> HDMI (0x40)
> <7>[ 3305.279462] [drm:intel_dump_pipe_config [i915]] cpu_transcoder:
> B, pipe bpp: 36, dithering: 0
> <7>[ 3305.279491] [drm:intel_dump_pipe_config [i915]] audio: 1,
> infoframes: 1
> <7>[ 3305.279519] [drm:intel_dump_pipe_config [i915]] requested mode:
> <7>[ 3305.279525] [drm:drm_mode_debug_printmodeline] Modeline
> 0:"1920x1080" 60 148500 1920 2008 2052 2200 1080 1083 1088 1125 0x48
> 0x9
> <7>[ 3305.279555] [drm:intel_dump_pipe_config [i915]] adjusted mode:
> <7>[ 3305.279561] [drm:drm_mode_debug_printmodeline] Modeline
> 0:"1920x1080" 60 148500 1920 2008 2052 2200 1080 1083 1088 1125 0x48
> 0x9
> <7>[ 3305.279592] [drm:intel_dump_pipe_config [i915]] crtc timings:
> 148500 1920 2008 2052 2200 1080 1083 1088 1125, type: 0x48 flags: 0x9
> <7>[ 3305.279621] [drm:intel_dump_pipe_config [i915]] port clock:
> 222750, pipe src size: 1920x1080, pixel rate 148500
> <7>[ 3305.279651] [drm:intel_dump_pipe_config [i915]] pch pfit: pos:
> 0x00000000, size: 0x00000000, disabled
> <7>[ 3305.279680] [drm:intel_dump_pipe_config [i915]] ips: 0, double
> wide: 0
> <7>[ 3305.279708] [drm:intel_dump_pipe_config [i915]] dpll_hw_state:
> wrpll: 0xb0210414 spll: 0x0
> <7>[ 3305.279736] [drm:intel_dump_pipe_config [i915]] planes on this
> crtc
> <7>[ 3305.279764] [drm:intel_dump_pipe_config [i915]]
> [PLANE:40:primary B] disabled, scaler_id = 0
> <7>[ 3305.279793] [drm:intel_dump_pipe_config [i915]]
> [PLANE:43:sprite B] disabled, scaler_id = 0
> <7>[ 3305.279822] [drm:intel_dump_pipe_config [i915]]
> [PLANE:48:cursor B] disabled, scaler_id = 0
> <7>[ 3305.280439] [drm:intel_find_shared_dpll [i915]] [CRTC:51:pipe
> B] allocated WRPLL 1
> <7>[ 3305.280465] [drm:intel_reference_shared_dpll [i915]] using
> WRPLL 1 for pipe B
> <7>[ 3305.280519] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:65:CRT]
> <7>[ 3305.280544] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:66:DDI B]
> <7>[ 3305.280570] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:73:DDI C]
> <7>[ 3305.280595] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:76:DDI D]
> <7>[ 3305.280620] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:78:DP-MST A]
> <7>[ 3305.280645] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:79:DP-MST B]
> <7>[ 3305.280671] [drm:intel_atomic_commit_tail [i915]]
> [ENCODER:80:DP-MST C]
> <7>[ 3305.280696] [drm:verify_single_dpll_state.isra.88 [i915]] WRPLL
> 1
> <7>[ 3305.280722] [drm:verify_single_dpll_state.isra.88 [i915]] WRPLL
> 2
> <7>[ 3305.280747] [drm:verify_single_dpll_state.isra.88 [i915]] SPLL
> <7>[ 3305.280773] [drm:verify_single_dpll_state.isra.88 [i915]] LCPLL
> 810
> <7>[ 3305.280798] [drm:verify_single_dpll_state.isra.88 [i915]] LCPLL
> 1350
> <7>[ 3305.280823] [drm:verify_single_dpll_state.isra.88 [i915]] LCPLL
> 2700
> <7>[ 3305.280857] [drm:intel_enable_shared_dpll [i915]] enable WRPLL
> 1 (active 2, on? 0) for crtc 51
> <7>[ 3305.280883] [drm:intel_enable_shared_dpll [i915]] enabling
> WRPLL 1
> <7>[ 3305.280992] [drm:intel_enable_pipe [i915]] enabling pipe B
> <7>[ 3305.281048] [drm:intel_audio_codec_enable [i915]] ELD on
> [CONNECTOR:67:HDMI-A-1], [ENCODER:66:DDI B]
> <7>[ 3305.281086] [drm:hsw_audio_codec_enable [i915]] Enable audio
> codec on pipe B, 32 bytes ELD
> <7>[ 3305.281120] [drm:audio_config_hdmi_pixel_clock [i915]]
> Configuring HDMI audio for pixel clock 148500 (0x00090000)
> <7>[ 3305.281146] [drm:hsw_audio_config_update [i915]] using
> automatic N
> <7>[ 3305.319297] [drm:__intel_fbc_disable [i915]] Disabling FBC on
> pipe A
> <7>[ 3305.319328] [drm:verify_connector_state.isra.87 [i915]]
> [CONNECTOR:67:HDMI-A-1]
> <7>[ 3305.319362] [drm:intel_atomic_commit_tail [i915]] [CRTC:51:pipe
> B]
> <7>[ 3305.319396] [drm:verify_single_dpll_state.isra.88 [i915]] WRPLL
> 1
> <7>[ 3305.319794] [drm:intel_fbc_enable [i915]] reserved 11796480
> bytes of contiguous stolen space for FBC, threshold: 1
> <7>[ 3305.319818] [drm:intel_fbc_enable [i915]] Enabling FBC on pipe
> A
> 
> And that I think makes the difference between rpm working and not.

Then I suppose it's better to debug the current solution instead of
writing a new one, especially since the current one is part of lib/ and
may be used by other tests.


> -Chris
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx