Created attachment 110053 [details] fastboot-fails.txt I've tried to improve boot timing related to screen detection (see https://bugs.freedesktop.org/show_bug.cgi?id=86733 ) by setting i915.fastboot=1 however, that doesn't seem to make any difference. This is Intel NUC haswell, should i915.fastboot=1 improve bootspeed / reduce boot flicker / work? Attaching a drm log. If there is anything else you want to know, let me know and I'll be happy to provide info.
[ 0.476617] [drm] VT-d active for gfx access [ 0.484163] [drm] DMAR active, disabling use of stolen memory so it is not going to be able to preserve the BIOS framebuffer and mode. But [ 0.489956] [drm:intel_modeset_readout_hw_state] [CRTC:7] hw state readout: disabled [ 0.489962] [drm:intel_modeset_readout_hw_state] [CRTC:11] hw state readout: disabled [ 0.489965] [drm:intel_modeset_readout_hw_state] [CRTC:15] hw state readout: disabled [ 0.489968] [drm:intel_modeset_readout_hw_state] WRPLL 1 hw state readout: refcount 0, on 0 [ 0.489971] [drm:intel_modeset_readout_hw_state] WRPLL 2 hw state readout: refcount 0, on 0 [ 0.489974] [drm:intel_modeset_readout_hw_state] [ENCODER:17:TMDS-17] hw state readout: disabled, pipe A [ 0.489977] [drm:intel_modeset_readout_hw_state] [ENCODER:22:TMDS-22] hw state readout: disabled, pipe A [ 0.489979] [drm:intel_modeset_readout_hw_state] [ENCODER:24:DP MST-24] hw state readout: disabled, pipe A [ 0.489981] [drm:intel_modeset_readout_hw_state] [ENCODER:25:DP MST-25] hw state readout: disabled, pipe B [ 0.489983] [drm:intel_modeset_readout_hw_state] [ENCODER:26:DP MST-26] hw state readout: disabled, pipe C [ 0.489986] [drm:intel_modeset_readout_hw_state] [CONNECTOR:18:HDMI-A-1] hw state readout: disabled [ 0.489988] [drm:intel_modeset_readout_hw_state] [CONNECTOR:23:DP-1] hw state readout: disabled [ 0.489991] [drm:intel_modeset_readout_hw_state] [CONNECTOR:27:HDMI-A-2] hw state readout: disabled everything is regarded as off anyway.
(In reply to Chris Wilson from comment #1) > [ 0.476617] [drm] VT-d active for gfx access > [ 0.484163] [drm] DMAR active, disabling use of stolen memory > > so it is not going to be able to preserve the BIOS framebuffer and mode. But > > [ 0.489956] [drm:intel_modeset_readout_hw_state] [CRTC:7] hw state > readout: disabled > [ 0.489962] [drm:intel_modeset_readout_hw_state] [CRTC:11] hw state > readout: disabled > [ 0.489965] [drm:intel_modeset_readout_hw_state] [CRTC:15] hw state > readout: disabled > [ 0.489968] [drm:intel_modeset_readout_hw_state] WRPLL 1 hw state > readout: refcount 0, on 0 > [ 0.489971] [drm:intel_modeset_readout_hw_state] WRPLL 2 hw state > readout: refcount 0, on 0 > [ 0.489974] [drm:intel_modeset_readout_hw_state] [ENCODER:17:TMDS-17] hw > state readout: disabled, pipe A > [ 0.489977] [drm:intel_modeset_readout_hw_state] [ENCODER:22:TMDS-22] hw > state readout: disabled, pipe A > [ 0.489979] [drm:intel_modeset_readout_hw_state] [ENCODER:24:DP MST-24] > hw state readout: disabled, pipe A > [ 0.489981] [drm:intel_modeset_readout_hw_state] [ENCODER:25:DP MST-25] > hw state readout: disabled, pipe B > [ 0.489983] [drm:intel_modeset_readout_hw_state] [ENCODER:26:DP MST-26] > hw state readout: disabled, pipe C > [ 0.489986] [drm:intel_modeset_readout_hw_state] [CONNECTOR:18:HDMI-A-1] > hw state readout: disabled > [ 0.489988] [drm:intel_modeset_readout_hw_state] [CONNECTOR:23:DP-1] hw > state readout: disabled > [ 0.489991] [drm:intel_modeset_readout_hw_state] [CONNECTOR:27:HDMI-A-2] > hw state readout: disabled > > everything is regarded as off anyway. Hm I see. Well the Dell monitor I have is eager to go into "sleep mode" and thus not actually showing the firmware bootsplash often. I did the following - unplug DP cable, plug it in and quickly poweron the NUC. This way the firmware/uefi splash is shown, then i get loads of screen flicker then I boot to Basic System and dmesg is printed on the screen... but dracut fails to complete the boot as things are not detected (e.g. disk/by-uuid not present) anyway i ran udevadm trigger to fix that, complete boot and I'll see if the dmesg / fastboot improves.
Created attachment 110055 [details] better-fastboot.txt Attaching a new bootlog this one did have DP active at boot time, and should be possible to be picked up by i915.fastboot.
That's detecting it, and we have: [ 0.496891] [drm:ironlake_get_plane_config] pipe/plane 0/0 with fb: size=1920x1080@32, offset=0, pitch 7680, size 0x7e9000 However VT'd is still active and so we cannot use stolen memory (which is where the BIOS fb is) so we still have to redo the modeset, and as you noted elsewhere we have a 0.8s of i2c native defer whilst probing the DP.
I don't see a bug here.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.