Created attachment 1976945 [details] dmesg from Rawhide using ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 consoleblank=0 mitigations=auto preempt=full drm.debug=0x1e log_buf_len=1M Original Summary: complete I/O lockup on boot on Eaglelake 8086:2e32 with various cmdline combinations of consoleblank=/mitigations=/preempt= Description of problem: Typical dmesg tail (without drm.debug=0x1e log_buf_len=1M): [ 8.253842] pci 0000:00:00.0: detected gtt size: 2097152K total, 262144K mappable [ 8.254160] pci 0000:00:00.0: detected 32768K stolen memory [ 8.254261] i915 0000:00:02.0: vgaarb: deactivate vga console [ 8.254465] Console: switching to colour dummy device 80x25 [ 8.301562] [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.0 on minor 0 [ 8.459373] fbcon: i915drmfb (fb0) is primary device [ 8.459638] Console: switching to colour frame buffer device 210x65 [ 8.464110] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device [ 8.695036] r8169 0000:01:00.0 eth0: Link is Up - 1Gbps/Full - flow control rx/tx [ 19.418090] i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] flip_done timed out [ 29.658090] i915 0000:00:02.0: [drm] *ERROR* [CRTC:60:pipe B] flip_done timed out [ 29.658109] i915 0000:00:02.0: [drm] *ERROR* pipe A underrun [ 40.410096] i915 0000:00:02.0: [drm] *ERROR* flip_done timed out [ 40.410103] i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] commit wait timed out [ 50.650091] i915 0000:00:02.0: [drm] *ERROR* flip_done timed out [ 50.650098] i915 0000:00:02.0: [drm] *ERROR* [CRTC:60:pipe B] commit wait timed out Version-Release number of selected component (if applicable): How reproducible: 0: always with ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 1. More often than not with 6.2.15-300.fc38 and 6.3.12-200.fc38 on F38 2. according to table with 6.3.12-200.fc38 and 6.5rc2 on Rawhide: Using Dual Displays: Bad: consoleblank=0 consoleblank=0 mitigations=auto consoleblank=0 mitigations=auto preempt=full consoleblank=0 mitigations=off consoleblank=0 preempt=full consoleblank=0 preempt=full mitigations=auto consoleblank=0 preempt=full mitigations=off consoleblank=0 preempt=none mitigations=auto consoleblank=0 preempt=none mitigations=off mitigations=auto mitigations=auto preempt=full mitigations=auto preempt=none mitigations=off preempt=full preempt=full preempt=none OK: consoleblank=0 preempt=none mitigations=off mitigations=off preempt=none (none of consoleblank, mitigations, preempt) Using DVI display Only: Bad: consoleblank=0 consoleblank=0 mitigations=auto consoleblank=0 mitigations=auto preempt=none consoleblank=0 mitigations=off preempt=full consoleblank=0 preempt=full consoleblank=0 preempt=none mitigations=auto preempt=full mitigations=off preempt=full mitigations=off preempt=none OK: consoleblank=0 mitigations=auto preempt=full consoleblank=0 mitigations=off consoleblank=0 mitigations=off preempt=none mitigations=auto preempt=none (none of consoleblank, mitigations, preempt) Steps to Reproduce: 1. Connect displays to motherboard's DVI and VGA ports 2. Try to boot Actual results: 1-DRM/CRTC errors 2-primary display flashes two white cursors at high rate on black screen 3-secondary display sleeps due to "out of range" 4-no keyboard response 5-remote login functional Expected results: 1-Xorg/KDM/Plasma-X11 work normally 2-keyboard is normal/responsive Additional info: 1-normal operation with two one or two displays in F37/6.3.x, Tumblweed/6.3.9, Trixie/6.3.x, Mageia/6.3 2-# pinxi -GSaz --vs --zl --hostname pinxi 3.3.28-02 (2023-07-18) System: Host: big41 Kernel: 6.3.12-100.fc37.x86_64 arch: x86_64 bits: 64 compiler: gcc v: 2.38-27.fc37 clocksource: tsc available: hpet,acpi_pm parameters: ro root=LABEL=<filter> ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 noresume consoleblank=0 preempt=full mitigations=off Desktop: KDE Plasma v: 5.27.4 tk: Qt v: 5.15.9 wm: kwin_x11 vt: 7 dm: 1: KDM 2: XDM Distro: Fedora release 37 (Thirty Seven) Graphics: Device-1: Intel 4 Series Integrated Graphics vendor: Biostar Microtech Intl Corp driver: i915 v: kernel arch: Gen-5 process: Intel 45nm built: 2008 ports: active: HDMI-A-1,VGA-1 empty: DP-1 bus-ID: 00:02.0 chip-ID: 8086:2e32 class-ID: 0300 Display: x11 server: X.Org v: 1.20.14 with: Xwayland v: 22.1.9 compositor: kwin_x11 driver: X: loaded: modesetting unloaded: fbdev,vesa dri: crocus gpu: i915 display-ID: :0 screens: 1 Screen-1: 0 s-res: 3600x1200 s-dpi: 120 s-size: 762x254mm (30.00x10.00") s-diag: 803mm (31.62") Monitor-1: HDMI-A-1 mapped: HDMI-1 pos: primary,left model: NEC EA243WM serial: <filter> built: 2011 res: 1920x1200 hz: 60 dpi: 94 gamma: 1.2 size: 519x324mm (20.43x12.76") diag: 612mm (24.1") ratio: 16:10 modes: max: 1920x1200 min: 640x480 Monitor-2: VGA-1 pos: right model: Dell P2213 serial: <filter> built: 2012 res: 1680x1050 hz: 60 dpi: 90 gamma: 1.2 size: 473x296mm (18.62x11.65") diag: 558mm (22") ratio: 16:10 modes: max: 1680x1050 min: 720x400 API: OpenGL v: 2.1 Mesa 23.0.3 renderer: Mesa Intel G41 (ELK) direct-render: Yes #
Updates apparently fixed this on 38, but same problem is now on 39 with 6.4.16 and 6.5.5 kernels, and even fewer cmdline option combinations succeed: Next-last failure was with ipv6.disable=1 net.ifnames=0 selinux=0 mitigations=off Last succeed was with ipv6.disable=1 net.ifnames=0 selinux=0 ibt=off Last failure was with ipv6.disable=1 net.ifnames=0 selinux=0 video=1440x900@60 # journalctl --no-host | tail Oct 08 11:34:44 kernel: i915 0000:00:02.0: [drm] *ERROR* [CONNECTOR:64:HDMI-A-1] commit wait timed out Oct 08 11:34:49 kdm[546]: X server is stuck in D state; leaving it alone Oct 08 11:34:49 kdm[546]: X server for display :0 cannot be started, session disabled Oct 08 11:34:54 kernel: i915 0000:00:02.0: [drm] *ERROR* flip_done timed out Oct 08 11:34:54 kernel: i915 0000:00:02.0: [drm] *ERROR* [PLANE:31:primary A] commit wait timed out Oct 08 11:35:04 kernel: i915 0000:00:02.0: [drm] *ERROR* flip_done timed out Oct 08 11:35:04 kernel: i915 0000:00:02.0: [drm] *ERROR* [PLANE:46:primary B] commit wait timed out Oct 08 11:35:14 kernel: i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] flip_done timed out Oct 08 11:35:25 kernel: i915 0000:00:02.0: [drm] *ERROR* [CRTC:60:pipe B] flip_done timed out Oct 08 11:35:25 kernel: i915 0000:00:02.0: [drm] *ERROR* pipe A underrun # uname -r 6.5.5-300.fc39.x86_64 #
Problem is now gone on F39: # cat /proc/cmdline ... ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 consoleblank=0 mitigations=off ibt=off video=1440x900@60 # uname -a Linux big41 6.6.3-200.fc39.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 19:11:52 UTC 2023 x86_64 GNU/Linux #
But it's back in F38: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux # cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 audit=0 noresume consoleblank=0 mitigations=off selinux=0 # Omit mitigations=off and it's OK. Mitigations=auto is also bad: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux # cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 audit=0 noresume consoleblank=0 mitigations=auto selinux=0 # dmesg | tail [ 8.520381] i915 0000:00:02.0: vgaarb: deactivate vga console [ 8.520577] Console: switching to colour dummy device 80x25 [ 8.548068] [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.0 on minor 0 [ 8.730188] fbcon: i915drmfb (fb0) is primary device [ 8.730417] Console: switching to colour frame buffer device 210x65 [ 8.730431] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device [ 9.991983] r8169 0000:01:00.0 eth0: Link is Up - 1Gbps/Full - flow control rx/tx [ 20.103015] i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] flip_done timed out [ 30.343015] i915 0000:00:02.0: [drm] *ERROR* [CRTC:60:pipe B] flip_done timed out [ 30.343033] i915 0000:00:02.0: [drm] *ERROR* pipe A underrun # The following was OK: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux [root@big41 ~]# cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 audit=0 noresume consoleblank=0 mitigations=auto # So it seems the problem isn't what the parameters are, but more like the total cmdline length, because the following was also OK: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux [root@big41 ~]# cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 noresume mitigations=off selinux=0 # And again not: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux [root@big41 ~]# cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 noresume mitigations=off selinux=0 video=1440x900@60 [root@big41 ~]# dmesg | tail [ 8.261285] [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.0 on minor 0 [ 8.442207] fbcon: i915drmfb (fb0) is primary device [ 8.442355] Console: switching to colour frame buffer device 180x56 [ 8.442364] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device [ 9.987429] r8169 0000:01:00.0 eth0: Link is Up - 1Gbps/Full - flow control rx/tx [ 19.591958] i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] flip_done timed out [ 29.831963] i915 0000:00:02.0: [drm] *ERROR* [CRTC:60:pipe B] flip_done timed out [ 29.831983] i915 0000:00:02.0: [drm] *ERROR* pipe A underrun [ 40.583966] i915 0000:00:02.0: [drm] *ERROR* flip_done timed out [ 40.583974] i915 0000:00:02.0: [drm] *ERROR* [CRTC:45:pipe A] commit wait timed out # The same cmdline as in comment #2, 101, is also OK in F38: # uname -a Linux big41 6.6.3-100.fc38.x86_64 #1 SMP PREEMPT_DYNAMIC Tue Nov 28 20:36:17 UTC 2023 x86_64 GNU/Linux # cat /proc/cmdline ro root=LABEL=i256p20f38 ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 consoleblank=0 mitigations=off ibt=off # Character count of 103 is OK, but 107 is not, so I tried these two: ro root=LABEL=i256p29f39 ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 consoleblank=0 preempt=no 0123456 = fail ro root=LABEL=i256p29f39 ipv6.disable=1 net.ifnames=0 audit=0 selinux=0 consoleblank=0 preempt=no 012345 = succeed So, the success limit appears to be 104 characters, at least in F39.
6.7.9 on F39 and 6.8.9 on F40 don't reproduce this.