Re: [stable regression] 5.14.12 breaks graphics on an Intel GM45 laptop
From: Mateusz Jończyk
Date: Fri Oct 15 2021 - 09:27:45 EST
Continuing in https://gitlab.freedesktop.org/drm/intel/-/issues/4297
W dniu 15.10.2021 o 14:51, Jani Nikula pisze:
> On Fri, 15 Oct 2021, Imre Deak <imre.deak@xxxxxxxxx> wrote:
>> On Fri, Oct 15, 2021 at 11:23:56AM +0300, Imre Deak wrote:
>>> On Fri, Oct 15, 2021 at 10:20:15AM +0300, Jani Nikula wrote:
>>>> On Thu, 14 Oct 2021, mat.jonczyk <mat.jonczyk@xxxxx> wrote:
>>>>> Hello,
>>>>>
>>>>> Kernel 5.14.12 breaks GPU support on an Dell Satellite E6500 laptop with GM45 integrated graphics. The screen does not update after
>>>>> the i965 module gets loaded. Same problem happens on 5.15.0-rc5.
>>>>>
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: pci 0000:00:00.0: Intel GM45 Chipset
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: pci 0000:00:00.0: detected gtt size: 2097152K total, 262144K mappable
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: pci 0000:00:00.0: detected 32768K stolen memory
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: input: Integrated_Webcam_2M: Integrate as
>>>>> /devices/pci0000:00/0000:00:1a.7/usb1/1-6/1-6:1.0/input/input22
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: usbcore: registered new interface driver uvcvideo
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: checking generic (e0000000 3c0000) vs hw (f6c00000 400000)
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: checking generic (e0000000 3c0000) vs hw (e0000000 10000000)
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: fb0: switching to inteldrmfb from VESA VGA
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Console: switching to colour dummy device 80x25
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: i915 0000:00:02.0: vgaarb: deactivate vga console
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes:
>>>>> olddecodes=io+mem,decodes=io+mem:owns=io+mem
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: usb 3-1.3: New USB device found, idVendor=413c, idProduct=8156, bcdDevice= 4.56
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: usb 3-1.3: New USB device strings: Mfr=1, Product=2, SerialNumber=0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: usb 3-1.3: Product: Dell Wireless 370 Bluetooth Mini-card
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: usb 3-1.3: Manufacturer: Dell Computer Corp
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: invalid opcode: 0000 [#1] SMP PTI
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: CPU: 1 PID: 303 Comm: systemd-udevd Not tainted 5.14.12-051412-generic #202110130438
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Hardware name: Dell Inc. Latitude E6500 /0NY667, BIOS A29 06/04/2013
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RIP: 0010:intel_dp_sync_state+0xe0/0xf0 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Code: 85 69 ff ff ff 48 8b 9f 60 01 00 00 80 bb 73 01 00 00 00 0f 85 64 ff ff ff 48 8d bb 60 01
>>>>> 00 00 e8 a5 cb ff ff e9 53 ff ff ff <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 0f 1f 44 00 00 55
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RSP: 0000:ffff9d6c006bf870 EFLAGS: 00010282
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RAX: ffffffffffffffff RBX: ffff9096c7b84000 RCX: ffff9096c76544e0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff9096c7b841a6
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RBP: ffff9d6c006bf890 R08: 00000000ffffff92 R09: ffffffffc0797464
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R10: 0000000000000002 R11: ffff9096c7b84df0 R12: 0000000000000000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R13: 0000000000000000 R14: 0000000000000004 R15: ffff9096c7b84000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: FS: 00007fc523e0a880(0000) GS:ffff9096dbd00000(0000) knlGS:0000000000000000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: CR2: 0000559a4efd6110 CR3: 000000010570e000 CR4: 00000000000006e0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Call Trace:
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: intel_modeset_readout_hw_state+0x2a4/0xb90 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? __cond_resched+0x1a/0x50
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: intel_modeset_setup_hw_state+0x17f/0x650 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? drm_modeset_lock_all_ctx+0x1a8/0x3c0 [drm]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? drm_warn_on_modeset_not_all_locked.part.0+0x5e/0x90 [drm]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: intel_modeset_init_nogem+0x293/0x4d0 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? intel_irq_postinstall+0x470/0x5b0 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: i915_driver_probe+0x1ab/0x440 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? mutex_lock+0x13/0x40
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: i915_pci_probe+0x58/0x140 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: local_pci_probe+0x48/0x80
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: pci_device_probe+0x105/0x1d0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: really_probe+0x20c/0x410
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: __driver_probe_device+0x109/0x180
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: driver_probe_device+0x23/0x90
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: __driver_attach+0xac/0x1b0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? __device_attach_driver+0xe0/0xe0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: bus_for_each_dev+0x7e/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: driver_attach+0x1e/0x20
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: bus_add_driver+0x135/0x1f0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: driver_register+0x95/0xf0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: __pci_register_driver+0x68/0x70
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: i915_init+0x66/0x86 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? 0xffffffffc0c80000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: do_one_initcall+0x48/0x1d0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? kmem_cache_alloc_trace+0x159/0x2c0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: do_init_module+0x62/0x290
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: load_module+0xaa3/0xb30
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: __do_sys_finit_module+0xbf/0x120
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: __x64_sys_finit_module+0x18/0x20
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: do_syscall_64+0x5c/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? exit_to_user_mode_prepare+0x37/0xb0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? syscall_exit_to_user_mode+0x27/0x50
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? __x64_sys_mmap+0x33/0x40
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? do_syscall_64+0x69/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? __x64_sys_lseek+0x18/0x20
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? do_syscall_64+0x69/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? do_syscall_64+0x69/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? do_syscall_64+0x69/0xc0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? common_interrupt+0x5a/0xa0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ? asm_common_interrupt+0x8/0x40
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: entry_SYSCALL_64_after_hwframe+0x44/0xae
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RIP: 0033:0x7fc52438c89d
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Code: 00 c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca
>>>>> 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d c3 f5 0c 00 f7 d8 64 89 01 48
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RSP: 002b:00007ffee98db388 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RAX: ffffffffffffffda RBX: 0000559a4f001eb0 RCX: 00007fc52438c89d
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RDX: 0000000000000000 RSI: 00007fc524269ded RDI: 0000000000000017
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RBP: 0000000000020000 R08: 0000000000000000 R09: 0000000000000000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R10: 0000000000000017 R11: 0000000000000246 R12: 00007fc524269ded
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R13: 0000000000000000 R14: 0000559a4edadd00 R15: 0000559a4f001eb0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Modules linked in: dell_smm_hwmon(+) snd_hda_codec_generic(+) snd_hda_intel kvm(+)
>>>>> snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec snd_hda_core snd_hwdep snd_pcm dell_wmi ledtrig_audio dell_smbios dcdbas
>>>>> uvcvideo cdc_mbim sparse_keymap dell_wmi_descriptor wmi_bmof snd_seq_midi cdc_wdm videobuf2_vmalloc snd_seq_midi_event
>>>>> serio_raw videobuf2_memops i915(+) snd_rawmidi cdc_ncm iwldvm videobuf2_v4l2 snd_seq cdc_ether mac80211 videobuf2_common
>>>>> qcserial usbnet usb_wwan mii libarc4 ttm videodev usbserial snd_seq_device pcmcia input_leds mc drm_kms_helper iwlwifi snd_timer
>>>>> joydev cec yenta_socket rc_core pcmcia_rsrc pcmcia_core i2c_algo_bit fb_sys_fops syscopyarea snd sysfillrect sysimgblt cfg80211
>>>>> soundcore mac_hid sch_fq_codel msr parport_pc ppdev lp drm parport ip_tables x_tables autofs4 wacom hid_generic usbhid gpio_ich
>>>>> hid sdhci_pci firewire_ohci cqhci ahci i2c_i801 firewire_core xhci_pci psmouse lpc_ich libahci i2c_smbus sdhci crc_itu_t xhci_pci_renesas
>>>>> e1000e wmi video
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: ---[ end trace 80e312eaeb64de58 ]---
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RIP: 0010:intel_dp_sync_state+0xe0/0xf0 [i915]
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: Code: 85 69 ff ff ff 48 8b 9f 60 01 00 00 80 bb 73 01 00 00 00 0f 85 64 ff ff ff 48 8d bb 60 01
>>>>> 00 00 e8 a5 cb ff ff e9 53 ff ff ff <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 0f 1f 44 00 00 55
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RSP: 0000:ffff9d6c006bf870 EFLAGS: 00010282
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RAX: ffffffffffffffff RBX: ffff9096c7b84000 RCX: ffff9096c76544e0
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff9096c7b841a6
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: RBP: ffff9d6c006bf890 R08: 00000000ffffff92 R09: ffffffffc0797464
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R10: 0000000000000002 R11: ffff9096c7b84df0 R12: 0000000000000000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: R13: 0000000000000000 R14: 0000000000000004 R15: ffff9096c7b84000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: FS: 00007fc523e0a880(0000) GS:ffff9096dbd00000(0000) knlGS:0000000000000000
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>>>> paź 14 23:05:34 mjonczyk-laptop kernel: CR2: 0000559a4efd6110 CR3: 000000010570e000 CR4: 00000000000006e0
>>>>> paź 14 23:05:35 mjonczyk-laptop kernel: snd_hda_codec_idt hdaudioC0D0: autoconfig for 92HD71B7X: line_outs=1
>>>>> (0xf/0x0/0x0/0x0/0x0) type:line
>>>>> paź 14 23:05:35 mjonczyk-laptop kernel: snd_hda_codec_idt hdaudioC0D0: speaker_outs=1 (0xd/0x0/0x0/0x0/0x0)
>>>>>
>>>>> The most likely culprit is "drm/i915/tc: Fix TypeC port init/resume time sanitization", so I'm CCing its authors.
>>>>> I'll test tomorrow with this patch reverted. Please excuse me for writing this in HTML format, but my main machine is temporarily
>>>>> broken.
>>>> Imre, intel_dp_sync_state() is now called on g4x DP encoders where
>>>> encoder->get_hw_state() returns false. Seems like a fragile change
>>>> that's not mentioned in the commit message or ->sync_state
>>>> documentation. Could be the culprit, though the backtrace isn't
>>>> conclusive.
>>> Yes, the above is the culprit commit, thanks for the report and decoding
>>> the crash log. I'll follow up with the fix shortly.
>> Mat,
>>
>> actually a bit puzzled what's going on. The crash happens on an UD2
>> undefined instruction and couldn't figure out how we could end up
>> running such code.
> That's what I thought too, hence "the backtrace isn't conclusive".
>
> Did 5.14.11 work?
>
> Would be interested in the results with the commit reverted to confirm.
>
>
> BR,
> Jani.
>
>
>
>> The only relevant change wrt. the crash in my patch is to read out the
>> DPCD on a disabled encoder, however this shouldn't be a problem. It's
>> not required though so to make the pre-DDI code uniform with DDI platforms
>> I sent
>> https://patchwork.freedesktop.org/series/95878/
>>
>> Could you please give this patch a go?
>>
>> Could you provide a full dmesg log booting with drm.debug=0x10e and
>> capturing the crash? (If you have a working VGA console, or ssh access
>> to the machine, you could boot with
>> modprobe.blacklist=i915,snd_hda_intel 3
>> and modprobe i915 manually).
>>
>> Could you also provide your i915.ko binary?
>>
>> Probably better to keep track of this issue on a ticket you could open
>> at
>> https://gitlab.freedesktop.org/drm/intel/-/issues/new
>>
>> Thanks,
>> Imre
>>
>>>> BR,
>>>> Jani.
>>>>
>>>>
>>>>
>>>>
>>>>> I am testing this on Ubuntu 20.04 using kernels from Ubuntu mainline PPA:
>>>>> https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.14.12/
>>>>> https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.15-rc5/
>>>>>
>>>>> Greetings,
>>>>> Mateusz Jończyk
>>>> --
>>>> Jani Nikula, Intel Open Source Graphics Center