Fix kernel panic at boot on dual GFX systems

[Kai-Heng Feng]

[Impact]
On dual GFX systems, if amdgpu fails to probe a device, vgaarb for
another GFX triggers a NULL pointer dereference and freeze the system.

[Fix]
Defer VGA client registering so resources are handled properly when
amdgpu probe fails.

[Test]
With the patch applied, no more kernel panic when vgaarb is changing the
VGA mode.

[Where problems could occur]
VGA clients won't take effect until a successful probe, so unless there
is very subtle bug in vgaarb or vgaswitcheroo, it's very unlikely to
introduce any regression.

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-hirsute' to 'verification-done-hirsute'. If the problem still exists, change the tag 'verification-needed-hirsute' to 'verification-failed-hirsute'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-focal' to 'verification-done-focal'. If the problem still exists, change the tag 'verification-needed-focal' to 'verification-failed-focal'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 5.11.0-18.19

---------------
linux (5.11.0-18.19) hirsute; urgency=medium

  * hirsute/linux: 5.11.0-18.19 -proposed tracker (LP: #1927578)

  * Packaging resync (LP: #1786013)
    - update dkms package versions

  * Introduce the 465 driver series, fabric-manager, and libnvidia-nscq
    (LP: #1925522)
    - debian/dkms-versions -- add NVIDIA 465 and migrate 450 to 460

  * linux-image-5.0.0-35-generic breaks checkpointing of container
    (LP: #1857257)
    - SAUCE: overlayfs: fix incorrect mnt_id of files opened from map_files

  * Hirsute update: v5.11.17 upstream stable release (LP: #1927535)
    - vhost-vdpa: protect concurrent access to vhost device iotlb
    - Revert "UBUNTU: SAUCE: ovl: Restore vm_file value when lower fs mmap fails"
    - ovl: fix reference counting in ovl_mmap error path
    - coda: fix reference counting in coda_file_mmap error path
    - amd/display: allow non-linear multi-planar formats
    - drm/amdgpu: reserve fence slot to update page table
    - drm/amdgpu: fix GCR_GENERAL_CNTL offset for dimgrey_cavefish
    - gpio: omap: Save and restore sysconfig
    - KEYS: trusted: Fix TPM reservation for seal/unseal
    - vdpa/mlx5: Set err = -ENOMEM in case dma_map_sg_attrs fails
    - pinctrl: lewisburg: Update number of pins in community
    - block: return -EBUSY when there are open partitions in blkdev_reread_part
    - pinctrl: core: Show pin numbers for the controllers with base = 0
    - arm64: dts: allwinner: Revert SD card CD GPIO for Pine64-LTS
    - bpf: Allow variable-offset stack access
    - bpf: Refactor and streamline bounds check into helper
    - bpf: Tighten speculative pointer arithmetic mask
    - perf/x86/intel/uncore: Remove uncore extra PCI dev HSWEP_PCI_PCU_3
    - perf/x86/kvm: Fix Broadwell Xeon stepping in isolation_ucodes[]
    - perf auxtrace: Fix potential NULL pointer dereference
    - perf map: Fix error return code in maps__clone()
    - HID: google: add don USB id
    - HID: asus: Add support for 2021 ASUS N-Key keyboard
    - HID: alps: fix error return code in alps_input_configured()
    - HID cp2112: fix support for multiple gpiochips
    - HID: wacom: Assign boolean values to a bool variable
    - soc: qcom: geni: shield geni_icc_get() for ACPI boot
    - dmaengine: xilinx: dpdma: Fix descriptor issuing on video group
    - dmaengine: xilinx: dpdma: Fix race condition in done IRQ
    - ARM: dts: Fix swapped mmc order for omap3
    - m68k: fix flatmem memory model setup
    - net: geneve: check skb is large enough for IPv4/IPv6 header
    - dmaengine: tegra20: Fix runtime PM imbalance on error
    - s390/entry: save the caller of psw_idle
    - arm64: kprobes: Restore local irqflag if kprobes is cancelled
    - xen-netback: Check for hotplug-status existence before watching
    - cavium/liquidio: Fix duplicate argument
    - csky: change a Kconfig symbol name to fix e1000 build error
    - ia64: fix discontig.c section mismatches
    - ia64: tools: remove duplicate definition of ia64_mf() on ia64
    - x86/crash: Fix crash_setup_memmap_entries() out-of-bounds access
    - net: hso: fix NULL-deref on disconnect regression
    - USB: CDC-ACM...

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux-oem-5.10 - 5.10.0-1029.30

---------------
linux-oem-5.10 (5.10.0-1029.30) focal; urgency=medium

  * focal/linux-oem-5.10: 5.10.0-1029.30 -proposed tracker (LP: #1930076)

  * CVE-2021-33200
    - bpf: Wrap aux data inside bpf_sanitize_info container
    - bpf: Fix mask direction swap upon off reg sign change
    - bpf: No need to simulate speculative domain for immediates

linux-oem-5.10 (5.10.0-1028.29) focal; urgency=medium

  * focal/linux-oem-5.10: 5.10.0-1028.29 -proposed tracker (LP: #1929167)

  * Packaging resync (LP: #1786013)
    - update dkms package versions

  * TGL-H system NV GPU fallen off the bus after resuming from s2idle with the
    external display connected via docking station (LP: #1929166)
    - SAUCE: ACPI: avoid NVIDIA GPU fallen with an _OSI string

  * AX201 BT will cause system could not enter S0i3 (LP: #1928047)
    - drm/i915: Tweaked Wa_14010685332 for all PCHs

  * Realtek USB hubs in Dell WD19SC/DC/TB fail to work after exiting s2idle
    (LP: #1928242)
    - USB: Verify the port status when timeout happens during port suspend
    - Revert "USB: Add reset-resume quirk for WD19's Realtek Hub"

  * Support mic-mute on Dell's platform (LP: #1928750)
    - ASoC: rt715: add main capture switch and main capture volume
    - ASoC: rt715: remove kcontrols which no longer be used
    - ASoC: rt715: modification for code simplicity
    - platform/x86: Move all dell drivers to their own subdirectory
    - SAUCE: platform/x86: dell-privacy: Add support for Dell hardware privacy
    - SAUCE: ASoC: rt715:add micmute led state control supports
    - [Config] Update configs for Dell's E-Privacy

linux-oem-5.10 (5.10.0-1027.28) focal; urgency=medium

  * focal/linux-oem-5.10: 5.10.0-1027.28 -proposed tracker (LP: #1927620)

  * Introduce the 465 driver series, fabric-manager, and libnvidia-nscq
    (LP: #1925522)
    - debian/dkms-versions -- add NVIDIA 465 and migrate 450 to 460

  * Packaging resync (LP: #1786013)
    - update dkms package versions

  * On TGL platforms screen shows garbage when browsing website by scrolling
    mouse (LP: #1926579)
    - SAUCE: drm/i915/display: Disable PSR2 if TGL Display stepping is B1 from A0

  * Add s2idle support on AMD Renoir and Cezanne (LP: #1927067)
    - drm/amd/display: setup system context in dm_init
    - drm/amd/display: add S/G support for Renoir
    - drm/amdgpu: drop extra drm_kms_helper_poll_enable/disable calls
    - drm/amdgpu: use runpm flag rather than fbcon for kfd runtime suspend (v2)
    - drm/amdgpu: reset runpm flag if device suspend fails
    - drm/amdgpu: add s0i3 capacity check for s0i3 routine (v2)
    - drm/amdgpu: add amdgpu_gfx_state_change_set() set gfx power change entry
      (v2)
    - drm/amdgpu: update amdgpu device suspend/resume sequence for s0i3 support
    - drm/amd/pm: add gfx_state_change_set() for rn gfx power switch (v2)
    - drm/amdgpu: add judgement for suspend/resume sequence
    - drm/amdgpu/pm: no need GPU status set since
      mmnbif_gpu_BIF_DOORBELL_FENCE_CNTL added in FSDL
    - drm/amdgpu: fix shutdown and poweroff process failed with s0ix
    - drm/amdgpu: Only check for S0ix if AMD_PMC ...

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 5.11.0-18.19+21.10.1

---------------
linux (5.11.0-18.19+21.10.1) impish; urgency=medium

  * impish/linux: 5.11.0-18.19+21.10.1 -proposed tracker (LP: #1927560)

  * Packaging resync (LP: #1786013)
    - [Packaging] update update.conf
    - update dkms package versions

  [ Ubuntu: 5.11.0-18.19 ]

  * hirsute/linux: 5.11.0-18.19 -proposed tracker (LP: #1927578)
  * Packaging resync (LP: #1786013)
    - update dkms package versions
  * Introduce the 465 driver series, fabric-manager, and libnvidia-nscq
    (LP: #1925522)
    - debian/dkms-versions -- add NVIDIA 465 and migrate 450 to 460
  * linux-image-5.0.0-35-generic breaks checkpointing of container
    (LP: #1857257)
    - SAUCE: overlayfs: fix incorrect mnt_id of files opened from map_files
  * Hirsute update: v5.11.17 upstream stable release (LP: #1927535)
    - vhost-vdpa: protect concurrent access to vhost device iotlb
    - Revert "UBUNTU: SAUCE: ovl: Restore vm_file value when lower fs mmap fails"
    - ovl: fix reference counting in ovl_mmap error path
    - coda: fix reference counting in coda_file_mmap error path
    - amd/display: allow non-linear multi-planar formats
    - drm/amdgpu: reserve fence slot to update page table
    - drm/amdgpu: fix GCR_GENERAL_CNTL offset for dimgrey_cavefish
    - gpio: omap: Save and restore sysconfig
    - KEYS: trusted: Fix TPM reservation for seal/unseal
    - vdpa/mlx5: Set err = -ENOMEM in case dma_map_sg_attrs fails
    - pinctrl: lewisburg: Update number of pins in community
    - block: return -EBUSY when there are open partitions in blkdev_reread_part
    - pinctrl: core: Show pin numbers for the controllers with base = 0
    - arm64: dts: allwinner: Revert SD card CD GPIO for Pine64-LTS
    - bpf: Allow variable-offset stack access
    - bpf: Refactor and streamline bounds check into helper
    - bpf: Tighten speculative pointer arithmetic mask
    - perf/x86/intel/uncore: Remove uncore extra PCI dev HSWEP_PCI_PCU_3
    - perf/x86/kvm: Fix Broadwell Xeon stepping in isolation_ucodes[]
    - perf auxtrace: Fix potential NULL pointer dereference
    - perf map: Fix error return code in maps__clone()
    - HID: google: add don USB id
    - HID: asus: Add support for 2021 ASUS N-Key keyboard
    - HID: alps: fix error return code in alps_input_configured()
    - HID cp2112: fix support for multiple gpiochips
    - HID: wacom: Assign boolean values to a bool variable
    - soc: qcom: geni: shield geni_icc_get() for ACPI boot
    - dmaengine: xilinx: dpdma: Fix descriptor issuing on video group
    - dmaengine: xilinx: dpdma: Fix race condition in done IRQ
    - ARM: dts: Fix swapped mmc order for omap3
    - m68k: fix flatmem memory model setup
    - net: geneve: check skb is large enough for IPv4/IPv6 header
    - dmaengine: tegra20: Fix runtime PM imbalance on error
    - s390/entry: save the caller of psw_idle
    - arm64: kprobes: Restore local irqflag if kprobes is cancelled
    - xen-netback: Check for hotplug-status existence before watching
    - cavium/liquidio: Fix duplicate argument
    - csky: change a Kconfig symbol name to fix e1000 build error
    - ia64: fix discontig.c section mis...