Hi After we updated the kernel in Debian bullseye from 5.10.162 to 5.10.178, we got a report from Jared Epp in https://bugs.debian.org/1035779 that a Windows Guest VM no longer booted, and Kernel reporting: [ 51.576266] BUG: kernel NULL pointer dereference, address: 0000000000000000 [ 51.576269] #PF: supervisor read access in kernel mode [ 51.576270] #PF: error_code(0x0000) - not-present page [ 51.576271] PGD 0 P4D 0 [ 51.576273] Oops: 0000 [#1] SMP NOPTI [ 51.576275] CPU: 6 PID: 2209 Comm: CPU 0/KVM Not tainted 5.10.0-22-amd64 #1 Debian 5.10.178-3 [ 51.576276] Hardware name: ASUS System Product Name/CROSSHAIR VI HERO, BIOS 8701 02/08/2023 [ 51.576280] RIP: 0010:find_first_bit+0x19/0x40 [ 51.576281] Code: 5d 41 5c 41 5d 41 5e 41 5f c3 cc cc cc cc cc cc cc 49 89 f0 48 85 f6 74 28 31 c0 eb 0d 48 83 c0 40 48 83 c7 08 4c 39 c0 73 17 <48> 8b 17 48 85 d2 74 eb f3 48 0f bc d2 48 01 d0 49 39 c0 4c 0f 47 [ 51.576282] RSP: 0018:ffffa99ac3a23a30 EFLAGS: 00010246 [ 51.576283] RAX: 0000000000000000 RBX: ffffa99ac38a5000 RCX: 0000000000000000 [ 51.576283] RDX: 0000000000000000 RSI: 0000000000000120 RDI: 0000000000000000 [ 51.576284] RBP: 0000000000000000 R08: 0000000000000120 R09: ffff94e2c1ae72a8 [ 51.576284] R10: 000000000000000f R11: 0000000000000000 R12: ffff94e2c1ae72a8 [ 51.576285] R13: 0000000000000323 R14: 0000000000000003 R15: 0000000000000006 [ 51.576286] FS: 0000000000000000(0053) GS:ffff94e89e980000(002b) knlGS:fffff8033f006000 [ 51.576286] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 51.576287] CR2: 0000000000000000 CR3: 000000018e4ee000 CR4: 0000000000750ee0 [ 51.576287] PKRU: 55555554 [ 51.576288] Call Trace: [ 51.576307] kvm_make_vcpus_request_mask+0x38/0xf0 [kvm] [ 51.576319] kvm_hv_flush_tlb+0x147/0x370 [kvm] [ 51.576328] ? kvm_page_track_is_active+0x12/0x50 [kvm] [ 51.576336] ? make_spte+0x146/0x260 [kvm] [ 51.576344] ? mmu_spte_update+0x11/0x1c0 [kvm] [ 51.576351] ? set_spte+0xee/0x140 [kvm] [ 51.576358] ? mmu_set_spte+0x327/0x4a0 [kvm] [ 51.576365] ? kvm_release_pfn_clean+0x22/0x40 [kvm] [ 51.576372] ? direct_page_fault+0x223/0xa20 [kvm] [ 51.576374] ? svm_get_segment+0x18/0x100 [kvm_amd] [ 51.576382] ? kvm_get_cs_db_l_bits+0x35/0x70 [kvm] [ 51.576383] ? svm_get_segment+0x18/0x100 [kvm_amd] [ 51.576390] ? kvm_get_cs_db_l_bits+0x35/0x70 [kvm] [ 51.576398] kvm_hv_hypercall+0x176/0x580 [kvm] [ 51.576401] ? get_cpu_vendor+0x40/0xa0 [ 51.576403] ? native_load_tr_desc+0x67/0x70 [ 51.576411] kvm_arch_vcpu_ioctl_run+0xbe8/0x1740 [kvm] [ 51.576419] kvm_vcpu_ioctl+0x21e/0x5b0 [kvm] [ 51.576422] __x64_sys_ioctl+0x8b/0xc0 [ 51.576424] do_syscall_64+0x33/0x80 [ 51.576426] entry_SYSCALL_64_after_hwframe+0x61/0xc6 [ 51.576428] RIP: 0033:0x7fad816f2237 [ 51.576429] Code: 00 00 00 48 8b 05 59 cc 0d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 29 cc 0d 00 f7 d8 64 89 01 48 [ 51.576429] RSP: 002b:00007fad7ce65508 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 [ 51.576430] RAX: ffffffffffffffda RBX: 000000000000ae80 RCX: 00007fad816f2237 [ 51.576431] RDX: 0000000000000000 RSI: 000000000000ae80 RDI: 000000000000001c [ 51.576431] RBP: 000055a3e17511c0 R08: 000055a3df109848 R09: 000055a3df5335c0 [ 51.576432] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 [ 51.576432] R13: 000055a3df54fbc0 R14: 00007fad7ce657c0 R15: 0000000000802000 [ 51.576434] Modules linked in: xt_nat veth nft_chain_nat xt_MASQUERADE nf_nat nf_conntrack_netlink xfrm_user xfrm_algo br_netfilter vhost_net vhost vhost_iotlb tap tun bridge stp llc overlay ip6t_REJECT nf_reject_ipv6 xt_hl ip6_tables ip6t_rt ipt_REJECT nf_reject_ipv4 xt_multiport nft_limit snd_hda_codec_realtek snd_hda_codec_generic ledtrig_audio snd_hda_codec_hdmi snd_hda_intel nls_ascii snd_intel_dspcfg nls_cp437 soundwire_intel vfat soundwire_generic_allocation fat snd_soc_core snd_compress soundwire_cadence snd_hda_codec edac_mce_amd xt_limit xt_addrtype kvm_amd snd_hda_core xt_tcpudp snd_hwdep eeepc_wmi kvm soundwire_bus xpad xt_conntrack cdc_acm joydev asus_wmi ff_memless snd_pcm nf_conntrack battery sparse_keymap snd_timer nf_defrag_ipv6 rfkill nf_defrag_ipv4 irqbypass nft_compat snd video rapl efi_pstore wmi_bmof pcspkr ccp soundcore k10temp sp5100_tco nft_counter watchdog sg tpm_crb tpm_tis tpm_tis_core tpm rng_core acpi_cpufreq evdev nf_tables libcrc32c nfnetlink msr fuse [ 51.576464] configfs efivarfs ip_tables x_tables autofs4 ext4 crc16 mbcache jbd2 crc32c_generic dm_crypt dm_mod hid_logitech_hidpp hid_logitech_dj amdgpu sr_mod cdrom gpu_sched sd_mod hid_generic ttm crc32_pclmul crc32c_intel usbhid hid drm_kms_helper ahci cec libahci ghash_clmulni_intel xhci_pci libata xhci_hcd drm nvme aesni_intel mxm_wmi igb libaes usbcore crypto_simd nvme_core cryptd scsi_mod glue_helper i2c_piix4 dca ptp pps_core t10_pi i2c_algo_bit crc_t10dif crct10dif_generic usb_common crct10dif_pclmul crct10dif_common wmi gpio_amdpt gpio_generic button [ 51.576484] CR2: 0000000000000000 [ 51.576485] ---[ end trace acfac62cc884c67c ]--- In fact this was similar to https://forum.proxmox.com/threads/with-latest-5-15-104-1-pve-windows-server-vm-freeze-stuck.125294/ and could be workarounded by dropping the '+hv-tlbflush' flag. Jared confirmed that applying 6470accc7ba9 ("KVM: x86: hyper-v: Avoid calling kvm_make_vcpus_request_mask() with vcpu_mask==NULL") did solve the issue. (Same commit was applied in 5.15.105). Can you pick it for 5.10.y? Regards, Salvatore
