On Tue, Mar 31, 2020 at 07:50:03PM -0700, Cameron Kaiser wrote: > HPT kernel. When removing the kvm_hv module, the kernel goes bananas. Any > ideas? I had to hard-power-down the machine to get control. I don't have a > regression range for this. > > # uname -a > Linux censored.floodgap.com 5.5.10-200.fc31.ppc64le #1 SMP Wed Mar 18 14:11:55 UTC 2020 ppc64le ppc64le ppc64le GNU/Linux > # lsmod | grep kvm > kvm_hv 237807 0 > kvm_pr 144641 0 > kvm 365360 2 kvm_hv,kvm_pr > # dmesg | grep -i mmu > [ 0.000000] dt-cpu-ftrs: final cpu/mmu features = 0x0001f86f8f5fb1a7 0x3c006041 > [ 0.000000] hash-mmu: Page sizes from device-tree: > [ 0.000000] hash-mmu: base_shift=12: shift=12, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=0 > [ 0.000000] hash-mmu: base_shift=12: shift=16, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=7 > [ 0.000000] hash-mmu: base_shift=12: shift=24, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=56 > [ 0.000000] hash-mmu: base_shift=16: shift=16, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=1 > [ 0.000000] hash-mmu: base_shift=16: shift=24, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=8 > [ 0.000000] hash-mmu: base_shift=24: shift=24, sllp=0x0100, avpnm=0x00000001, tlbiel=0, penc=0 > [ 0.000000] hash-mmu: base_shift=34: shift=34, sllp=0x0120, avpnm=0x000007ff, tlbiel=0, penc=3 > [ 0.000000] hash-mmu: Partition table (____ptrval____) > [ 0.000000] hash-mmu: Initializing hash mmu with SLB > [ 0.000000] mmu_features = 0x7c006001 > [ 0.000000] hash-mmu: ppc64_pft_size = 0x0 > [ 0.000000] hash-mmu: htab_hash_mask = 0x3ffff > [ 0.158070] IOMMU table initialized, virtual merging enabled > [ 0.180907] pci 0000:01:00.0: Adding to iommu group 0 > [ 0.180913] pci 0000:01:00.1: Adding to iommu group 0 > [ 0.180928] pci 0001:01:00.0: Adding to iommu group 1 > [ 0.180942] pci 0003:01:00.0: Adding to iommu group 2 > [ 0.180957] pci 0004:01:00.0: Adding to iommu group 3 > [ 0.180963] pci 0004:01:00.1: Adding to iommu group 3 > [ 0.180978] pci 0005:02:00.0: Adding to iommu group 4 > [ 0.180993] pci 0030:01:00.0: Adding to iommu group 5 > [ 0.181007] pci 0031:01:00.0: Adding to iommu group 6 > [ 0.181021] pci 0033:01:00.0: Adding to iommu group 7 > [ 0.449652] iommu: Default domain type: Translated > # modprobe -r kvm_hv > (system becomes unstable) > > Backtrace and dmesg: > > BUG: Kernel NULL pointer dereference on read at 0x00000008 > Faulting instruction address: 0xc00000000088a114 > Oops: Kernel access of bad area, sig: 11 [#1] > LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA PowerNV > Modules linked in: kvm_hv(-) kvm_pr kvm xt_CHECKSUM xt_MASQUERADE xt_conntrack ipt_REJECT nf_reject_ipv4 ip6table_mangle ip6table_nat iptable_mangle iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter tun bridge stp llc rfkill sunrpc vmx_crypto at24 regmap_i2c snd_hda_codec_hdmi snd_usb_audio snd_hda_intel snd_intel_dspcfg snd_hda_codec joydev snd_usbmidi_lib snd_rawmidi snd_hda_core mc snd_hwdep snd_seq crct10dif_vpmsum snd_seq_device ofpart snd_pcm ipmi_powernv powernv_flash ipmi_devintf mtd ipmi_msghandler opal_prd rtc_opal snd_timer i2c_opal snd soundcore ip_tables xfs libcrc32c amdgpu mfd_core gpu_sched i2c_algo_bit ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm firewire_ohci firewire_core nvme crc32c_vpmsum tg3 nvme_core crc_itu_t drm_panel_orientation_quirks i2c_core fuse > CPU: 7 PID: 2268 Comm: modprobe Not tainted 5.5.10-200.fc31.ppc64le #1 > NIP: c00000000088a114 LR: c00000000088a110 CTR: c000000000529520 > REGS: c0000003dc147970 TRAP: 0300 Not tainted (5.5.10-200.fc31.ppc64le) > MSR: 9000000000009033 <SF,HV,EE,ME,IR,DR,RI,LE> CR: 24008288 XER: 20040000 > CFAR: c00000000000dec4 DAR: 0000000000000008 DSISR: 40000000 IRQMASK: 1 > GPR00: c00000000088a110 c0000003dc147c00 c000000001e11100 0000000000000000 > GPR04: 0000000000000000 0000000000000006 0000000000000000 c0000003dac2be00 > GPR08: 0000000000000000 0000000000000001 0000000080000007 c0080000067019c0 > GPR12: c000000000529520 c0000003ffff7f00 0000000125c687c0 0000000125c687b8 > GPR16: 0000000125c687b0 0000000125c687a8 0000000125c68308 0000000000000000 > GPR20: 0000000125c90068 0000000000000000 00007fffe9aa8c58 00007fffe9aa8c58 > GPR24: 00000100052d02a0 0000000000000000 00000100052d1db0 00000100052d1e18 > GPR28: c00000000201a9f0 0000000000000000 c008000006714fd0 0000000000000000 > NIP [c00000000088a114] percpu_ref_kill_and_confirm+0x44/0x160 > LR [c00000000088a110] percpu_ref_kill_and_confirm+0x40/0x160 > Call Trace: > [c0000003dc147c00] [c00000000076de6c] selinux_capable+0xbc/0x220 (unreliable) > [c0000003dc147c80] [c00000000052984c] memunmap_pages+0x32c/0x3f0 > [c0000003dc147d00] [c0080000067002b8] kvmppc_uvmem_free+0x30/0x80 [kvm_hv] > [c0000003dc147d30] [c0080000066f050c] kvmppc_book3s_exit_hv+0x24/0xc8 [kvm_hv] > [c0000003dc147d60] [c00000000027c884] sys_delete_module+0x224/0x3b0 > [c0000003dc147e20] [c00000000000b278] system_call+0x5c/0x68 You have kvmppc_uvmem_free in the backtrace. You probably need either to apply http://patchwork.ozlabs.org/patch/1258499/ "KVM: PPC: Book3S HV: Skip kvmppc_uvmem_free if Ultravisor is not supported" or else turn off CONFIG_PPC_UV. Paul.
