Message-ID: <4bf6fcae-20e7-3eae-83ec-51fb52110487@oracle.com>
Date:   Tue, 5 Jan 2021 00:05:04 +0100
From:   "Maciej S. Szmigiero" <maciej.szmigiero@...cle.com>
To:     Ben Gardon <bgardon@...gle.com>,
        Sean Christopherson <seanjc@...gle.com>
Cc:     linux-kernel <linux-kernel@...r.kernel.org>,
        kvm <kvm@...r.kernel.org>,
        Cannon Matthews <cannonmatthews@...gle.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Peter Xu <peterx@...hat.com>, Peter Shier <pshier@...gle.com>,
        Peter Feiner <pfeiner@...gle.com>,
        Junaid Shahid <junaids@...gle.com>,
        Jim Mattson <jmattson@...gle.com>,
        Yulei Zhang <yulei.kernel@...il.com>,
        Wanpeng Li <kernellwp@...il.com>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        "Xiao Guangrong <xiaoguangrong.eric"@gmail.com
Subject: reproducible BUG() in kvm_mmu_get_root() in TDP MMU

Hi,

I am hitting a reproducible BUG() with KVM TDP MMU.

The reproducer based on set_memory_region_test.c from KVM selftests
is available here:
https://gist.github.com/maciejsszmigiero/890218151c242d99f63ea0825334c6c0

The test simply moves a memslot a bit back and forth on the host
while the guest is concurrently writing around the area being
moved.

The code runs fine on the default KVM MMU but triggers a BUG() when
TDP MMU is enabled by adding "tdp_mmu=1" kvm module parameter.

The backtrace is:
[ 1308.455120] kernel BUG at arch/x86/kvm/mmu/mmu_internal.h:100!
[ 1308.524951] invalid opcode: 0000 [#1] SMP PTI
[ 1308.577080] CPU: 92 PID: 18675 Comm: memslot_move_te Not tainted 5.11.0-rc2+ #80
[ 1308.665617] Hardware name: Oracle Corporation ORACLE SERVER X7-2c/SERVER MODULE ASSY, , BIOS 46070300 12/20/2019
[ 1308.787438] RIP: 0010:kvm_tdp_mmu_get_vcpu_root_hpa+0x10c/0x120 [kvm]
[ 1308.864587] Code: db 74 1c b8 00 00 00 80 48 03 43 40 72 1e 48 c7 c2 00 00 00 80 48 2b 15 92 0a 1d d3 48 01 d0 5b 41 5c 41 5d 41 5e 41 5f 5d c3 <0f> 0b 48 8b 15 eb e8 3c d3 eb e7 66 0f 1f 84 00 00 00 00 00 0f 1f
[ 1309.089393] RSP: 0018:ffffa65affa73d10 EFLAGS: 00010246
[ 1309.151922] RAX: 0000000000000000 RBX: ffff9b46829bac78 RCX: 0000000000000000
[ 1309.237334] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffa65ada1bd000
[ 1309.322744] RBP: ffffa65affa73d38 R08: 0000000000000000 R09: ffff9b454e443200
[ 1309.408156] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000001794
[ 1309.493567] R13: ffffa65ada1bd000 R14: ffff9b454e443040 R15: ffffa65ada1d2418
[ 1309.578977] FS:  00007fdb0430b700(0000) GS:ffff9ba3bfa00000(0000) knlGS:0000000000000000
[ 1309.675833] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 1309.744605] CR2: 0000000000000000 CR3: 0000006090046006 CR4: 00000000007726e0
[ 1309.830018] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 1309.915428] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 1310.000837] PKRU: 55555554
[ 1310.033199] Call Trace:
[ 1310.062445]  kvm_mmu_load+0x29e/0x480 [kvm]
[ 1310.112542]  vcpu_enter_guest+0x112d/0x15b0 [kvm]
[ 1310.168865]  ? vmx_vcpu_load+0x2e/0x40 [kvm_intel]
[ 1310.226201]  kvm_arch_vcpu_ioctl_run+0xf9/0x580 [kvm]
[ 1310.286685]  kvm_vcpu_ioctl+0x247/0x600 [kvm]
[ 1310.338838]  ? tick_program_event+0x44/0x70
[ 1310.388888]  ? __audit_syscall_entry+0xdd/0x130
[ 1310.443101]  __x64_sys_ioctl+0x92/0xd0
[ 1310.487946]  do_syscall_64+0x37/0x50
[ 1310.530711]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 1310.591158] RIP: 0033:0x7fdb44a06307
[ 1310.633925] Code: 44 00 00 48 8b 05 69 1b 2d 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 39 1b 2d 00 f7 d8 64 89 01 48
[ 1310.858726] RSP: 002b:00007fdb0430ae78 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[ 1310.949338] RAX: ffffffffffffffda RBX: 00000000019662f0 RCX: 00007fdb44a06307
[ 1311.034747] RDX: 0000000000000000 RSI: 000000000000ae80 RDI: 0000000000000007
[ 1311.120159] RBP: 0000000001965000 R08: 000000000040b2ff R09: 0000000000000000
[ 1311.205567] R10: 00007fdb0430a2a0 R11: 0000000000000246 R12: 0000000000000000
[ 1311.291738] R13: 0000000001965000 R14: 0000000000000000 R15: 00007fdb0430b700
[ 1311.377873] Modules linked in: kvm_intel kvm xt_comment xt_owner ip6t_rpfilter ip6t_REJECT nf_reject_ipv6 ipt_REJECT nf_reject_ipv4 xt_conntrack ebtable_nat ebtable_broute ip6table_nat ip6table_mangle ip6table_security ip6table_raw iptable_nat nf_nat iptable_mangle iptable_security iptable_raw nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter rpcrdma ib_isert iscsi_target_mod ib_iser ib_srpt target_core_mod ib_srp scsi_transport_srp ib_ipoib rdma_ucm ib_umad iw_cxgb4 rdma_cm iw_cm ib_cm intel_rapl_msr intel_rapl_common skx_edac nfit libnvdimm x86_pkg_temp_thermal intel_powerclamp coretemp bnxt_re ib_uverbs mgag200 ib_core drm_kms_helper cec drm iTCO_wdt iTCO_vendor_support sg irqbypass pcspkr syscopyarea sysfillrect sysimgblt i2c_i801 ioatdma fb_sys_fops joydev i2c_algo_bit i2c_smbus lpc_ich intel_pch_thermal dca ip_tables vfat fat xfs sd_mod t10_pi be2iscsi bnx2i cnic uio cxgb4i cxgb4 tls cxgb3i cxgb3 mdio libcxgbi
[ 1311.377953]  libcxgb qla4xxx iscsi_boot_sysfs crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel crypto_simd cryptd glue_helper bnxt_en wmi sunrpc dm_mirror dm_region_hash dm_log dm_mod iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi [last unloaded: kvm]
[ 1312.712917] ---[ end trace 4716cc8fd037784d ]---
[ 1312.884672] RIP: 0010:kvm_tdp_mmu_get_vcpu_root_hpa+0x10c/0x120 [kvm]
[ 1312.962622] Code: db 74 1c b8 00 00 00 80 48 03 43 40 72 1e 48 c7 c2 00 00 00 80 48 2b 15 92 0a 1d d3 48 01 d0 5b 41 5c 41 5d 41 5e 41 5f 5d c3 <0f> 0b 48 8b 15 eb e8 3c d3 eb e7 66 0f 1f 84 00 00 00 00 00 0f 1f
[ 1313.189000] RSP: 0018:ffffa65affa73d10 EFLAGS: 00010246
[ 1313.252321] RAX: 0000000000000000 RBX: ffff9b46829bac78 RCX: 0000000000000000
[ 1313.338522] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffa65ada1bd000
[ 1313.424727] RBP: ffffa65affa73d38 R08: 0000000000000000 R09: ffff9b454e443200
[ 1313.510932] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000001794
[ 1313.597140] R13: ffffa65ada1bd000 R14: ffff9b454e443040 R15: ffffa65ada1d2418
[ 1313.683343] FS:  00007fdb0430b700(0000) GS:ffff9ba3bfa00000(0000) knlGS:0000000000000000
[ 1313.780987] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 1313.850556] CR2: 0000000000000000 CR3: 0000006090046006 CR4: 00000000007726e0
[ 1313.936759] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 1314.022964] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 1314.109171] PKRU: 55555554
[ 1314.142325] Kernel panic - not syncing: Fatal exception
[ 1314.205755] Kernel Offset: 0x11a00000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
[ 1315.367254] ---[ end Kernel panic - not syncing: Fatal exception ]---

It looks like there might be an inbalance of kvm_mmu_get_root()
and kvm_mmu_put_root() somewhere but I couldn't really nail it down.

I've tried with and without "KVM: x86/mmu: Bug fixes and cleanups in
get_mmio_spte()" series applied, doesn't make any difference.

Thanks,
Maciej

Powered by blists - more mailing lists