linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org,
	Krish Sadhukhan <krish.sadhukhan@oracle.com>,
	Liran Alon <liran.alon@oracle.com>,
	Paolo Bonzini <pbonzini@redhat.com>
Subject: [PATCH 4.17 07/63] KVM: VMX: Mark VMXArea with revision_id of physical CPU even when eVMCS enabled
Date: Mon, 23 Jul 2018 14:24:13 +0200	[thread overview]
Message-ID: <20180723122446.654794759@linuxfoundation.org> (raw)
In-Reply-To: <20180723122446.351334162@linuxfoundation.org>

4.17-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Liran Alon <liran.alon@oracle.com>

commit 2307af1c4b2e0ad886f30e31739845322cbd328b upstream.

When eVMCS is enabled, all VMCS allocated to be used by KVM are marked
with revision_id of KVM_EVMCS_VERSION instead of revision_id reported
by MSR_IA32_VMX_BASIC.

However, even though not explictly documented by TLFS, VMXArea passed
as VMXON argument should still be marked with revision_id reported by
physical CPU.

This issue was found by the following setup:
* L0 = KVM which expose eVMCS to it's L1 guest.
* L1 = KVM which consume eVMCS reported by L0.
This setup caused the following to occur:
1) L1 execute hardware_enable().
2) hardware_enable() calls kvm_cpu_vmxon() to execute VMXON.
3) L0 intercept L1 VMXON and execute handle_vmon() which notes
vmxarea->revision_id != VMCS12_REVISION and therefore fails with
nested_vmx_failInvalid() which sets RFLAGS.CF.
4) L1 kvm_cpu_vmxon() don't check RFLAGS.CF for failure and therefore
hardware_enable() continues as usual.
5) L1 hardware_enable() then calls ept_sync_global() which executes
INVEPT.
6) L0 intercept INVEPT and execute handle_invept() which notes
!vmx->nested.vmxon and thus raise a #UD to L1.
7) Raised #UD caused L1 to panic.

Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Cc: stable@vger.kernel.org
Fixes: 773e8a0425c923bc02668a2d6534a5ef5a43cc69
Signed-off-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 arch/x86/kvm/vmx.c |   27 +++++++++++++++++++++------
 1 file changed, 21 insertions(+), 6 deletions(-)

--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -4110,11 +4110,7 @@ static __init int setup_vmcs_config(stru
 	vmcs_conf->order = get_order(vmcs_conf->size);
 	vmcs_conf->basic_cap = vmx_msr_high & ~0x1fff;
 
-	/* KVM supports Enlightened VMCS v1 only */
-	if (static_branch_unlikely(&enable_evmcs))
-		vmcs_conf->revision_id = KVM_EVMCS_VERSION;
-	else
-		vmcs_conf->revision_id = vmx_msr_low;
+	vmcs_conf->revision_id = vmx_msr_low;
 
 	vmcs_conf->pin_based_exec_ctrl = _pin_based_exec_control;
 	vmcs_conf->cpu_based_exec_ctrl = _cpu_based_exec_control;
@@ -4184,7 +4180,13 @@ static struct vmcs *alloc_vmcs_cpu(int c
 		return NULL;
 	vmcs = page_address(pages);
 	memset(vmcs, 0, vmcs_config.size);
-	vmcs->revision_id = vmcs_config.revision_id; /* vmcs revision id */
+
+	/* KVM supports Enlightened VMCS v1 only */
+	if (static_branch_unlikely(&enable_evmcs))
+		vmcs->revision_id = KVM_EVMCS_VERSION;
+	else
+		vmcs->revision_id = vmcs_config.revision_id;
+
 	return vmcs;
 }
 
@@ -4343,6 +4345,19 @@ static __init int alloc_kvm_area(void)
 			return -ENOMEM;
 		}
 
+		/*
+		 * When eVMCS is enabled, alloc_vmcs_cpu() sets
+		 * vmcs->revision_id to KVM_EVMCS_VERSION instead of
+		 * revision_id reported by MSR_IA32_VMX_BASIC.
+		 *
+		 * However, even though not explictly documented by
+		 * TLFS, VMXArea passed as VMXON argument should
+		 * still be marked with revision_id reported by
+		 * physical CPU.
+		 */
+		if (static_branch_unlikely(&enable_evmcs))
+			vmcs->revision_id = vmcs_config.revision_id;
+
 		per_cpu(vmxarea, cpu) = vmcs;
 	}
 	return 0;



  parent reply	other threads:[~2018-07-23 12:26 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-23 12:24 [PATCH 4.17 00/63] 4.17.10-stable review Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 01/63] scsi: sd_zbc: Fix variable type and bogus comment Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 02/63] scsi: qla2xxx: Fix inconsistent DMA mem alloc/free Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 03/63] scsi: qla2xxx: Fix kernel crash due to late workqueue allocation Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 04/63] scsi: qla2xxx: Fix NULL pointer dereference for fcport search Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 06/63] KVM: irqfd: fix race between EPOLLHUP and irq_bypass_register_consumer Greg Kroah-Hartman
2018-07-23 12:24 ` Greg Kroah-Hartman [this message]
2018-07-23 12:24 ` [PATCH 4.17 08/63] x86/kvm/vmx: dont read current->thread.{fs,gs}base of legacy tasks Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 11/63] x86/events/intel/ds: Fix bts_interrupt_threshold alignment Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 12/63] x86/MCE: Remove min interval polling limitation Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 13/63] fat: fix memory allocation failure handling of match_strdup() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 14/63] ALSA: rawmidi: Change resized buffers atomically Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 15/63] ALSA: hda/realtek - Add Panasonic CF-SZ6 headset jack quirk Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 16/63] ALSA: hda/realtek - Yet another Clevo P950 quirk entry Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 17/63] ALSA: hda: add mute led support for HP ProBook 455 G5 Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 18/63] ARCv2: [plat-hsdk]: Save accl reg pair by default Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 19/63] ARC: Fix CONFIG_SWAP Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 20/63] ARC: configs: Remove CONFIG_INITRAMFS_SOURCE from defconfigs Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 21/63] ARC: mm: allow mprotect to make stack mappings executable Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 22/63] mm: memcg: fix use after free in mem_cgroup_iter() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 23/63] mm/huge_memory.c: fix data loss when splitting a file pmd Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 24/63] cpufreq: intel_pstate: Register when ACPI PCCH is present Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 25/63] vfio/pci: Fix potential Spectre v1 Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 26/63] vfio/spapr: Use IOMMU pageshift rather than pagesize Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 27/63] stop_machine: Disable preemption when waking two stopper threads Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 28/63] powerpc/powernv: Fix save/restore of SPRG3 on entry/exit from stop (idle) Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 31/63] Revert "drm/amd/display: Dont return ddc result and read_bytes in same return value" Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 32/63] drm/nouveau: Remove bogus crtc check in pmops_runtime_idle Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 33/63] drm/nouveau: Use drm_connector_list_iter_* for iterating connectors Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 34/63] drm/nouveau: Avoid looping through fake MST connectors Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 36/63] ipv4: Return EINVAL when ping_group_range sysctl doesnt map to user ns Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 37/63] ipv6: fix useless rol32 call on hash Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 38/63] ipv6: ila: select CONFIG_DST_CACHE Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 39/63] lib/rhashtable: consider param->min_size when setting initial table size Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 40/63] net: diag: Dont double-free TCP_NEW_SYN_RECV sockets in tcp_abort Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 41/63] net: Dont copy pfmemalloc flag in __copy_skb_header() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 42/63] skbuff: Unconditionally copy pfmemalloc in __skb_clone() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 43/63] net/ipv4: Set oif in fib_compute_spec_dst Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 44/63] net/ipv6: Do not allow device only routes via the multipath API Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 45/63] net: phy: fix flag masking in __set_phy_supported Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 46/63] ptp: fix missing break in switch Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 48/63] rhashtable: add restart routine in rhashtable_free_and_destroy() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 49/63] sch_fq_codel: zero q->flows_cnt when fq_codel_init fails Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 50/63] tg3: Add higher cpu clock for 5762 Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 51/63] net: ip6_gre: get ipv6hdr after skb_cow_head() Greg Kroah-Hartman
2018-07-23 12:24 ` [PATCH 4.17 52/63] sctp: introduce sctp_dst_mtu Greg Kroah-Hartman
2018-07-25 19:19   ` Marcelo Ricardo Leitner
2018-07-23 12:24 ` [PATCH 4.17 53/63] sctp: fix the issue that pathmtu may be set lower than MINSEGMENT Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 54/63] hv_netvsc: Fix napi reschedule while receive completion is busy Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 55/63] net: aquantia: vlan unicast address list correct handling Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 56/63] net/mlx4_en: Dont reuse RX page when XDP is set Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 57/63] net: systemport: Fix CRC forwarding check for SYSTEMPORT Lite Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 58/63] ipv6: make DAD fail with enhanced DAD when nonce length differs Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 59/63] net: usb: asix: replace mii_nway_restart in resume path Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 60/63] alpha: fix osf_wait4() breakage Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 61/63] drm_mode_create_lease_ioctl(): fix open-coded filp_clone_open() Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 62/63] cxl_getfile(): fix double-iput() on alloc_file() failures Greg Kroah-Hartman
2018-07-23 12:25 ` [PATCH 4.17 63/63] xhci: Fix perceived dead host due to runtime suspend race with event handler Greg Kroah-Hartman
2018-07-24  7:38 ` [PATCH 4.17 00/63] 4.17.10-stable review Naresh Kamboju
2018-07-24  8:58   ` Greg Kroah-Hartman
2018-07-24 15:58 ` Guenter Roeck
2018-07-25  7:46   ` Greg Kroah-Hartman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180723122446.654794759@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=krish.sadhukhan@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=liran.alon@oracle.com \
    --cc=pbonzini@redhat.com \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).