Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1357496

[PATCH 4.4 10/50] KVM: MMU: fix ept=0/pte.u=1/pte.w=0/CR0.WP=0/CR4.SMEP=1/EFER.NX=0 combo

From Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Newsgroups linux.kernel
Subject [PATCH 4.4 10/50] KVM: MMU: fix ept=0/pte.u=1/pte.w=0/CR0.WP=0/CR4.SMEP=1/EFER.NX=0 combo
Date 2016-03-14 19:10 +0100
Message-ID <rcF6k-7mS-67@gated-at.bofh.it> (permalink)
References <rcEWC-73F-7@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw

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

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

From: Paolo Bonzini <pbonzini@redhat.com>

commit 844a5fe219cf472060315971e15cbf97674a3324 upstream.

Yes, all of these are needed. :) This is admittedly a bit odd, but
kvm-unit-tests access.flat tests this if you run it with "-cpu host"
and of course ept=0.

KVM runs the guest with CR0.WP=1, so it must handle supervisor writes
specially when pte.u=1/pte.w=0/CR0.WP=0.  Such writes cause a fault
when U=1 and W=0 in the SPTE, but they must succeed because CR0.WP=0.
When KVM gets the fault, it sets U=0 and W=1 in the shadow PTE and
restarts execution.  This will still cause a user write to fault, while
supervisor writes will succeed.  User reads will fault spuriously now,
and KVM will then flip U and W again in the SPTE (U=1, W=0).  User reads
will be enabled and supervisor writes disabled, going back to the
originary situation where supervisor writes fault spuriously.

When SMEP is in effect, however, U=0 will enable kernel execution of
this page.  To avoid this, KVM also sets NX=1 in the shadow PTE together
with U=0.  If the guest has not enabled NX, the result is a continuous
stream of page faults due to the NX bit being reserved.

The fix is to force EFER.NX=1 even if the CPU is taking care of the EFER
switch.  (All machines with SMEP have the CPU_LOAD_IA32_EFER vm-entry
control, so they do not use user-return notifiers for EFER---if they did,
EFER.NX would be forced to the same value as the host).

There is another bug in the reserved bit check, which I've split to a
separate patch for easier application to stable kernels.

Cc: Andy Lutomirski <luto@amacapital.net>
Reviewed-by: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Fixes: f6577a5fa15d82217ca73c74cd2dcbc0f6c781dd
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 Documentation/virtual/kvm/mmu.txt |    3 ++-
 arch/x86/kvm/vmx.c                |   36 +++++++++++++++++++++++-------------
 2 files changed, 25 insertions(+), 14 deletions(-)

--- a/Documentation/virtual/kvm/mmu.txt
+++ b/Documentation/virtual/kvm/mmu.txt
@@ -358,7 +358,8 @@ In the first case there are two addition
 - if CR4.SMEP is enabled: since we've turned the page into a kernel page,
   the kernel may now execute it.  We handle this by also setting spte.nx.
   If we get a user fetch or read fault, we'll change spte.u=1 and
-  spte.nx=gpte.nx back.
+  spte.nx=gpte.nx back.  For this to work, KVM forces EFER.NX to 1 when
+  shadow paging is in use.
 - if CR4.SMAP is disabled: since the page has been changed to a kernel
   page, it can not be reused when CR4.SMAP is enabled. We set
   CR4.SMAP && !CR0.WP into shadow page's role to avoid this case. Note,
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -1792,26 +1792,31 @@ static void reload_tss(void)
 
 static bool update_transition_efer(struct vcpu_vmx *vmx, int efer_offset)
 {
-	u64 guest_efer;
-	u64 ignore_bits;
+	u64 guest_efer = vmx->vcpu.arch.efer;
+	u64 ignore_bits = 0;
 
-	guest_efer = vmx->vcpu.arch.efer;
+	if (!enable_ept) {
+		/*
+		 * NX is needed to handle CR0.WP=1, CR4.SMEP=1.  Testing
+		 * host CPUID is more efficient than testing guest CPUID
+		 * or CR4.  Host SMEP is anyway a requirement for guest SMEP.
+		 */
+		if (boot_cpu_has(X86_FEATURE_SMEP))
+			guest_efer |= EFER_NX;
+		else if (!(guest_efer & EFER_NX))
+			ignore_bits |= EFER_NX;
+	}
 
 	/*
-	 * NX is emulated; LMA and LME handled by hardware; SCE meaningless
-	 * outside long mode
+	 * LMA and LME handled by hardware; SCE meaningless outside long mode.
 	 */
-	ignore_bits = EFER_NX | EFER_SCE;
+	ignore_bits |= EFER_SCE;
 #ifdef CONFIG_X86_64
 	ignore_bits |= EFER_LMA | EFER_LME;
 	/* SCE is meaningful only in long mode on Intel */
 	if (guest_efer & EFER_LMA)
 		ignore_bits &= ~(u64)EFER_SCE;
 #endif
-	guest_efer &= ~ignore_bits;
-	guest_efer |= host_efer & ignore_bits;
-	vmx->guest_msrs[efer_offset].data = guest_efer;
-	vmx->guest_msrs[efer_offset].mask = ~ignore_bits;
 
 	clear_atomic_switch_msr(vmx, MSR_EFER);
 
@@ -1822,16 +1827,21 @@ static bool update_transition_efer(struc
 	 */
 	if (cpu_has_load_ia32_efer ||
 	    (enable_ept && ((vmx->vcpu.arch.efer ^ host_efer) & EFER_NX))) {
-		guest_efer = vmx->vcpu.arch.efer;
 		if (!(guest_efer & EFER_LMA))
 			guest_efer &= ~EFER_LME;
 		if (guest_efer != host_efer)
 			add_atomic_switch_msr(vmx, MSR_EFER,
 					      guest_efer, host_efer);
 		return false;
-	}
+	} else {
+		guest_efer &= ~ignore_bits;
+		guest_efer |= host_efer & ignore_bits;
 
-	return true;
+		vmx->guest_msrs[efer_offset].data = guest_efer;
+		vmx->guest_msrs[efer_offset].mask = ~ignore_bits;
+
+		return true;
+	}
 }
 
 static unsigned long segment_base(u16 selector)

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 4.4 00/50] 4.4.6-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:10 +0100
  [PATCH 4.4 11/50] KVM: MMU: fix reserved bit check for ept=0/CR0.WP=0/CR4.SMEP=1/EFER.NX=0 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:10 +0100
  [PATCH 4.4 47/50] MIPS: smp.c: Fix uninitialised temp_foreign_map Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:10 +0100
  [PATCH 4.4 10/50] KVM: MMU: fix ept=0/pte.u=1/pte.w=0/CR0.WP=0/CR4.SMEP=1/EFER.NX=0 combo Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:10 +0100
  [PATCH 4.4 14/50] tracing: Fix check for cpu online when event is disabled Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:10 +0100
  [PATCH 4.4 39/50] Revert "drm/radeon/pm: adjust display configuration after powerstate" Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 30/50] mac80211: minstrel_ht: set default tx aggregation timeout to 0 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 21/50] ASoC: wm8958: Fix enum ctl accesses in a wrong type Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 27/50] iwlwifi: mvm: inc pending frames counter also when txing non-sta Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 43/50] userfaultfd: dont block on the last VM updates at exit time Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 32/50] mac80211: check PN correctly for GCMP-encrypted fragmented MPDUs Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 26/50] can: gs_usb: fixed disconnect bug by removing erroneous use of kfree() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 40/50] powerpc: Fix dedotify for binutils >= 2.26 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 28/50] mac80211: minstrel: Change expected throughput unit back to Kbps Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 33/50] mac80211: Fix Public Action frame RX in AP mode Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 41/50] powerpc/powernv: Add a kmsg_dumper that flushes console output on panic Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 29/50] mac80211: fix use of uninitialised values in RX aggregation Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 22/50] ovl: ignore lower entries when checking purity of non-directory entries Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 04/50] ARM: OMAP2+: hwmod: Introduce ti,no-idle dt property Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 45/50] ovl: fix getcwd() failure after unsuccessful rmdir Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 23/50] ovl: fix working on distributed fs as lower layer Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 03/50] ARM: dts: dra7: do not gate cpsw clock due to errata i877 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 02/50] ARM: mvebu: fix overlap of Crypto SRAM with PCIe memory window Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 24/50] wext: fix message delay/ordering Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 06/50] kvm: cap halt polling at exactly halt_poll_ns Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 44/50] ovl: copy new uid/gid into overlayfs runtime inode Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 35/50] x86/mm: Fix slow_virt_to_phys() for X86_PAE again Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 42/50] powerpc/powernv: Fix OPAL_CONSOLE_FLUSH prototype and usages Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 34/50] gpu: ipu-v3: Do not bail out on missing optional port nodes Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:20 +0100
  [PATCH 4.4 15/50] dmaengine: at_xdmac: fix residue computation Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:30 +0100
  [PATCH 4.4 25/50] cfg80211/wext: fix message ordering Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:30 +0100
  [PATCH 4.4 20/50] ASoC: wm8994: Fix enum ctl accesses in a wrong type Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:30 +0100
  [PATCH 4.4 19/50] ASoC: samsung: Use IRQ safe spin lock calls Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-14 19:30 +0100
  Re: [PATCH 4.4 00/50] 4.4.6-stable review Shuah Khan <shuahkh@osg.samsung.com> - 2016-03-15 00:20 +0100
    Re: [PATCH 4.4 00/50] 4.4.6-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-16 16:50 +0100
  Re: [PATCH 4.4 00/50] 4.4.6-stable review Guenter Roeck <linux@roeck-us.net> - 2016-03-15 03:40 +0100
    Re: [PATCH 4.4 00/50] 4.4.6-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-03-16 16:50 +0100

csiph-web