[PATCH v5 0/6] Delay VERW

[PATCH v5 0/6] Delay VERW

[Pawan Gupta]

v5:
- Added comment to SYM_CODE_START_NOALIGN(mds_verw_sel) explaining VERW
  operand is in code segment so that VERW at works with KPTI. (Josh/Borislav).
- Fixed changelog for patch 1/6. (Borislav)
- Clarify CLEAR_CPU_BUFFERS macro documentation. (Josh)
- KVM: Move the check to skip FB_CLEAR_CTRL optimization for guests when
  X86_FEATURE_CLEAR_CPU_BUF is set. (Josh)
- Rebased to v6.7

v4: https://lore.kernel.org/all/20231027-delay-verw-v4-0-9a3622d4bcf7@linux.intel.com/
- Fill unused part of mds_verw_sel cacheline with int3. (Andrew)
- Fix the formatting in documentation (0-day).
- s/inspite/in spite/ (Sean).
- Explicitly skip FB_CLEAR optimization when MDS affected (Sean).

v3: https://lore.kernel.org/r/20231025-delay-verw-v3-0-52663677ee35@linux.intel.com
- Use .entry.text section for VERW memory operand. (Andrew/PeterZ)
- Fix the duplicate header inclusion. (Chao)

v2: https://lore.kernel.org/r/20231024-delay-verw-v2-0-f1881340c807@linux.intel.com
- Removed the extra EXEC_VERW macro layers. (Sean)
- Move NOPL before VERW. (Sean)
- s/USER_CLEAR_CPU_BUFFERS/CLEAR_CPU_BUFFERS/. (Josh/Dave)
- Removed the comments before CLEAR_CPU_BUFFERS. (Josh)
- Remove CLEAR_CPU_BUFFERS from NMI returning to kernel and document the
  reason. (Josh/Dave)
- Reformat comment in md_clear_update_mitigation(). (Josh)
- Squash "x86/bugs: Cleanup mds_user_clear" patch. (Nikolay)
- s/GUEST_CLEAR_CPU_BUFFERS/CLEAR_CPU_BUFFERS/. (Josh)
- Added a patch from Sean to use CFLAGS.CF for VMLAUNCH/VMRESUME
  selection. This facilitates a single CLEAR_CPU_BUFFERS location for both
  VMLAUNCH and VMRESUME. (Sean)

v1: https://lore.kernel.org/r/20231020-delay-verw-v1-0-cff54096326d@linux.intel.com

Hi,

Legacy instruction VERW was overloaded by some processors to clear
micro-architectural CPU buffers as a mitigation of CPU bugs. This series
moves VERW execution to a later point in exit-to-user path. This is
needed because in some cases it may be possible for kernel data to be
accessed after VERW in arch_exit_to_user_mode(). Such accesses may put
data into MDS affected CPU buffers, for example:

  1. Kernel data accessed by an NMI between VERW and return-to-user can
     remain in CPU buffers (since NMI returning to kernel does not
     execute VERW to clear CPU buffers).
  2. Alyssa reported that after VERW is executed,
     CONFIG_GCC_PLUGIN_STACKLEAK=y scrubs the stack used by a system
     call. Memory accesses during stack scrubbing can move kernel stack
     contents into CPU buffers.
  3. When caller saved registers are restored after a return from
     function executing VERW, the kernel stack accesses can remain in
     CPU buffers(since they occur after VERW).

Although these cases are less practical to exploit, moving VERW closer
to ring transition reduces the attack surface.

Overview of the series:

Patch 1: Prepares VERW macros for use in asm.
Patch 2: Adds macros to 64-bit entry/exit points.
Patch 3: Adds macros to 32-bit entry/exit points.
Patch 4: Enables the new macros.
Patch 5: Uses CFLAGS.CF for VMLAUNCH/VMRESUME selection.
Patch 6: Adds macro to VMenter.

Below is some performance data collected on a Skylake client
compared with previous implementation:

Baseline: v6.6-rc5

| Test               | Configuration          | v1   | v3   |
| ------------------ | ---------------------- | ---- | ---- |
| build-linux-kernel | defconfig              | 1.00 | 1.00 |
| hackbench          | 32 - Process           | 1.02 | 1.06 |
| nginx              | Short Connection - 500 | 1.01 | 1.04 |

Cc: linux-kernel@vger.kernel.org
Cc: linux-doc@vger.kernel.org
Cc: kvm@vger.kernel.org
Cc: Alyssa Milburn <alyssa.milburn@linux.intel.com>
Cc: Daniel Sneddon <daniel.sneddon@linux.intel.com>
Cc: antonio.gomez.iglesias@linux.intel.com
To: Thomas Gleixner <tglx@linutronix.de>
To: Ingo Molnar <mingo@redhat.com>
To: Borislav Petkov <bp@alien8.de>
To: Dave Hansen <dave.hansen@linux.intel.com>
To: x86@kernel.org
To: "H. Peter Anvin" <hpa@zytor.com>
To: Peter Zijlstra <peterz@infradead.org>
To: Josh Poimboeuf <jpoimboe@kernel.org>
To: Andy Lutomirski <luto@kernel.org>
To: Jonathan Corbet <corbet@lwn.net>
To: Sean Christopherson <seanjc@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
To: tony.luck@intel.com
To: ak@linux.intel.com
To: tim.c.chen@linux.intel.com
To: Andrew Cooper <andrew.cooper3@citrix.com>
To: Nikolay Borisov <nik.borisov@suse.com>

Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
Pawan Gupta (5):
      x86/bugs: Add asm helpers for executing VERW
      x86/entry_64: Add VERW just before userspace transition
      x86/entry_32: Add VERW just before userspace transition
      x86/bugs: Use ALTERNATIVE() instead of mds_user_clear static key
      KVM: VMX: Move VERW closer to VMentry for MDS mitigation

Sean Christopherson (1):
      KVM: VMX: Use BT+JNC, i.e. EFLAGS.CF to select VMRESUME vs. VMLAUNCH

 Documentation/arch/x86/mds.rst       | 38 +++++++++++++++++++++++++-----------
 arch/x86/entry/entry.S               | 22 +++++++++++++++++++++
 arch/x86/entry/entry_32.S            |  3 +++
 arch/x86/entry/entry_64.S            | 11 +++++++++++
 arch/x86/entry/entry_64_compat.S     |  1 +
 arch/x86/include/asm/cpufeatures.h   |  2 +-
 arch/x86/include/asm/entry-common.h  |  1 -
 arch/x86/include/asm/nospec-branch.h | 27 +++++++++++++------------
 arch/x86/kernel/cpu/bugs.c           | 15 ++++++--------
 arch/x86/kernel/nmi.c                |  3 ---
 arch/x86/kvm/vmx/run_flags.h         |  7 +++++--
 arch/x86/kvm/vmx/vmenter.S           |  9 ++++++---
 arch/x86/kvm/vmx/vmx.c               | 20 ++++++++++++++++---
 13 files changed, 114 insertions(+), 45 deletions(-)
---
base-commit: 0dd3ee31125508cd67f7e7172247f05b7fd1753a
change-id: 20231011-delay-verw-d0474986b2c3

[PATCH v5 1/6] x86/bugs: Add asm helpers for executing VERW

[Pawan Gupta]

MDS mitigation requires clearing the CPU buffers before returning to
user. This needs to be done late in the exit-to-user path. Current
location of VERW leaves a possibility of kernel data ending up in CPU
buffers for memory accesses done after VERW such as:

  1. Kernel data accessed by an NMI between VERW and return-to-user can
     remain in CPU buffers since NMI returning to kernel does not
     execute VERW to clear CPU buffers.
  2. Alyssa reported that after VERW is executed,
     CONFIG_GCC_PLUGIN_STACKLEAK=y scrubs the stack used by a system
     call. Memory accesses during stack scrubbing can move kernel stack
     contents into CPU buffers.
  3. When caller saved registers are restored after a return from
     function executing VERW, the kernel stack accesses can remain in
     CPU buffers(since they occur after VERW).

To fix this VERW needs to be moved very late in exit-to-user path.

In preparation for moving VERW to entry/exit asm code, create macros
that can be used in asm. Also make VERW patching depend on a new feature
flag X86_FEATURE_CLEAR_CPU_BUF.

Reported-by: Alyssa Milburn <alyssa.milburn@intel.com>
Suggested-by: Andrew Cooper <andrew.cooper3@citrix.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 arch/x86/entry/entry.S               | 22 ++++++++++++++++++++++
 arch/x86/include/asm/cpufeatures.h   |  2 +-
 arch/x86/include/asm/nospec-branch.h | 15 +++++++++++++++
 3 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/arch/x86/entry/entry.S b/arch/x86/entry/entry.S
index 8c8d38f0cb1d..bd8e77c5a375 100644
--- a/arch/x86/entry/entry.S
+++ b/arch/x86/entry/entry.S
@@ -6,6 +6,9 @@
 #include <linux/export.h>
 #include <linux/linkage.h>
 #include <asm/msr-index.h>
+#include <asm/unwind_hints.h>
+#include <asm/segment.h>
+#include <asm/cache.h>
 
 .pushsection .noinstr.text, "ax"
 
@@ -20,3 +23,22 @@ SYM_FUNC_END(entry_ibpb)
 EXPORT_SYMBOL_GPL(entry_ibpb);
 
 .popsection
+
+/*
+ * Defines the VERW operand that is disguised as entry code so that
+ * it can be referenced with KPTI enabled. This ensures VERW can be
+ * used late in exit-to-user path after page tables are switched.
+ */
+.pushsection .entry.text, "ax"
+
+.align L1_CACHE_BYTES, 0xcc
+SYM_CODE_START_NOALIGN(mds_verw_sel)
+	UNWIND_HINT_UNDEFINED
+	ANNOTATE_NOENDBR
+	.word __KERNEL_DS
+.align L1_CACHE_BYTES, 0xcc
+SYM_CODE_END(mds_verw_sel);
+/* For KVM */
+EXPORT_SYMBOL_GPL(mds_verw_sel);
+
+.popsection
diff --git a/arch/x86/include/asm/cpufeatures.h b/arch/x86/include/asm/cpufeatures.h
index 4af140cf5719..79a7e81b9458 100644
--- a/arch/x86/include/asm/cpufeatures.h
+++ b/arch/x86/include/asm/cpufeatures.h
@@ -308,10 +308,10 @@
 #define X86_FEATURE_SMBA		(11*32+21) /* "" Slow Memory Bandwidth Allocation */
 #define X86_FEATURE_BMEC		(11*32+22) /* "" Bandwidth Monitoring Event Configuration */
 #define X86_FEATURE_USER_SHSTK		(11*32+23) /* Shadow stack support for user mode applications */
-
 #define X86_FEATURE_SRSO		(11*32+24) /* "" AMD BTB untrain RETs */
 #define X86_FEATURE_SRSO_ALIAS		(11*32+25) /* "" AMD BTB untrain RETs through aliasing */
 #define X86_FEATURE_IBPB_ON_VMEXIT	(11*32+26) /* "" Issue an IBPB only on VMEXIT */
+#define X86_FEATURE_CLEAR_CPU_BUF	(11*32+27) /* "" Clear CPU buffers using VERW */
 
 /* Intel-defined CPU features, CPUID level 0x00000007:1 (EAX), word 12 */
 #define X86_FEATURE_AVX_VNNI		(12*32+ 4) /* AVX VNNI instructions */
diff --git a/arch/x86/include/asm/nospec-branch.h b/arch/x86/include/asm/nospec-branch.h
index f93e9b96927a..4ea4c310db52 100644
--- a/arch/x86/include/asm/nospec-branch.h
+++ b/arch/x86/include/asm/nospec-branch.h
@@ -315,6 +315,21 @@
 #endif
 .endm
 
+/*
+ * Macros to execute VERW instruction that mitigate transient data sampling
+ * attacks such as MDS. On affected systems a microcode update overloaded VERW
+ * instruction to also clear the CPU buffers. VERW clobbers CFLAGS.ZF.
+ *
+ * Note: Only the memory operand variant of VERW clears the CPU buffers.
+ */
+.macro EXEC_VERW
+	verw _ASM_RIP(mds_verw_sel)
+.endm
+
+.macro CLEAR_CPU_BUFFERS
+	ALTERNATIVE "", __stringify(EXEC_VERW), X86_FEATURE_CLEAR_CPU_BUF
+.endm
+
 #else /* __ASSEMBLY__ */
 
 #define ANNOTATE_RETPOLINE_SAFE					\

-- 
2.34.1

[PATCH v5 2/6] x86/entry_64: Add VERW just before userspace transition

[Pawan Gupta]

Mitigation for MDS is to use VERW instruction to clear any secrets in
CPU Buffers. Any memory accesses after VERW execution can still remain
in CPU buffers. It is safer to execute VERW late in return to user path
to minimize the window in which kernel data can end up in CPU buffers.
There are not many kernel secrets to be had after SWITCH_TO_USER_CR3.

Add support for deploying VERW mitigation after user register state is
restored. This helps minimize the chances of kernel data ending up into
CPU buffers after executing VERW.

Note that the mitigation at the new location is not yet enabled.

  Corner case not handled
  =======================
  Interrupts returning to kernel don't clear CPUs buffers since the
  exit-to-user path is expected to do that anyways. But, there could be
  a case when an NMI is generated in kernel after the exit-to-user path
  has cleared the buffers. This case is not handled and NMI returning to
  kernel don't clear CPU buffers because:

  1. It is rare to get an NMI after VERW, but before returning to userspace.
  2. For an unprivileged user, there is no known way to make that NMI
     less rare or target it.
  3. It would take a large number of these precisely-timed NMIs to mount
     an actual attack.  There's presumably not enough bandwidth.
  4. The NMI in question occurs after a VERW, i.e. when user state is
     restored and most interesting data is already scrubbed. Whats left
     is only the data that NMI touches, and that may or may not be of
     any interest.

Suggested-by: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 arch/x86/entry/entry_64.S        | 11 +++++++++++
 arch/x86/entry/entry_64_compat.S |  1 +
 2 files changed, 12 insertions(+)

diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
index de6469dffe3a..bdb17fad5d04 100644
--- a/arch/x86/entry/entry_64.S
+++ b/arch/x86/entry/entry_64.S
@@ -161,6 +161,7 @@ syscall_return_via_sysret:
 SYM_INNER_LABEL(entry_SYSRETQ_unsafe_stack, SYM_L_GLOBAL)
 	ANNOTATE_NOENDBR
 	swapgs
+	CLEAR_CPU_BUFFERS
 	sysretq
 SYM_INNER_LABEL(entry_SYSRETQ_end, SYM_L_GLOBAL)
 	ANNOTATE_NOENDBR
@@ -601,6 +602,7 @@ SYM_INNER_LABEL(swapgs_restore_regs_and_return_to_usermode, SYM_L_GLOBAL)
 	/* Restore RDI. */
 	popq	%rdi
 	swapgs
+	CLEAR_CPU_BUFFERS
 	jmp	.Lnative_iret
 
 
@@ -712,6 +714,8 @@ native_irq_return_ldt:
 	 */
 	popq	%rax				/* Restore user RAX */
 
+	CLEAR_CPU_BUFFERS
+
 	/*
 	 * RSP now points to an ordinary IRET frame, except that the page
 	 * is read-only and RSP[31:16] are preloaded with the userspace
@@ -1438,6 +1442,12 @@ nmi_restore:
 	std
 	movq	$0, 5*8(%rsp)		/* clear "NMI executing" */
 
+	/*
+	 * Skip CLEAR_CPU_BUFFERS here, since it only helps in rare cases like
+	 * NMI in kernel after user state is restored. For an unprivileged user
+	 * these conditions are hard to meet.
+	 */
+
 	/*
 	 * iretq reads the "iret" frame and exits the NMI stack in a
 	 * single instruction.  We are returning to kernel mode, so this
@@ -1455,6 +1465,7 @@ SYM_CODE_START(entry_SYSCALL32_ignore)
 	UNWIND_HINT_END_OF_STACK
 	ENDBR
 	mov	$-ENOSYS, %eax
+	CLEAR_CPU_BUFFERS
 	sysretl
 SYM_CODE_END(entry_SYSCALL32_ignore)
 
diff --git a/arch/x86/entry/entry_64_compat.S b/arch/x86/entry/entry_64_compat.S
index de94e2e84ecc..eabf48c4d4b4 100644
--- a/arch/x86/entry/entry_64_compat.S
+++ b/arch/x86/entry/entry_64_compat.S
@@ -270,6 +270,7 @@ SYM_INNER_LABEL(entry_SYSRETL_compat_unsafe_stack, SYM_L_GLOBAL)
 	xorl	%r9d, %r9d
 	xorl	%r10d, %r10d
 	swapgs
+	CLEAR_CPU_BUFFERS
 	sysretl
 SYM_INNER_LABEL(entry_SYSRETL_compat_end, SYM_L_GLOBAL)
 	ANNOTATE_NOENDBR

-- 
2.34.1

[PATCH v5 3/6] x86/entry_32: Add VERW just before userspace transition

[Pawan Gupta]

As done for entry_64, add support for executing VERW late in exit to
user path for 32-bit mode.

Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 arch/x86/entry/entry_32.S | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index c73047bf9f4b..fba427646805 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -885,6 +885,7 @@ SYM_FUNC_START(entry_SYSENTER_32)
 	BUG_IF_WRONG_CR3 no_user_check=1
 	popfl
 	popl	%eax
+	CLEAR_CPU_BUFFERS
 
 	/*
 	 * Return back to the vDSO, which will pop ecx and edx.
@@ -954,6 +955,7 @@ restore_all_switch_stack:
 
 	/* Restore user state */
 	RESTORE_REGS pop=4			# skip orig_eax/error_code
+	CLEAR_CPU_BUFFERS
 .Lirq_return:
 	/*
 	 * ARCH_HAS_MEMBARRIER_SYNC_CORE rely on IRET core serialization
@@ -1146,6 +1148,7 @@ SYM_CODE_START(asm_exc_nmi)
 
 	/* Not on SYSENTER stack. */
 	call	exc_nmi
+	CLEAR_CPU_BUFFERS
 	jmp	.Lnmi_return
 
 .Lnmi_from_sysenter_stack:

-- 
2.34.1

[PATCH v5 4/6] x86/bugs: Use ALTERNATIVE() instead of mds_user_clear static key

[Pawan Gupta]

The VERW mitigation at exit-to-user is enabled via a static branch
mds_user_clear. This static branch is never toggled after boot, and can
be safely replaced with an ALTERNATIVE() which is convenient to use in
asm.

Switch to ALTERNATIVE() to use the VERW mitigation late in exit-to-user
path. Also remove the now redundant VERW in exc_nmi() and
arch_exit_to_user_mode().

Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 Documentation/arch/x86/mds.rst       | 38 +++++++++++++++++++++++++-----------
 arch/x86/include/asm/entry-common.h  |  1 -
 arch/x86/include/asm/nospec-branch.h | 12 ------------
 arch/x86/kernel/cpu/bugs.c           | 15 ++++++--------
 arch/x86/kernel/nmi.c                |  3 ---
 arch/x86/kvm/vmx/vmx.c               |  2 +-
 6 files changed, 34 insertions(+), 37 deletions(-)

diff --git a/Documentation/arch/x86/mds.rst b/Documentation/arch/x86/mds.rst
index e73fdff62c0a..c58c72362911 100644
--- a/Documentation/arch/x86/mds.rst
+++ b/Documentation/arch/x86/mds.rst
@@ -95,6 +95,9 @@ The kernel provides a function to invoke the buffer clearing:
 
     mds_clear_cpu_buffers()
 
+Also macro CLEAR_CPU_BUFFERS can be used in ASM late in exit-to-user path.
+Other than CFLAGS.ZF, this macro doesn't clobber any registers.
+
 The mitigation is invoked on kernel/userspace, hypervisor/guest and C-state
 (idle) transitions.
 
@@ -138,17 +141,30 @@ Mitigation points
 
    When transitioning from kernel to user space the CPU buffers are flushed
    on affected CPUs when the mitigation is not disabled on the kernel
-   command line. The migitation is enabled through the static key
-   mds_user_clear.
-
-   The mitigation is invoked in prepare_exit_to_usermode() which covers
-   all but one of the kernel to user space transitions.  The exception
-   is when we return from a Non Maskable Interrupt (NMI), which is
-   handled directly in do_nmi().
-
-   (The reason that NMI is special is that prepare_exit_to_usermode() can
-    enable IRQs.  In NMI context, NMIs are blocked, and we don't want to
-    enable IRQs with NMIs blocked.)
+   command line. The mitigation is enabled through the feature flag
+   X86_FEATURE_CLEAR_CPU_BUF.
+
+   The mitigation is invoked just before transitioning to userspace after
+   user registers are restored. This is done to minimize the window in
+   which kernel data could be accessed after VERW e.g. via an NMI after
+   VERW.
+
+   **Corner case not handled**
+   Interrupts returning to kernel don't clear CPUs buffers since the
+   exit-to-user path is expected to do that anyways. But, there could be
+   a case when an NMI is generated in kernel after the exit-to-user path
+   has cleared the buffers. This case is not handled and NMI returning to
+   kernel don't clear CPU buffers because:
+
+   1. It is rare to get an NMI after VERW, but before returning to userspace.
+   2. For an unprivileged user, there is no known way to make that NMI
+      less rare or target it.
+   3. It would take a large number of these precisely-timed NMIs to mount
+      an actual attack.  There's presumably not enough bandwidth.
+   4. The NMI in question occurs after a VERW, i.e. when user state is
+      restored and most interesting data is already scrubbed. Whats left
+      is only the data that NMI touches, and that may or may not be of
+      any interest.
 
 
 2. C-State transition
diff --git a/arch/x86/include/asm/entry-common.h b/arch/x86/include/asm/entry-common.h
index ce8f50192ae3..7e523bb3d2d3 100644
--- a/arch/x86/include/asm/entry-common.h
+++ b/arch/x86/include/asm/entry-common.h
@@ -91,7 +91,6 @@ static inline void arch_exit_to_user_mode_prepare(struct pt_regs *regs,
 
 static __always_inline void arch_exit_to_user_mode(void)
 {
-	mds_user_clear_cpu_buffers();
 	amd_clear_divider();
 }
 #define arch_exit_to_user_mode arch_exit_to_user_mode
diff --git a/arch/x86/include/asm/nospec-branch.h b/arch/x86/include/asm/nospec-branch.h
index 4ea4c310db52..0a8fa023a804 100644
--- a/arch/x86/include/asm/nospec-branch.h
+++ b/arch/x86/include/asm/nospec-branch.h
@@ -544,7 +544,6 @@ DECLARE_STATIC_KEY_FALSE(switch_to_cond_stibp);
 DECLARE_STATIC_KEY_FALSE(switch_mm_cond_ibpb);
 DECLARE_STATIC_KEY_FALSE(switch_mm_always_ibpb);
 
-DECLARE_STATIC_KEY_FALSE(mds_user_clear);
 DECLARE_STATIC_KEY_FALSE(mds_idle_clear);
 
 DECLARE_STATIC_KEY_FALSE(switch_mm_cond_l1d_flush);
@@ -576,17 +575,6 @@ static __always_inline void mds_clear_cpu_buffers(void)
 	asm volatile("verw %[ds]" : : [ds] "m" (ds) : "cc");
 }
 
-/**
- * mds_user_clear_cpu_buffers - Mitigation for MDS and TAA vulnerability
- *
- * Clear CPU buffers if the corresponding static key is enabled
- */
-static __always_inline void mds_user_clear_cpu_buffers(void)
-{
-	if (static_branch_likely(&mds_user_clear))
-		mds_clear_cpu_buffers();
-}
-
 /**
  * mds_idle_clear_cpu_buffers - Mitigation for MDS vulnerability
  *
diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
index bb0ab8466b91..48d049cd74e7 100644
--- a/arch/x86/kernel/cpu/bugs.c
+++ b/arch/x86/kernel/cpu/bugs.c
@@ -111,9 +111,6 @@ DEFINE_STATIC_KEY_FALSE(switch_mm_cond_ibpb);
 /* Control unconditional IBPB in switch_mm() */
 DEFINE_STATIC_KEY_FALSE(switch_mm_always_ibpb);
 
-/* Control MDS CPU buffer clear before returning to user space */
-DEFINE_STATIC_KEY_FALSE(mds_user_clear);
-EXPORT_SYMBOL_GPL(mds_user_clear);
 /* Control MDS CPU buffer clear before idling (halt, mwait) */
 DEFINE_STATIC_KEY_FALSE(mds_idle_clear);
 EXPORT_SYMBOL_GPL(mds_idle_clear);
@@ -252,7 +249,7 @@ static void __init mds_select_mitigation(void)
 		if (!boot_cpu_has(X86_FEATURE_MD_CLEAR))
 			mds_mitigation = MDS_MITIGATION_VMWERV;
 
-		static_branch_enable(&mds_user_clear);
+		setup_force_cpu_cap(X86_FEATURE_CLEAR_CPU_BUF);
 
 		if (!boot_cpu_has(X86_BUG_MSBDS_ONLY) &&
 		    (mds_nosmt || cpu_mitigations_auto_nosmt()))
@@ -356,7 +353,7 @@ static void __init taa_select_mitigation(void)
 	 * For guests that can't determine whether the correct microcode is
 	 * present on host, enable the mitigation for UCODE_NEEDED as well.
 	 */
-	static_branch_enable(&mds_user_clear);
+	setup_force_cpu_cap(X86_FEATURE_CLEAR_CPU_BUF);
 
 	if (taa_nosmt || cpu_mitigations_auto_nosmt())
 		cpu_smt_disable(false);
@@ -424,7 +421,7 @@ static void __init mmio_select_mitigation(void)
 	 */
 	if (boot_cpu_has_bug(X86_BUG_MDS) || (boot_cpu_has_bug(X86_BUG_TAA) &&
 					      boot_cpu_has(X86_FEATURE_RTM)))
-		static_branch_enable(&mds_user_clear);
+		setup_force_cpu_cap(X86_FEATURE_CLEAR_CPU_BUF);
 	else
 		static_branch_enable(&mmio_stale_data_clear);
 
@@ -484,12 +481,12 @@ static void __init md_clear_update_mitigation(void)
 	if (cpu_mitigations_off())
 		return;
 
-	if (!static_key_enabled(&mds_user_clear))
+	if (!boot_cpu_has(X86_FEATURE_CLEAR_CPU_BUF))
 		goto out;
 
 	/*
-	 * mds_user_clear is now enabled. Update MDS, TAA and MMIO Stale Data
-	 * mitigation, if necessary.
+	 * X86_FEATURE_CLEAR_CPU_BUF is now enabled. Update MDS, TAA and MMIO
+	 * Stale Data mitigation, if necessary.
 	 */
 	if (mds_mitigation == MDS_MITIGATION_OFF &&
 	    boot_cpu_has_bug(X86_BUG_MDS)) {
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c
index 17e955ab69fe..3082cf24b69e 100644
--- a/arch/x86/kernel/nmi.c
+++ b/arch/x86/kernel/nmi.c
@@ -563,9 +563,6 @@ DEFINE_IDTENTRY_RAW(exc_nmi)
 	}
 	if (this_cpu_dec_return(nmi_state))
 		goto nmi_restart;
-
-	if (user_mode(regs))
-		mds_user_clear_cpu_buffers();
 }
 
 #if IS_ENABLED(CONFIG_KVM_INTEL)
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index be20a60047b1..bdcf2c041e0c 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -7229,7 +7229,7 @@ static noinstr void vmx_vcpu_enter_exit(struct kvm_vcpu *vcpu,
 	/* L1D Flush includes CPU buffer clear to mitigate MDS */
 	if (static_branch_unlikely(&vmx_l1d_should_flush))
 		vmx_l1d_flush(vcpu);
-	else if (static_branch_unlikely(&mds_user_clear))
+	else if (cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF))
 		mds_clear_cpu_buffers();
 	else if (static_branch_unlikely(&mmio_stale_data_clear) &&
 		 kvm_arch_has_assigned_device(vcpu->kvm))

-- 
2.34.1

[PATCH v5 5/6] KVM: VMX: Use BT+JNC, i.e. EFLAGS.CF to select VMRESUME vs. VMLAUNCH

[Pawan Gupta]

From: Sean Christopherson <seanjc@google.com>

Use EFLAGS.CF instead of EFLAGS.ZF to track whether to use VMRESUME versus
VMLAUNCH.  Freeing up EFLAGS.ZF will allow doing VERW, which clobbers ZF,
for MDS mitigations as late as possible without needing to duplicate VERW
for both paths.

Reviewed-by: Nikolay Borisov <nik.borisov@suse.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 arch/x86/kvm/vmx/run_flags.h | 7 +++++--
 arch/x86/kvm/vmx/vmenter.S   | 6 +++---
 2 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/arch/x86/kvm/vmx/run_flags.h b/arch/x86/kvm/vmx/run_flags.h
index edc3f16cc189..6a9bfdfbb6e5 100644
--- a/arch/x86/kvm/vmx/run_flags.h
+++ b/arch/x86/kvm/vmx/run_flags.h
@@ -2,7 +2,10 @@
 #ifndef __KVM_X86_VMX_RUN_FLAGS_H
 #define __KVM_X86_VMX_RUN_FLAGS_H
 
-#define VMX_RUN_VMRESUME	(1 << 0)
-#define VMX_RUN_SAVE_SPEC_CTRL	(1 << 1)
+#define VMX_RUN_VMRESUME_SHIFT		0
+#define VMX_RUN_SAVE_SPEC_CTRL_SHIFT	1
+
+#define VMX_RUN_VMRESUME		BIT(VMX_RUN_VMRESUME_SHIFT)
+#define VMX_RUN_SAVE_SPEC_CTRL		BIT(VMX_RUN_SAVE_SPEC_CTRL_SHIFT)
 
 #endif /* __KVM_X86_VMX_RUN_FLAGS_H */
diff --git a/arch/x86/kvm/vmx/vmenter.S b/arch/x86/kvm/vmx/vmenter.S
index be275a0410a8..b3b13ec04bac 100644
--- a/arch/x86/kvm/vmx/vmenter.S
+++ b/arch/x86/kvm/vmx/vmenter.S
@@ -139,7 +139,7 @@ SYM_FUNC_START(__vmx_vcpu_run)
 	mov (%_ASM_SP), %_ASM_AX
 
 	/* Check if vmlaunch or vmresume is needed */
-	test $VMX_RUN_VMRESUME, %ebx
+	bt   $VMX_RUN_VMRESUME_SHIFT, %ebx
 
 	/* Load guest registers.  Don't clobber flags. */
 	mov VCPU_RCX(%_ASM_AX), %_ASM_CX
@@ -161,8 +161,8 @@ SYM_FUNC_START(__vmx_vcpu_run)
 	/* Load guest RAX.  This kills the @regs pointer! */
 	mov VCPU_RAX(%_ASM_AX), %_ASM_AX
 
-	/* Check EFLAGS.ZF from 'test VMX_RUN_VMRESUME' above */
-	jz .Lvmlaunch
+	/* Check EFLAGS.CF from the VMX_RUN_VMRESUME bit test above. */
+	jnc .Lvmlaunch
 
 	/*
 	 * After a successful VMRESUME/VMLAUNCH, control flow "magically"

-- 
2.34.1

[PATCH v5 6/6] KVM: VMX: Move VERW closer to VMentry for MDS mitigation

[Pawan Gupta]

During VMentry VERW is executed to mitigate MDS. After VERW, any memory
access like register push onto stack may put host data in MDS affected
CPU buffers. A guest can then use MDS to sample host data.

Although likelihood of secrets surviving in registers at current VERW
callsite is less, but it can't be ruled out. Harden the MDS mitigation
by moving the VERW mitigation late in VMentry path.

Note that VERW for MMIO Stale Data mitigation is unchanged because of
the complexity of per-guest conditional VERW which is not easy to handle
that late in asm with no GPRs available. If the CPU is also affected by
MDS, VERW is unconditionally executed late in asm regardless of guest
having MMIO access.

Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
---
 arch/x86/kvm/vmx/vmenter.S |  3 +++
 arch/x86/kvm/vmx/vmx.c     | 20 +++++++++++++++++---
 2 files changed, 20 insertions(+), 3 deletions(-)

diff --git a/arch/x86/kvm/vmx/vmenter.S b/arch/x86/kvm/vmx/vmenter.S
index b3b13ec04bac..139960deb736 100644
--- a/arch/x86/kvm/vmx/vmenter.S
+++ b/arch/x86/kvm/vmx/vmenter.S
@@ -161,6 +161,9 @@ SYM_FUNC_START(__vmx_vcpu_run)
 	/* Load guest RAX.  This kills the @regs pointer! */
 	mov VCPU_RAX(%_ASM_AX), %_ASM_AX
 
+	/* Clobbers EFLAGS.ZF */
+	CLEAR_CPU_BUFFERS
+
 	/* Check EFLAGS.CF from the VMX_RUN_VMRESUME bit test above. */
 	jnc .Lvmlaunch
 
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index bdcf2c041e0c..8defba8e417b 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -387,6 +387,17 @@ static __always_inline void vmx_enable_fb_clear(struct vcpu_vmx *vmx)
 
 static void vmx_update_fb_clear_dis(struct kvm_vcpu *vcpu, struct vcpu_vmx *vmx)
 {
+	/*
+	 * FB_CLEAR_CTRL is to optimize VERW latency in guests when host is
+	 * affected by MMIO Stale Data, but not by MDS/TAA. When
+	 * X86_FEATURE_CLEAR_CPU_BUF is enabled, system is likely affected by
+	 * MDS/TAA. Skip the optimization for such a case.
+	 */
+	if (cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF)) {
+		vmx->disable_fb_clear = false;
+		return;
+	}
+
 	vmx->disable_fb_clear = (host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
 				!boot_cpu_has_bug(X86_BUG_MDS) &&
 				!boot_cpu_has_bug(X86_BUG_TAA);
@@ -7226,11 +7237,14 @@ static noinstr void vmx_vcpu_enter_exit(struct kvm_vcpu *vcpu,
 
 	guest_state_enter_irqoff();
 
-	/* L1D Flush includes CPU buffer clear to mitigate MDS */
+	/*
+	 * L1D Flush includes CPU buffer clear to mitigate MDS, but VERW
+	 * mitigation for MDS is done late in VMentry and is still
+	 * executed in spite of L1D Flush. This is because an extra VERW
+	 * should not matter much after the big hammer L1D Flush.
+	 */
 	if (static_branch_unlikely(&vmx_l1d_should_flush))
 		vmx_l1d_flush(vcpu);
-	else if (cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF))
-		mds_clear_cpu_buffers();
 	else if (static_branch_unlikely(&mmio_stale_data_clear) &&
 		 kvm_arch_has_assigned_device(vcpu->kvm))
 		mds_clear_cpu_buffers();

-- 
2.34.1

Re: [PATCH v5 6/6] KVM: VMX: Move VERW closer to VMentry for MDS mitigation

[Sean Christopherson]

On Thu, Jan 11, 2024, Pawan Gupta wrote:
> diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> index bdcf2c041e0c..8defba8e417b 100644
> --- a/arch/x86/kvm/vmx/vmx.c
> +++ b/arch/x86/kvm/vmx/vmx.c
> @@ -387,6 +387,17 @@ static __always_inline void vmx_enable_fb_clear(struct vcpu_vmx *vmx)
>  
>  static void vmx_update_fb_clear_dis(struct kvm_vcpu *vcpu, struct vcpu_vmx *vmx)
>  {
> +	/*
> +	 * FB_CLEAR_CTRL is to optimize VERW latency in guests when host is
> +	 * affected by MMIO Stale Data, but not by MDS/TAA. When
> +	 * X86_FEATURE_CLEAR_CPU_BUF is enabled, system is likely affected by
> +	 * MDS/TAA. Skip the optimization for such a case.

This is unnecessary speculation (ha!), and it'll also be confusing for many readers
as the code below explicitly checks for MDS/TAA.  We have no idea why the host
admin forced the mitigation to be enabled, and it doesn't matter.  The important
thing to capture is that the intent is to keep the mitigation enabled when it
was forcefully enabled, that should be self-explanatory and doesn't require
speculating on _why_ the mitigation was forced on.

> +	 */
> +	if (cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF)) {
> +		vmx->disable_fb_clear = false;
> +		return;
> +	}
> +
>  	vmx->disable_fb_clear = (host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
>  				!boot_cpu_has_bug(X86_BUG_MDS) &&
>  				!boot_cpu_has_bug(X86_BUG_TAA);

I would rather include the X86_FEATURE_CLEAR_CPU_BUF check along with all the
other checks, and then add a common early return. E.g.

	/*
	 * Disable VERW's behavior of clearing CPU buffers for the guest if the
	 * CPU isn't affected MDS/TAA, and the host hasn't forcefully enabled
	 * the mitigation.  Disabing the clearing provides a performance boost
	 * for guests that aren't aware that manually clearing CPU buffers is
	 * unnecessary, at the cost of MSR accesses on VM-Entry and VM-Exit.
	 */
	vmx->disable_fb_clear = !cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF) &&
				(host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
				!boot_cpu_has_bug(X86_BUG_MDS) &&
				!boot_cpu_has_bug(X86_BUG_TAA);

	if (!vmx->disable_fb_clear)
		return;

Re: [PATCH v5 6/6] KVM: VMX: Move VERW closer to VMentry for MDS mitigation

[Pawan Gupta]

On Thu, Jan 11, 2024 at 08:45:13AM -0800, Sean Christopherson wrote:
> On Thu, Jan 11, 2024, Pawan Gupta wrote:
> > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
> > index bdcf2c041e0c..8defba8e417b 100644
> > --- a/arch/x86/kvm/vmx/vmx.c
> > +++ b/arch/x86/kvm/vmx/vmx.c
> > @@ -387,6 +387,17 @@ static __always_inline void vmx_enable_fb_clear(struct vcpu_vmx *vmx)
> >  
> >  static void vmx_update_fb_clear_dis(struct kvm_vcpu *vcpu, struct vcpu_vmx *vmx)
> >  {
> > +	/*
> > +	 * FB_CLEAR_CTRL is to optimize VERW latency in guests when host is
> > +	 * affected by MMIO Stale Data, but not by MDS/TAA. When
> > +	 * X86_FEATURE_CLEAR_CPU_BUF is enabled, system is likely affected by
> > +	 * MDS/TAA. Skip the optimization for such a case.
> 
> This is unnecessary speculation (ha!), and it'll also be confusing for many readers
> as the code below explicitly checks for MDS/TAA.  We have no idea why the host
> admin forced the mitigation to be enabled, and it doesn't matter.  The important
> thing to capture is that the intent is to keep the mitigation enabled when it
> was forcefully enabled, that should be self-explanatory and doesn't require
> speculating on _why_ the mitigation was forced on.

Agree.

> > +	 */
> > +	if (cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF)) {
> > +		vmx->disable_fb_clear = false;
> > +		return;
> > +	}
> > +
> >  	vmx->disable_fb_clear = (host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
> >  				!boot_cpu_has_bug(X86_BUG_MDS) &&
> >  				!boot_cpu_has_bug(X86_BUG_TAA);
> 
> I would rather include the X86_FEATURE_CLEAR_CPU_BUF check along with all the
> other checks, and then add a common early return. E.g.
> 
> 	/*
> 	 * Disable VERW's behavior of clearing CPU buffers for the guest if the
> 	 * CPU isn't affected MDS/TAA, and the host hasn't forcefully enabled
> 	 * the mitigation.  Disabing the clearing provides a performance boost
> 	 * for guests that aren't aware that manually clearing CPU buffers is
> 	 * unnecessary, at the cost of MSR accesses on VM-Entry and VM-Exit.
> 	 */
> 	vmx->disable_fb_clear = !cpu_feature_enabled(X86_FEATURE_CLEAR_CPU_BUF) &&
> 				(host_arch_capabilities & ARCH_CAP_FB_CLEAR_CTRL) &&
> 				!boot_cpu_has_bug(X86_BUG_MDS) &&
> 				!boot_cpu_has_bug(X86_BUG_TAA);
> 
> 	if (!vmx->disable_fb_clear)
> 		return;

This is better. Thanks.