FAILED: patch "[PATCH] KVM: x86: Don't (re)check L1 intercepts when completing" failed to apply to 6.1-stable tree

PATCH In review [email protected] 3 participants 3 messages Oct 10, 2025 – Oct 17, 2025

Subsystems (3)

Stable branch x86 KVM x86

lore

Share & tools

lore

[email protected]

author maintainer

2025-10-10 12:35 UTC #0 lore raw

The patch below does not apply to the 6.1-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y
git checkout FETCH_HEAD
git cherry-pick -x e750f85391286a4c8100275516973324b621a269
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2025101005-married-company-a3fc@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^..

Possible dependencies:

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From e750f85391286a4c8100275516973324b621a269 Mon Sep 17 00:00:00 2001
From: Sean Christopherson <[email protected]>
Date: Tue, 15 Jul 2025 12:06:38 -0700
Subject: [PATCH] KVM: x86: Don't (re)check L1 intercepts when completing
userspace I/O

When completing emulation of instruction that generated a userspace exit for I/O, don't recheck L1 intercepts as KVM has already finished that phase of instruction execution, i.e. has already committed to allowing L2 to perform I/O. If L1 (or host userspace) modifies the I/O permission bitmaps during the exit to userspace, KVM will treat the access as being intercepted despite already having emulated the I/O access.

Pivot on EMULTYPE_NO_DECODE to detect that KVM is completing emulation. Of the three users of EMULTYPE_NO_DECODE, only complete_emulated_io() (the intended "recipient") can reach the code in question. gp_interception()'s use is mutually exclusive with is_guest_mode(), and complete_emulated_insn_gp() unconditionally pairs EMULTYPE_NO_DECODE with EMULTYPE_SKIP.

The bad behavior was detected by a syzkaller program that toggles port I/O interception during the userspace I/O exit, ultimately resulting in a WARN on vcpu->arch.pio.count being non-zero due to KVM no completing emulation of the I/O instruction.

  WARNING: CPU: 23 PID: 1083 at arch/x86/kvm/x86.c:8039 emulator_pio_in_out+0x154/0x170 [kvm]
  Modules linked in: kvm_intel kvm irqbypass
  CPU: 23 UID: 1000 PID: 1083 Comm: repro Not tainted 6.16.0-rc5-c1610d2d66b1-next-vm #74 NONE
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:emulator_pio_in_out+0x154/0x170 [kvm]
  PKRU: 55555554
  Call Trace:
   <TASK>
   kvm_fast_pio+0xd6/0x1d0 [kvm]
   vmx_handle_exit+0x149/0x610 [kvm_intel]
   kvm_arch_vcpu_ioctl_run+0xda8/0x1ac0 [kvm]
   kvm_vcpu_ioctl+0x244/0x8c0 [kvm]
   __x64_sys_ioctl+0x8a/0xd0
   do_syscall_64+0x5d/0xc60
   entry_SYSCALL_64_after_hwframe+0x4b/0x53
   </TASK>

Reported-by [email protected]

Closes https://lore.kernel.org/all/[email protected] open here

Fixes 8a76d7f25f8f ("KVM: x86: Add x86 callback for intercept check")

Cc [email protected]

Cc Jim Mattson <[email protected]>

Link https://lore.kernel.org/r/[email protected] open here

Signed-off-by Sean Christopherson <[email protected]>

arch/x86/kvm/emulate.c +4 -5

index 1349e278cd2a..542d3664afa3 100644
@@ -5107,12 +5107,11 @@ void init_decode_cache(struct x86_emulate_ctxt *ctxt)
 	ctxt->mem_read.end = 0;
 }
 
-int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
+int x86_emulate_insn(struct x86_emulate_ctxt *ctxt, bool check_intercepts)
 {
 	const struct x86_emulate_ops *ops = ctxt->ops;
 	int rc = X86EMUL_CONTINUE;
 	int saved_dst_type = ctxt->dst.type;
-	bool is_guest_mode = ctxt->ops->is_guest_mode(ctxt);
 
 	ctxt->mem_read.pos = 0;

@@ -5160,7 +5159,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 				fetch_possible_mmx_operand(&ctxt->dst);
 		}
 
-		if (unlikely(is_guest_mode) && ctxt->intercept) {
+		if (unlikely(check_intercepts) && ctxt->intercept) {
 			rc = emulator_check_intercept(ctxt, ctxt->intercept,
 						      X86_ICPT_PRE_EXCEPT);
 			if (rc != X86EMUL_CONTINUE)

@@ -5189,7 +5188,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 				goto done;
 		}
 
-		if (unlikely(is_guest_mode) && (ctxt->d & Intercept)) {
+		if (unlikely(check_intercepts) && (ctxt->d & Intercept)) {
 			rc = emulator_check_intercept(ctxt, ctxt->intercept,
 						      X86_ICPT_POST_EXCEPT);
 			if (rc != X86EMUL_CONTINUE)

@@ -5243,7 +5242,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 
 special_insn:
 
-	if (unlikely(is_guest_mode) && (ctxt->d & Intercept)) {
+	if (unlikely(check_intercepts) && (ctxt->d & Intercept)) {
 		rc = emulator_check_intercept(ctxt, ctxt->intercept,
 					      X86_ICPT_POST_MEMACCESS);
 		if (rc != X86EMUL_CONTINUE)

arch/x86/kvm/kvm_emulate.h +1 -2

index c1df5acfacaf..7b5ddb787a25 100644
@@ -235,7 +235,6 @@ struct x86_emulate_ops {
 	void (*set_nmi_mask)(struct x86_emulate_ctxt *ctxt, bool masked);
 
 	bool (*is_smm)(struct x86_emulate_ctxt *ctxt);
-	bool (*is_guest_mode)(struct x86_emulate_ctxt *ctxt);
 	int (*leave_smm)(struct x86_emulate_ctxt *ctxt);
 	void (*triple_fault)(struct x86_emulate_ctxt *ctxt);
 	int (*set_xcr)(struct x86_emulate_ctxt *ctxt, u32 index, u64 xcr);

@@ -521,7 +520,7 @@ bool x86_page_table_writing_insn(struct x86_emulate_ctxt *ctxt);
 #define EMULATION_RESTART 1
 #define EMULATION_INTERCEPTED 2
 void init_decode_cache(struct x86_emulate_ctxt *ctxt);
-int x86_emulate_insn(struct x86_emulate_ctxt *ctxt);
+int x86_emulate_insn(struct x86_emulate_ctxt *ctxt, bool check_intercepts);
 int emulator_task_switch(struct x86_emulate_ctxt *ctxt,
 			 u16 tss_selector, int idt_index, int reason,
 			 bool has_error_code, u32 error_code);

arch/x86/kvm/x86.c +8 -7

index a1c49bc681c4..79057622fa76 100644
@@ -8470,11 +8470,6 @@ static bool emulator_is_smm(struct x86_emulate_ctxt *ctxt)
 	return is_smm(emul_to_vcpu(ctxt));
 }
 
-static bool emulator_is_guest_mode(struct x86_emulate_ctxt *ctxt)
-{
-	return is_guest_mode(emul_to_vcpu(ctxt));
-}
-
 #ifndef CONFIG_KVM_SMM
 static int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
 {

@@ -8558,7 +8553,6 @@ static const struct x86_emulate_ops emulate_ops = {
 	.guest_cpuid_is_intel_compatible = emulator_guest_cpuid_is_intel_compatible,
 	.set_nmi_mask        = emulator_set_nmi_mask,
 	.is_smm              = emulator_is_smm,
-	.is_guest_mode       = emulator_is_guest_mode,
 	.leave_smm           = emulator_leave_smm,
 	.triple_fault        = emulator_triple_fault,
 	.set_xcr             = emulator_set_xcr,

@@ -9143,7 +9137,14 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa,
 		ctxt->exception.address = 0;
 	}
 
-	r = x86_emulate_insn(ctxt);
+	/*
+	 * Check L1's instruction intercepts when emulating instructions for
+	 * L2, unless KVM is re-emulating a previously decoded instruction,
+	 * e.g. to complete userspace I/O, in which case KVM has already
+	 * checked the intercepts.
+	 */
+	r = x86_emulate_insn(ctxt, is_guest_mode(vcpu) &&
+				   !(emulation_type & EMULTYPE_NO_DECODE));
 
 	if (r == EMULATION_INTERCEPTED)
 		return 1;

Sasha Levin

maintainer

2025-10-13 15:11 UTC #1 lore raw

PATCH 1/1 KVM: x86: Don't (re)check L1 intercepts when completing userspace I/O

From: Sean Christopherson <[email protected]>

[ Upstream commit e750f85391286a4c8100275516973324b621a269 ]

  WARNING: CPU: 23 PID: 1083 at arch/x86/kvm/x86.c:8039 emulator_pio_in_out+0x154/0x170 [kvm]
  Modules linked in: kvm_intel kvm irqbypass
  CPU: 23 UID: 1000 PID: 1083 Comm: repro Not tainted 6.16.0-rc5-c1610d2d66b1-next-vm #74 NONE
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:emulator_pio_in_out+0x154/0x170 [kvm]
  PKRU: 55555554
  Call Trace:
   <TASK>
   kvm_fast_pio+0xd6/0x1d0 [kvm]
   vmx_handle_exit+0x149/0x610 [kvm_intel]
   kvm_arch_vcpu_ioctl_run+0xda8/0x1ac0 [kvm]
   kvm_vcpu_ioctl+0x244/0x8c0 [kvm]
   __x64_sys_ioctl+0x8a/0xd0
   do_syscall_64+0x5d/0xc60
   entry_SYSCALL_64_after_hwframe+0x4b/0x53
   </TASK>

Reported-by [email protected]

Closes https://lore.kernel.org/all/[email protected] open here

Fixes 8a76d7f25f8f ("KVM: x86: Add x86 callback for intercept check")

Cc [email protected]

Cc Jim Mattson <[email protected]>

Link https://lore.kernel.org/r/[email protected] open here

Signed-off-by Sean Christopherson <[email protected]>

[ is_guest_mode() was open coded ]

Signed-off-by Sasha Levin <[email protected]>

---
 arch/x86/kvm/emulate.c     | 11 ++++-------
 arch/x86/kvm/kvm_emulate.h |  2 +-
 arch/x86/kvm/x86.c         |  9 ++++++++-
 3 files changed, 13 insertions(+), 9 deletions(-)

arch/x86/kvm/emulate.c +4 -7

index 4a43261d25a2a..0cb97e570f63e 100644
@@ -5478,12 +5478,11 @@ void init_decode_cache(struct x86_emulate_ctxt *ctxt)
 	ctxt->mem_read.end = 0;
 }
 
-int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
+int x86_emulate_insn(struct x86_emulate_ctxt *ctxt, bool check_intercepts)
 {
 	const struct x86_emulate_ops *ops = ctxt->ops;
 	int rc = X86EMUL_CONTINUE;
 	int saved_dst_type = ctxt->dst.type;
-	unsigned emul_flags;
 
 	ctxt->mem_read.pos = 0;

@@ -5497,8 +5496,6 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 		rc = emulate_ud(ctxt);
 		goto done;
 	}
-
-	emul_flags = ctxt->ops->get_hflags(ctxt);
 	if (unlikely(ctxt->d &
 		     (No64|Undefined|Sse|Mmx|Intercept|CheckPerm|Priv|Prot|String))) {
 		if ((ctxt->mode == X86EMUL_MODE_PROT64 && (ctxt->d & No64)) ||

@@ -5532,7 +5529,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 				fetch_possible_mmx_operand(&ctxt->dst);
 		}
 
-		if (unlikely(emul_flags & X86EMUL_GUEST_MASK) && ctxt->intercept) {
+		if (unlikely(check_intercepts) && ctxt->intercept) {
 			rc = emulator_check_intercept(ctxt, ctxt->intercept,
 						      X86_ICPT_PRE_EXCEPT);
 			if (rc != X86EMUL_CONTINUE)

@@ -5561,7 +5558,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 				goto done;
 		}
 
-		if (unlikely(emul_flags & X86EMUL_GUEST_MASK) && (ctxt->d & Intercept)) {
+		if (unlikely(check_intercepts) && (ctxt->d & Intercept)) {
 			rc = emulator_check_intercept(ctxt, ctxt->intercept,
 						      X86_ICPT_POST_EXCEPT);
 			if (rc != X86EMUL_CONTINUE)

@@ -5615,7 +5612,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
 
 special_insn:
 
-	if (unlikely(emul_flags & X86EMUL_GUEST_MASK) && (ctxt->d & Intercept)) {
+	if (unlikely(check_intercepts) && (ctxt->d & Intercept)) {
 		rc = emulator_check_intercept(ctxt, ctxt->intercept,
 					      X86_ICPT_POST_MEMACCESS);
 		if (rc != X86EMUL_CONTINUE)

arch/x86/kvm/kvm_emulate.h +1 -1

index 89246446d6aa9..034cafba2cc62 100644
@@ -517,7 +517,7 @@ bool x86_page_table_writing_insn(struct x86_emulate_ctxt *ctxt);
 #define EMULATION_RESTART 1
 #define EMULATION_INTERCEPTED 2
 void init_decode_cache(struct x86_emulate_ctxt *ctxt);
-int x86_emulate_insn(struct x86_emulate_ctxt *ctxt);
+int x86_emulate_insn(struct x86_emulate_ctxt *ctxt, bool check_intercepts);
 int emulator_task_switch(struct x86_emulate_ctxt *ctxt,
 			 u16 tss_selector, int idt_index, int reason,
 			 bool has_error_code, u32 error_code);

arch/x86/kvm/x86.c +8 -1

index 11ca05d830e72..b349e1fbf9dbc 100644
@@ -8976,7 +8976,14 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa,
 		ctxt->exception.address = 0;
 	}
 
-	r = x86_emulate_insn(ctxt);
+	/*
+	 * Check L1's instruction intercepts when emulating instructions for
+	 * L2, unless KVM is re-emulating a previously decoded instruction,
+	 * e.g. to complete userspace I/O, in which case KVM has already
+	 * checked the intercepts.
+	 */
+	r = x86_emulate_insn(ctxt, is_guest_mode(vcpu) &&
+				   !(emulation_type & EMULTYPE_NO_DECODE));
 
 	if (r == EMULATION_INTERCEPTED)
 		return 1;

Signature · 1 line

2.51.0

Sean Christopherson

maintainer Acked-by on v1 1/1

2025-10-17 22:02 UTC #2 lore raw

Re: [PATCH 6.1.y] KVM: x86: Don't (re)check L1 intercepts when completing userspace I/O

Quoted 49 lines · Sasha Levin

From: Sean Christopherson <[email protected]>
[ Upstream commit e750f85391286a4c8100275516973324b621a269 ]
When completing emulation of instruction that generated a userspace exit for I/O, don't recheck L1 intercepts as KVM has already finished that phase of instruction execution, i.e. has already committed to allowing L2 to perform I/O. If L1 (or host userspace) modifies the I/O permission bitmaps during the exit to userspace, KVM will treat the access as being intercepted despite already having emulated the I/O access.
Pivot on EMULTYPE_NO_DECODE to detect that KVM is completing emulation. Of the three users of EMULTYPE_NO_DECODE, only complete_emulated_io() (the intended "recipient") can reach the code in question. gp_interception()'s use is mutually exclusive with is_guest_mode(), and complete_emulated_insn_gp() unconditionally pairs EMULTYPE_NO_DECODE with EMULTYPE_SKIP.
The bad behavior was detected by a syzkaller program that toggles port I/O interception during the userspace I/O exit, ultimately resulting in a WARN on vcpu->arch.pio.count being non-zero due to KVM no completing emulation of the I/O instruction.
  WARNING: CPU: 23 PID: 1083 at arch/x86/kvm/x86.c:8039 emulator_pio_in_out+0x154/0x170 [kvm]
  Modules linked in: kvm_intel kvm irqbypass
  CPU: 23 UID: 1000 PID: 1083 Comm: repro Not tainted 6.16.0-rc5-c1610d2d66b1-next-vm #74 NONE
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:emulator_pio_in_out+0x154/0x170 [kvm]
  PKRU: 55555554
  Call Trace:
   <TASK>
   kvm_fast_pio+0xd6/0x1d0 [kvm]
   vmx_handle_exit+0x149/0x610 [kvm_intel]
   kvm_arch_vcpu_ioctl_run+0xda8/0x1ac0 [kvm]
   kvm_vcpu_ioctl+0x244/0x8c0 [kvm]
   __x64_sys_ioctl+0x8a/0xd0
   do_syscall_64+0x5d/0xc60
   entry_SYSCALL_64_after_hwframe+0x4b/0x53
   </TASK>
Reported-by [email protected]
Closes https://lore.kernel.org/all/[email protected] open here
Fixes 8a76d7f25f8f ("KVM: x86: Add x86 callback for intercept check")
Cc [email protected]
Cc Jim Mattson <[email protected]>
Link https://lore.kernel.org/r/[email protected] open here
Signed-off-by Sean Christopherson <[email protected]>
[ is_guest_mode() was open coded ]
Signed-off-by Sasha Levin <[email protected]>

Quoted 1 line

---

Acked-by Sean Christopherson <[email protected]>

Thanks Sasha!