From: Sasha Levin <sashal@kernel.org>
To: stable@vger.kernel.org
Cc: David Matlack <dmatlack@google.com>,
Sean Christopherson <seanjc@google.com>,
Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.10.y 4/5] KVM: Introduce vcpu->wants_to_run
Date: Thu, 30 Jul 2026 15:29:36 -0400 [thread overview]
Message-ID: <20260730192937.3123765-4-sashal@kernel.org> (raw)
In-Reply-To: <20260730192937.3123765-1-sashal@kernel.org>
From: David Matlack <dmatlack@google.com>
[ Upstream commit a6816314af5749cd88944bfdceb270c627cdf348 ]
Introduce vcpu->wants_to_run to indicate when a vCPU is in its core run
loop, i.e. when the vCPU is running the KVM_RUN ioctl and immediate_exit
was not set.
Replace all references to vcpu->run->immediate_exit with
!vcpu->wants_to_run to avoid TOCTOU races with userspace. For example, a
malicious userspace could invoked KVM_RUN with immediate_exit=true and
then after KVM reads it to set wants_to_run=false, flip it to false.
This would result in the vCPU running in KVM_RUN with
wants_to_run=false. This wouldn't cause any real bugs today but is a
dangerous landmine.
Signed-off-by: David Matlack <dmatlack@google.com>
Link: https://lore.kernel.org/r/20240503181734.1467938-2-dmatlack@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Stable-dep-of: e800decd9c0a ("KVM: x86: Only reset TSC Deadline Timer in apic_timer_expired on KVM_RUN")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
arch/arm64/kvm/arm.c | 2 +-
arch/mips/kvm/mips.c | 2 +-
arch/powerpc/kvm/powerpc.c | 2 +-
arch/s390/kvm/kvm-s390.c | 2 +-
arch/x86/kvm/x86.c | 4 ++--
include/linux/kvm_host.h | 1 +
virt/kvm/kvm_main.c | 3 +++
7 files changed, 10 insertions(+), 6 deletions(-)
diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index afe8be2fef88..309a925c2e7a 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -686,7 +686,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
return ret;
}
- if (run->immediate_exit)
+ if (!vcpu->wants_to_run)
return -EINTR;
vcpu_load(vcpu);
diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 3d6a7f5827b1..e70b8a5793e0 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -464,7 +464,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
vcpu->mmio_needed = 0;
}
- if (vcpu->run->immediate_exit)
+ if (!vcpu->wants_to_run)
goto out;
lose_fpu(1);
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 0f5ebec660a7..103d389b59b5 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1832,7 +1832,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
kvm_sigset_activate(vcpu);
- if (run->immediate_exit)
+ if (!vcpu->wants_to_run)
r = -EINTR;
else
r = kvmppc_vcpu_run(vcpu);
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index f6c27b44766f..15cd2c22ed00 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -4373,7 +4373,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
struct kvm_run *kvm_run = vcpu->run;
int rc;
- if (kvm_run->immediate_exit)
+ if (!vcpu->wants_to_run)
return -EINTR;
if (kvm_run->kvm_valid_regs & ~KVM_SYNC_S390_VALID_FIELDS ||
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index f9cfba9ec644..1b8860a1c7c2 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -9648,7 +9648,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
kvm_load_guest_fpu(vcpu);
if (unlikely(vcpu->arch.mp_state == KVM_MP_STATE_UNINITIALIZED)) {
- if (kvm_run->immediate_exit) {
+ if (!vcpu->wants_to_run) {
r = -EINTR;
goto out;
}
@@ -9692,7 +9692,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
} else
WARN_ON(vcpu->arch.pio.count || vcpu->mmio_needed);
- if (kvm_run->immediate_exit)
+ if (!vcpu->wants_to_run)
r = -EINTR;
else
r = vcpu_run(vcpu);
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 1e6a377af368..5b9c303876a5 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -318,6 +318,7 @@ struct kvm_vcpu {
bool dy_eligible;
} spin_loop;
#endif
+ bool wants_to_run;
bool preempted;
bool ready;
struct kvm_vcpu_arch arch;
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index b340be2d81a3..ae961d487f94 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -3366,7 +3366,10 @@ static long kvm_vcpu_ioctl(struct file *filp,
synchronize_rcu();
put_pid(oldpid);
}
+ vcpu->wants_to_run = !READ_ONCE(vcpu->run->immediate_exit);
r = kvm_arch_vcpu_ioctl_run(vcpu);
+ vcpu->wants_to_run = false;
+
trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
break;
}
--
2.53.0
next prev parent reply other threads:[~2026-07-30 19:29 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-25 6:53 FAILED: patch "[PATCH] KVM: x86: Only reset TSC Deadline Timer in apic_timer_expired" failed to apply to 5.10-stable tree gregkh
2026-07-30 19:29 ` [PATCH 5.10.y 1/5] KVM: x86: Drop @vcpu parameter from kvm_x86_ops.hwapic_isr_update() Sasha Levin
2026-07-30 19:29 ` [PATCH 5.10.y 2/5] KVM: x86: Check for in-kernel xAPIC when querying APICv for directed yield Sasha Levin
2026-07-30 19:29 ` [PATCH 5.10.y 3/5] KVM: x86: Move "apicv_active" into "struct kvm_lapic" Sasha Levin
2026-07-30 19:29 ` Sasha Levin [this message]
2026-07-30 19:29 ` [PATCH 5.10.y 5/5] KVM: x86: Only reset TSC Deadline Timer in apic_timer_expired on KVM_RUN Sasha Levin
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=20260730192937.3123765-4-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=dmatlack@google.com \
--cc=seanjc@google.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