[PATCH] x86/fpu: Ensure XFD state on signal delivery

[PATCH] x86/fpu: Ensure XFD state on signal delivery

[Chang S. Bae]

Sean reported [1] the following splat when running KVM tests:

   WARNING: CPU: 232 PID: 15391 at xfd_validate_state+0x65/0x70
   Call Trace:
    <TASK>
    fpu__clear_user_states+0x9c/0x100
    arch_do_signal_or_restart+0x142/0x210
    exit_to_user_mode_loop+0x55/0x100
    do_syscall_64+0x205/0x2c0
    entry_SYSCALL_64_after_hwframe+0x4b/0x53

Chao further identified [2] a reproducible scenarios involving signal
delivery: a non-AMX task is preempted by an AMX-enabled task which
modifies the XFD MSR.

When the non-AMX task resumes and reloads XSTATE with init values,
a warning is triggered due to a mismatch between fpstate::xfd and the
CPU's current XFD state. fpu__clear_user_states() does not currently
re-synchronize the XFD state after such preemption.

Invoke xfd_update_state() which detects and corrects the mismatch if the
dynamic feature is enabled.

This also benefits the sigreturn path, as fpu__restore_sig() may call
fpu__clear_user_states() when the sigframe is inaccessible.

Fixes: 672365477ae8a ("x86/fpu: Update XFD state where required")
Reported-by: Sean Christopherson <seanjc@google.com>
Closes: https://lore.kernel.org/lkml/aDCo_SczQOUaB2rS@google.com [1]
Tested-by: Chao Gao <chao.gao@intel.com>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/aDWbctO%2FRfTGiCg3@intel.com [2]
---
 arch/x86/kernel/fpu/core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index ea138583dd92..5fa782a2ae7c 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -800,6 +800,9 @@ void fpu__clear_user_states(struct fpu *fpu)
 	    !fpregs_state_valid(fpu, smp_processor_id()))
 		os_xrstor_supervisor(fpu->fpstate);
 
+	/* Ensure XFD state is in sync before reloading XSTATE */
+	xfd_update_state(fpu->fpstate);
+
 	/* Reset user states in registers. */
 	restore_fpregs_from_init_fpstate(XFEATURE_MASK_USER_RESTORE);
 
-- 
2.48.1

Re: [PATCH] x86/fpu: Ensure XFD state on signal delivery

[Chao Gao]

On Mon, Jun 09, 2025 at 05:16:59PM -0700, Chang S. Bae wrote:
>Sean reported [1] the following splat when running KVM tests:
>
>   WARNING: CPU: 232 PID: 15391 at xfd_validate_state+0x65/0x70
>   Call Trace:
>    <TASK>
>    fpu__clear_user_states+0x9c/0x100
>    arch_do_signal_or_restart+0x142/0x210
>    exit_to_user_mode_loop+0x55/0x100
>    do_syscall_64+0x205/0x2c0
>    entry_SYSCALL_64_after_hwframe+0x4b/0x53
>
>Chao further identified [2] a reproducible scenarios involving signal
>delivery: a non-AMX task is preempted by an AMX-enabled task which
>modifies the XFD MSR.
>
>When the non-AMX task resumes and reloads XSTATE with init values,
>a warning is triggered due to a mismatch between fpstate::xfd and the
>CPU's current XFD state. fpu__clear_user_states() does not currently
>re-synchronize the XFD state after such preemption.
>
>Invoke xfd_update_state() which detects and corrects the mismatch if the
>dynamic feature is enabled.
>
>This also benefits the sigreturn path, as fpu__restore_sig() may call
>fpu__clear_user_states() when the sigframe is inaccessible.
>
>Fixes: 672365477ae8a ("x86/fpu: Update XFD state where required")
>Reported-by: Sean Christopherson <seanjc@google.com>
>Closes: https://lore.kernel.org/lkml/aDCo_SczQOUaB2rS@google.com [1]
>Tested-by: Chao Gao <chao.gao@intel.com>
>Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
>Cc: stable@vger.kernel.org
>Link: https://lore.kernel.org/all/aDWbctO%2FRfTGiCg3@intel.com [2]

Reviewed-by: Chao Gao <chao.gao@intel.com>

Thanks for looking into this issue.