[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <9b01758f-4441-4f4a-9803-400a94a48e5f@redhat.com>
Date: Mon, 22 Dec 2025 23:18:32 +0100
From: Paolo Bonzini <pbonzini@...hat.com>
To: linux-kernel@...r.kernel.org, kvm@...r.kernel.org
Subject: Re: [PATCH] KVM: x86: kvm_fpu_get() is fpregs_lock_and_load()
On 12/22/25 22:33, Paolo Bonzini wrote:
> The only difference is usage of switch_fpu_return() vs.
> fpregs_restore_userregs(). In turn, these are only different
> if there is no FPU at all, but KVM requires one. Therefore use the
> pre-made export---the code is simpler and there is no functional change.
>
> Signed-off-by: Paolo Bonzini <pbonzini@...hat.com>
> ---
> arch/x86/kernel/fpu/core.c | 2 +-
> arch/x86/kvm/fpu.h | 6 +-----
> 2 files changed, 2 insertions(+), 6 deletions(-)
>
> diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
> index 3ab27fb86618..8ded41b023a2 100644
> --- a/arch/x86/kernel/fpu/core.c
> +++ b/arch/x86/kernel/fpu/core.c
> @@ -858,7 +858,6 @@ void switch_fpu_return(void)
>
> fpregs_restore_userregs();
> }
> -EXPORT_SYMBOL_FOR_KVM(switch_fpu_return);
Oops, unlike in my tree (patch to be sent tomorrow after I write a
testcase) there's still an occurrence of switch_fpu_return() in
vcpu_enter_guest().
Paolo
> void fpregs_lock_and_load(void)
> {
> @@ -877,6 +876,7 @@ void fpregs_lock_and_load(void)
>
> fpregs_assert_state_consistent();
> }
> +EXPORT_SYMBOL_FOR_KVM(fpregs_lock_and_load);
>
> #ifdef CONFIG_X86_DEBUG_FPU
> /*
> diff --git a/arch/x86/kvm/fpu.h b/arch/x86/kvm/fpu.h
> index f898781b6a06..b6a03d8fa8af 100644
> --- a/arch/x86/kvm/fpu.h
> +++ b/arch/x86/kvm/fpu.h
> @@ -149,11 +149,7 @@ static inline void _kvm_write_mmx_reg(int reg, const u64 *data)
>
> static inline void kvm_fpu_get(void)
> {
> - fpregs_lock();
> -
> - fpregs_assert_state_consistent();
> - if (test_thread_flag(TIF_NEED_FPU_LOAD))
> - switch_fpu_return();
> + fpregs_lock_and_load();
> }
>
> static inline void kvm_fpu_put(void)
Powered by blists - more mailing lists