public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Yang Weijiang <weijiang.yang@intel.com>
To: pbonzini@redhat.com, seanjc@google.com, x86@kernel.org,
	kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	rick.p.edgecombe@intel.com
Cc: weijiang.yang@intel.com, Thomas Gleixner <tglx@linutronix.de>
Subject: [PATCH 05/19] x86/fpu: Add helper for modifying xstate
Date: Thu, 16 Jun 2022 04:46:29 -0400	[thread overview]
Message-ID: <20220616084643.19564-6-weijiang.yang@intel.com> (raw)
In-Reply-To: <20220616084643.19564-1-weijiang.yang@intel.com>

From: Rick Edgecombe <rick.p.edgecombe@intel.com>

Just like user xfeatures, supervisor xfeatures can be active in the
registers or present in the task FPU buffer. If the registers are
active, the registers can be modified directly. If the registers are
not active, the modification must be performed on the task FPU buffer.

When the state is not active, the kernel could perform modifications
directly to the buffer. But in order for it to do that, it needs
to know where in the buffer the specific state it wants to modify is
located. Doing this is not robust against optimizations that compact
the FPU buffer, as each access would require computing where in the
buffer it is.

The easiest way to modify supervisor xfeature data is to force restore
the registers and write directly to the MSRs. Often times this is just fine
anyway as the registers need to be restored before returning to userspace.
Do this for now, leaving buffer writing optimizations for the future.

Add a new function fpregs_lock_and_load() that can simultaneously call
fpregs_lock() and do this restore. Also perform some extra sanity
checks in this function since this will be used in non-fpu focused code.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Signed-off-by: Yang Weijiang <weijiang.yang@intel.com>

---
v2:
 - Drop optimization of writing directly the buffer, and change API
   accordingly.
 - fpregs_lock_and_load() suggested by tglx
 - Some commit log verbiage from dhansen

v1:
 - New patch.

 arch/x86/include/asm/fpu/api.h |  7 ++++++-
 arch/x86/kernel/fpu/core.c     | 19 +++++++++++++++++++
 2 files changed, 25 insertions(+), 1 deletion(-)

diff --git a/arch/x86/include/asm/fpu/api.h b/arch/x86/include/asm/fpu/api.h
index 6b0f31fb53f7..4f34812b4dd5 100644
--- a/arch/x86/include/asm/fpu/api.h
+++ b/arch/x86/include/asm/fpu/api.h
@@ -82,6 +82,12 @@ static inline void fpregs_unlock(void)
 		preempt_enable();
 }
 
+/*
+ * Lock and load the fpu state into the registers, if they are not already
+ * loaded.
+ */
+void fpu_lock_and_load(void);
+
 #ifdef CONFIG_X86_DEBUG_FPU
 extern void fpregs_assert_state_consistent(void);
 #else
@@ -163,5 +169,4 @@ static inline bool fpstate_is_confidential(struct fpu_guest *gfpu)
 
 /* prctl */
 extern long fpu_xstate_prctl(int option, unsigned long arg2);
-
 #endif /* _ASM_X86_FPU_API_H */
diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index 0531d6a06df5..4d250dba1619 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -756,6 +756,25 @@ void switch_fpu_return(void)
 }
 EXPORT_SYMBOL_GPL(switch_fpu_return);
 
+void fpu_lock_and_load(void)
+{
+	/*
+	 * fpregs_lock() only disables preemption (mostly). So modifing state
+	 * in an interrupt could screw up some in progress fpregs operation,
+	 * but appear to work. Warn about it.
+	 */
+	WARN_ON_ONCE(!irq_fpu_usable());
+	WARN_ON_ONCE(current->flags & PF_KTHREAD);
+
+	fpregs_lock();
+
+	fpregs_assert_state_consistent();
+
+	if (test_thread_flag(TIF_NEED_FPU_LOAD))
+		fpregs_restore_userregs();
+}
+EXPORT_SYMBOL_GPL(fpu_lock_and_load);
+
 #ifdef CONFIG_X86_DEBUG_FPU
 /*
  * If current FPU state according to its tracking (loaded FPU context on this
-- 
2.27.0


  parent reply	other threads:[~2022-06-16  8:49 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-16  8:46 [PATCH 00/19] Refresh queued CET virtualization series Yang Weijiang
2022-06-16  8:46 ` [PATCH 01/19] x86/cet/shstk: Add Kconfig option for Shadow Stack Yang Weijiang
2022-06-16  8:46 ` [PATCH 02/19] x86/cpufeatures: Add CPU feature flags for shadow stacks Yang Weijiang
2022-06-16  8:46 ` [PATCH 03/19] x86/cpufeatures: Enable CET CR4 bit for shadow stack Yang Weijiang
2022-06-16 10:24   ` Peter Zijlstra
2022-06-16 17:12     ` Edgecombe, Rick P
2022-06-17 11:38       ` Peter Zijlstra
2022-06-17 21:18         ` Edgecombe, Rick P
2022-06-16 10:25   ` Peter Zijlstra
2022-06-16 17:36     ` Edgecombe, Rick P
2022-06-16  8:46 ` [PATCH 04/19] x86/fpu/xstate: Introduce CET MSR and XSAVES supervisor states Yang Weijiang
2022-06-16 10:27   ` Peter Zijlstra
2022-06-16 17:12     ` Edgecombe, Rick P
2022-06-16  8:46 ` Yang Weijiang [this message]
2022-06-16  8:46 ` [PATCH 06/19] KVM: x86: Report XSS as an MSR to be saved if there are supported features Yang Weijiang
2022-06-16  8:46 ` [PATCH 07/19] KVM: x86: Refresh CPUID on writes to MSR_IA32_XSS Yang Weijiang
2022-06-16  8:46 ` [PATCH 08/19] KVM: x86: Load guest fpu state when accessing MSRs managed by XSAVES Yang Weijiang
2022-06-16  8:46 ` [PATCH 09/19] KVM: x86: Add #CP support in guest exception classification Yang Weijiang
2022-06-16  8:46 ` [PATCH 10/19] KVM: VMX: Introduce CET VMCS fields and flags Yang Weijiang
2022-06-16  8:46 ` [PATCH 11/19] KVM: x86: Add fault checks for CR4.CET Yang Weijiang
2022-06-16  8:46 ` [PATCH 12/19] KVM: VMX: Emulate reads and writes to CET MSRs Yang Weijiang
2022-06-16  8:46 ` [PATCH 13/19] KVM: VMX: Add a synthetic MSR to allow userspace VMM to access GUEST_SSP Yang Weijiang
2022-06-16  8:46 ` [PATCH 14/19] KVM: x86: Report CET MSRs as to-be-saved if CET is supported Yang Weijiang
2022-06-16  8:46 ` [PATCH 15/19] KVM: x86: Save/Restore GUEST_SSP to/from SMM state save area Yang Weijiang
2022-06-16  8:46 ` [PATCH 16/19] KVM: x86: Enable CET virtualization for VMX and advertise CET to userspace Yang Weijiang
2022-06-16 10:59   ` Peter Zijlstra
2022-06-16 15:27     ` Yang, Weijiang
2022-06-25  6:55     ` Yang, Weijiang
2022-06-16  8:46 ` [PATCH 17/19] KVM: VMX: Pass through CET MSRs to the guest when supported Yang Weijiang
2022-06-16  8:46 ` [PATCH 18/19] KVM: nVMX: Enable CET support for nested VMX Yang Weijiang
2022-06-16  8:46 ` [PATCH 19/19] KVM: x86: Enable supervisor IBT support for guest Yang Weijiang
2022-06-16 11:05   ` Peter Zijlstra
2022-06-16 11:19   ` Peter Zijlstra
2022-06-16 15:56     ` Yang, Weijiang
2022-06-16  9:10 ` [PATCH 00/19] Refresh queued CET virtualization series Christoph Hellwig
2022-06-16 11:25   ` Peter Zijlstra
2022-06-16 10:12 ` Peter Zijlstra
2022-06-16 10:21   ` Paolo Bonzini
2022-06-16 14:18     ` Peter Zijlstra
2022-06-16 15:06       ` Yang, Weijiang
2022-06-16 15:28       ` Paolo Bonzini
2022-06-18  6:43         ` Yang, Weijiang
2022-07-14 19:36           ` Sean Christopherson
2022-07-15 15:04             ` Yang, Weijiang
2022-07-15 15:58               ` Sean Christopherson

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=20220616084643.19564-6-weijiang.yang@intel.com \
    --to=weijiang.yang@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.com \
    --cc=rick.p.edgecombe@intel.com \
    --cc=seanjc@google.com \
    --cc=tglx@linutronix.de \
    --cc=x86@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