All of lore.kernel.org
 help / color / mirror / Atom feed
From: "tip-bot2 for Yu-cheng Yu" <tip-bot2@linutronix.de>
To: linux-tip-commits@vger.kernel.org
Cc: "Yu-cheng Yu" <yu-cheng.yu@intel.com>,
	Borislav Petkov <bp@suse.de>, x86 <x86@kernel.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: [tip: x86/fpu] x86/fpu/xstate: Preserve supervisor states for the slow path in __fpu__restore_sig()
Date: Sat, 16 May 2020 15:10:16 -0000	[thread overview]
Message-ID: <158964181633.17951.13893473131310795449.tip-bot2@tip-bot2> (raw)
In-Reply-To: <20200512145444.15483-10-yu-cheng.yu@intel.com>

The following commit has been merged into the x86/fpu branch of tip:

Commit-ID:     98265c17efa9f2279c59262cd27679aca12e0bb8
Gitweb:        https://git.kernel.org/tip/98265c17efa9f2279c59262cd27679aca12e0bb8
Author:        Yu-cheng Yu <yu-cheng.yu@intel.com>
AuthorDate:    Tue, 12 May 2020 07:54:43 -07:00
Committer:     Borislav Petkov <bp@suse.de>
CommitterDate: Sat, 16 May 2020 12:09:11 +02:00

x86/fpu/xstate: Preserve supervisor states for the slow path in __fpu__restore_sig()

The signal return code is responsible for taking an XSAVE buffer
present in user memory and loading it into the hardware registers. This
operation only affects user XSAVE state and never affects supervisor
state.

The fast path through this code simply points XRSTOR directly at the
user buffer. However, since user memory is not guaranteed to be always
mapped, this XRSTOR can fail. If it fails, the signal return code falls
back to a slow path which can tolerate page faults.

That slow path copies the xfeatures one by one out of the user buffer
into the task's fpu state area. However, by being in a context where it
can handle page faults, the code can also schedule.

The lazy-fpu-load code would think it has an up-to-date fpstate and
would fail to save the supervisor state when scheduling the task out.
When scheduling back in, it would likely restore stale supervisor state.

To fix that, preserve supervisor state before the slow path.  Modify
copy_user_to_fpregs_zeroing() so that if it fails, fpregs are not zeroed,
and there is no need for fpregs_deactivate() and supervisor states are
preserved.

Move set_thread_flag(TIF_NEED_FPU_LOAD) to the slow path.  Without doing
this, the fast path also needs supervisor states to be saved first.

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200512145444.15483-10-yu-cheng.yu@intel.com
---
 arch/x86/kernel/fpu/signal.c | 53 ++++++++++++++++++-----------------
 1 file changed, 28 insertions(+), 25 deletions(-)

diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c
index 77e5c2e..6184fe7 100644
--- a/arch/x86/kernel/fpu/signal.c
+++ b/arch/x86/kernel/fpu/signal.c
@@ -262,19 +262,23 @@ sanitize_restored_user_xstate(union fpregs_state *state,
 static int copy_user_to_fpregs_zeroing(void __user *buf, u64 xbv, int fx_only)
 {
 	u64 init_bv;
+	int r;
 
 	if (use_xsave()) {
 		if (fx_only) {
 			init_bv = xfeatures_mask_user() & ~XFEATURE_MASK_FPSSE;
 
-			copy_kernel_to_xregs(&init_fpstate.xsave, init_bv);
-			return copy_user_to_fxregs(buf);
+			r = copy_user_to_fxregs(buf);
+			if (!r)
+				copy_kernel_to_xregs(&init_fpstate.xsave, init_bv);
+			return r;
 		} else {
 			init_bv = xfeatures_mask_user() & ~xbv;
 
-			if (unlikely(init_bv))
+			r = copy_user_to_xregs(buf, xbv);
+			if (!r && unlikely(init_bv))
 				copy_kernel_to_xregs(&init_fpstate.xsave, init_bv);
-			return copy_user_to_xregs(buf, xbv);
+			return r;
 		}
 	} else if (use_fxsr()) {
 		return copy_user_to_fxregs(buf);
@@ -327,28 +331,10 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size)
 		}
 	}
 
-	/*
-	 * The current state of the FPU registers does not matter. By setting
-	 * TIF_NEED_FPU_LOAD unconditionally it is ensured that the our xstate
-	 * is not modified on context switch and that the xstate is considered
-	 * to be loaded again on return to userland (overriding last_cpu avoids
-	 * the optimisation).
-	 */
-	set_thread_flag(TIF_NEED_FPU_LOAD);
-	__fpu_invalidate_fpregs_state(fpu);
-
 	if ((unsigned long)buf_fx % 64)
 		fx_only = 1;
-	/*
-	 * For 32-bit frames with fxstate, copy the fxstate so it can be
-	 * reconstructed later.
-	 */
-	if (ia32_fxstate) {
-		ret = __copy_from_user(&env, buf, sizeof(env));
-		if (ret)
-			goto err_out;
-		envp = &env;
-	} else {
+
+	if (!ia32_fxstate) {
 		/*
 		 * Attempt to restore the FPU registers directly from user
 		 * memory. For that to succeed, the user access cannot cause
@@ -365,10 +351,27 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size)
 			fpregs_unlock();
 			return 0;
 		}
-		fpregs_deactivate(fpu);
 		fpregs_unlock();
+	} else {
+		/*
+		 * For 32-bit frames with fxstate, copy the fxstate so it can
+		 * be reconstructed later.
+		 */
+		ret = __copy_from_user(&env, buf, sizeof(env));
+		if (ret)
+			goto err_out;
+		envp = &env;
 	}
 
+	/*
+	 * The current state of the FPU registers does not matter. By setting
+	 * TIF_NEED_FPU_LOAD unconditionally it is ensured that the our xstate
+	 * is not modified on context switch and that the xstate is considered
+	 * to be loaded again on return to userland (overriding last_cpu avoids
+	 * the optimisation).
+	 */
+	set_thread_flag(TIF_NEED_FPU_LOAD);
+	__fpu_invalidate_fpregs_state(fpu);
 
 	if (use_xsave() && !fx_only) {
 		u64 init_bv = xfeatures_mask_user() & ~user_xfeatures;

  reply	other threads:[~2020-05-16 15:10 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-12 14:54 [PATCH v4 00/10] Support XSAVES supervisor states Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 01/10] x86/fpu/xstate: Rename validate_xstate_header() to validate_user_xstate_header() Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Fenghua Yu
2020-05-12 14:54 ` [PATCH v4 02/10] x86/fpu/xstate: Define new macros for supervisor and user xstates Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Fenghua Yu
2020-05-12 14:54 ` [PATCH v4 03/10] x86/fpu/xstate: Separate user and supervisor xfeatures mask Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 04/10] x86/fpu/xstate: Introduce XSAVES supervisor states Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 05/10] x86/fpu/xstate: Define new functions for clearing fpregs and xstates Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Fenghua Yu
2021-05-24 16:34     ` Andy Lutomirski
2021-05-25 17:44       ` Yu, Yu-cheng
2021-05-25 18:00         ` Thomas Gleixner
2022-11-29 11:19           ` Ivan Zahariev
2022-11-29 18:16             ` Dave Hansen
2022-12-01 12:58               ` Ivan Zahariev
2022-12-01 14:04                 ` Dave Hansen
2020-05-12 14:54 ` [PATCH v4 06/10] x86/fpu/xstate: Update sanitize_restored_xstate() for supervisor xstates Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 07/10] x86/fpu/xstate: Update copy_kernel_to_xregs_err() for XSAVES supervisor states Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] x86/fpu/xstate: Update copy_kernel_to_xregs_err() for " tip-bot2 for Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 08/10] x86/fpu: Introduce copy_supervisor_to_kernel() Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Yu-cheng Yu
2020-05-12 14:54 ` [PATCH v4 09/10] x86/fpu/xstate: Preserve supervisor states for slow path of __fpu__restore_sig() Yu-cheng Yu
2020-05-16 15:10   ` tip-bot2 for Yu-cheng Yu [this message]
2020-05-12 14:54 ` [PATCH v4 10/10] x86/fpu/xstate: Restore supervisor states for signal return Yu-cheng Yu
2020-05-16 15:10   ` [tip: x86/fpu] " tip-bot2 for Yu-cheng Yu

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=158964181633.17951.13893473131310795449.tip-bot2@tip-bot2 \
    --to=tip-bot2@linutronix.de \
    --cc=bp@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-tip-commits@vger.kernel.org \
    --cc=x86@kernel.org \
    --cc=yu-cheng.yu@intel.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.