[PATCH] [1/2] i386/x86-64: Fix x87 information leak between processes

[PATCH] [1/2] i386/x86-64: Fix x87 information leak between processes

[Andi Kleen]

AMD K7/K8 CPUs only save/restore the FOP/FIP/FDP x87 registers in FXSAVE
when an exception is pending.  This means the value leak through
context switches and allow processes to observe some x87 instruction 
state of other processes.

This was actually documented by AMD, but nobody recognized it as
being different from Intel before.

The fix first adds an optimization: instead of unconditionally 
calling FNCLEX after each FXSAVE test if ES is pending and skip
it when not needed. Then do a x87 load from a kernel variable to 
clear FOP/FIP/FDP.

This means other processes always will only see a constant value
defined by the kernel in their FP state.

I took some pain to make sure to chose a variable that's already
in L1 during context switch to make the overhead of this low.

Also alternative() is used to patch away the new code on CPUs
who don't need it.

Patch for both i386/x86-64.

The problem was discovered originally by Jan Beulich. Richard
Brunner provided the basic code for the workarounds, with contribution
from Jan.

This is CVE-2006-1056

Cc: richard.brunner@amd.com
Cc: jbeulich@novell.com

Signed-off-by: Andi Kleen <ak@suse.de>

Index: linux/include/asm-x86_64/i387.h
===================================================================
--- linux.orig/include/asm-x86_64/i387.h
+++ linux/include/asm-x86_64/i387.h
@@ -72,6 +72,23 @@ extern int set_fpregs(struct task_struct
 #define set_fpu_swd(t,val) ((t)->thread.i387.fxsave.swd = (val))
 #define set_fpu_fxsr_twd(t,val) ((t)->thread.i387.fxsave.twd = (val))
 
+#define X87_FSW_ES (1 << 7)	/* Exception Summary */
+
+/* AMD CPUs don't save/restore FDP/FIP/FOP unless an exception
+   is pending. Clear the x87 state here by setting it to fixed
+   values. The kernel data segment can be sometimes 0 and sometimes
+   new user value. Both should be ok.
+   Use the PDA as safe address because it should be already in L1. */
+static inline void clear_fpu_state(struct i387_fxsave_struct *fx)
+{
+	if (unlikely(fx->swd & X87_FSW_ES))
+		 asm volatile("fnclex");
+	alternative_input(ASM_NOP8 ASM_NOP2,
+	     	     "    emms\n"		/* clear stack tags */
+	     	     "    fildl %%gs:0",	/* load to clear state */
+		     X86_FEATURE_FXSAVE_LEAK);
+}
+
 static inline int restore_fpu_checking(struct i387_fxsave_struct *fx) 
 { 
 	int err;
@@ -119,6 +136,7 @@ static inline int save_i387_checking(str
 #endif
 	if (unlikely(err))
 		__clear_user(fx, sizeof(struct i387_fxsave_struct));
+	/* No need to clear here because the caller clears USED_MATH */
 	return err;
 } 
 
@@ -149,7 +167,7 @@ static inline void __fxsave_clear(struct
 				"i" (offsetof(__typeof__(*tsk),
 					      thread.i387.fxsave)));
 #endif
-	__asm__ __volatile__("fnclex");
+	clear_fpu_state(&tsk->thread.i387.fxsave);
 }
 
 static inline void kernel_fpu_begin(void)
Index: linux/include/asm-i386/i387.h
===================================================================
--- linux.orig/include/asm-i386/i387.h
+++ linux/include/asm-i386/i387.h
@@ -13,6 +13,7 @@
 
 #include <linux/sched.h>
 #include <linux/init.h>
+#include <linux/kernel_stat.h>
 #include <asm/processor.h>
 #include <asm/sigcontext.h>
 #include <asm/user.h>
@@ -38,17 +39,38 @@ extern void init_fpu(struct task_struct 
 extern void kernel_fpu_begin(void);
 #define kernel_fpu_end() do { stts(); preempt_enable(); } while(0)
 
+/* We need a safe address that is cheap to find and that is already
+   in L1 during context switch. The best choices are unfortunately
+   different for UP and SMP */
+#ifdef CONFIG_SMP
+#define safe_address (__per_cpu_offset[0])
+#else
+#define safe_address (kstat_cpu(0).cpustat.user)
+#endif
+
 /*
  * These must be called with preempt disabled
  */
 static inline void __save_init_fpu( struct task_struct *tsk )
 {
+	/* Use more nops than strictly needed in case the compiler
+	   varies code */
 	alternative_input(
-		"fnsave %1 ; fwait ;" GENERIC_NOP2,
-		"fxsave %1 ; fnclex",
+		"fnsave %[fx] ;fwait;" GENERIC_NOP8 GENERIC_NOP4,
+		"fxsave %[fx]\n"
+		"bt $7,%[fsw] ; jc 1f ; fnclex\n1:",
 		X86_FEATURE_FXSR,
-		"m" (tsk->thread.i387.fxsave)
-		:"memory");
+		[fx] "m" (tsk->thread.i387.fxsave),
+		[fsw] "m" (tsk->thread.i387.fxsave.swd) : "memory");
+	/* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception
+	   is pending.  Clear the x87 state here by setting it to fixed
+   	   values. __per_cpu_offset[0] is a random variable that should be in L1 */
+	alternative_input(
+		GENERIC_NOP8 GENERIC_NOP2,
+		"emms\n\t"	  	/* clear stack tags */
+		"fildl %[addr]", 	/* set F?P to defined value */
+		X86_FEATURE_FXSAVE_LEAK,
+		[addr] "m" (safe_address));
 	task_thread_info(tsk)->status &= ~TS_USEDFPU;
 }
 
Index: linux/arch/i386/kernel/cpu/amd.c
===================================================================
--- linux.orig/arch/i386/kernel/cpu/amd.c
+++ linux/arch/i386/kernel/cpu/amd.c
@@ -207,6 +207,8 @@ static void __init init_amd(struct cpuin
 		set_bit(X86_FEATURE_K7, c->x86_capability); 
 		break;
 	}
+	if (c->x86 >= 6)
+		set_bit(X86_FEATURE_FXSAVE_LEAK, c->x86_capability);
 
 	display_cacheinfo(c);
 
Index: linux/arch/x86_64/kernel/setup.c
===================================================================
--- linux.orig/arch/x86_64/kernel/setup.c
+++ linux/arch/x86_64/kernel/setup.c
@@ -928,6 +928,10 @@ static int __init init_amd(struct cpuinf
 	if (c->x86 == 15 && ((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58))
 		set_bit(X86_FEATURE_REP_GOOD, &c->x86_capability);
 
+	/* Enable workaround for FXSAVE leak */
+	if (c->x86 >= 6)
+		set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
+
 	r = get_model_name(c);
 	if (!r) { 
 		switch (c->x86) { 
Index: linux/include/asm-i386/cpufeature.h
===================================================================
--- linux.orig/include/asm-i386/cpufeature.h
+++ linux/include/asm-i386/cpufeature.h
@@ -71,6 +71,7 @@
 #define X86_FEATURE_P4		(3*32+ 7) /* P4 */
 #define X86_FEATURE_CONSTANT_TSC (3*32+ 8) /* TSC ticks at a constant rate */
 #define X86_FEATURE_UP		(3*32+ 9) /* smp kernel running on up */
+#define X86_FEATURE_FXSAVE_LEAK (3*32+10) /* FXSAVE leaks FOP/FIP/FOP */
 
 /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
 #define X86_FEATURE_XMM3	(4*32+ 0) /* Streaming SIMD Extensions-3 */
Index: linux/include/asm-x86_64/cpufeature.h
===================================================================
--- linux.orig/include/asm-x86_64/cpufeature.h
+++ linux/include/asm-x86_64/cpufeature.h
@@ -64,6 +64,7 @@
 #define X86_FEATURE_REP_GOOD	(3*32+ 4) /* rep microcode works well on this CPU */
 #define X86_FEATURE_CONSTANT_TSC (3*32+5) /* TSC runs at constant rate */
 #define X86_FEATURE_SYNC_RDTSC  (3*32+6)  /* RDTSC syncs CPU core */
+#define X86_FEATURE_FXSAVE_LEAK (3*32+7)  /* FIP/FOP/FDP leaks through FXSAVE */
 
 /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
 #define X86_FEATURE_XMM3	(4*32+ 0) /* Streaming SIMD Extensions-3 */
Index: linux/arch/x86_64/kernel/process.c
===================================================================
--- linux.orig/arch/x86_64/kernel/process.c
+++ linux/arch/x86_64/kernel/process.c
@@ -575,8 +575,10 @@ __switch_to(struct task_struct *prev_p, 
 	prev->userrsp = read_pda(oldrsp); 
 	write_pda(oldrsp, next->userrsp); 
 	write_pda(pcurrent, next_p); 
+
 	/* This must be here to ensure both math_state_restore() and
-	   kernel_fpu_begin() work consistently. */
+	   kernel_fpu_begin() work consistently. 
+	   And the AMD workaround requires it to be after DS reload. */
 	unlazy_fpu(prev_p);
 	write_pda(kernelstack,
 		  task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);

Re: [PATCH] [1/2] i386/x86-64: Fix x87 information leak between processes

[Matt Heler]

Patch says 1/2 . Is there another patch that comes with this ? Or is vger 
lagging again ? 

On Wednesday 19 April 2006 5:36 pm, Andi Kleen wrote:
> AMD K7/K8 CPUs only save/restore the FOP/FIP/FDP x87 registers in FXSAVE
> when an exception is pending.  This means the value leak through
> context switches and allow processes to observe some x87 instruction
> state of other processes.
>
> This was actually documented by AMD, but nobody recognized it as
> being different from Intel before.
>
> The fix first adds an optimization: instead of unconditionally
> calling FNCLEX after each FXSAVE test if ES is pending and skip
> it when not needed. Then do a x87 load from a kernel variable to
> clear FOP/FIP/FDP.
>
> This means other processes always will only see a constant value
> defined by the kernel in their FP state.
>
> I took some pain to make sure to chose a variable that's already
> in L1 during context switch to make the overhead of this low.
>
> Also alternative() is used to patch away the new code on CPUs
> who don't need it.
>
> Patch for both i386/x86-64.
>
> The problem was discovered originally by Jan Beulich. Richard
> Brunner provided the basic code for the workarounds, with contribution
> from Jan.
>
> This is CVE-2006-1056
>
> Cc: richard.brunner@amd.com
> Cc: jbeulich@novell.com
>
> Signed-off-by: Andi Kleen <ak@suse.de>
>
> Index: linux/include/asm-x86_64/i387.h
> ===================================================================
> --- linux.orig/include/asm-x86_64/i387.h
> +++ linux/include/asm-x86_64/i387.h
> @@ -72,6 +72,23 @@ extern int set_fpregs(struct task_struct
>  #define set_fpu_swd(t,val) ((t)->thread.i387.fxsave.swd = (val))
>  #define set_fpu_fxsr_twd(t,val) ((t)->thread.i387.fxsave.twd = (val))
>
> +#define X87_FSW_ES (1 << 7)	/* Exception Summary */
> +
> +/* AMD CPUs don't save/restore FDP/FIP/FOP unless an exception
> +   is pending. Clear the x87 state here by setting it to fixed
> +   values. The kernel data segment can be sometimes 0 and sometimes
> +   new user value. Both should be ok.
> +   Use the PDA as safe address because it should be already in L1. */
> +static inline void clear_fpu_state(struct i387_fxsave_struct *fx)
> +{
> +	if (unlikely(fx->swd & X87_FSW_ES))
> +		 asm volatile("fnclex");
> +	alternative_input(ASM_NOP8 ASM_NOP2,
> +	     	     "    emms\n"		/* clear stack tags */
> +	     	     "    fildl %%gs:0",	/* load to clear state */
> +		     X86_FEATURE_FXSAVE_LEAK);
> +}
> +
>  static inline int restore_fpu_checking(struct i387_fxsave_struct *fx)
>  {
>  	int err;
> @@ -119,6 +136,7 @@ static inline int save_i387_checking(str
>  #endif
>  	if (unlikely(err))
>  		__clear_user(fx, sizeof(struct i387_fxsave_struct));
> +	/* No need to clear here because the caller clears USED_MATH */
>  	return err;
>  }
>
> @@ -149,7 +167,7 @@ static inline void __fxsave_clear(struct
>  				"i" (offsetof(__typeof__(*tsk),
>  					      thread.i387.fxsave)));
>  #endif
> -	__asm__ __volatile__("fnclex");
> +	clear_fpu_state(&tsk->thread.i387.fxsave);
>  }
>
>  static inline void kernel_fpu_begin(void)
> Index: linux/include/asm-i386/i387.h
> ===================================================================
> --- linux.orig/include/asm-i386/i387.h
> +++ linux/include/asm-i386/i387.h
> @@ -13,6 +13,7 @@
>
>  #include <linux/sched.h>
>  #include <linux/init.h>
> +#include <linux/kernel_stat.h>
>  #include <asm/processor.h>
>  #include <asm/sigcontext.h>
>  #include <asm/user.h>
> @@ -38,17 +39,38 @@ extern void init_fpu(struct task_struct
>  extern void kernel_fpu_begin(void);
>  #define kernel_fpu_end() do { stts(); preempt_enable(); } while(0)
>
> +/* We need a safe address that is cheap to find and that is already
> +   in L1 during context switch. The best choices are unfortunately
> +   different for UP and SMP */
> +#ifdef CONFIG_SMP
> +#define safe_address (__per_cpu_offset[0])
> +#else
> +#define safe_address (kstat_cpu(0).cpustat.user)
> +#endif
> +
>  /*
>   * These must be called with preempt disabled
>   */
>  static inline void __save_init_fpu( struct task_struct *tsk )
>  {
> +	/* Use more nops than strictly needed in case the compiler
> +	   varies code */
>  	alternative_input(
> -		"fnsave %1 ; fwait ;" GENERIC_NOP2,
> -		"fxsave %1 ; fnclex",
> +		"fnsave %[fx] ;fwait;" GENERIC_NOP8 GENERIC_NOP4,
> +		"fxsave %[fx]\n"
> +		"bt $7,%[fsw] ; jc 1f ; fnclex\n1:",
>  		X86_FEATURE_FXSR,
> -		"m" (tsk->thread.i387.fxsave)
> -		:"memory");
> +		[fx] "m" (tsk->thread.i387.fxsave),
> +		[fsw] "m" (tsk->thread.i387.fxsave.swd) : "memory");
> +	/* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception
> +	   is pending.  Clear the x87 state here by setting it to fixed
> +   	   values. __per_cpu_offset[0] is a random variable that should be in
> L1 */ +	alternative_input(
> +		GENERIC_NOP8 GENERIC_NOP2,
> +		"emms\n\t"	  	/* clear stack tags */
> +		"fildl %[addr]", 	/* set F?P to defined value */
> +		X86_FEATURE_FXSAVE_LEAK,
> +		[addr] "m" (safe_address));
>  	task_thread_info(tsk)->status &= ~TS_USEDFPU;
>  }
>
> Index: linux/arch/i386/kernel/cpu/amd.c
> ===================================================================
> --- linux.orig/arch/i386/kernel/cpu/amd.c
> +++ linux/arch/i386/kernel/cpu/amd.c
> @@ -207,6 +207,8 @@ static void __init init_amd(struct cpuin
>  		set_bit(X86_FEATURE_K7, c->x86_capability);
>  		break;
>  	}
> +	if (c->x86 >= 6)
> +		set_bit(X86_FEATURE_FXSAVE_LEAK, c->x86_capability);
>
>  	display_cacheinfo(c);
>
> Index: linux/arch/x86_64/kernel/setup.c
> ===================================================================
> --- linux.orig/arch/x86_64/kernel/setup.c
> +++ linux/arch/x86_64/kernel/setup.c
> @@ -928,6 +928,10 @@ static int __init init_amd(struct cpuinf
>  	if (c->x86 == 15 && ((level >= 0x0f48 && level < 0x0f50) || level >=
> 0x0f58)) set_bit(X86_FEATURE_REP_GOOD, &c->x86_capability);
>
> +	/* Enable workaround for FXSAVE leak */
> +	if (c->x86 >= 6)
> +		set_bit(X86_FEATURE_FXSAVE_LEAK, &c->x86_capability);
> +
>  	r = get_model_name(c);
>  	if (!r) {
>  		switch (c->x86) {
> Index: linux/include/asm-i386/cpufeature.h
> ===================================================================
> --- linux.orig/include/asm-i386/cpufeature.h
> +++ linux/include/asm-i386/cpufeature.h
> @@ -71,6 +71,7 @@
>  #define X86_FEATURE_P4		(3*32+ 7) /* P4 */
>  #define X86_FEATURE_CONSTANT_TSC (3*32+ 8) /* TSC ticks at a constant rate
> */ #define X86_FEATURE_UP		(3*32+ 9) /* smp kernel running on up */
> +#define X86_FEATURE_FXSAVE_LEAK (3*32+10) /* FXSAVE leaks FOP/FIP/FOP */
>
>  /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
>  #define X86_FEATURE_XMM3	(4*32+ 0) /* Streaming SIMD Extensions-3 */
> Index: linux/include/asm-x86_64/cpufeature.h
> ===================================================================
> --- linux.orig/include/asm-x86_64/cpufeature.h
> +++ linux/include/asm-x86_64/cpufeature.h
> @@ -64,6 +64,7 @@
>  #define X86_FEATURE_REP_GOOD	(3*32+ 4) /* rep microcode works well on this
> CPU */ #define X86_FEATURE_CONSTANT_TSC (3*32+5) /* TSC runs at constant
> rate */ #define X86_FEATURE_SYNC_RDTSC  (3*32+6)  /* RDTSC syncs CPU core
> */ +#define X86_FEATURE_FXSAVE_LEAK (3*32+7)  /* FIP/FOP/FDP leaks through
> FXSAVE */
>
>  /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
>  #define X86_FEATURE_XMM3	(4*32+ 0) /* Streaming SIMD Extensions-3 */
> Index: linux/arch/x86_64/kernel/process.c
> ===================================================================
> --- linux.orig/arch/x86_64/kernel/process.c
> +++ linux/arch/x86_64/kernel/process.c
> @@ -575,8 +575,10 @@ __switch_to(struct task_struct *prev_p,
>  	prev->userrsp = read_pda(oldrsp);
>  	write_pda(oldrsp, next->userrsp);
>  	write_pda(pcurrent, next_p);
> +
>  	/* This must be here to ensure both math_state_restore() and
> -	   kernel_fpu_begin() work consistently. */
> +	   kernel_fpu_begin() work consistently.
> +	   And the AMD workaround requires it to be after DS reload. */
>  	unlazy_fpu(prev_p);
>  	write_pda(kernelstack,
>  		  task_stack_page(next_p) + THREAD_SIZE - PDA_STACKOFFSET);
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: [discuss] Re: [PATCH] [1/2] i386/x86-64: Fix x87 information leak between processes

[Andi Kleen]

On Thursday 20 April 2006 08:28, Matt Heler wrote:
> Patch says 1/2 . Is there another patch that comes with this ? Or is vger 
> lagging again ? 

The other patch was a unrelated x86-64 only patch which I normally
not cc to linux-kernel, but only to discuss@x86-64.org. You 
can check it out in the archives.

-Andi