[PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[David Vrabel]

From: David Vrabel <david.vrabel@citrix.com>

In 32 bit guests, if a userspace process has %eax == -ERESTARTSYS
(-512) or -ERESTARTNOINTR (-513) when it is interrupted by an event
/and/ the process has a pending signal then %eip (and %eax) are
corrupted when returning to the main process after handling the
signal.  The application may then crash with SIGSEGV or a SIGILL or it
may have subtly incorrect behaviour (depending on what instruction it
returned to).

The occurs because handle_signal() is incorrectly thinking that there
is a system call that needs to restarted so it adjusts %eip and %eax
to re-execute the system call instruction (even though user space had
not done a system call).

If %eax == -514 (-ERESTARTNOHAND (-514) or -ERESTART_RESTARTBLOCK
(-516) then handle_signal() only corrupted %eax (by setting it to
-EINTR).  This may cause the application to crash or have incorrect
behaviour.

handle_signal() assumes that regs->orig_ax >= 0 means a system call so
any kernel entry point that is not for a system call must push a
negative value for orig_ax.  For example, for physical interrupts on
bare metal the inverse of the vector is pushed and page_fault() sets
regs->orig_ax to -1, overwriting the hardware provided error code.

xen_hypervisor_callback() was incorrectly pushing 0 for orig_ax
instead of -1.

Classic Xen kernels pushed %eax which works as %eax cannot be both
non-negative and -RESTARTSYS (etc.), but using -1 is consistent with
other non-system call entry points.

There were similar bugs in xen_failsafe_callback(), if the fault was
corrected and normal return path was used.  64 bit guests would push 0
which is broken.  32 bit guests would push %eax which is safe (see
previous paragraph), but for consistency this is also changed to -1.

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Jan Beulich <JBeulich@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Cc: stable@vger.kernel.org
---
 arch/x86/kernel/entry_32.S |    4 ++--
 arch/x86/kernel/entry_64.S |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index 2c63407..6a19e66 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -1042,7 +1042,7 @@ ENTRY(xen_sysenter_target)
 
 ENTRY(xen_hypervisor_callback)
 	CFI_STARTPROC
-	pushl_cfi $0
+	pushl_cfi $-1 /* orig_ax = -1 => not a system call */
 	SAVE_ALL
 	TRACE_IRQS_OFF
 
@@ -1078,7 +1078,7 @@ ENDPROC(xen_hypervisor_callback)
 # We distinguish between categories by maintaining a status value in EAX.
 ENTRY(xen_failsafe_callback)
 	CFI_STARTPROC
-	pushl_cfi %eax
+	pushl_cfi $-1  /* orig_ax = -1 => not a system call */
 	movl $1,%eax
 1:	mov 4(%esp),%ds
 2:	mov 8(%esp),%es
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index cdc790c..430b1fc 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -1451,7 +1451,7 @@ ENTRY(xen_failsafe_callback)
 	CFI_RESTORE r11
 	addq $0x30,%rsp
 	CFI_ADJUST_CFA_OFFSET -0x30
-	pushq_cfi $0
+	pushq_cfi $-1 /* orig_ax = -1 => not a system call */
 	SAVE_ALL
 	jmp error_exit
 	CFI_ENDPROC
-- 
1.7.2.5

Re: [PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[Konrad Rzeszutek Wilk]

On Wed, Oct 17, 2012 at 02:29:40PM +0100, David Vrabel wrote:
> From: David Vrabel <david.vrabel@citrix.com>
> 
> In 32 bit guests, if a userspace process has %eax == -ERESTARTSYS
> (-512) or -ERESTARTNOINTR (-513) when it is interrupted by an event
> /and/ the process has a pending signal then %eip (and %eax) are
> corrupted when returning to the main process after handling the
> signal.  The application may then crash with SIGSEGV or a SIGILL or it
> may have subtly incorrect behaviour (depending on what instruction it
> returned to).
> 
> The occurs because handle_signal() is incorrectly thinking that there
> is a system call that needs to restarted so it adjusts %eip and %eax
> to re-execute the system call instruction (even though user space had
> not done a system call).
> 
> If %eax == -514 (-ERESTARTNOHAND (-514) or -ERESTART_RESTARTBLOCK
> (-516) then handle_signal() only corrupted %eax (by setting it to
> -EINTR).  This may cause the application to crash or have incorrect
> behaviour.
> 
> handle_signal() assumes that regs->orig_ax >= 0 means a system call so
> any kernel entry point that is not for a system call must push a
> negative value for orig_ax.  For example, for physical interrupts on
> bare metal the inverse of the vector is pushed and page_fault() sets
> regs->orig_ax to -1, overwriting the hardware provided error code.
> 
> xen_hypervisor_callback() was incorrectly pushing 0 for orig_ax
> instead of -1.
> 
> Classic Xen kernels pushed %eax which works as %eax cannot be both
> non-negative and -RESTARTSYS (etc.), but using -1 is consistent with
> other non-system call entry points.
> 
> There were similar bugs in xen_failsafe_callback(), if the fault was
> corrected and normal return path was used.  64 bit guests would push 0
> which is broken.  32 bit guests would push %eax which is safe (see
> previous paragraph), but for consistency this is also changed to -1.
> 
> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
> Acked-by: Jan Beulich <JBeulich@suse.com>
> Acked-by: Ian Campbell <ian.campbell@citrix.com>
> Cc: stable@vger.kernel.org

Applied.

> ---
>  arch/x86/kernel/entry_32.S |    4 ++--
>  arch/x86/kernel/entry_64.S |    2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
> index 2c63407..6a19e66 100644
> --- a/arch/x86/kernel/entry_32.S
> +++ b/arch/x86/kernel/entry_32.S
> @@ -1042,7 +1042,7 @@ ENTRY(xen_sysenter_target)
>  
>  ENTRY(xen_hypervisor_callback)
>  	CFI_STARTPROC
> -	pushl_cfi $0
> +	pushl_cfi $-1 /* orig_ax = -1 => not a system call */
>  	SAVE_ALL
>  	TRACE_IRQS_OFF
>  
> @@ -1078,7 +1078,7 @@ ENDPROC(xen_hypervisor_callback)
>  # We distinguish between categories by maintaining a status value in EAX.
>  ENTRY(xen_failsafe_callback)
>  	CFI_STARTPROC
> -	pushl_cfi %eax
> +	pushl_cfi $-1  /* orig_ax = -1 => not a system call */
>  	movl $1,%eax
>  1:	mov 4(%esp),%ds
>  2:	mov 8(%esp),%es
> diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
> index cdc790c..430b1fc 100644
> --- a/arch/x86/kernel/entry_64.S
> +++ b/arch/x86/kernel/entry_64.S
> @@ -1451,7 +1451,7 @@ ENTRY(xen_failsafe_callback)
>  	CFI_RESTORE r11
>  	addq $0x30,%rsp
>  	CFI_ADJUST_CFA_OFFSET -0x30
> -	pushq_cfi $0
> +	pushq_cfi $-1 /* orig_ax = -1 => not a system call */
>  	SAVE_ALL
>  	jmp error_exit
>  	CFI_ENDPROC
> -- 
> 1.7.2.5

Re: [PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[Jan Beulich]

>>> On 17.10.12 at 15:29, David Vrabel <david.vrabel@citrix.com> wrote:
> From: David Vrabel <david.vrabel@citrix.com>
> 
> In 32 bit guests, if a userspace process has %eax == -ERESTARTSYS
> (-512) or -ERESTARTNOINTR (-513) when it is interrupted by an event
> /and/ the process has a pending signal then %eip (and %eax) are
> corrupted when returning to the main process after handling the
> signal.  The application may then crash with SIGSEGV or a SIGILL or it
> may have subtly incorrect behaviour (depending on what instruction it
> returned to).
> 
> The occurs because handle_signal() is incorrectly thinking that there
> is a system call that needs to restarted so it adjusts %eip and %eax
> to re-execute the system call instruction (even though user space had
> not done a system call).
> 
> If %eax == -514 (-ERESTARTNOHAND (-514) or -ERESTART_RESTARTBLOCK
> (-516) then handle_signal() only corrupted %eax (by setting it to
> -EINTR).  This may cause the application to crash or have incorrect
> behaviour.
> 
> handle_signal() assumes that regs->orig_ax >= 0 means a system call so
> any kernel entry point that is not for a system call must push a
> negative value for orig_ax.  For example, for physical interrupts on
> bare metal the inverse of the vector is pushed and page_fault() sets
> regs->orig_ax to -1, overwriting the hardware provided error code.
> 
> xen_hypervisor_callback() was incorrectly pushing 0 for orig_ax
> instead of -1.
> 
> Classic Xen kernels pushed %eax which works as %eax cannot be both
> non-negative and -RESTARTSYS (etc.), but using -1 is consistent with
> other non-system call entry points.
> 
> There were similar bugs in xen_failsafe_callback(), if the fault was
> corrected and normal return path was used.  64 bit guests would push 0
> which is broken.  32 bit guests would push %eax which is safe (see
> previous paragraph), but for consistency this is also changed to -1.
> 
> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
> Acked-by: Jan Beulich <JBeulich@suse.com>
> Acked-by: Ian Campbell <ian.campbell@citrix.com>
> Cc: stable@vger.kernel.org 
> ---
>  arch/x86/kernel/entry_32.S |    4 ++--
>  arch/x86/kernel/entry_64.S |    2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
> index 2c63407..6a19e66 100644
> --- a/arch/x86/kernel/entry_32.S
> +++ b/arch/x86/kernel/entry_32.S
> @@ -1042,7 +1042,7 @@ ENTRY(xen_sysenter_target)
>  
>  ENTRY(xen_hypervisor_callback)
>  	CFI_STARTPROC
> -	pushl_cfi $0
> +	pushl_cfi $-1 /* orig_ax = -1 => not a system call */
>  	SAVE_ALL
>  	TRACE_IRQS_OFF
>  
> @@ -1078,7 +1078,7 @@ ENDPROC(xen_hypervisor_callback)
>  # We distinguish between categories by maintaining a status value in EAX.
>  ENTRY(xen_failsafe_callback)
>  	CFI_STARTPROC
> -	pushl_cfi %eax
> +	pushl_cfi $-1  /* orig_ax = -1 => not a system call */

While making this apply to the 2.6.18 tree, I noticed that you
replaced the wrong push here, thus causing register corruption.
Just like on the 64-bit side, the one that needs fixing is the one
right before the SAVE_ALL (and hence it's again not just for
consistency, as zero is being pushed there too).

Jan

>  	movl $1,%eax
>  1:	mov 4(%esp),%ds
>  2:	mov 8(%esp),%es
> diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
> index cdc790c..430b1fc 100644
> --- a/arch/x86/kernel/entry_64.S
> +++ b/arch/x86/kernel/entry_64.S
> @@ -1451,7 +1451,7 @@ ENTRY(xen_failsafe_callback)
>  	CFI_RESTORE r11
>  	addq $0x30,%rsp
>  	CFI_ADJUST_CFA_OFFSET -0x30
> -	pushq_cfi $0
> +	pushq_cfi $-1 /* orig_ax = -1 => not a system call */
>  	SAVE_ALL
>  	jmp error_exit
>  	CFI_ENDPROC
> -- 
> 1.7.2.5
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org 
> http://lists.xen.org/xen-devel 

Re: [PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[David Vrabel]

On 19/10/12 16:29, Jan Beulich wrote:
>>>> On 17.10.12 at 15:29, David Vrabel <david.vrabel@citrix.com> wrote:
>> From: David Vrabel <david.vrabel@citrix.com>
>>
>> In 32 bit guests, if a userspace process has %eax == -ERESTARTSYS
>> (-512) or -ERESTARTNOINTR (-513) when it is interrupted by an event
>> /and/ the process has a pending signal then %eip (and %eax) are
>> corrupted when returning to the main process after handling the
>> signal.  The application may then crash with SIGSEGV or a SIGILL or it
>> may have subtly incorrect behaviour (depending on what instruction it
>> returned to).
>>
[...]
>> diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
>> index 2c63407..6a19e66 100644
>> --- a/arch/x86/kernel/entry_32.S
>> +++ b/arch/x86/kernel/entry_32.S
>> @@ -1042,7 +1042,7 @@ ENTRY(xen_sysenter_target)
>>  
>>  ENTRY(xen_hypervisor_callback)
>>  	CFI_STARTPROC
>> -	pushl_cfi $0
>> +	pushl_cfi $-1 /* orig_ax = -1 => not a system call */
>>  	SAVE_ALL
>>  	TRACE_IRQS_OFF
>>  
>> @@ -1078,7 +1078,7 @@ ENDPROC(xen_hypervisor_callback)
>>  # We distinguish between categories by maintaining a status value in EAX.
>>  ENTRY(xen_failsafe_callback)
>>  	CFI_STARTPROC
>> -	pushl_cfi %eax
>> +	pushl_cfi $-1  /* orig_ax = -1 => not a system call */
> 
> While making this apply to the 2.6.18 tree, I noticed that you
> replaced the wrong push here, thus causing register corruption.
> Just like on the 64-bit side, the one that needs fixing is the one
> right before the SAVE_ALL (and hence it's again not just for
> consistency, as zero is being pushed there too).

Oops.

We would have liked to test this path but could not see how to.  Do you
have any ideas?

David

Re: [PATCHv2] xen/x86: don't corrupt %eip when returning from a signal handler

[Jan Beulich]

>>> On 19.10.12 at 17:44, David Vrabel <david.vrabel@citrix.com> wrote:
> On 19/10/12 16:29, Jan Beulich wrote:
>>>>> On 17.10.12 at 15:29, David Vrabel <david.vrabel@citrix.com> wrote:
>>> @@ -1078,7 +1078,7 @@ ENDPROC(xen_hypervisor_callback)
>>>  # We distinguish between categories by maintaining a status value in EAX.
>>>  ENTRY(xen_failsafe_callback)
>>>  	CFI_STARTPROC
>>> -	pushl_cfi %eax
>>> +	pushl_cfi $-1  /* orig_ax = -1 => not a system call */
>> 
>> While making this apply to the 2.6.18 tree, I noticed that you
>> replaced the wrong push here, thus causing register corruption.
>> Just like on the 64-bit side, the one that needs fixing is the one
>> right before the SAVE_ALL (and hence it's again not just for
>> consistency, as zero is being pushed there too).
> 
> Oops.
> 
> We would have liked to test this path but could not see how to.  Do you
> have any ideas?

I'm not aware of a way to reliably trigger this without adding
assisting code to the kernel.

Jan