diff mbox series

[for_v29,v2,1/5] x86/sgx: vdso: Make __vdso_sgx_enter_enclave() callable from C code

Message ID 20200330180811.31381-2-sean.j.christopherson@intel.com (mailing list archive)
State New, archived
Headers show
Series x86/sgx: Make vDSO callable from C | expand

Commit Message

Sean Christopherson March 30, 2020, 6:08 p.m. UTC 
Make __vdso_sgx_enter_enclave() callable from C by preserving %rbx
and taking @leaf in %rcx instead of %rax.  Being able to invoke the vDSO
from C reduces the overhead of runtimes that are tightly coupled with
their enclaves, e.g. that can rely on the enclave to save and restore
non-volatile registers, as the runtime doesn't need an assembly wrapper
to preserve non-volatile registers and/or shuffle stack arguments.

Note, both %rcx and %rbx are consumed by EENTER/ERESUME, i.e. consuming
them doesn't violate the primary tenet of __vdso_sgx_enter_enclave()
that "thou shalt not restrict how information is exchanged between an
enclave and its host process".

Suggested-by: Nathaniel McCallum <npmccallum@redhat.com>
Cc: Cedric Xing <cedric.xing@intel.com>
Cc: Jethro Beekman <jethro@fortanix.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: linux-sgx@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
---
 arch/x86/entry/vdso/vsgx_enter_enclave.S | 30 ++++++++++++++----------
 1 file changed, 18 insertions(+), 12 deletions(-)

Comments

Jarkko Sakkinen March 30, 2020, 9:04 p.m. UTC | #1 
On Mon, Mar 30, 2020 at 11:08:07AM -0700, Sean Christopherson wrote:
> Make __vdso_sgx_enter_enclave() callable from C by preserving %rbx
> and taking @leaf in %rcx instead of %rax.  Being able to invoke the vDSO
> from C reduces the overhead of runtimes that are tightly coupled with
> their enclaves, e.g. that can rely on the enclave to save and restore
> non-volatile registers, as the runtime doesn't need an assembly wrapper
> to preserve non-volatile registers and/or shuffle stack arguments.
> 
> Note, both %rcx and %rbx are consumed by EENTER/ERESUME, i.e. consuming
> them doesn't violate the primary tenet of __vdso_sgx_enter_enclave()
> that "thou shalt not restrict how information is exchanged between an
> enclave and its host process".
> 
> Suggested-by: Nathaniel McCallum <npmccallum@redhat.com>
> Cc: Cedric Xing <cedric.xing@intel.com>
> Cc: Jethro Beekman <jethro@fortanix.com>
> Cc: Andy Lutomirski <luto@amacapital.net>
> Cc: linux-sgx@vger.kernel.org
> Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
> ---
>  arch/x86/entry/vdso/vsgx_enter_enclave.S | 30 ++++++++++++++----------
>  1 file changed, 18 insertions(+), 12 deletions(-)
> 
> diff --git a/arch/x86/entry/vdso/vsgx_enter_enclave.S b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> index 34cee2b0ef09..c56064fb36bc 100644
> --- a/arch/x86/entry/vdso/vsgx_enter_enclave.S
> +++ b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> @@ -17,22 +17,22 @@
>  
>  /**
>   * __vdso_sgx_enter_enclave() - Enter an SGX enclave
> + * @rdi:	Pass-through value for RDI
> + * @rsi:	Pass-through value for RSI
> + * @rdx:	Pass-through value for RDX
>   * @leaf:	ENCLU leaf, must be EENTER or ERESUME
> + * @r8:		Pass-through value for R8
> + * @r9:		Pass-through value for R9
>   * @tcs:	TCS, must be non-NULL
>   * @e:		Optional struct sgx_enclave_exception instance
>   * @handler:	Optional enclave exit handler
>   *
> - * **Important!**  __vdso_sgx_enter_enclave() is **NOT** compliant with the
> - * x86-64 ABI, i.e. cannot be called from standard C code.
> - *
> - * Input ABI:
> - *  @leaf	%eax
> - *  @tcs	8(%rsp)
> - *  @e 		0x10(%rsp)
> - *  @handler	0x18(%rsp)
> - *
> - * Output ABI:
> - *  @ret	%eax
> + * **Important!**  __vdso_sgx_enter_enclave() does not ensure full compliance

I'd simply put **NOTE** here instead of **Important!** as it is more
common.

> + * with the x86-64 ABI, e.g. doesn't explicitly clear EFLAGS.DF after EEXIT.
> + * Except for non-volatile general purpose registers, preserving/setting state
> + * in accordance with the x86-64 ABI is the responsibility of the enclave and
> + * its runtime, i.e. __vdso_sgx_enter_enclave() cannot be called from C code
> + * without careful consideration by both the enclave and its runtime.

Instead "e.g. doesn't explcitly clear EFLAGS.DF after EEXIT" (which is
somewhat confusing statement) paragraph should be replaced with a simple
enumerated list of differences.

Something might be left out but that's cool. Just do your best and it
can refined over time to be more exact.

>   *
>   * All general purpose registers except RAX, RBX and RCX are passed as-is to
>   * the enclave. RAX, RBX and RCX are consumed by EENTER and ERESUME and are
> @@ -71,7 +71,9 @@
>   */
>  #ifdef SGX_KERNEL_DOC
>  /* C-style function prototype to coerce kernel-doc into parsing the comment. */
> -int __vdso_sgx_enter_enclave(int leaf, void *tcs,
> +int __vdso_sgx_enter_enclave(unsigned long rdi, unsigned long rsi,
> +			     unsigned long rdx, unsigned int leaf,
> +			     unsigned long r8,  unsigned long r9, void *tcs,
>  			     struct sgx_enclave_exception *e,
>  			     sgx_enclave_exit_handler_t handler);
>  #endif
> @@ -83,7 +85,10 @@ SYM_FUNC_START(__vdso_sgx_enter_enclave)
>  	.cfi_rel_offset		%rbp, 0
>  	mov	%rsp, %rbp
>  	.cfi_def_cfa_register	%rbp
> +	push	%rbx
> +	.cfi_rel_offset		%rbx, -8
>  
> +	mov	%ecx, %eax
>  .Lenter_enclave:
>  	/* EENTER <= leaf <= ERESUME */
>  	cmp	$EENTER, %eax
> @@ -109,6 +114,7 @@ SYM_FUNC_START(__vdso_sgx_enter_enclave)
>  	jne	.Linvoke_userspace_handler
>  
>  .Lout:
> +	pop	%rbx
>  	leave
>  	.cfi_def_cfa		%rsp, 8
>  	ret
> -- 
> 2.24.1
>
Sean Christopherson April 17, 2020, 3:05 p.m. UTC | #2 
On Tue, Mar 31, 2020 at 12:04:46AM +0300, Jarkko Sakkinen wrote:
> On Mon, Mar 30, 2020 at 11:08:07AM -0700, Sean Christopherson wrote:
> > Make __vdso_sgx_enter_enclave() callable from C by preserving %rbx
> > and taking @leaf in %rcx instead of %rax.  Being able to invoke the vDSO
> > from C reduces the overhead of runtimes that are tightly coupled with
> > their enclaves, e.g. that can rely on the enclave to save and restore
> > non-volatile registers, as the runtime doesn't need an assembly wrapper
> > to preserve non-volatile registers and/or shuffle stack arguments.
> > 
> > Note, both %rcx and %rbx are consumed by EENTER/ERESUME, i.e. consuming
> > them doesn't violate the primary tenet of __vdso_sgx_enter_enclave()
> > that "thou shalt not restrict how information is exchanged between an
> > enclave and its host process".
> > 
> > Suggested-by: Nathaniel McCallum <npmccallum@redhat.com>
> > Cc: Cedric Xing <cedric.xing@intel.com>
> > Cc: Jethro Beekman <jethro@fortanix.com>
> > Cc: Andy Lutomirski <luto@amacapital.net>
> > Cc: linux-sgx@vger.kernel.org
> > Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
> > ---
> >  arch/x86/entry/vdso/vsgx_enter_enclave.S | 30 ++++++++++++++----------
> >  1 file changed, 18 insertions(+), 12 deletions(-)
> > 
> > diff --git a/arch/x86/entry/vdso/vsgx_enter_enclave.S b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > index 34cee2b0ef09..c56064fb36bc 100644
> > --- a/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > +++ b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > @@ -17,22 +17,22 @@
> >  
> >  /**
> >   * __vdso_sgx_enter_enclave() - Enter an SGX enclave
> > + * @rdi:	Pass-through value for RDI
> > + * @rsi:	Pass-through value for RSI
> > + * @rdx:	Pass-through value for RDX
> >   * @leaf:	ENCLU leaf, must be EENTER or ERESUME
> > + * @r8:		Pass-through value for R8
> > + * @r9:		Pass-through value for R9
> >   * @tcs:	TCS, must be non-NULL
> >   * @e:		Optional struct sgx_enclave_exception instance
> >   * @handler:	Optional enclave exit handler
> >   *
> > - * **Important!**  __vdso_sgx_enter_enclave() is **NOT** compliant with the
> > - * x86-64 ABI, i.e. cannot be called from standard C code.
> > - *
> > - * Input ABI:
> > - *  @leaf	%eax
> > - *  @tcs	8(%rsp)
> > - *  @e 		0x10(%rsp)
> > - *  @handler	0x18(%rsp)
> > - *
> > - * Output ABI:
> > - *  @ret	%eax
> > + * **Important!**  __vdso_sgx_enter_enclave() does not ensure full compliance
> 
> I'd simply put **NOTE** here instead of **Important!** as it is more
> common.
> 
> > + * with the x86-64 ABI, e.g. doesn't explicitly clear EFLAGS.DF after EEXIT.
> > + * Except for non-volatile general purpose registers, preserving/setting state
> > + * in accordance with the x86-64 ABI is the responsibility of the enclave and
> > + * its runtime, i.e. __vdso_sgx_enter_enclave() cannot be called from C code
> > + * without careful consideration by both the enclave and its runtime.
> 
> Instead "e.g. doesn't explcitly clear EFLAGS.DF after EEXIT" (which is
> somewhat confusing statement) paragraph should be replaced with a simple
> enumerated list of differences.

I don't think the list is that simple, e.g. there is a lot of state that
is defined by the ABI that isn't touched, IMO talking about that state will
add confusion.

I also don't understand what's confusing about stating EFLAGS.DF isn't
cleared.

> Something might be left out but that's cool. Just do your best and it
> can refined over time to be more exact.
> 
> >   *
> >   * All general purpose registers except RAX, RBX and RCX are passed as-is to
> >   * the enclave. RAX, RBX and RCX are consumed by EENTER and ERESUME and are
> > @@ -71,7 +71,9 @@
> >   */
> >  #ifdef SGX_KERNEL_DOC
> >  /* C-style function prototype to coerce kernel-doc into parsing the comment. */
> > -int __vdso_sgx_enter_enclave(int leaf, void *tcs,
> > +int __vdso_sgx_enter_enclave(unsigned long rdi, unsigned long rsi,
> > +			     unsigned long rdx, unsigned int leaf,
> > +			     unsigned long r8,  unsigned long r9, void *tcs,
> >  			     struct sgx_enclave_exception *e,
> >  			     sgx_enclave_exit_handler_t handler);
> >  #endif
> > @@ -83,7 +85,10 @@ SYM_FUNC_START(__vdso_sgx_enter_enclave)
> >  	.cfi_rel_offset		%rbp, 0
> >  	mov	%rsp, %rbp
> >  	.cfi_def_cfa_register	%rbp
> > +	push	%rbx
> > +	.cfi_rel_offset		%rbx, -8
> >  
> > +	mov	%ecx, %eax
> >  .Lenter_enclave:
> >  	/* EENTER <= leaf <= ERESUME */
> >  	cmp	$EENTER, %eax
> > @@ -109,6 +114,7 @@ SYM_FUNC_START(__vdso_sgx_enter_enclave)
> >  	jne	.Linvoke_userspace_handler
> >  
> >  .Lout:
> > +	pop	%rbx
> >  	leave
> >  	.cfi_def_cfa		%rsp, 8
> >  	ret
> > -- 
> > 2.24.1
> >
Jarkko Sakkinen April 17, 2020, 6:57 p.m. UTC | #3 
On Fri, Apr 17, 2020 at 08:05:01AM -0700, Sean Christopherson wrote:
> On Tue, Mar 31, 2020 at 12:04:46AM +0300, Jarkko Sakkinen wrote:
> > On Mon, Mar 30, 2020 at 11:08:07AM -0700, Sean Christopherson wrote:
> > > Make __vdso_sgx_enter_enclave() callable from C by preserving %rbx
> > > and taking @leaf in %rcx instead of %rax.  Being able to invoke the vDSO
> > > from C reduces the overhead of runtimes that are tightly coupled with
> > > their enclaves, e.g. that can rely on the enclave to save and restore
> > > non-volatile registers, as the runtime doesn't need an assembly wrapper
> > > to preserve non-volatile registers and/or shuffle stack arguments.
> > > 
> > > Note, both %rcx and %rbx are consumed by EENTER/ERESUME, i.e. consuming
> > > them doesn't violate the primary tenet of __vdso_sgx_enter_enclave()
> > > that "thou shalt not restrict how information is exchanged between an
> > > enclave and its host process".
> > > 
> > > Suggested-by: Nathaniel McCallum <npmccallum@redhat.com>
> > > Cc: Cedric Xing <cedric.xing@intel.com>
> > > Cc: Jethro Beekman <jethro@fortanix.com>
> > > Cc: Andy Lutomirski <luto@amacapital.net>
> > > Cc: linux-sgx@vger.kernel.org
> > > Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
> > > ---
> > >  arch/x86/entry/vdso/vsgx_enter_enclave.S | 30 ++++++++++++++----------
> > >  1 file changed, 18 insertions(+), 12 deletions(-)
> > > 
> > > diff --git a/arch/x86/entry/vdso/vsgx_enter_enclave.S b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > > index 34cee2b0ef09..c56064fb36bc 100644
> > > --- a/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > > +++ b/arch/x86/entry/vdso/vsgx_enter_enclave.S
> > > @@ -17,22 +17,22 @@
> > >  
> > >  /**
> > >   * __vdso_sgx_enter_enclave() - Enter an SGX enclave
> > > + * @rdi:	Pass-through value for RDI
> > > + * @rsi:	Pass-through value for RSI
> > > + * @rdx:	Pass-through value for RDX
> > >   * @leaf:	ENCLU leaf, must be EENTER or ERESUME
> > > + * @r8:		Pass-through value for R8
> > > + * @r9:		Pass-through value for R9
> > >   * @tcs:	TCS, must be non-NULL
> > >   * @e:		Optional struct sgx_enclave_exception instance
> > >   * @handler:	Optional enclave exit handler
> > >   *
> > > - * **Important!**  __vdso_sgx_enter_enclave() is **NOT** compliant with the
> > > - * x86-64 ABI, i.e. cannot be called from standard C code.
> > > - *
> > > - * Input ABI:
> > > - *  @leaf	%eax
> > > - *  @tcs	8(%rsp)
> > > - *  @e 		0x10(%rsp)
> > > - *  @handler	0x18(%rsp)
> > > - *
> > > - * Output ABI:
> > > - *  @ret	%eax
> > > + * **Important!**  __vdso_sgx_enter_enclave() does not ensure full compliance
> > 
> > I'd simply put **NOTE** here instead of **Important!** as it is more
> > common.
> > 
> > > + * with the x86-64 ABI, e.g. doesn't explicitly clear EFLAGS.DF after EEXIT.
> > > + * Except for non-volatile general purpose registers, preserving/setting state
> > > + * in accordance with the x86-64 ABI is the responsibility of the enclave and
> > > + * its runtime, i.e. __vdso_sgx_enter_enclave() cannot be called from C code
> > > + * without careful consideration by both the enclave and its runtime.
> > 
> > Instead "e.g. doesn't explcitly clear EFLAGS.DF after EEXIT" (which is
> > somewhat confusing statement) paragraph should be replaced with a simple
> > enumerated list of differences.
> 
> I don't think the list is that simple, e.g. there is a lot of state that
> is defined by the ABI that isn't touched, IMO talking about that state will
> add confusion.
> 
> I also don't understand what's confusing about stating EFLAGS.DF isn't
> cleared.

Too much time was has passed since the last version, and too many
patches have been reviewed since.

Please just refresh the way you feel best and I will make conclusions
based on that. It's faster that way in the end.

/Jarkko
diff mbox series

Patch

diff --git a/arch/x86/entry/vdso/vsgx_enter_enclave.S b/arch/x86/entry/vdso/vsgx_enter_enclave.S
index 34cee2b0ef09..c56064fb36bc 100644
--- a/arch/x86/entry/vdso/vsgx_enter_enclave.S
+++ b/arch/x86/entry/vdso/vsgx_enter_enclave.S
@@ -17,22 +17,22 @@ 
 
 /**
  * __vdso_sgx_enter_enclave() - Enter an SGX enclave
+ * @rdi:	Pass-through value for RDI
+ * @rsi:	Pass-through value for RSI
+ * @rdx:	Pass-through value for RDX
  * @leaf:	ENCLU leaf, must be EENTER or ERESUME
+ * @r8:		Pass-through value for R8
+ * @r9:		Pass-through value for R9
  * @tcs:	TCS, must be non-NULL
  * @e:		Optional struct sgx_enclave_exception instance
  * @handler:	Optional enclave exit handler
  *
- * **Important!**  __vdso_sgx_enter_enclave() is **NOT** compliant with the
- * x86-64 ABI, i.e. cannot be called from standard C code.
- *
- * Input ABI:
- *  @leaf	%eax
- *  @tcs	8(%rsp)
- *  @e 		0x10(%rsp)
- *  @handler	0x18(%rsp)
- *
- * Output ABI:
- *  @ret	%eax
+ * **Important!**  __vdso_sgx_enter_enclave() does not ensure full compliance
+ * with the x86-64 ABI, e.g. doesn't explicitly clear EFLAGS.DF after EEXIT.
+ * Except for non-volatile general purpose registers, preserving/setting state
+ * in accordance with the x86-64 ABI is the responsibility of the enclave and
+ * its runtime, i.e. __vdso_sgx_enter_enclave() cannot be called from C code
+ * without careful consideration by both the enclave and its runtime.
  *
  * All general purpose registers except RAX, RBX and RCX are passed as-is to
  * the enclave. RAX, RBX and RCX are consumed by EENTER and ERESUME and are
@@ -71,7 +71,9 @@ 
  */
 #ifdef SGX_KERNEL_DOC
 /* C-style function prototype to coerce kernel-doc into parsing the comment. */
-int __vdso_sgx_enter_enclave(int leaf, void *tcs,
+int __vdso_sgx_enter_enclave(unsigned long rdi, unsigned long rsi,
+			     unsigned long rdx, unsigned int leaf,
+			     unsigned long r8,  unsigned long r9, void *tcs,
 			     struct sgx_enclave_exception *e,
 			     sgx_enclave_exit_handler_t handler);
 #endif
@@ -83,7 +85,10 @@  SYM_FUNC_START(__vdso_sgx_enter_enclave)
 	.cfi_rel_offset		%rbp, 0
 	mov	%rsp, %rbp
 	.cfi_def_cfa_register	%rbp
+	push	%rbx
+	.cfi_rel_offset		%rbx, -8
 
+	mov	%ecx, %eax
 .Lenter_enclave:
 	/* EENTER <= leaf <= ERESUME */
 	cmp	$EENTER, %eax
@@ -109,6 +114,7 @@  SYM_FUNC_START(__vdso_sgx_enter_enclave)
 	jne	.Linvoke_userspace_handler
 
 .Lout:
+	pop	%rbx
 	leave
 	.cfi_def_cfa		%rsp, 8
 	ret