x86-64: fix #GP generation in assembly code

When guest use of sysenter (64-bit PV guest) or syscall (32-bit PV
guest) gets converted into a GP fault (due to no callback having got
registered), we must
- honor the GP fault handler's request the keep enabled or mask event
  delivery
- not allow TBF_EXCEPTION to remain set past the generation of the
  (guest) exception in the vCPU's trap_bounce.flags, as that would
  otherwise allow for the next exception occurring in guest mode,
  should it happen to get handled in Xen itself, to nevertheless get
  bounced to the guest kernel.

Also, just like compat mode syscall handling already did, native mode
sysenter handling should, when converting to #GP, subtract 2 from the
RIP present in the frame so that the guest's GP fault handler would
see the fault pointing to the offending instruction instead of past it.

Finally, since those exception generating code blocks needed to be
modified anyway, convert them to make use of UNLIKELY_{START,END}().

[ This bug is security vulnerability, XSA-8 / CVE-2012-0218. ]

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Keir Fraser <keir@xen.org>
Committed-by: Jan Beulich <jbeulich@suse.com>

xen-unstable changeset:   25200:80f4113be500 25204:569d6f05e1ef
Committed-by: Ian Jackson <ian.jackson@eu.citrix.com>

3 files changed, 27 insertions, 20 deletions

diff --git a/xen/arch/x86/x86_64/asm-offsets.c b/xen/arch/x86/x86_64/asm-offsets.c
index d8024bb169..d95739a182 100644
--- a/xen/arch/x86/x86_64/asm-offsets.c
+++ b/xen/arch/x86/x86_64/asm-offsets.c
@@ -90,6 +90,8 @@ void __dummy__(void)
            arch.guest_context.trap_ctxt[TRAP_gp_fault].address);
     OFFSET(VCPU_gp_fault_sel, struct vcpu,
            arch.guest_context.trap_ctxt[TRAP_gp_fault].cs);
+    OFFSET(VCPU_gp_fault_flags, struct vcpu,
+           arch.guest_context.trap_ctxt[TRAP_gp_fault].flags);
     OFFSET(VCPU_kernel_sp, struct vcpu, arch.guest_context.kernel_sp);
     OFFSET(VCPU_kernel_ss, struct vcpu, arch.guest_context.kernel_ss);
     OFFSET(VCPU_guest_context_flags, struct vcpu, arch.guest_context.flags);
diff --git a/xen/arch/x86/x86_64/compat/entry.S b/xen/arch/x86/x86_64/compat/entry.S
index 8208b6b54f..85bf42f2fb 100644
--- a/xen/arch/x86/x86_64/compat/entry.S
+++ b/xen/arch/x86/x86_64/compat/entry.S
@@ -214,6 +214,7 @@ compat_failsafe_callback:
 ENTRY(compat_post_handle_exception)
         testb $TBF_EXCEPTION,TRAPBOUNCE_flags(%rdx)
         jz    compat_test_all_events
+.Lcompat_bounce_exception:
         call  compat_create_bounce_frame
         movb  $0,TRAPBOUNCE_flags(%rdx)
         jmp   compat_test_all_events
@@ -226,19 +227,20 @@ ENTRY(compat_syscall)
         leaq  VCPU_trap_bounce(%rbx),%rdx
         testl $~3,%esi
         leal  (,%rcx,TBF_INTERRUPT),%ecx
-        jz    2f
-1:      movq  %rax,TRAPBOUNCE_eip(%rdx)
-        movw  %si,TRAPBOUNCE_cs(%rdx)
-        movb  %cl,TRAPBOUNCE_flags(%rdx)
-        call  compat_create_bounce_frame
-        jmp   compat_test_all_events
-2:      movl  $TRAP_gp_fault,UREGS_entry_vector(%rsp)
+UNLIKELY_START(z, compat_syscall_gpf)
+        movl  $TRAP_gp_fault,UREGS_entry_vector(%rsp)
         subl  $2,UREGS_rip(%rsp)
-        movq  VCPU_gp_fault_addr(%rbx),%rax
-        movzwl VCPU_gp_fault_sel(%rbx),%esi
-        movb  $(TBF_EXCEPTION|TBF_EXCEPTION_ERRCODE|TBF_INTERRUPT),%cl
         movl  $0,TRAPBOUNCE_error_code(%rdx)
-        jmp   1b
+        movl  VCPU_gp_fault_addr(%rbx),%eax
+        movzwl VCPU_gp_fault_sel(%rbx),%esi
+        testb $4,VCPU_gp_fault_flags(%rbx)
+        setnz %cl
+        leal  TBF_EXCEPTION|TBF_EXCEPTION_ERRCODE(,%rcx,TBF_INTERRUPT),%ecx
+UNLIKELY_END(compat_syscall_gpf)
+        movq  %rax,TRAPBOUNCE_eip(%rdx)
+        movw  %si,TRAPBOUNCE_cs(%rdx)
+        movb  %cl,TRAPBOUNCE_flags(%rdx)
+        jmp   .Lcompat_bounce_exception
 
 ENTRY(compat_sysenter)
         cmpl  $TRAP_gp_fault,UREGS_entry_vector(%rsp)
diff --git a/xen/arch/x86/x86_64/entry.S b/xen/arch/x86/x86_64/entry.S
index c8db00cb38..1f389fa502 100644
--- a/xen/arch/x86/x86_64/entry.S
+++ b/xen/arch/x86/x86_64/entry.S
@@ -289,19 +289,21 @@ sysenter_eflags_saved:
         leaq  VCPU_trap_bounce(%rbx),%rdx
         testq %rax,%rax
         leal  (,%rcx,TBF_INTERRUPT),%ecx
-        jz    2f
-1:      movq  VCPU_domain(%rbx),%rdi
+UNLIKELY_START(z, sysenter_gpf)
+        movl  $TRAP_gp_fault,UREGS_entry_vector(%rsp)
+        subq  $2,UREGS_rip(%rsp)
+        movl  %eax,TRAPBOUNCE_error_code(%rdx)
+        movq  VCPU_gp_fault_addr(%rbx),%rax
+        testb $4,VCPU_gp_fault_flags(%rbx)
+        setnz %cl
+        leal  TBF_EXCEPTION|TBF_EXCEPTION_ERRCODE(,%rcx,TBF_INTERRUPT),%ecx
+UNLIKELY_END(sysenter_gpf)
+        movq  VCPU_domain(%rbx),%rdi
         movq  %rax,TRAPBOUNCE_eip(%rdx)
         movb  %cl,TRAPBOUNCE_flags(%rdx)
         testb $1,DOMAIN_is_32bit_pv(%rdi)
         jnz   compat_sysenter
-        call  create_bounce_frame
-        jmp   test_all_events
-2:      movl  %eax,TRAPBOUNCE_error_code(%rdx)
-        movq  VCPU_gp_fault_addr(%rbx),%rax
-        movb  $(TBF_EXCEPTION|TBF_EXCEPTION_ERRCODE|TBF_INTERRUPT),%cl
-        movl  $TRAP_gp_fault,UREGS_entry_vector(%rsp)
-        jmp   1b
+        jmp   .Lbounce_exception
 
 ENTRY(int80_direct_trap)
         pushq $0
@@ -493,6 +495,7 @@ handle_exception_saved:
         jnz   compat_post_handle_exception
         testb $TBF_EXCEPTION,TRAPBOUNCE_flags(%rdx)
         jz    test_all_events
+.Lbounce_exception:
         call  create_bounce_frame
         movb  $0,TRAPBOUNCE_flags(%rdx)
         jmp   test_all_events