From: Andrew Cooper <andrew.cooper3@citrix.com>
Subject: x86/spec-ctrl: Support IBPB-on-entry

We are going to need this to mitigate Branch Type Confusion on AMD/Hygon CPUs,
but as we've talked about using it in other cases too, arrange to support it
generally.  However, this is also very expensive in some cases, so we're going
to want per-domain controls.

Introduce SCF_ist_ibpb and SCF_entry_ibpb controls, adding them to the IST and
DOM masks as appropriate.  Also introduce X86_FEATURE_IBPB_ENTRY_{PV,HVM} to
to patch the code blocks.

For SVM, the STGI is serialising enough to protect against Spectre-v1 attacks,
so no "else lfence" is necessary.  VT-x will use use the MSR host load list,
so doesn't need any code in the VMExit path.

For the IST path, we can't safely check CPL==0 to skip a flush, as we might
have hit an entry path before it's IBPB.  As IST hitting Xen is rare, flush
irrespective of CPL.  A later path, SCF_ist_sc_msr, provides Spectre-v1
safety.

For the PV paths, we know we're interrupting CPL>0, while for the INTR paths,
we can safely check CPL==0.  Only flush when interrupting guest context.

An "else lfence" is needed for safety, but we want to be able to skip it on
unaffected CPUs, so the block wants to be an alternative, which means the
lfence has to be inline rather than UNLIKELY() (the replacement block doesn't
have displacements fixed up for anything other than the first instruction).

As with SPEC_CTRL_ENTRY_FROM_INTR_IST, %rdx is 0 on entry so rely on this to
shrink the logic marginally.  Update the comments to specify this new
dependency.

This is part of XSA-407.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

diff --git a/xen/arch/x86/hvm/svm/entry.S b/xen/arch/x86/hvm/svm/entry.S
index 4ae55a2ef605..0ff4008060fa 100644
--- a/xen/arch/x86/hvm/svm/entry.S
+++ b/xen/arch/x86/hvm/svm/entry.S
@@ -97,7 +97,19 @@ __UNLIKELY_END(nsvm_hap)
 
         GET_CURRENT(bx)
 
-        /* SPEC_CTRL_ENTRY_FROM_SVM    Req: %rsp=regs/cpuinfo         Clob: acd */
+        /* SPEC_CTRL_ENTRY_FROM_SVM    Req: %rsp=regs/cpuinfo, %rdx=0 Clob: acd */
+
+        .macro svm_vmexit_cond_ibpb
+            testb  $SCF_entry_ibpb, CPUINFO_xen_spec_ctrl(%rsp)
+            jz     .L_skip_ibpb
+
+            mov    $MSR_PRED_CMD, %ecx
+            mov    $PRED_CMD_IBPB, %eax
+            wrmsr
+.L_skip_ibpb:
+	.endm
+        ALTERNATIVE "", svm_vmexit_cond_ibpb, X86_FEATURE_IBPB_ENTRY_HVM
+
         ALTERNATIVE "", DO_OVERWRITE_RSB, X86_FEATURE_SC_RSB_HVM
 
         .macro svm_vmexit_spec_ctrl
@@ -114,6 +126,10 @@ __UNLIKELY_END(nsvm_hap)
         ALTERNATIVE "", svm_vmexit_spec_ctrl, X86_FEATURE_SC_MSR_HVM
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
+        /*
+         * STGI is executed unconditionally, and is sufficiently serialising
+         * to safely resolve any Spectre-v1 concerns in the above logic.
+         */
         stgi
 GLOBAL(svm_stgi_label)
         mov  %rsp,%rdi
diff --git a/xen/arch/x86/hvm/vmx/vmcs.c b/xen/arch/x86/hvm/vmx/vmcs.c
index f9f9bc18cdbc..dd817cee4e69 100644
--- a/xen/arch/x86/hvm/vmx/vmcs.c
+++ b/xen/arch/x86/hvm/vmx/vmcs.c
@@ -1345,6 +1345,10 @@ static int construct_vmcs(struct vcpu *v)
         rc = vmx_add_msr(v, MSR_FLUSH_CMD, FLUSH_CMD_L1D,
                          VMX_MSR_GUEST_LOADONLY);
 
+    if ( !rc && (d->arch.spec_ctrl_flags & SCF_entry_ibpb) )
+        rc = vmx_add_msr(v, MSR_PRED_CMD, PRED_CMD_IBPB,
+                         VMX_MSR_HOST);
+
  out:
     vmx_vmcs_exit(v);
 
diff --git a/xen/arch/x86/x86_64/compat/entry.S b/xen/arch/x86/x86_64/compat/entry.S
index 0cfe95314249..5c999271e617 100644
--- a/xen/arch/x86/x86_64/compat/entry.S
+++ b/xen/arch/x86/x86_64/compat/entry.S
@@ -20,7 +20,7 @@ ENTRY(entry_int82)
         movl  $HYPERCALL_VECTOR, 4(%rsp)
         SAVE_ALL compat=1 /* DPL1 gate, restricted to 32bit PV guests only. */
 
-        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         CR4_PV32_RESTORE
@@ -216,7 +216,7 @@ ENTRY(cstar_enter)
         movl  $TRAP_syscall, 4(%rsp)
         SAVE_ALL
 
-        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         GET_STACK_END(bx)
diff --git a/xen/arch/x86/x86_64/entry.S b/xen/arch/x86/x86_64/entry.S
index 9bfc5964a911..3c8593325606 100644
--- a/xen/arch/x86/x86_64/entry.S
+++ b/xen/arch/x86/x86_64/entry.S
@@ -260,7 +260,7 @@ ENTRY(lstar_enter)
         movl  $TRAP_syscall, 4(%rsp)
         SAVE_ALL
 
-        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         GET_STACK_END(bx)
@@ -299,7 +299,7 @@ GLOBAL(sysenter_eflags_saved)
         movl  $TRAP_syscall, 4(%rsp)
         SAVE_ALL
 
-        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         GET_STACK_END(bx)
@@ -351,7 +351,7 @@ ENTRY(int80_direct_trap)
         movl  $0x80, 4(%rsp)
         SAVE_ALL
 
-        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_PV /* Req: %rsp=regs/cpuinfo, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         GET_STACK_END(bx)
@@ -618,7 +618,7 @@ ENTRY(common_interrupt)
 
         GET_STACK_END(14)
 
-        SPEC_CTRL_ENTRY_FROM_INTR /* Req: %rsp=regs, %r14=end, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_INTR /* Req: %rsp=regs, %r14=end, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         mov   STACK_CPUINFO_FIELD(xen_cr3)(%r14), %rcx
@@ -652,7 +652,7 @@ GLOBAL(handle_exception)
 
         GET_STACK_END(14)
 
-        SPEC_CTRL_ENTRY_FROM_INTR /* Req: %rsp=regs, %r14=end, Clob: acd */
+        SPEC_CTRL_ENTRY_FROM_INTR /* Req: %rsp=regs, %r14=end, %rdx=0, Clob: acd */
         /* WARNING! `ret`, `call *`, `jmp *` not safe before this point. */
 
         mov   STACK_CPUINFO_FIELD(xen_cr3)(%r14), %rcx
diff --git a/xen/include/asm-x86/cpufeatures.h b/xen/include/asm-x86/cpufeatures.h
index f7488d3ccbfa..b233e5835fb5 100644
--- a/xen/include/asm-x86/cpufeatures.h
+++ b/xen/include/asm-x86/cpufeatures.h
@@ -39,6 +39,8 @@ XEN_CPUFEATURE(XEN_LBR,           X86_SYNTH(22)) /* Xen uses MSR_DEBUGCTL.LBR */
 XEN_CPUFEATURE(SC_VERW_IDLE,      X86_SYNTH(25)) /* VERW used by Xen for idle */
 XEN_CPUFEATURE(XEN_SHSTK,         X86_SYNTH(26)) /* Xen uses CET Shadow Stacks */
 XEN_CPUFEATURE(XEN_IBT,           X86_SYNTH(27)) /* Xen uses CET Indirect Branch Tracking */
+XEN_CPUFEATURE(IBPB_ENTRY_PV,     X86_SYNTH(28)) /* MSR_PRED_CMD used by Xen for PV */
+XEN_CPUFEATURE(IBPB_ENTRY_HVM,    X86_SYNTH(29)) /* MSR_PRED_CMD used by Xen for HVM */
 
 /* Bug words follow the synthetic words. */
 #define X86_NR_BUG 1
diff --git a/xen/include/asm-x86/spec_ctrl.h b/xen/include/asm-x86/spec_ctrl.h
index fd8162ca9ab9..10cd0cd2518f 100644
--- a/xen/include/asm-x86/spec_ctrl.h
+++ b/xen/include/asm-x86/spec_ctrl.h
@@ -34,6 +34,8 @@
 #define SCF_ist_sc_msr (1 << 1)
 #define SCF_ist_rsb    (1 << 2)
 #define SCF_verw       (1 << 3)
+#define SCF_ist_ibpb   (1 << 4)
+#define SCF_entry_ibpb (1 << 5)
 
 /*
  * The IST paths (NMI/#MC) can interrupt any arbitrary context.  Some
@@ -46,13 +48,13 @@
  * These are the controls to inhibit on the S3 resume path until microcode has
  * been reloaded.
  */
-#define SCF_IST_MASK (SCF_ist_sc_msr)
+#define SCF_IST_MASK (SCF_ist_sc_msr | SCF_ist_ibpb)
 
 /*
  * Some speculative protections are per-domain.  These settings are merged
  * into the top-of-stack block in the context switch path.
  */
-#define SCF_DOM_MASK (SCF_verw)
+#define SCF_DOM_MASK (SCF_verw | SCF_entry_ibpb)
 
 #ifndef __ASSEMBLY__
 
diff --git a/xen/include/asm-x86/spec_ctrl_asm.h b/xen/include/asm-x86/spec_ctrl_asm.h
index 15e24cde00d1..9eb4ad9ab71d 100644
--- a/xen/include/asm-x86/spec_ctrl_asm.h
+++ b/xen/include/asm-x86/spec_ctrl_asm.h
@@ -88,6 +88,35 @@
  *  - SPEC_CTRL_EXIT_TO_{SVM,VMX}
  */
 
+.macro DO_SPEC_CTRL_COND_IBPB maybexen:req
+/*
+ * Requires %rsp=regs (also cpuinfo if !maybexen)
+ * Requires %r14=stack_end (if maybexen), %rdx=0
+ * Clobbers %rax, %rcx, %rdx
+ *
+ * Conditionally issue IBPB if SCF_entry_ibpb is active.  In the maybexen
+ * case, we can safely look at UREGS_cs to skip taking the hit when
+ * interrupting Xen.
+ */
+    .if \maybexen
+        testb  $SCF_entry_ibpb, STACK_CPUINFO_FIELD(spec_ctrl_flags)(%r14)
+        jz     .L\@_skip
+        testb  $3, UREGS_cs(%rsp)
+    .else
+        testb  $SCF_entry_ibpb, CPUINFO_xen_spec_ctrl(%rsp)
+    .endif
+    jz     .L\@_skip
+
+    mov     $MSR_PRED_CMD, %ecx
+    mov     $PRED_CMD_IBPB, %eax
+    wrmsr
+    jmp     .L\@_done
+
+.L\@_skip:
+    lfence
+.L\@_done:
+.endm
+
 .macro DO_OVERWRITE_RSB tmp=rax
 /*
  * Requires nothing
@@ -225,12 +254,16 @@
 
 /* Use after an entry from PV context (syscall/sysenter/int80/int82/etc). */
 #define SPEC_CTRL_ENTRY_FROM_PV                                         \
+    ALTERNATIVE "", __stringify(DO_SPEC_CTRL_COND_IBPB maybexen=0),     \
+        X86_FEATURE_IBPB_ENTRY_PV;                                      \
     ALTERNATIVE "", DO_OVERWRITE_RSB, X86_FEATURE_SC_RSB_PV;            \
     ALTERNATIVE "", __stringify(DO_SPEC_CTRL_ENTRY maybexen=0),         \
         X86_FEATURE_SC_MSR_PV
 
 /* Use in interrupt/exception context.  May interrupt Xen or PV context. */
 #define SPEC_CTRL_ENTRY_FROM_INTR                                       \
+    ALTERNATIVE "", __stringify(DO_SPEC_CTRL_COND_IBPB maybexen=1),     \
+        X86_FEATURE_IBPB_ENTRY_PV;                                      \
     ALTERNATIVE "", DO_OVERWRITE_RSB, X86_FEATURE_SC_RSB_PV;            \
     ALTERNATIVE "", __stringify(DO_SPEC_CTRL_ENTRY maybexen=1),         \
         X86_FEATURE_SC_MSR_PV
@@ -254,11 +287,23 @@
  * Requires %rsp=regs, %r14=stack_end, %rdx=0
  * Clobbers %rax, %rbx, %rcx, %rdx
  *
- * This is logical merge of DO_OVERWRITE_RSB and DO_SPEC_CTRL_ENTRY
- * maybexen=1, but with conditionals rather than alternatives.
+ * This is logical merge of:
+ *    DO_SPEC_CTRL_COND_IBPB maybexen=0
+ *    DO_OVERWRITE_RSB
+ *    DO_SPEC_CTRL_ENTRY maybexen=1
+ * but with conditionals rather than alternatives.
  */
     movzbl STACK_CPUINFO_FIELD(spec_ctrl_flags)(%r14), %ebx
 
+    test    $SCF_ist_ibpb, %bl
+    jz      .L\@_skip_ibpb
+
+    mov     $MSR_PRED_CMD, %ecx
+    mov     $PRED_CMD_IBPB, %eax
+    wrmsr
+
+.L\@_skip_ibpb:
+
     test $SCF_ist_rsb, %bl
     jz .L\@_skip_rsb