[PATCH RFC v12 15/21] pvh: Set up more PV stuff in set_info_guest

[George Dunlap <george.dunlap@eu.citrix.com>]

Allow the guest to set up a few more things when bringing up a vcpu.

This includes cr3 and gs_base.

Also set up wallclock, and only initialize a vcpu once.

Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Mukesh Rathor <mukesh.rathor@oracle.com>
CC: Jan Beulich <jan.beulich@suse.com>
CC: Tim Deegan <tim@xen.org>
CC: Keir Fraser <keir@xen.org>
---
 xen/arch/x86/domain.c             |   27 ++++++++++++++++++++++++++-
 xen/arch/x86/hvm/vmx/vmx.c        |   34 ++++++++++++++++++++++++++++++++++
 xen/include/asm-x86/hvm/hvm.h     |    8 ++++++++
 xen/include/asm-x86/hvm/vmx/vmx.h |    2 ++
 xen/include/public/arch-x86/xen.h |    4 ++++
 5 files changed, 74 insertions(+), 1 deletion(-)

diff --git a/xen/arch/x86/domain.c b/xen/arch/x86/domain.c
index 02a6479..d046790 100644
--- a/xen/arch/x86/domain.c
+++ b/xen/arch/x86/domain.c
@@ -737,7 +737,31 @@ int arch_set_info_guest(
     if ( has_hvm_container_vcpu(v) )
     {
         hvm_set_info_guest(v);
-        goto out;
+
+        if ( is_hvm_vcpu(v) || v->is_initialised )
+            goto out;
+
+        /* PVH */
+        if ( c.nat->ctrlreg[1] )
+            return -EINVAL;
+
+        ASSERT(!compat);
+
+        cr3_gfn = xen_cr3_to_pfn(c.nat->ctrlreg[3]);
+        cr3_page = get_page_from_gfn(d, cr3_gfn, NULL, P2M_ALLOC);
+
+        v->arch.cr3 = page_to_maddr(cr3_page);
+        v->arch.hvm_vcpu.guest_cr[3] = c.nat->ctrlreg[3];
+        
+        if ( paging_mode_enabled(d) )
+            paging_update_paging_modes(v);
+
+        update_cr3(v);
+
+        if ( (rc = pvh_vcpu_boot_set_info(v, c.nat)) != 0 )
+            return rc;
+
+        goto pvh_skip_pv_stuff;
     }
 
     init_int80_direct_trap(v);
@@ -947,6 +971,7 @@ int arch_set_info_guest(
 
     update_cr3(v);
 
+ pvh_skip_pv_stuff:
     if ( v->vcpu_id == 0 )
         update_domain_wallclock_time(d);
 
diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c
index c2156af..b3f933a 100644
--- a/xen/arch/x86/hvm/vmx/vmx.c
+++ b/xen/arch/x86/hvm/vmx/vmx.c
@@ -122,6 +122,39 @@ static int vmx_vcpu_initialise(struct vcpu *v)
     return 0;
 }
 
+/*
+ * Set vmcs fields during boot of a vcpu. Called from arch_set_info_guest.
+ *
+ * Boot vcpu call is from tools via:
+ *     do_domctl -> XEN_DOMCTL_setvcpucontext -> arch_set_info_guest
+ *
+ * Secondary vcpu's are brought up by the guest itself via:
+ *     do_vcpu_op -> VCPUOP_initialise -> arch_set_info_guest
+ *     (In case of linux, the call comes from cpu_initialize_context()).
+ *
+ * Note, PVH save/restore is expected to happen the HVM way, ie,
+ *        do_domctl -> XEN_DOMCTL_sethvmcontext -> hvm_load/save
+ * and not get here.
+ *
+ * PVH 32bitfixme: this function needs to be modified for 32bit guest.
+ */
+int vmx_pvh_vcpu_boot_set_info(struct vcpu *v,
+                               struct vcpu_guest_context *ctxtp)
+{
+    if ( ctxtp->ldt_base || ctxtp->ldt_ents ||
+         ctxtp->user_regs.cs || ctxtp->user_regs.ss || ctxtp->user_regs.es ||
+         ctxtp->user_regs.ds || ctxtp->user_regs.fs || ctxtp->user_regs.gs ||
+         *ctxtp->gdt_frames || ctxtp->gdt_ents ||
+         ctxtp->fs_base || ctxtp->gs_base_user )
+        return -EINVAL;
+
+    vmx_vmcs_enter(v);
+    __vmwrite(GUEST_GS_BASE, ctxtp->gs_base_kernel);
+    vmx_vmcs_exit(v);
+
+    return 0;
+}
+
 static void vmx_vcpu_destroy(struct vcpu *v)
 {
     vmx_destroy_vmcs(v);
@@ -1606,6 +1639,7 @@ static struct hvm_function_table __initdata vmx_function_table = {
     .handle_eoi           = vmx_handle_eoi,
     .nhvm_hap_walk_L1_p2m = nvmx_hap_walk_L1_p2m,
     .read_selector        = vmx_read_selector,
+    .pvh_vcpu_boot_set_info = vmx_pvh_vcpu_boot_set_info,
 };
 
 const struct hvm_function_table * __init start_vmx(void)
diff --git a/xen/include/asm-x86/hvm/hvm.h b/xen/include/asm-x86/hvm/hvm.h
index 401fa4c..d81a744 100644
--- a/xen/include/asm-x86/hvm/hvm.h
+++ b/xen/include/asm-x86/hvm/hvm.h
@@ -194,6 +194,8 @@ struct hvm_function_table {
                                 uint8_t *p2m_acc, bool_t access_r,
                                 bool_t access_w, bool_t access_x);
     u16 (*read_selector)(struct vcpu *v, enum x86_segment seg);
+    int (*pvh_vcpu_boot_set_info)(struct vcpu *v,
+                                  struct vcpu_guest_context *ctxtp);
 };
 
 extern struct hvm_function_table hvm_funcs;
@@ -327,6 +329,12 @@ static inline unsigned long hvm_get_shadow_gs_base(struct vcpu *v)
     return hvm_funcs.get_shadow_gs_base(v);
 }
 
+static inline int pvh_vcpu_boot_set_info(struct vcpu *v,
+                                         struct vcpu_guest_context *ctxtp)
+{
+    return hvm_funcs.pvh_vcpu_boot_set_info(v, ctxtp);
+}
+
 #define is_viridian_domain(_d)                                             \
  (is_hvm_domain(_d) && ((_d)->arch.hvm_domain.params[HVM_PARAM_VIRIDIAN]))
 
diff --git a/xen/include/asm-x86/hvm/vmx/vmx.h b/xen/include/asm-x86/hvm/vmx/vmx.h
index c33b9f9..623e4ce 100644
--- a/xen/include/asm-x86/hvm/vmx/vmx.h
+++ b/xen/include/asm-x86/hvm/vmx/vmx.h
@@ -458,6 +458,8 @@ void ept_walk_table(struct domain *d, unsigned long gfn);
 void setup_ept_dump(void);
 
 void update_guest_eip(void);
+int  vmx_pvh_vcpu_boot_set_info(struct vcpu *v,
+                                struct vcpu_guest_context *ctxtp);
 
 int alloc_p2m_hap_data(struct p2m_domain *p2m);
 void free_p2m_hap_data(struct p2m_domain *p2m);
diff --git a/xen/include/public/arch-x86/xen.h b/xen/include/public/arch-x86/xen.h
index b7f6a51..4f12f50 100644
--- a/xen/include/public/arch-x86/xen.h
+++ b/xen/include/public/arch-x86/xen.h
@@ -150,6 +150,10 @@ typedef uint64_t tsc_timestamp_t; /* RDTSC timestamp */
 /*
  * The following is all CPU context. Note that the fpu_ctxt block is filled 
  * in by FXSAVE if the CPU has feature FXSR; otherwise FSAVE is used.
+ *
+ * PVH 64bit: In the vcpu boot path, for vmcs context, only gs_base_kernel
+ *            is honored. Other fields like gdt, ldt, and selectors must be
+ *            zeroed. See vmx_pvh_vcpu_boot_set_info.
  */
 struct vcpu_guest_context {
     /* FPU registers come first so they can be aligned for FXSAVE/FXRSTOR. */
-- 
1.7.9.5