Nested paging working to the point where we manage to get through a puppy boot

main changes are to appropriately set permissions all the way down the page table path
when demand creating the direct map nested page tables

diff --git a/palacios/src/palacios/vmm_direct_paging.c b/palacios/src/palacios/vmm_direct_paging.c
index 821d7b3..65dc470 100644 (file)
--- a/palacios/src/palacios/vmm_direct_paging.c
+++ b/palacios/src/palacios/vmm_direct_paging.c
@@ -69,6 +69,9 @@ int v3_handle_nested_pagefault(struct guest_info * info, addr_t fault_addr, pf_e
     // THIS IS VERY BAD
     v3_vm_cpu_mode_t mode = LONG;
 
+
+    PrintDebug("Nested PageFault: fault_addr=%p, error_code=%u\n",(void*)fault_addr, *(uint_t *)&error_code);
+
     switch(mode) {
        case REAL:
        case PROTECTED:

diff --git a/palacios/src/palacios/vmm_direct_paging_64.h b/palacios/src/palacios/vmm_direct_paging_64.h
index d79693b..2b9f470 100644 (file)
--- a/palacios/src/palacios/vmm_direct_paging_64.h
+++ b/palacios/src/palacios/vmm_direct_paging_64.h
@@ -27,6 +27,13 @@
 #include <palacios/vm_guest_mem.h>
 #include <palacios/vm_guest.h>
 
+#define DEBUG_NESTED_PAGING 1
+
+#ifndef DEBUG_NESTED_PAGING
+#undef PrintDebug
+#define PrintDebug(fmt, args...)
+#endif
+
 
 static inline int handle_passthrough_pagefault_64(struct guest_info * info, 
                                                     addr_t fault_addr, 
@@ -43,6 +50,8 @@ static inline int handle_passthrough_pagefault_64(struct guest_info * info,
     int pte_index = PTE64_INDEX(fault_addr);
 
 
+    
+
     struct v3_shadow_region * region =  v3_get_shadow_region(info, fault_addr);
   
     if ((region == NULL) || 
@@ -53,6 +62,7 @@ static inline int handle_passthrough_pagefault_64(struct guest_info * info,
     }
 
     host_addr = v3_get_shadow_addr(region, fault_addr);
+    //
 
     // Lookup the correct PML address based on the PAGING MODE
     if (info->shdw_pg_mode == SHADOW_PAGING) {
@@ -65,8 +75,11 @@ static inline int handle_passthrough_pagefault_64(struct guest_info * info,
     if (pml[pml_index].present == 0) {
        pdpe = (pdpe64_t *)create_generic_pt_page();
    
-       pml[pml_index].present = 1;
        // Set default PML Flags...
+       pml[pml_index].present = 1;
+        pml[pml_index].writable = 1;
+        pml[pml_index].user_page = 1;
+
        pml[pml_index].pdp_base_addr = PAGE_BASE_ADDR((addr_t)V3_PAddr(pdpe));    
     } else {
        pdpe = V3_VAddr((void*)BASE_TO_PAGE_ADDR(pml[pml_index].pdp_base_addr));
@@ -76,8 +89,11 @@ static inline int handle_passthrough_pagefault_64(struct guest_info * info,
     if (pdpe[pdpe_index].present == 0) {
        pde = (pde64_t *)create_generic_pt_page();
        
-       pdpe[pdpe_index].present = 1;
        // Set default PDPE Flags...
+       pdpe[pdpe_index].present = 1;
+       pdpe[pdpe_index].writable = 1;
+       pdpe[pdpe_index].user_page = 1;
+
        pdpe[pdpe_index].pd_base_addr = PAGE_BASE_ADDR((addr_t)V3_PAddr(pde));    
     } else {
        pde = V3_VAddr((void*)BASE_TO_PAGE_ADDR(pdpe[pdpe_index].pd_base_addr));