X-Git-Url: http://v3vee.org/palacios/gitweb/gitweb.cgi?a=blobdiff_plain;f=palacios%2Fsrc%2Fpalacios%2Fvmm_shadow_paging.c;h=a2f09f9544cc9d924b97efbb09e59d29c66b4014;hb=684409c4ad29c0f06167bebf2bb51e21374a6840;hp=c9546e4d30bfece0df576771d146f085bba7e568;hpb=a109eb919a162bd7de58d62020801bc2e633be50;p=palacios.git diff --git a/palacios/src/palacios/vmm_shadow_paging.c b/palacios/src/palacios/vmm_shadow_paging.c index c9546e4..a2f09f9 100644 --- a/palacios/src/palacios/vmm_shadow_paging.c +++ b/palacios/src/palacios/vmm_shadow_paging.c @@ -1,209 +1,443 @@ #include + #include #include +#include -extern struct vmm_os_hooks * os_hooks; int init_shadow_page_state(struct shadow_page_state * state) { state->guest_mode = PDE32; state->shadow_mode = PDE32; - state->guest_cr3.r_reg = 0; - state->shadow_cr3.r_reg = 0; + state->guest_cr3 = 0; + state->shadow_cr3 = 0; return 0; } - -int wholesale_update_shadow_page_state(struct guest_info * guest_info) { - unsigned i, j; - pde32_t * guest_pde; - pde32_t * shadow_pde; +int handle_shadow_pagefault(struct guest_info * info, addr_t fault_addr, pf_error_t error_code) { + + if (info->mem_mode == PHYSICAL_MEM) { + // If paging is not turned on we need to handle the special cases + return handle_special_page_fault(info, fault_addr, error_code); + } else if (info->mem_mode == VIRTUAL_MEM) { + + switch (info->cpu_mode) { + case PROTECTED: + return handle_shadow_pagefault32(info, fault_addr, error_code); + break; + case PROTECTED_PAE: + case LONG: + // currently not handled + return -1; + break; + default: + return -1; + } + } else { + PrintDebug("Invalid Memory mode\n"); + return -1; + } +} - struct shadow_page_state * state = &(guest_info->shdw_pg_state); +int handle_shadow_pagefault32(struct guest_info * info, addr_t fault_addr, pf_error_t error_code) { + pde32_t * guest_pde = NULL; + pde32_t * shadow_pde = (pde32_t *)CR3_TO_PDE32(info->shdw_pg_state.shadow_cr3); + addr_t guest_cr3 = CR3_TO_PDE32(info->shdw_pg_state.guest_cr3); + pt_access_status_t guest_pde_access; + pt_access_status_t shadow_pde_access; + pde32_t * guest_pde_entry = NULL; + pde32_t * shadow_pde_entry = (pde32_t *)&(shadow_pde[PDE32_INDEX(fault_addr)]); - // For now, we'll only work with PDE32 - if (state->guest_mode != PDE32) { + if (guest_pa_to_host_va(info, guest_cr3, (addr_t*)&guest_pde) == -1) { + PrintDebug("Invalid Guest PDE Address: 0x%x\n", guest_cr3); return -1; } - shadow_pde = (pde32_t *)(CR3_TO_PDE32(state->shadow_cr3.e_reg.low)); - if (host_pa_to_host_va(CR3_TO_PDE32(state->guest_cr3.e_reg.low), (addr_t*)&guest_pde) != 0) { - return -1; + guest_pde_entry = (pde32_t *)&(guest_pde[PDE32_INDEX(fault_addr)]); + + // Check the guest page permissions + guest_pde_access = can_access_pde32(guest_pde, fault_addr, error_code); + + if (guest_pde_access != PT_ACCESS_OK) { + // inject page fault to the guest (Guest PDE fault) + + info->ctrl_regs.cr2 = fault_addr; + raise_exception_with_error(info, PF_EXCEPTION, *(uint_t *)&error_code); + + PrintDebug("Injecting PDE pf to guest\n"); + return 0; } - // Delete the current page table - delete_page_tables_pde32(shadow_pde); + shadow_pde_access = can_access_pde32(shadow_pde, fault_addr, error_code); - shadow_pde = os_hooks->allocate_pages(1); - state->shadow_cr3.e_reg.low = (addr_t)shadow_pde; + if (shadow_pde_access == PT_ENTRY_NOT_PRESENT) { + pte32_t * shadow_pte = NULL; - state->shadow_mode = PDE32; + V3_AllocPages(shadow_pte, 1); + memset(shadow_pte, 0, PAGE_SIZE); + + shadow_pde_entry->pt_base_addr = PD32_BASE_ADDR(shadow_pte); + - for (i = 0; i < MAX_PDE32_ENTRIES; i++) { - shadow_pde[i] = guest_pde[i]; + shadow_pde_entry->present = 1; + shadow_pde_entry->user_page = guest_pde_entry->user_page; + + // VMM Specific options + shadow_pde_entry->write_through = 0; + shadow_pde_entry->cache_disable = 0; + shadow_pde_entry->global_page = 0; + // + + guest_pde_entry->accessed = 1; + + if (guest_pde_entry->large_page == 0) { + shadow_pde_entry->writable = guest_pde_entry->writable; + } else { + /* + * Check the Intel manual because we are ignoring Large Page issues here + * Also be wary of hooked pages + */ + + PrintDebug("Large PAge!!!\n"); + return -1; - // The shadow can be identical to the guest if it's not present - if (!shadow_pde[i].present) { - continue; } - if (shadow_pde[i].large_pages) { - // large page - just map it through shadow map to generate its physical location - addr_t guest_addr = PAGE_ADDR(shadow_pde[i].pt_base_addr); - addr_t host_addr; - shadow_region_t * ent; + } else if (shadow_pde_access == PT_WRITE_ERROR) { + + // + // Page Directory Entry marked read-only + // + + PrintDebug("Shadow Paging Write Error\n"); + return -1; + } else if (shadow_pde_access == PT_USER_ERROR) { + + // + // Page Directory Entry marked non-user + // + + PrintDebug("Shadow Paging User access error\n"); + return -1; + } else if (shadow_pde_access == PT_ACCESS_OK) { + pte32_t * shadow_pte = (pte32_t *)PDE32_T_ADDR((*shadow_pde_entry)); + pte32_t * guest_pte = NULL; + + // Page Table Entry fault + + if (guest_pa_to_host_va(info, PDE32_T_ADDR((*guest_pde_entry)), (addr_t*)&guest_pte) == -1) { + PrintDebug("Invalid Guest PTE Address: 0x%x\n", PDE32_T_ADDR((*guest_pde_entry))); + // Machine check the guest - ent = get_shadow_region_by_addr(&(guest_info->mem_map), guest_addr); + raise_exception(info, MC_EXCEPTION); - if (!ent) { - // FIXME Panic here - guest is trying to map to physical memory - // it does not own in any way! + return 0; + } + + + if (handle_shadow_pte32_fault(info, fault_addr, error_code, shadow_pte, guest_pte) == -1) { + PrintDebug("Error handling Page fault caused by PTE\n"); + return -1; + } + + } else { + // Unknown error raise page fault in guest + info->ctrl_regs.cr2 = fault_addr; + raise_exception_with_error(info, PF_EXCEPTION, *(uint_t *)&error_code); + + // For debugging we will return an error here for the time being, + // this probably shouldn't ever happen + PrintDebug("Unknown Error occurred\n"); + PrintDebug("Manual Says to inject page fault into guest\n"); + return -1; + } + + //PrintDebugPageTables(shadow_pde); + PrintDebug("Returning end of PDE function\n"); + return 0; +} + + + +/* + * We assume the the guest pte pointer has already been translated to a host virtual address + */ +int handle_shadow_pte32_fault(struct guest_info * info, + addr_t fault_addr, + pf_error_t error_code, + pte32_t * shadow_pte, + pte32_t * guest_pte) { + + pt_access_status_t guest_pte_access; + pt_access_status_t shadow_pte_access; + pte32_t * guest_pte_entry = (pte32_t *)&(guest_pte[PTE32_INDEX(fault_addr)]);; + pte32_t * shadow_pte_entry = (pte32_t *)&(shadow_pte[PTE32_INDEX(fault_addr)]); + + + // Check the guest page permissions + guest_pte_access = can_access_pte32(guest_pte, fault_addr, error_code); + + + if (guest_pte_access != PT_ACCESS_OK) { + // Inject page fault into the guest + + info->ctrl_regs.cr2 = fault_addr; + raise_exception_with_error(info, PF_EXCEPTION, *(uint_t *)&error_code); + + PrintDebug("Access error injecting pf to guest\n"); + return 0; + } + + + shadow_pte_access = can_access_pte32(shadow_pte, fault_addr, error_code); + + if (shadow_pte_access == PT_ACCESS_OK) { + // Inconsistent state... + // Guest Re-Entry will flush page tables and everything should now work + PrintDebug("Inconsistent state... Guest re-entry should flush tlb\n"); + return 0; + } else if (shadow_pte_access == PT_ENTRY_NOT_PRESENT) { + addr_t shadow_pa; + addr_t guest_pa = PTE32_T_ADDR((*guest_pte_entry)); + + // Page Table Entry Not Present + + host_region_type_t host_page_type = get_shadow_addr_type(info, guest_pa); + + if (host_page_type == HOST_REGION_INVALID) { + // Inject a machine check in the guest + + raise_exception(info, MC_EXCEPTION); + + PrintDebug("Invalid Guest Address in page table (0x%x)\n", guest_pa); + return 0; + + } else if (host_page_type == HOST_REGION_PHYSICAL_MEMORY) { + + shadow_pa = get_shadow_addr(info, guest_pa); + + shadow_pte_entry->page_base_addr = PT32_BASE_ADDR(shadow_pa); + + shadow_pte_entry->present = guest_pte_entry->present; + shadow_pte_entry->user_page = guest_pte_entry->user_page; + + //set according to VMM policy + shadow_pte_entry->write_through = 0; + shadow_pte_entry->cache_disable = 0; + shadow_pte_entry->global_page = 0; + // + + guest_pte_entry->accessed = 1; + + if (guest_pte_entry->dirty == 1) { + shadow_pte_entry->writable = guest_pte_entry->writable; + } else if ((guest_pte_entry->dirty == 0) && (error_code.write == 1)) { + shadow_pte_entry->writable = guest_pte_entry->writable; + guest_pte_entry->dirty = 1; + } else if ((guest_pte_entry->dirty = 0) && (error_code.write == 0)) { + shadow_pte_entry->writable = 0; + } + } else { + // Page fault handled by hook functions + if (handle_special_page_fault(info, fault_addr, error_code) == -1) { + PrintDebug("Special Page fault handler returned error for address: %x\n", fault_addr); return -1; } + } - // FIXME Bounds check here to see if it's trying to trick us - - switch (ent->host_type) { - case HOST_REGION_PHYSICAL_MEMORY: - // points into currently allocated physical memory, so we just - // set up the shadow to point to the mapped location - if (guest_pa_to_host_pa(guest_info, guest_addr, &host_addr)) { - // Panic here - return -1; + } else if ((shadow_pte_access == PT_WRITE_ERROR) && + (guest_pte_entry->dirty == 0)) { + guest_pte_entry->dirty = 1; + shadow_pte_entry->writable = guest_pte_entry->writable; + + PrintDebug("Shadow PTE Write Error\n"); + + return 0; + } else { + // Inject page fault into the guest + + info->ctrl_regs.cr2 = fault_addr; + raise_exception_with_error(info, PF_EXCEPTION, *(uint_t *)&error_code); + + PrintDebug("PTE Page fault fell through... Not sure if this should ever happen\n"); + PrintDebug("Manual Says to inject page fault into guest\n"); + return -1; + } + + PrintDebug("Returning end of function\n"); + return 0; +} + + + +addr_t create_new_shadow_pt32(struct guest_info * info) { + void * host_pde = 0; + + V3_AllocPages(host_pde, 1); + memset(host_pde, 0, PAGE_SIZE); + + return (addr_t)host_pde; +} + + + +/* Currently Does not work with Segmentation!!! */ +int handle_shadow_invlpg(struct guest_info * info) { + if (info->mem_mode != VIRTUAL_MEM) { + // Paging must be turned on... + // should handle with some sort of fault I think + PrintDebug("ERROR: INVLPG called in non paged mode\n"); + return -1; + } + + + if (info->cpu_mode == PROTECTED) { + char instr[15]; + int ret; + int index = 0; + + ret = read_guest_va_memory(info, get_addr_linear(info, info->rip, &(info->segments.cs)), 15, instr); + if (ret != 15) { + PrintDebug("Could not read instruction 0x%x (ret=%d)\n", info->rip, ret); + return -1; + } + + + /* Can INVLPG work with Segments?? */ + while (is_prefix_byte(instr[index])) { + index++; + } + + + if ((instr[index] == (uchar_t)0x0f) && + (instr[index + 1] == (uchar_t)0x01)) { + + addr_t first_operand; + addr_t second_operand; + operand_type_t addr_type; + + index += 2; + + addr_type = decode_operands32(&(info->vm_regs), instr + index, &index, &first_operand, &second_operand, REG32); + + if (addr_type == MEM_OPERAND) { + pde32_t * shadow_pd = (pde32_t *)CR3_TO_PDE32(info->shdw_pg_state.shadow_cr3); + pde32_t * shadow_pde_entry = (pde32_t *)&shadow_pd[PDE32_INDEX(first_operand)]; + + //PrintDebug("PDE Index=%d\n", PDE32_INDEX(first_operand)); + //PrintDebug("FirstOperand = %x\n", first_operand); + + if (shadow_pde_entry->large_page == 1) { + shadow_pde_entry->present = 0; + } else { + if (shadow_pde_entry->present == 1) { + pte32_t * shadow_pt = (pte32_t *)PDE32_T_ADDR((*shadow_pde_entry)); + pte32_t * shadow_pte_entry = (pte32_t *)&shadow_pt[PTE32_INDEX(first_operand)]; + + shadow_pte_entry->present = 0; + } } - shadow_pde[i].pt_base_addr = PAGE_ALIGNED_ADDR(host_addr); - // FIXME set vmm_info bits here - break; - case HOST_REGION_UNALLOCATED: - // points to physical memory that is *allowed* but that we - // have not yet allocated. We mark as not present and set a - // bit to remind us to allocate it later - shadow_pde[i].present = 0; - // FIXME Set vminfo bits here so that we know that we will be - // allocating it later - break; - case HOST_REGION_NOTHING: - // points to physical memory that is NOT ALLOWED. - // We will mark it as not present and set a bit to remind - // us that it's bad later and insert a GPF then - shadow_pde[i].present = 0; - break; - case HOST_REGION_MEMORY_MAPPED_DEVICE: - case HOST_REGION_REMOTE: - case HOST_REGION_SWAPPED: - default: - // Panic. Currently unhandled + info->rip += index; + + } else { + PrintDebug("Invalid Operand type\n"); return -1; - break; } } else { - pte32_t * guest_pte; - pte32_t * shadow_pte; - addr_t guest_addr; - addr_t guest_pte_host_addr; - shadow_region_t * ent; + PrintDebug("invalid Instruction Opcode\n"); + PrintTraceMemDump(instr, 15); + return -1; + } + } - // small page - set PDE and follow down to the child table - shadow_pde[i] = guest_pde[i]; + return 0; +} - guest_addr = PAGE_ADDR(guest_pde[i].pt_base_addr); - // Allocate a new second level page table for the shadow - shadow_pte = os_hooks->allocate_pages(1); - // make our first level page table in the shadow point to it - shadow_pde[i].pt_base_addr = PAGE_ALIGNED_ADDR(shadow_pte); - - ent = get_shadow_region_by_addr(&(guest_info->mem_map), guest_addr); - +/* Deprecated */ +/* +addr_t setup_shadow_pt32(struct guest_info * info, addr_t virt_cr3) { + addr_t cr3_guest_addr = CR3_TO_PDE32(virt_cr3); + pde32_t * guest_pde; + pde32_t * host_pde = NULL; + int i; + + // Setup up guest_pde to point to the PageDir in host addr + if (guest_pa_to_host_va(info, cr3_guest_addr, (addr_t*)&guest_pde) == -1) { + return 0; + } + + V3_AllocPages(host_pde, 1); + memset(host_pde, 0, PAGE_SIZE); - /* JRL: This is bad.... */ - // For now the guest Page Table must always be mapped to host physical memory - /* If we swap out a page table or if it isn't present for some reason, this turns real ugly */ + for (i = 0; i < MAX_PDE32_ENTRIES; i++) { + if (guest_pde[i].present == 1) { + addr_t pt_host_addr; + addr_t host_pte; - if ((!ent) || (ent->host_type != HOST_REGION_PHYSICAL_MEMORY)) { - // FIXME Panic here - guest is trying to map to physical memory - // it does not own in any way! - return -1; + if (guest_pa_to_host_va(info, PDE32_T_ADDR(guest_pde[i]), &pt_host_addr) == -1) { + return 0; } - // Address of the relevant second level page table in the guest - if (guest_pa_to_host_pa(guest_info, guest_addr, &guest_pte_host_addr)) { - // Panic here - return -1; + if ((host_pte = setup_shadow_pte32(info, pt_host_addr)) == 0) { + return 0; } + host_pde[i].present = 1; + host_pde[i].pt_base_addr = PD32_BASE_ADDR(host_pte); + + // + // Set Page DIR flags + // + } + } + + PrintDebugPageTables(host_pde); - // host_addr now contains the host physical address for the guest's 2nd level page table - // Now we transform it to relevant virtual address - guest_pte = os_hooks->paddr_to_vaddr((void *)guest_pte_host_addr); + return (addr_t)host_pde; +} - // Now we walk through the second level guest page table - // and clone it into the shadow - for (j = 0; j < MAX_PTE32_ENTRIES; j++) { - shadow_pte[j] = guest_pte[j]; - addr_t guest_addr = PAGE_ADDR(shadow_pte[j].page_base_addr); - - shadow_region_t * ent; - ent = get_shadow_region_by_addr(&(guest_info->mem_map), guest_addr); - - if (!ent) { - // FIXME Panic here - guest is trying to map to physical memory - // it does not own in any way! - return -1; - } +addr_t setup_shadow_pte32(struct guest_info * info, addr_t pt_host_addr) { + pte32_t * guest_pte = (pte32_t *)pt_host_addr; + pte32_t * host_pte = NULL; + int i; - switch (ent->host_type) { - case HOST_REGION_PHYSICAL_MEMORY: - { - addr_t host_addr; - - // points into currently allocated physical memory, so we just - // set up the shadow to point to the mapped location - if (guest_pa_to_host_pa(guest_info, guest_addr, &host_addr)) { - // Panic here - return -1; - } - - shadow_pte[j].page_base_addr = PAGE_ALIGNED_ADDR(host_addr); - // FIXME set vmm_info bits here - break; - } - case HOST_REGION_UNALLOCATED: - // points to physical memory that is *allowed* but that we - // have not yet allocated. We mark as not present and set a - // bit to remind us to allocate it later - shadow_pte[j].present = 0; - // FIXME Set vminfo bits here so that we know that we will be - // allocating it later - break; - case HOST_REGION_NOTHING: - // points to physical memory that is NOT ALLOWED. - // We will mark it as not present and set a bit to remind - // us that it's bad later and insert a GPF then - shadow_pte[j].present = 0; - break; - case HOST_REGION_MEMORY_MAPPED_DEVICE: - case HOST_REGION_REMOTE: - case HOST_REGION_SWAPPED: - default: - // Panic. Currently unhandled - return -1; - break; - } + V3_AllocPages(host_pte, 1); + memset(host_pte, 0, PAGE_SIZE); + + for (i = 0; i < MAX_PTE32_ENTRIES; i++) { + if (guest_pte[i].present == 1) { + addr_t guest_pa = PTE32_T_ADDR(guest_pte[i]); + shadow_mem_type_t page_type; + addr_t host_pa = 0; + + page_type = get_shadow_addr_type(info, guest_pa); + + if (page_type == HOST_REGION_PHYSICAL_MEMORY) { + host_pa = get_shadow_addr(info, guest_pa); + } else { + + // + // Setup various memory types + // } + + host_pte[i].page_base_addr = PT32_BASE_ADDR(host_pa); + host_pte[i].present = 1; } } - return 0; + + return (addr_t)host_pte; } - +*/