/* * Copyright (c) 2006-2022, RT-Thread Development Team * * SPDX-License-Identifier: Apache-2.0 * * Change Logs: * Date Author Notes * 2022-11-14 WangXiaoyao the first version */ /** * @brief Virtual Address Space */ #include #include #include #include #define DBG_TAG "mm.aspace" #define DBG_LVL DBG_INFO #include #include "avl_adpt.h" #include "mm_aspace.h" #include "mm_fault.h" #include "mm_flag.h" #include "mm_page.h" #include "mm_private.h" #include #include static void *_find_free(rt_aspace_t aspace, void *prefer, rt_size_t req_size, void *limit_start, rt_size_t limit_size, mm_flag_t flags); static void _varea_uninstall(rt_varea_t varea); struct rt_aspace rt_kernel_space; int _init_lock(rt_aspace_t aspace) { int err; MM_PGTBL_LOCK_INIT(aspace); err = rt_mutex_init(&aspace->bst_lock, "aspace", RT_IPC_FLAG_FIFO); return err; } rt_err_t rt_aspace_init(rt_aspace_t aspace, void *start, rt_size_t length, void *pgtbl) { int err = RT_EOK; if (pgtbl) { aspace->page_table = pgtbl; aspace->start = start; aspace->size = length; err = _aspace_bst_init(aspace); if (err == RT_EOK) { /** * It has the side effect that lock will be added to object * system management. So it must be paired with a detach once * the initialization return successfully. */ err = _init_lock(aspace); } } else { err = -RT_EINVAL; } return err; } rt_aspace_t rt_aspace_create(void *start, rt_size_t length, void *pgtbl) { rt_aspace_t aspace = NULL; int err; RT_ASSERT(length <= 0 - (rt_size_t)start); aspace = (rt_aspace_t)rt_malloc(sizeof(*aspace)); if (aspace) { rt_memset(aspace, 0, sizeof(*aspace)); err = rt_aspace_init(aspace, start, length, pgtbl); if (err != RT_EOK) { LOG_W("%s(%p, %lx, %p): failed with code %d\n", __func__, start, length, pgtbl, err); rt_free(aspace); aspace = RT_NULL; } } return aspace; } void rt_aspace_detach(rt_aspace_t aspace) { WR_LOCK(aspace); rt_varea_t varea = ASPACE_VAREA_FIRST(aspace); while (varea) { rt_varea_t prev = varea; _varea_uninstall(varea); varea = ASPACE_VAREA_NEXT(varea); if (!(prev->flag & MMF_STATIC_ALLOC)) { rt_free(prev); } } WR_UNLOCK(aspace); rt_mutex_detach(&aspace->bst_lock); } void rt_aspace_delete(rt_aspace_t aspace) { RT_ASSERT(aspace); rt_aspace_detach(aspace); rt_free(aspace); } static int _do_named_map(rt_aspace_t aspace, void *vaddr, rt_size_t length, rt_size_t offset, rt_size_t attr) { LOG_D("%s: va %p length %p", __func__, vaddr, length); int err = RT_EOK; /* it's ensured by caller that (void*)end will not overflow */ void *end = vaddr + length; void *phyaddr = (void *)(offset << MM_PAGE_SHIFT); while (vaddr != end) { /* TODO try to map with huge TLB, when flag & HUGEPAGE */ rt_size_t pgsz = ARCH_PAGE_SIZE; void *ret = rt_hw_mmu_map(aspace, vaddr, phyaddr, pgsz, attr); if (ret == RT_NULL) { err = -RT_ERROR; break; } vaddr += pgsz; phyaddr += pgsz; } if (err == RT_EOK) rt_hw_tlb_invalidate_range(aspace, end - length, length, ARCH_PAGE_SIZE); return err; } rt_inline void _do_page_fault(struct rt_aspace_fault_msg *msg, rt_size_t off, void *vaddr, rt_mem_obj_t mem_obj, rt_varea_t varea) { msg->off = off; msg->fault_vaddr = vaddr; msg->fault_op = MM_FAULT_OP_READ; msg->fault_type = MM_FAULT_TYPE_PAGE_FAULT; msg->response.status = -1; msg->response.vaddr = 0; msg->response.size = 0; mem_obj->on_page_fault(varea, msg); } int _varea_map_with_msg(rt_varea_t varea, struct rt_aspace_fault_msg *msg) { int err = -RT_ERROR; if (msg->response.status == MM_FAULT_STATUS_OK) { /** * the page returned by handler is not checked * cause no much assumption can make on it */ void *store = msg->response.vaddr; rt_size_t store_sz = msg->response.size; if (msg->fault_vaddr + store_sz > varea->start + varea->size) { LOG_W("%s: too much (0x%lx) of buffer on vaddr %p is provided", __func__, store_sz, msg->fault_vaddr); } else { void *map; void *v_addr = msg->fault_vaddr; void *p_addr = store + PV_OFFSET; map = rt_hw_mmu_map(varea->aspace, v_addr, p_addr, store_sz, varea->attr); if (!map) { LOG_W("%s: MMU mapping failed for va %p to %p of %lx", __func__, msg->fault_vaddr, store + PV_OFFSET, store_sz); } else { rt_hw_tlb_invalidate_range(varea->aspace, v_addr, store_sz, ARCH_PAGE_SIZE); err = RT_EOK; } } } else if (msg->response.status == MM_FAULT_STATUS_OK_MAPPED) { if (rt_hw_mmu_v2p(varea->aspace, msg->fault_vaddr) == ARCH_MAP_FAILED) { LOG_W("%s: no page is mapped on %p", __func__, msg->fault_vaddr); } else { err = RT_EOK; } } else { LOG_W("%s: failed on va %p inside varea %p(%s)", __func__, msg->fault_vaddr, varea, varea->mem_obj->get_name ? varea->mem_obj->get_name(varea) : "unknow"); } return err; } /* allocate memory page for mapping range */ static int _do_prefetch(rt_aspace_t aspace, rt_varea_t varea, void *start, rt_size_t size) { int err = RT_EOK; /* it's ensured by caller that start & size ara page-aligned */ void *end = start + size; void *vaddr = start; rt_size_t off = varea->offset + ((start - varea->start) >> ARCH_PAGE_SHIFT); while (vaddr != end) { /* TODO try to map with huge TLB, when flag & HUGEPAGE */ struct rt_aspace_fault_msg msg; _do_page_fault(&msg, off, vaddr, varea->mem_obj, varea); if (_varea_map_with_msg(varea, &msg)) break; /** * It's hard to identify the mapping pattern on a customized handler * So we terminate the prefetch process on that case */ if (msg.response.status == MM_FAULT_STATUS_OK_MAPPED) break; vaddr += msg.response.size; off += msg.response.size >> ARCH_PAGE_SHIFT; } return err; } /* caller must hold the aspace lock */ static int _varea_install(rt_aspace_t aspace, rt_varea_t varea, rt_mm_va_hint_t hint) { void *alloc_va; int err = RT_EOK; /** * find a suitable va range. * even though this is sleepable, it's still ok for startup routine */ alloc_va = _find_free(aspace, hint->prefer, hint->map_size, hint->limit_start, hint->limit_range_size, hint->flags); /* TODO try merge surrounding regions to optimize memory footprint */ if (alloc_va != RT_NULL) { varea->start = alloc_va; _aspace_bst_insert(aspace, varea); } else { err = -RT_ENOSPC; } return err; } static inline void _varea_post_install(rt_varea_t varea, rt_aspace_t aspace, rt_size_t attr, rt_size_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { varea->aspace = aspace; varea->attr = attr; varea->mem_obj = mem_obj; varea->flag = flags; varea->offset = offset; varea->frames = NULL; if (varea->mem_obj && varea->mem_obj->on_varea_open) varea->mem_obj->on_varea_open(varea); } /** * restore context modified by varea install * caller must NOT hold the aspace lock */ static void _varea_uninstall(rt_varea_t varea) { rt_aspace_t aspace = varea->aspace; if (varea->mem_obj && varea->mem_obj->on_varea_close) varea->mem_obj->on_varea_close(varea); rt_hw_mmu_unmap(aspace, varea->start, varea->size); rt_hw_tlb_invalidate_range(aspace, varea->start, varea->size, ARCH_PAGE_SIZE); rt_varea_pgmgr_pop_all(varea); WR_LOCK(aspace); _aspace_bst_remove(aspace, varea); WR_UNLOCK(aspace); } static int _mm_aspace_map(rt_aspace_t aspace, rt_varea_t varea, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { int err = RT_EOK; WR_LOCK(aspace); /** * @brief .prefer & .map_size are scratched from varea which setup by caller * .limit_start & .limit_range_size have default to be in range of aspace * .flags is from parameter, and will be fill in varea if install successfully */ struct rt_mm_va_hint hint = {.prefer = varea->start, .map_size = varea->size, .limit_start = aspace->start, .limit_range_size = aspace->size, .flags = flags}; if (mem_obj->hint_free) { /* mem object can control mapping range and so by modifing hint */ mem_obj->hint_free(&hint); } /* try to allocate a virtual address region for varea */ err = _varea_install(aspace, varea, &hint); WR_UNLOCK(aspace); if (err == RT_EOK) { /* fill in varea data */ _varea_post_install(varea, aspace, attr, flags, mem_obj, offset); if (MMF_TEST_CNTL(flags, MMF_PREFETCH)) { /* do the MMU & TLB business */ err = _do_prefetch(aspace, varea, varea->start, varea->size); if (err) { /* restore data structure and MMU */ _varea_uninstall(varea); } } } return err; } rt_varea_t _varea_create(void *start, rt_size_t size) { rt_varea_t varea; varea = (rt_varea_t)rt_malloc(sizeof(struct rt_varea)); if (varea) { varea->start = start; varea->size = size; } return varea; } #define _IS_OVERFLOW(start, length) ((length) > (0ul - (uintptr_t)(start))) #define _IS_OVERSIZE(start, length, limit_s, limit_sz) (((length) + (rt_size_t)((start) - (limit_start))) > (limit_size)) static inline int _not_in_range(void *start, rt_size_t length, void *limit_start, rt_size_t limit_size) { if (start != RT_NULL) LOG_D("%s: [%p : %p] [%p : %p]", __func__, start, length, limit_start, limit_size); /* assuming (base + length) will not overflow except (0) */ return start != RT_NULL ? (_IS_OVERFLOW(start, length) || start < limit_start || _IS_OVERSIZE(start, length, limit_start, limit_size)) : length > limit_size; } static inline int _not_align(void *start, rt_size_t length, rt_size_t mask) { return (start != RT_NULL) && (((uintptr_t)start & mask) || (length & mask)); } static inline int _not_support(rt_size_t flags) { rt_size_t support_ops = (MMF_PREFETCH | MMF_MAP_FIXED | MMF_TEXT | MMF_STATIC_ALLOC | MMF_REQUEST_ALIGN); return flags & ~(support_ops | _MMF_ALIGN_MASK); } int rt_aspace_map(rt_aspace_t aspace, void **addr, rt_size_t length, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { /* TODO check not in atomic context: irq, spinlock, local intr disable... */ int err; rt_varea_t varea; if (!aspace || !addr || !mem_obj || length == 0) { err = -RT_EINVAL; LOG_I("%s(%p, %p, %lx, %lx, %lx, %p, %lx): Invalid input", __func__, aspace, addr, length, attr, flags, mem_obj, offset); } else if (_not_in_range(*addr, length, aspace->start, aspace->size)) { err = -RT_EINVAL; LOG_I("%s(addr:%p, len:%lx): out of range", __func__, *addr, length); } else if (_not_support(flags)) { LOG_I("%s: no support flags 0x%lx", __func__, flags); err = -RT_ENOSYS; } else { /* allocate the varea and fill in start and size */ varea = _varea_create(*addr, length); if (varea) { err = _mm_aspace_map(aspace, varea, attr, flags, mem_obj, offset); if (err != RT_EOK) { rt_free(varea); } } else { LOG_W("%s: memory allocation failed", __func__); err = -RT_ENOMEM; } } if (err != RT_EOK) { *addr = NULL; } else { *addr = varea->start; } return err; } int rt_aspace_map_static(rt_aspace_t aspace, rt_varea_t varea, void **addr, rt_size_t length, rt_size_t attr, mm_flag_t flags, rt_mem_obj_t mem_obj, rt_size_t offset) { int err; if (!aspace || !varea || !addr || !mem_obj || length == 0 || _not_in_range(*addr, length, aspace->start, aspace->size)) { err = -RT_EINVAL; LOG_W("%s: Invalid input", __func__); } else if (_not_support(flags)) { LOG_W("%s: no support flags", __func__); err = -RT_ENOSYS; } else { varea->size = length; varea->start = *addr; flags |= MMF_STATIC_ALLOC; err = _mm_aspace_map(aspace, varea, attr, flags, mem_obj, offset); } if (err != RT_EOK) { *addr = NULL; } else { *addr = varea->start; } return err; } int _mm_aspace_map_phy(rt_aspace_t aspace, rt_varea_t varea, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; void *vaddr; if (!aspace || !hint || !hint->limit_range_size || !hint->map_size) { LOG_W("%s: Invalid input", __func__); err = -RT_EINVAL; } else if (_not_align(hint->prefer, hint->map_size, ARCH_PAGE_MASK)) { LOG_W("%s: not aligned", __func__); err = -RT_EINVAL; } else if (_not_in_range(hint->limit_start, hint->limit_range_size, aspace->start, aspace->size) || _not_in_range(hint->prefer, hint->map_size, aspace->start, aspace->size)) { LOG_W("%s: not in range", __func__); err = -RT_EINVAL; } else { WR_LOCK(aspace); err = _varea_install(aspace, varea, hint); WR_UNLOCK(aspace); if (err == RT_EOK) { _varea_post_install(varea, aspace, attr, hint->flags, NULL, pa_off); vaddr = varea->start; err = _do_named_map(aspace, varea->start, varea->size, (rt_size_t)pa_off, attr); if (err != RT_EOK) { _varea_uninstall(varea); } } } if (ret_va) { if (err == RT_EOK) *ret_va = vaddr; else *ret_va = RT_NULL; } return err; } int rt_aspace_map_phy(rt_aspace_t aspace, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; if (hint) { rt_varea_t varea = _varea_create(hint->prefer, hint->map_size); if (varea) { err = _mm_aspace_map_phy(aspace, varea, hint, attr, pa_off, ret_va); if (err != RT_EOK) { rt_free(varea); } } else { err = -RT_ENOMEM; } } else { err = -RT_EINVAL; } return err; } int rt_aspace_map_phy_static(rt_aspace_t aspace, rt_varea_t varea, rt_mm_va_hint_t hint, rt_size_t attr, rt_size_t pa_off, void **ret_va) { int err; if (varea && hint) { varea->start = hint->prefer; varea->size = hint->map_size; hint->flags |= (MMF_MAP_FIXED | MMF_STATIC_ALLOC); LOG_D("%s: start %p size %p phy at %p", __func__, varea->start, varea->size, pa_off << MM_PAGE_SHIFT); err = _mm_aspace_map_phy(aspace, varea, hint, attr, pa_off, ret_va); } else { err = -RT_EINVAL; } return err; } void _aspace_unmap(rt_aspace_t aspace, void *addr) { WR_LOCK(aspace); rt_varea_t varea = _aspace_bst_search(aspace, addr); WR_UNLOCK(aspace); if (varea == RT_NULL) { LOG_I("%s: No such entry found at %p\n", __func__, addr); } _varea_uninstall(varea); if (!(varea->flag & MMF_STATIC_ALLOC)) { rt_free(varea); } } int rt_aspace_unmap(rt_aspace_t aspace, void *addr) { if (!aspace) { LOG_I("%s: Invalid input", __func__); return -RT_EINVAL; } if (_not_in_range(addr, 1, aspace->start, aspace->size)) { LOG_I("%s: %lx not in range of aspace[%lx:%lx]", __func__, addr, aspace->start, aspace->start + aspace->size); return -RT_EINVAL; } _aspace_unmap(aspace, addr); return RT_EOK; } static inline void *_lower(void *a, void *b) { return a < b ? a : b; } static inline void *_align(void *va, rt_ubase_t align_mask) { return (void *)((rt_ubase_t)(va + ~align_mask) & align_mask); } static void *_ascending_search(rt_varea_t varea, rt_size_t req_size, rt_ubase_t align_mask, struct _mm_range limit) { void *ret = RT_NULL; while (varea && varea->start < limit.end) { void *candidate = varea->start + varea->size; candidate = _align(candidate, align_mask); if (candidate > limit.end || limit.end - candidate + 1 < req_size) break; rt_varea_t nx_va = ASPACE_VAREA_NEXT(varea); if (nx_va) { rt_size_t gap_size = _lower(limit.end, nx_va->start - 1) - candidate + 1; if (gap_size >= req_size) { ret = candidate; break; } } else { ret = candidate; } varea = nx_va; } return ret; } /** find suitable place in [limit_start, limit_end] */ static void *_find_head_and_asc_search(rt_aspace_t aspace, rt_size_t req_size, rt_ubase_t align_mask, struct _mm_range limit) { void *va = RT_NULL; rt_varea_t varea = _aspace_bst_search_exceed(aspace, limit.start); if (varea) { void *candidate = _align(limit.start, align_mask); rt_size_t gap_size = varea->start - candidate; if (gap_size >= req_size) { rt_varea_t former = _aspace_bst_search(aspace, limit.start); if (former) { candidate = _align(former->start + former->size, align_mask); gap_size = varea->start - candidate; if (gap_size >= req_size) va = candidate; else va = _ascending_search(varea, req_size, align_mask, limit); } else { va = candidate; } } else { va = _ascending_search(varea, req_size, align_mask, limit); } } else { void *candidate; rt_size_t gap_size; candidate = limit.start; candidate = _align(candidate, align_mask); gap_size = limit.end - candidate + 1; if (gap_size >= req_size) va = candidate; } return va; } static void *_find_free(rt_aspace_t aspace, void *prefer, rt_size_t req_size, void *limit_start, rt_size_t limit_size, mm_flag_t flags) { rt_varea_t varea = NULL; void *va = RT_NULL; struct _mm_range limit = {limit_start, limit_start + limit_size - 1}; rt_ubase_t align_mask = ~0ul; if (flags & MMF_REQUEST_ALIGN) { align_mask = ~((1 << MMF_GET_ALIGN(flags)) - 1); } if (prefer != RT_NULL) { /* if prefer and free, just return the prefer region */ prefer = _align(prefer, align_mask); struct _mm_range range = {prefer, prefer + req_size - 1}; varea = _aspace_bst_search_overlap(aspace, range); if (!varea) { va = prefer; } else if (flags & MMF_MAP_FIXED) { /* OVERLAP */ } else { /* search from `varea` in ascending order */ va = _ascending_search(varea, req_size, align_mask, limit); if (va == RT_NULL) { /* rewind to first range */ limit.end = varea->start - 1; va = _find_head_and_asc_search(aspace, req_size, align_mask, limit); } } } else { va = _find_head_and_asc_search(aspace, req_size, align_mask, limit); } return va; } int rt_aspace_load_page(rt_aspace_t aspace, void *addr, rt_size_t npage) { int err = RT_EOK; rt_varea_t varea; void *end = addr + (npage << ARCH_PAGE_SHIFT); WR_LOCK(aspace); varea = _aspace_bst_search(aspace, addr); WR_UNLOCK(aspace); if (!varea) { LOG_W("%s: varea not exist", __func__); err = -RT_ENOENT; } else if (addr >= end || (rt_size_t)addr & ARCH_PAGE_MASK || _not_in_range(addr, npage << ARCH_PAGE_SHIFT, varea->start, varea->size)) { LOG_W("%s: Unaligned parameter or out of range", __func__); err = -RT_EINVAL; } else { err = _do_prefetch(aspace, varea, addr, npage << ARCH_PAGE_SHIFT); } return err; } int rt_varea_map_page(rt_varea_t varea, void *vaddr, void *page) { int err = RT_EOK; void *page_pa = rt_kmem_v2p(page); if (!varea || !vaddr || !page) { LOG_W("%s(%p,%p,%p): invalid input", __func__, varea, vaddr, page); err = -RT_EINVAL; } else if (page_pa == ARCH_MAP_FAILED) { LOG_W("%s: page is not in kernel space", __func__); err = -RT_ERROR; } else if (_not_in_range(vaddr, ARCH_PAGE_SIZE, varea->start, varea->size)) { LOG_W("%s(%p,%lx): not in range of varea(%p,%lx)", __func__, vaddr, ARCH_PAGE_SIZE, varea->start, varea->size); err = -RT_EINVAL; } else { err = _do_named_map( varea->aspace, vaddr, ARCH_PAGE_SIZE, MM_PA_TO_OFF(page_pa), varea->attr ); } return err; } #define ALIGNED(addr) (!((rt_size_t)(addr) & ARCH_PAGE_MASK)) int rt_varea_map_range(rt_varea_t varea, void *vaddr, void *paddr, rt_size_t length) { int err; if (!varea || !vaddr || !paddr || !length || !ALIGNED(vaddr) || !ALIGNED(paddr) || !(ALIGNED(length))) { LOG_W("%s(%p,%p,%p,%lx): invalid input", __func__, varea, vaddr, paddr, length); err = -RT_EINVAL; } else if (_not_in_range(vaddr, length, varea->start, varea->size)) { LOG_W("%s(%p,%lx): not in range of varea(%p,%lx)", __func__, vaddr, length, varea->start, varea->size); err = -RT_EINVAL; } else { err = _do_named_map( varea->aspace, vaddr, length, MM_PA_TO_OFF(paddr), varea->attr ); } return err; } int rt_aspace_offload_page(rt_aspace_t aspace, void *addr, rt_size_t npage) { return -RT_ENOSYS; } int rt_aspace_control(rt_aspace_t aspace, void *addr, enum rt_mmu_cntl cmd) { int err; rt_varea_t varea; WR_LOCK(aspace); varea = _aspace_bst_search(aspace, addr); WR_UNLOCK(aspace); if (varea) { err = rt_hw_mmu_control(aspace, varea->start, varea->size, cmd); if (err == RT_EOK) { rt_hw_tlb_invalidate_range(aspace, varea->start, varea->size, ARCH_PAGE_SIZE); } } else { err = -RT_ENOENT; } return err; } int rt_aspace_traversal(rt_aspace_t aspace, int (*fn)(rt_varea_t varea, void *arg), void *arg) { rt_varea_t varea; WR_LOCK(aspace); varea = ASPACE_VAREA_FIRST(aspace); while (varea) { fn(varea, arg); varea = ASPACE_VAREA_NEXT(varea); } WR_UNLOCK(aspace); return 0; } static int _dump(rt_varea_t varea, void *arg) { if (varea->mem_obj && varea->mem_obj->get_name) { rt_kprintf("[%p - %p] %s\n", varea->start, varea->start + varea->size, varea->mem_obj->get_name(varea)); } else { rt_kprintf("[%p - %p] phy-map\n", varea->start, varea->start + varea->size); rt_kprintf("\t\\_ paddr = %p\n", varea->offset << MM_PAGE_SHIFT); } return 0; } void rt_aspace_print_all(rt_aspace_t aspace) { rt_aspace_traversal(aspace, _dump, NULL); }