函数逻辑报告

Linux Kernel

v5.5.9

Brick Technologies Co., Ltd

Source Code:mm\rmap.c Create Date:2022-07-27 16:26:28
Last Modify:2020-03-12 14:18:49 Copyright©Brick
首页 函数Tree
注解内核,赢得工具下载SCCTEnglish

函数名称：@arg: enum ttu_flags will be passed to this argument

函数原型：static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, unsigned long address, void *arg)

返回类型：bool

参数：

类型参数名称
struct page *page
struct vm_area_struct *vma
unsigned longaddress
void *arg
1369  mm等于The address space we belong to.
1370  struct page_vma_mapped_walk pvmw = {page = page, vma = vma, address = address, }
1377  bool ret = true
1379  flags等于arg
1382  如果flags按位与 munlock mode 且非Flags, see mm.h. 按位与VM_LOCKED的值则返回:true
1385  如果IS_ENABLED(CONFIG_FOO) evaluates to 1 if CONFIG_FOO is set to 'y' or 'm',* 0 otherwise.(CONFIG_MIGRATION)且flags按位与 migration mode 且is_zone_device_page(page)且非is_device_private_page(page)则返回:true
1389  如果flags按位与 split huge PMD if any 则
1390  split_huge_pmd_address(vma, address, flags & freeze pte under splitting thp , page)
1402  mmu_notifier_range_init( & range, MMU_NOTIFY_CLEAR, 0, vma, The address space we belong to. , address, 两数取小(The first byte after our end addresswithin vm_mm. , address + Returns the number of bytes in this potentially compound page. ))
1405  如果PageHuge(page)则
1410  adjust_range_if_pmd_sharing_possible(vma, & start, & end)
1413  mmu_notifier_invalidate_range_start( & range)
1415  当page_vma_mapped_walk - check if @pvmw->page is mapped in @pvmw->vma at*@pvmw->address*@pvmw: pointer to struct page_vma_mapped_walk. page, vma, address and flags* must be set. pmd, pte and ptl must be NULL.* Returns true if the page is mapped in the vma循环
1431  如果非flags按位与 ignore mlock 的值则
1432  如果Flags, see mm.h. 按位与VM_LOCKED则
1445  如果flags按位与 munlock mode 则继续下一循环
1450  VM_BUG_ON_PAGE(!pte, page)
1452  subpage等于page减page_to_pfn(page)加pte_pfn( * pte)
1453  address等于address
1455  如果PageHuge(page)则
1456  如果huge_pmd_unshare(mm, & address, pte)则
1483  如果IS_ENABLED(CONFIG_FOO) evaluates to 1 if CONFIG_FOO is set to 'y' or 'm',* 0 otherwise.(CONFIG_MIGRATION)且flags按位与 migration mode 且is_zone_device_page(page)则
1489  pteval等于ptep_get_and_clear(mm, address, pte)
1496  entry等于make_migration_entry(page, 0)
1497  swp_pte等于Convert the arch-independent representation of a swp_entry_t into the* arch-dependent pte representation.
1498  如果pte_soft_dirty(pteval)则swp_pte等于pte_swp_mksoft_dirty(swp_pte)
1500  set_pte_at(mm, address, pte, swp_pte)
1512  subpage等于page
1513  转到:discard
1516  如果非flags按位与 don't age 的值则
1517  如果ptep_clear_flush_young_notify(vma, address, pte)则
1519  ret = false
1520  page_vma_mapped_walk_done( & pvmw)
1521  退出
1526  flush_cache_page(vma, address, pte_pfn( * pte))
1527  如果should_defer_flush(mm, flags)则
1536  pteval等于ptep_get_and_clear(mm, address, pte)
1538  set_tlb_ubc_flush_pending(mm, The following only work if pte_present() is true.* Undefined behaviour if not..)
1539  否则
1540  pteval等于ptep_clear_flush(vma, address, pte)
1544  如果The following only work if pte_present() is true.* Undefined behaviour if not..则Dirty a page
1548  update_hiwater_rss(mm)
1550  如果PageHWPoison(page)且非flags按位与 corrupted page is recoverable 的值则
1551  pteval等于Convert the arch-independent representation of a swp_entry_t into the* arch-dependent pte representation.
1552  如果PageHuge(page)则
1557  否则
1562  否则如果Some architectures provide facilities to virtualization guests* so that they can flag allocated pages as unused. This allows the* host to transparently reclaim unused pages. This function returns* whether the pte's page is unused.且非userfaultfd_armed(vma)则
1573  dec_mm_counter(mm, mm_counter(page))
1575  mmu_notifier_invalidate_range(mm, address, address + PAGE_SIZE)
1577  否则如果IS_ENABLED(CONFIG_FOO) evaluates to 1 if CONFIG_FOO is set to 'y' or 'm',* 0 otherwise.(CONFIG_MIGRATION)且flags按位与 migration mode 按位或 freeze pte under splitting thp 的值则
1582  如果Some architectures support metadata associated with a page小于0则
1583  set_pte_at(mm, address, pte, pteval)
1584  ret = false
1585  page_vma_mapped_walk_done( & pvmw)
1586  退出
1594  entry等于make_migration_entry(subpage, pte_write(pteval))
1596  swp_pte等于Convert the arch-independent representation of a swp_entry_t into the* arch-dependent pte representation.
1597  如果pte_soft_dirty(pteval)则swp_pte等于pte_swp_mksoft_dirty(swp_pte)
1599  set_pte_at(mm, address, pte, swp_pte)
1604  否则如果PageAnon(page)则
1605  swp_entry_t entry = {val = page_private(subpage)}
1611  如果此条件成立可能性小(为编译器优化)(PageSwapBacked(page) != PageSwapCache(page))则
1622  如果非PageSwapBacked(page)则
1623  如果非PageDirty(page)则
1635  set_pte_at(mm, address, pte, pteval)
1636  SetPageSwapBacked(page)
1637  ret = false
1638  page_vma_mapped_walk_done( & pvmw)
1639  退出
1642  如果swap_duplicate(entry)小于0则
1643  set_pte_at(mm, address, pte, pteval)
1644  ret = false
1645  page_vma_mapped_walk_done( & pvmw)
1646  退出
1648  如果Some architectures support metadata associated with a page小于0则
1649  set_pte_at(mm, address, pte, pteval)
1650  ret = false
1651  page_vma_mapped_walk_done( & pvmw)
1652  退出
1654  如果链表为空则
1660  dec_mm_counter(mm, MM_ANONPAGES)
1661  inc_mm_counter(mm, MM_SWAPENTS)
1662  swp_pte等于Convert the arch-independent representation of a swp_entry_t into the* arch-dependent pte representation.
1663  如果pte_soft_dirty(pteval)则swp_pte等于pte_swp_mksoft_dirty(swp_pte)
1665  set_pte_at(mm, address, pte, swp_pte)
1667  mmu_notifier_invalidate_range(mm, address, address + PAGE_SIZE)
1669  否则
1680  dec_mm_counter(mm, Optimized variant when page is already known not to be PageAnon )
1682  discard :
1690  page_remove_rmap - take down pte mapping from a page*@page: page to remove mapping from*@compound: uncharge the page as compound or small page* The caller needs to hold the pte lock.
1691  put_page(page)
1694  mmu_notifier_invalidate_range_end( & range)
1696  返回:ret