]> www.pilppa.org Git - linux-2.6-omap-h63xx.git/blobdiff - mm/memory.c
Merge commit 'v2.6.28-rc2' into core/locking
[linux-2.6-omap-h63xx.git] / mm / memory.c
index b8fdf4e5e65b52a812413ffc4797edf4e056f42c..fc031d68327e5fad33130b15d50a020b7b9b31a8 100644 (file)
@@ -1129,12 +1129,17 @@ static inline int use_zero_page(struct vm_area_struct *vma)
        return !vma->vm_ops || !vma->vm_ops->fault;
 }
 
-int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
-               unsigned long start, int len, int write, int force,
+
+
+int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
+                    unsigned long start, int len, int flags,
                struct page **pages, struct vm_area_struct **vmas)
 {
        int i;
-       unsigned int vm_flags;
+       unsigned int vm_flags = 0;
+       int write = !!(flags & GUP_FLAGS_WRITE);
+       int force = !!(flags & GUP_FLAGS_FORCE);
+       int ignore = !!(flags & GUP_FLAGS_IGNORE_VMA_PERMISSIONS);
 
        if (len <= 0)
                return 0;
@@ -1158,7 +1163,9 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
                        pud_t *pud;
                        pmd_t *pmd;
                        pte_t *pte;
-                       if (write) /* user gate pages are read-only */
+
+                       /* user gate pages are read-only */
+                       if (!ignore && write)
                                return i ? : -EFAULT;
                        if (pg > TASK_SIZE)
                                pgd = pgd_offset_k(pg);
@@ -1190,8 +1197,9 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
                        continue;
                }
 
-               if (!vma || (vma->vm_flags & (VM_IO | VM_PFNMAP))
-                               || !(vm_flags & vma->vm_flags))
+               if (!vma ||
+                   (vma->vm_flags & (VM_IO | VM_PFNMAP)) ||
+                   (!ignore && !(vm_flags & vma->vm_flags)))
                        return i ? : -EFAULT;
 
                if (is_vm_hugetlb_page(vma)) {
@@ -1266,6 +1274,23 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
        } while (len);
        return i;
 }
+
+int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
+               unsigned long start, int len, int write, int force,
+               struct page **pages, struct vm_area_struct **vmas)
+{
+       int flags = 0;
+
+       if (write)
+               flags |= GUP_FLAGS_WRITE;
+       if (force)
+               flags |= GUP_FLAGS_FORCE;
+
+       return __get_user_pages(tsk, mm,
+                               start, len, flags,
+                               pages, vmas);
+}
+
 EXPORT_SYMBOL(get_user_pages);
 
 pte_t *get_locked_pte(struct mm_struct *mm, unsigned long addr,
@@ -1296,18 +1321,14 @@ static int insert_page(struct vm_area_struct *vma, unsigned long addr,
        pte_t *pte;
        spinlock_t *ptl;
 
-       retval = mem_cgroup_charge(page, mm, GFP_KERNEL);
-       if (retval)
-               goto out;
-
        retval = -EINVAL;
        if (PageAnon(page))
-               goto out_uncharge;
+               goto out;
        retval = -ENOMEM;
        flush_dcache_page(page);
        pte = get_locked_pte(mm, addr, &ptl);
        if (!pte)
-               goto out_uncharge;
+               goto out;
        retval = -EBUSY;
        if (!pte_none(*pte))
                goto out_unlock;
@@ -1323,8 +1344,6 @@ static int insert_page(struct vm_area_struct *vma, unsigned long addr,
        return retval;
 out_unlock:
        pte_unmap_unlock(pte, ptl);
-out_uncharge:
-       mem_cgroup_uncharge_page(page);
 out:
        return retval;
 }
@@ -1858,6 +1877,15 @@ gotten:
        new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address);
        if (!new_page)
                goto oom;
+       /*
+        * Don't let another task, with possibly unlocked vma,
+        * keep the mlocked page.
+        */
+       if (vma->vm_flags & VM_LOCKED) {
+               lock_page(old_page);    /* for LRU manipulation */
+               clear_page_mlock(old_page);
+               unlock_page(old_page);
+       }
        cow_user_page(new_page, old_page, address, vma);
        __SetPageUptodate(new_page);
 
@@ -1886,11 +1914,13 @@ gotten:
                 * thread doing COW.
                 */
                ptep_clear_flush_notify(vma, address, page_table);
-               set_pte_at(mm, address, page_table, entry);
-               update_mmu_cache(vma, address, entry);
-               lru_cache_add_active(new_page);
+               SetPageSwapBacked(new_page);
+               lru_cache_add_active_or_unevictable(new_page, vma);
                page_add_new_anon_rmap(new_page, vma, address);
 
+//TODO:  is this safe?  do_anonymous_page() does it this way.
+               set_pte_at(mm, address, page_table, entry);
+               update_mmu_cache(vma, address, entry);
                if (old_page) {
                        /*
                         * Only after switching the pte to the new page may
@@ -2288,16 +2318,17 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
                count_vm_event(PGMAJFAULT);
        }
 
+       mark_page_accessed(page);
+
+       lock_page(page);
+       delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
+
        if (mem_cgroup_charge(page, mm, GFP_KERNEL)) {
-               delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
                ret = VM_FAULT_OOM;
+               unlock_page(page);
                goto out;
        }
 
-       mark_page_accessed(page);
-       lock_page(page);
-       delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
-
        /*
         * Back out if somebody else already faulted in this pte.
         */
@@ -2324,7 +2355,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
        page_add_anon_rmap(page, vma, address);
 
        swap_free(entry);
-       if (vm_swap_full())
+       if (vm_swap_full() || (vma->vm_flags & VM_LOCKED) || PageMlocked(page))
                remove_exclusive_swap_page(page);
        unlock_page(page);
 
@@ -2382,7 +2413,8 @@ static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma,
        if (!pte_none(*page_table))
                goto release;
        inc_mm_counter(mm, anon_rss);
-       lru_cache_add_active(page);
+       SetPageSwapBacked(page);
+       lru_cache_add_active_or_unevictable(page, vma);
        page_add_new_anon_rmap(page, vma, address);
        set_pte_at(mm, address, page_table, entry);
 
@@ -2423,6 +2455,7 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
        struct page *page;
        pte_t entry;
        int anon = 0;
+       int charged = 0;
        struct page *dirty_page = NULL;
        struct vm_fault vmf;
        int ret;
@@ -2463,6 +2496,18 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
                                ret = VM_FAULT_OOM;
                                goto out;
                        }
+                       if (mem_cgroup_charge(page, mm, GFP_KERNEL)) {
+                               ret = VM_FAULT_OOM;
+                               page_cache_release(page);
+                               goto out;
+                       }
+                       charged = 1;
+                       /*
+                        * Don't let another task, with possibly unlocked vma,
+                        * keep the mlocked page.
+                        */
+                       if (vma->vm_flags & VM_LOCKED)
+                               clear_page_mlock(vmf.page);
                        copy_user_highpage(page, vmf.page, address, vma);
                        __SetPageUptodate(page);
                } else {
@@ -2497,11 +2542,6 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
 
        }
 
-       if (mem_cgroup_charge(page, mm, GFP_KERNEL)) {
-               ret = VM_FAULT_OOM;
-               goto out;
-       }
-
        page_table = pte_offset_map_lock(mm, pmd, address, &ptl);
 
        /*
@@ -2520,11 +2560,11 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
                entry = mk_pte(page, vma->vm_page_prot);
                if (flags & FAULT_FLAG_WRITE)
                        entry = maybe_mkwrite(pte_mkdirty(entry), vma);
-               set_pte_at(mm, address, page_table, entry);
                if (anon) {
-                        inc_mm_counter(mm, anon_rss);
-                        lru_cache_add_active(page);
-                        page_add_new_anon_rmap(page, vma, address);
+                       inc_mm_counter(mm, anon_rss);
+                       SetPageSwapBacked(page);
+                       lru_cache_add_active_or_unevictable(page, vma);
+                       page_add_new_anon_rmap(page, vma, address);
                } else {
                        inc_mm_counter(mm, file_rss);
                        page_add_file_rmap(page);
@@ -2533,11 +2573,14 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
                                get_page(dirty_page);
                        }
                }
+//TODO:  is this safe?  do_anonymous_page() does it this way.
+               set_pte_at(mm, address, page_table, entry);
 
                /* no need to invalidate: a not-present page won't be cached */
                update_mmu_cache(vma, address, entry);
        } else {
-               mem_cgroup_uncharge_page(page);
+               if (charged)
+                       mem_cgroup_uncharge_page(page);
                if (anon)
                        page_cache_release(page);
                else
@@ -2772,19 +2815,9 @@ int make_pages_present(unsigned long addr, unsigned long end)
        len = DIV_ROUND_UP(end, PAGE_SIZE) - addr/PAGE_SIZE;
        ret = get_user_pages(current, current->mm, addr,
                        len, write, 0, NULL, NULL);
-       if (ret < 0) {
-               /*
-                  SUS require strange return value to mlock
-                   - invalid addr generate to ENOMEM.
-                   - out of memory should generate EAGAIN.
-               */
-               if (ret == -EFAULT)
-                       ret = -ENOMEM;
-               else if (ret == -ENOMEM)
-                       ret = -EAGAIN;
+       if (ret < 0)
                return ret;
-       }
-       return ret == len ? 0 : -ENOMEM;
+       return ret == len ? 0 : -EFAULT;
 }
 
 #if !defined(__HAVE_ARCH_GATE_AREA)