*  go into icache. We cache the reference to task_struct upon lookup too.
  *  Eventually it should become a filesystem in its own. We don't use the
  *  rest of procfs anymore.
+ *
+ *
+ *  Changelog:
+ *  17-Jan-2005
+ *  Allan Bezerra
+ *  Bruna Moreira <bruna.moreira@indt.org.br>
+ *  Edjard Mota <edjard.mota@indt.org.br>
+ *  Ilias Biris <ilias.biris@indt.org.br>
+ *  Mauricio Lin <mauricio.lin@indt.org.br>
+ *
+ *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
+ *
+ *  A new process specific entry (smaps) included in /proc. It shows the
+ *  size of rss for each memory area. The maps entry lacks information
+ *  about physical memory size (rss) for each mapped file, i.e.,
+ *  rss information for executables and library files.
+ *  This additional information is useful for any tools that need to know
+ *  about physical memory consumption for a process specific library.
+ *
+ *  Changelog:
+ *  21-Feb-2005
+ *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
+ *  Pud inclusion in the page table walking.
+ *
+ *  ChangeLog:
+ *  10-Mar-2005
+ *  10LE Instituto Nokia de Tecnologia - INdT:
+ *  A better way to walks through the page table as suggested by Hugh Dickins.
+ *
+ *  Simo Piiroinen <simo.piiroinen@nokia.com>:
+ *  Smaps information related to shared, private, clean and dirty pages.
+ *
+ *  Paul Mundt <paul.mundt@nokia.com>:
+ *  Overall revision about smaps.
  */
 
 #include <asm/uaccess.h>
        PROC_TGID_NUMA_MAPS,
        PROC_TGID_MOUNTS,
        PROC_TGID_WCHAN,
+       PROC_TGID_SMAPS,
 #ifdef CONFIG_SCHEDSTATS
        PROC_TGID_SCHEDSTAT,
 #endif
        PROC_TID_NUMA_MAPS,
        PROC_TID_MOUNTS,
        PROC_TID_WCHAN,
+       PROC_TID_SMAPS,
 #ifdef CONFIG_SCHEDSTATS
        PROC_TID_SCHEDSTAT,
 #endif
        E(PROC_TGID_ROOT,      "root",    S_IFLNK|S_IRWXUGO),
        E(PROC_TGID_EXE,       "exe",     S_IFLNK|S_IRWXUGO),
        E(PROC_TGID_MOUNTS,    "mounts",  S_IFREG|S_IRUGO),
+       E(PROC_TGID_SMAPS,     "smaps",   S_IFREG|S_IRUGO),
 #ifdef CONFIG_SECURITY
        E(PROC_TGID_ATTR,      "attr",    S_IFDIR|S_IRUGO|S_IXUGO),
 #endif
        E(PROC_TID_ROOT,       "root",    S_IFLNK|S_IRWXUGO),
        E(PROC_TID_EXE,        "exe",     S_IFLNK|S_IRWXUGO),
        E(PROC_TID_MOUNTS,     "mounts",  S_IFREG|S_IRUGO),
+       E(PROC_TID_SMAPS,      "smaps",   S_IFREG|S_IRUGO),
 #ifdef CONFIG_SECURITY
        E(PROC_TID_ATTR,       "attr",    S_IFDIR|S_IRUGO|S_IXUGO),
 #endif
 };
 #endif
 
+extern struct seq_operations proc_pid_smaps_op;
+static int smaps_open(struct inode *inode, struct file *file)
+{
+       struct task_struct *task = proc_task(inode);
+       int ret = seq_open(file, &proc_pid_smaps_op);
+       if (!ret) {
+               struct seq_file *m = file->private_data;
+               m->private = task;
+       }
+       return ret;
+}
+
+static struct file_operations proc_smaps_operations = {
+       .open           = smaps_open,
+       .read           = seq_read,
+       .llseek         = seq_lseek,
+       .release        = seq_release,
+};
+
 extern struct seq_operations mounts_op;
 static int mounts_open(struct inode *inode, struct file *file)
 {
                case PROC_TGID_MOUNTS:
                        inode->i_fop = &proc_mounts_operations;
                        break;
+               case PROC_TID_SMAPS:
+               case PROC_TGID_SMAPS:
+                       inode->i_fop = &proc_smaps_operations;
+                       break;
 #ifdef CONFIG_SECURITY
                case PROC_TID_ATTR:
                        inode->i_nlink = 2;
 
 #include <linux/hugetlb.h>
 #include <linux/mount.h>
 #include <linux/seq_file.h>
+#include <linux/highmem.h>
 #include <linux/pagemap.h>
 #include <linux/mempolicy.h>
+
 #include <asm/elf.h>
 #include <asm/uaccess.h>
+#include <asm/tlbflush.h>
 #include "internal.h"
 
 char *task_mem(struct mm_struct *mm, char *buffer)
        seq_printf(m, "%*c", len, ' ');
 }
 
-static int show_map(struct seq_file *m, void *v)
+struct mem_size_stats
+{
+       unsigned long resident;
+       unsigned long shared_clean;
+       unsigned long shared_dirty;
+       unsigned long private_clean;
+       unsigned long private_dirty;
+};
+
+static int show_map_internal(struct seq_file *m, void *v, struct mem_size_stats *mss)
 {
        struct task_struct *task = m->private;
-       struct vm_area_struct *map = v;
-       struct mm_struct *mm = map->vm_mm;
-       struct file *file = map->vm_file;
-       int flags = map->vm_flags;
+       struct vm_area_struct *vma = v;
+       struct mm_struct *mm = vma->vm_mm;
+       struct file *file = vma->vm_file;
+       int flags = vma->vm_flags;
        unsigned long ino = 0;
        dev_t dev = 0;
        int len;
 
        if (file) {
-               struct inode *inode = map->vm_file->f_dentry->d_inode;
+               struct inode *inode = vma->vm_file->f_dentry->d_inode;
                dev = inode->i_sb->s_dev;
                ino = inode->i_ino;
        }
 
        seq_printf(m, "%08lx-%08lx %c%c%c%c %08lx %02x:%02x %lu %n",
-                       map->vm_start,
-                       map->vm_end,
+                       vma->vm_start,
+                       vma->vm_end,
                        flags & VM_READ ? 'r' : '-',
                        flags & VM_WRITE ? 'w' : '-',
                        flags & VM_EXEC ? 'x' : '-',
                        flags & VM_MAYSHARE ? 's' : 'p',
-                       map->vm_pgoff << PAGE_SHIFT,
+                       vma->vm_pgoff << PAGE_SHIFT,
                        MAJOR(dev), MINOR(dev), ino, &len);
 
        /*
         * Print the dentry name for named mappings, and a
         * special [heap] marker for the heap:
         */
-       if (map->vm_file) {
+       if (file) {
                pad_len_spaces(m, len);
-               seq_path(m, file->f_vfsmnt, file->f_dentry, "");
+               seq_path(m, file->f_vfsmnt, file->f_dentry, "\n");
        } else {
                if (mm) {
-                       if (map->vm_start <= mm->start_brk &&
-                                               map->vm_end >= mm->brk) {
+                       if (vma->vm_start <= mm->start_brk &&
+                                               vma->vm_end >= mm->brk) {
                                pad_len_spaces(m, len);
                                seq_puts(m, "[heap]");
                        } else {
-                               if (map->vm_start <= mm->start_stack &&
-                                       map->vm_end >= mm->start_stack) {
+                               if (vma->vm_start <= mm->start_stack &&
+                                       vma->vm_end >= mm->start_stack) {
 
                                        pad_len_spaces(m, len);
                                        seq_puts(m, "[stack]");
                }
        }
        seq_putc(m, '\n');
-       if (m->count < m->size)  /* map is copied successfully */
-               m->version = (map != get_gate_vma(task))? map->vm_start: 0;
+
+       if (mss)
+               seq_printf(m,
+                          "Size:          %8lu kB\n"
+                          "Rss:           %8lu kB\n"
+                          "Shared_Clean:  %8lu kB\n"
+                          "Shared_Dirty:  %8lu kB\n"
+                          "Private_Clean: %8lu kB\n"
+                          "Private_Dirty: %8lu kB\n",
+                          (vma->vm_end - vma->vm_start) >> 10,
+                          mss->resident >> 10,
+                          mss->shared_clean  >> 10,
+                          mss->shared_dirty  >> 10,
+                          mss->private_clean >> 10,
+                          mss->private_dirty >> 10);
+
+       if (m->count < m->size)  /* vma is copied successfully */
+               m->version = (vma != get_gate_vma(task))? vma->vm_start: 0;
        return 0;
 }
 
+static int show_map(struct seq_file *m, void *v)
+{
+       return show_map_internal(m, v, 0);
+}
+
+static void smaps_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
+                               unsigned long addr, unsigned long end,
+                               struct mem_size_stats *mss)
+{
+       pte_t *pte, ptent;
+       unsigned long pfn;
+       struct page *page;
+
+       pte = pte_offset_map(pmd, addr);
+       do {
+               ptent = *pte;
+               if (pte_none(ptent) || !pte_present(ptent))
+                       continue;
+
+               mss->resident += PAGE_SIZE;
+               pfn = pte_pfn(ptent);
+               if (!pfn_valid(pfn))
+                       continue;
+
+               page = pfn_to_page(pfn);
+               if (page_count(page) >= 2) {
+                       if (pte_dirty(ptent))
+                               mss->shared_dirty += PAGE_SIZE;
+                       else
+                               mss->shared_clean += PAGE_SIZE;
+               } else {
+                       if (pte_dirty(ptent))
+                               mss->private_dirty += PAGE_SIZE;
+                       else
+                               mss->private_clean += PAGE_SIZE;
+               }
+       } while (pte++, addr += PAGE_SIZE, addr != end);
+       pte_unmap(pte - 1);
+       cond_resched_lock(&vma->vm_mm->page_table_lock);
+}
+
+static inline void smaps_pmd_range(struct vm_area_struct *vma, pud_t *pud,
+                               unsigned long addr, unsigned long end,
+                               struct mem_size_stats *mss)
+{
+       pmd_t *pmd;
+       unsigned long next;
+
+       pmd = pmd_offset(pud, addr);
+       do {
+               next = pmd_addr_end(addr, end);
+               if (pmd_none_or_clear_bad(pmd))
+                       continue;
+               smaps_pte_range(vma, pmd, addr, next, mss);
+       } while (pmd++, addr = next, addr != end);
+}
+
+static inline void smaps_pud_range(struct vm_area_struct *vma, pgd_t *pgd,
+                               unsigned long addr, unsigned long end,
+                               struct mem_size_stats *mss)
+{
+       pud_t *pud;
+       unsigned long next;
+
+       pud = pud_offset(pgd, addr);
+       do {
+               next = pud_addr_end(addr, end);
+               if (pud_none_or_clear_bad(pud))
+                       continue;
+               smaps_pmd_range(vma, pud, addr, next, mss);
+       } while (pud++, addr = next, addr != end);
+}
+
+static inline void smaps_pgd_range(struct vm_area_struct *vma,
+                               unsigned long addr, unsigned long end,
+                               struct mem_size_stats *mss)
+{
+       pgd_t *pgd;
+       unsigned long next;
+
+       pgd = pgd_offset(vma->vm_mm, addr);
+       do {
+               next = pgd_addr_end(addr, end);
+               if (pgd_none_or_clear_bad(pgd))
+                       continue;
+               smaps_pud_range(vma, pgd, addr, next, mss);
+       } while (pgd++, addr = next, addr != end);
+}
+
+static int show_smap(struct seq_file *m, void *v)
+{
+       struct vm_area_struct *vma = v;
+       struct mm_struct *mm = vma->vm_mm;
+       struct mem_size_stats mss;
+
+       memset(&mss, 0, sizeof mss);
+
+       if (mm) {
+               spin_lock(&mm->page_table_lock);
+               smaps_pgd_range(vma, vma->vm_start, vma->vm_end, &mss);
+               spin_unlock(&mm->page_table_lock);
+       }
+
+       return show_map_internal(m, v, &mss);
+}
+
 static void *m_start(struct seq_file *m, loff_t *pos)
 {
        struct task_struct *task = m->private;
        unsigned long last_addr = m->version;
        struct mm_struct *mm;
-       struct vm_area_struct *map, *tail_map;
+       struct vm_area_struct *vma, *tail_vma;
        loff_t l = *pos;
 
        /*
         * We remember last_addr rather than next_addr to hit with
         * mmap_cache most of the time. We have zero last_addr at
-        * the begining and also after lseek. We will have -1 last_addr
-        * after the end of the maps.
+        * the beginning and also after lseek. We will have -1 last_addr
+        * after the end of the vmas.
         */
 
        if (last_addr == -1UL)
        if (!mm)
                return NULL;
 
-       tail_map = get_gate_vma(task);
+       tail_vma = get_gate_vma(task);
        down_read(&mm->mmap_sem);
 
        /* Start with last addr hint */
-       if (last_addr && (map = find_vma(mm, last_addr))) {
-               map = map->vm_next;
+       if (last_addr && (vma = find_vma(mm, last_addr))) {
+               vma = vma->vm_next;
                goto out;
        }
 
        /*
-        * Check the map index is within the range and do
+        * Check the vma index is within the range and do
         * sequential scan until m_index.
         */
-       map = NULL;
+       vma = NULL;
        if ((unsigned long)l < mm->map_count) {
-               map = mm->mmap;
-               while (l-- && map)
-                       map = map->vm_next;
+               vma = mm->mmap;
+               while (l-- && vma)
+                       vma = vma->vm_next;
                goto out;
        }
 
        if (l != mm->map_count)
-               tail_map = NULL; /* After gate map */
+               tail_vma = NULL; /* After gate vma */
 
 out:
-       if (map)
-               return map;
+       if (vma)
+               return vma;
 
-       /* End of maps has reached */
-       m->version = (tail_map != NULL)? 0: -1UL;
+       /* End of vmas has been reached */
+       m->version = (tail_vma != NULL)? 0: -1UL;
        up_read(&mm->mmap_sem);
        mmput(mm);
-       return tail_map;
+       return tail_vma;
 }
 
 static void m_stop(struct seq_file *m, void *v)
 {
        struct task_struct *task = m->private;
-       struct vm_area_struct *map = v;
-       if (map && map != get_gate_vma(task)) {
-               struct mm_struct *mm = map->vm_mm;
+       struct vm_area_struct *vma = v;
+       if (vma && vma != get_gate_vma(task)) {
+               struct mm_struct *mm = vma->vm_mm;
                up_read(&mm->mmap_sem);
                mmput(mm);
        }
 static void *m_next(struct seq_file *m, void *v, loff_t *pos)
 {
        struct task_struct *task = m->private;
-       struct vm_area_struct *map = v;
-       struct vm_area_struct *tail_map = get_gate_vma(task);
+       struct vm_area_struct *vma = v;
+       struct vm_area_struct *tail_vma = get_gate_vma(task);
 
        (*pos)++;
-       if (map && (map != tail_map) && map->vm_next)
-               return map->vm_next;
+       if (vma && (vma != tail_vma) && vma->vm_next)
+               return vma->vm_next;
        m_stop(m, v);
-       return (map != tail_map)? tail_map: NULL;
+       return (vma != tail_vma)? tail_vma: NULL;
 }
 
 struct seq_operations proc_pid_maps_op = {
        .show   = show_map
 };
 
+struct seq_operations proc_pid_smaps_op = {
+       .start  = m_start,
+       .next   = m_next,
+       .stop   = m_stop,
+       .show   = show_smap
+};
+
 #ifdef CONFIG_NUMA
 
 struct numa_maps {