vserver 1.9.3
[linux-2.6.git] / fs / proc / base.c
index e1c3a5d..434ecb4 100644 (file)
@@ -33,7 +33,6 @@
 #include <linux/security.h>
 #include <linux/ptrace.h>
 #include <linux/vs_network.h>
-#include <linux/vs_cvirt.h>
 
 /*
  * For hysterical raisins we keep the same inumbers as in the old procfs.
@@ -62,6 +61,9 @@ enum pid_directory_inos {
        PROC_TGID_MAPS,
        PROC_TGID_MOUNTS,
        PROC_TGID_WCHAN,
+#ifdef CONFIG_SCHEDSTATS
+       PROC_TGID_SCHEDSTAT,
+#endif
 #ifdef CONFIG_SECURITY
        PROC_TGID_ATTR,
        PROC_TGID_ATTR_CURRENT,
@@ -87,6 +89,9 @@ enum pid_directory_inos {
        PROC_TID_MAPS,
        PROC_TID_MOUNTS,
        PROC_TID_WCHAN,
+#ifdef CONFIG_SCHEDSTATS
+       PROC_TID_SCHEDSTAT,
+#endif
 #ifdef CONFIG_SECURITY
        PROC_TID_ATTR,
        PROC_TID_ATTR_CURRENT,
@@ -128,6 +133,9 @@ static struct pid_entry tgid_base_stuff[] = {
 #endif
 #ifdef CONFIG_KALLSYMS
        E(PROC_TGID_WCHAN,     "wchan",   S_IFREG|S_IRUGO),
+#endif
+#ifdef CONFIG_SCHEDSTATS
+       E(PROC_TGID_SCHEDSTAT, "schedstat", S_IFREG|S_IRUGO),
 #endif
        E(PROC_TGID_VX_INFO,   "vinfo",   S_IFREG|S_IRUGO),
        E(PROC_TGID_IP_INFO,   "ninfo",   S_IFREG|S_IRUGO),
@@ -152,6 +160,9 @@ static struct pid_entry tid_base_stuff[] = {
 #endif
 #ifdef CONFIG_KALLSYMS
        E(PROC_TID_WCHAN,      "wchan",   S_IFREG|S_IRUGO),
+#endif
+#ifdef CONFIG_SCHEDSTATS
+       E(PROC_TID_SCHEDSTAT, "schedstat",S_IFREG|S_IRUGO),
 #endif
        E(PROC_TID_VX_INFO,    "vinfo",   S_IFREG|S_IRUGO),
        E(PROC_TID_IP_INFO,    "ninfo",   S_IFREG|S_IRUGO),
@@ -190,7 +201,6 @@ static inline int proc_type(struct inode *inode)
 int proc_pid_stat(struct task_struct*,char*);
 int proc_pid_status(struct task_struct*,char*);
 int proc_pid_statm(struct task_struct*,char*);
-int proc_pid_cpu(struct task_struct*,char*);
 
 static int proc_fd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt)
 {
@@ -286,7 +296,8 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf
 #define MAY_PTRACE(task) \
        (task == current || \
        (task->parent == current && \
-       (task->ptrace & PT_PTRACED) &&  task->state == TASK_STOPPED && \
+       (task->ptrace & PT_PTRACED) && \
+        (task->state == TASK_STOPPED || task->state == TASK_TRACED) && \
         security_ptrace(current,task) == 0))
 
 static int may_ptrace_attach(struct task_struct *task)
@@ -339,6 +350,8 @@ static int proc_pid_cmdline(struct task_struct *task, char * buffer)
        struct mm_struct *mm = get_task_mm(task);
        if (!mm)
                goto out;
+       if (!mm->arg_end)
+               goto out;       /* Shh! No looking before we're done */
 
        len = mm->arg_end - mm->arg_start;
  
@@ -349,7 +362,7 @@ static int proc_pid_cmdline(struct task_struct *task, char * buffer)
 
        // If the nul at the end of args has been overwritten, then
        // assume application is using setproctitle(3).
-       if (res > 0 && buffer[res-1] != '\0') {
+       if (res > 0 && buffer[res-1] != '\0' && len < PAGE_SIZE) {
                len = strnlen(buffer, res);
                if (len < res) {
                    res = len;
@@ -407,6 +420,19 @@ static int proc_pid_wchan(struct task_struct *task, char *buffer)
 }
 #endif /* CONFIG_KALLSYMS */
 
+#ifdef CONFIG_SCHEDSTATS
+/*
+ * Provides /proc/PID/schedstat
+ */
+static int proc_pid_schedstat(struct task_struct *task, char *buffer)
+{
+       return sprintf(buffer, "%lu %lu %lu\n",
+                       task->sched_info.cpu_time,
+                       task->sched_info.run_delay,
+                       task->sched_info.pcnt);
+}
+#endif
+
 /************************************************************************/
 /*                       Here the fs part begins                        */
 /************************************************************************/
@@ -527,7 +553,6 @@ static ssize_t proc_info_read(struct file * file, char __user * buf,
        struct inode * inode = file->f_dentry->d_inode;
        unsigned long page;
        ssize_t length;
-       ssize_t end;
        struct task_struct *task = proc_task(inode);
 
        if (count > PROC_BLOCK_SIZE)
@@ -537,24 +562,10 @@ static ssize_t proc_info_read(struct file * file, char __user * buf,
 
        length = PROC_I(inode)->op.proc_read(task, (char*)page);
 
-       if (length < 0) {
-               free_page(page);
-               return length;
-       }
-       /* Static 4kB (or whatever) block capacity */
-       if (*ppos >= length) {
-               free_page(page);
-               return 0;
-       }
-       if (count + *ppos > length)
-               count = length - *ppos;
-       end = count + *ppos;
-       if (copy_to_user(buf, (char *) page + *ppos, count))
-               count = -EFAULT;
-       else
-               *ppos = end;
+       if (length >= 0)
+               length = simple_read_from_buffer(buf, count, ppos, (char *)page, length);
        free_page(page);
-       return count;
+       return length;
 }
 
 static struct file_operations proc_info_file_operations = {
@@ -777,10 +788,9 @@ static struct inode_operations proc_pid_link_inode_operations = {
        .follow_link    = proc_pid_follow_link
 };
 
-static int pid_alive(struct task_struct *p)
+static inline int pid_alive(struct task_struct *p)
 {
-       BUG_ON(p->pids[PIDTYPE_PID].pidptr != &p->pids[PIDTYPE_PID].pid);
-       return atomic_read(&p->pids[PIDTYPE_PID].pid.count);
+       return p->pids[PIDTYPE_PID].nr != 0;
 }
 
 #define NUMBUF 10
@@ -1185,7 +1195,6 @@ static ssize_t proc_pid_attr_read(struct file * file, char __user * buf,
        struct inode * inode = file->f_dentry->d_inode;
        unsigned long page;
        ssize_t length;
-       ssize_t end;
        struct task_struct *task = proc_task(inode);
 
        if (count > PAGE_SIZE)
@@ -1196,24 +1205,10 @@ static ssize_t proc_pid_attr_read(struct file * file, char __user * buf,
        length = security_getprocattr(task, 
                                      (char*)file->f_dentry->d_name.name, 
                                      (void*)page, count);
-       if (length < 0) {
-               free_page(page);
-               return length;
-       }
-       /* Static 4kB (or whatever) block capacity */
-       if (*ppos >= length) {
-               free_page(page);
-               return 0;
-       }
-       if (count + *ppos > length)
-               count = length - *ppos;
-       end = count + *ppos;
-       if (copy_to_user(buf, (char *) page + *ppos, count))
-               count = -EFAULT;
-       else
-               *ppos = end;
+       if (length >= 0)
+               length = simple_read_from_buffer(buf, count, ppos, (char *)page, length);
        free_page(page);
-       return count;
+       return length;
 }
 
 static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf,
@@ -1391,6 +1386,13 @@ static struct dentry *proc_pident_lookup(struct inode *dir,
                        inode->i_fop = &proc_info_file_operations;
                        ei->op.proc_read = proc_pid_wchan;
                        break;
+#endif
+#ifdef CONFIG_SCHEDSTATS
+               case PROC_TID_SCHEDSTAT:
+               case PROC_TGID_SCHEDSTAT:
+                       inode->i_fop = &proc_info_file_operations;
+                       ei->op.proc_read = proc_pid_schedstat;
+                       break;
 #endif
                case PROC_TID_VX_INFO:
                case PROC_TGID_VX_INFO:
@@ -1494,14 +1496,14 @@ static int proc_self_readlink(struct dentry *dentry, char __user *buffer,
                              int buflen)
 {
        char tmp[30];
-       sprintf(tmp, "%d", current->tgid);
+       sprintf(tmp, "%d", vx_map_pid(current->tgid));
        return vfs_readlink(dentry,buffer,buflen,tmp);
 }
 
 static int proc_self_follow_link(struct dentry *dentry, struct nameidata *nd)
 {
        char tmp[30];
-       sprintf(tmp, "%d", current->tgid);
+       sprintf(tmp, "%d", vx_map_pid(current->tgid));
        return vfs_follow_link(nd,tmp);
 }      
 
@@ -1553,6 +1555,7 @@ struct dentry *proc_pid_unhash(struct task_struct *p)
        
 void proc_pid_flush(struct dentry *proc_dentry)
 {
+       might_sleep();
        if(proc_dentry != NULL) {
                shrink_dcache_parent(proc_dentry);
                dput(proc_dentry);
@@ -1583,7 +1586,7 @@ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct
                d_add(dentry, inode);
                return NULL;
        }
-       tgid = vx_rmap_tgid(current->vx_info, name_to_int(dentry));
+       tgid = name_to_int(dentry);
        if (tgid == ~0U)
                goto out;
 
@@ -1595,14 +1598,13 @@ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct
        if (!task)
                goto out;
 
-       inode = NULL;
-       if (vx_check(vx_task_xid(task), VX_WATCH|VX_IDENT))
-               inode = proc_pid_make_inode(dir->i_sb, task, PROC_TGID_INO);
+       if (!vx_check(vx_task_xid(task), VX_WATCH|VX_IDENT))
+               goto out_drop_task;
+
+       inode = proc_pid_make_inode(dir->i_sb, task, PROC_TGID_INO);
+       if (!inode)
+               goto out_drop_task;
 
-       if (!inode) {
-               put_task_struct(task);
-               goto out;
-       }
        inode->i_mode = S_IFDIR|S_IRUGO|S_IXUGO;
        inode->i_op = &proc_tgid_base_inode_operations;
        inode->i_fop = &proc_tgid_base_operations;
@@ -1627,6 +1629,8 @@ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct
                goto out;
        }
        return NULL;
+out_drop_task:
+       put_task_struct(task);
 out:
        return ERR_PTR(-ENOENT);
 }
@@ -1639,11 +1643,11 @@ static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry
        struct inode *inode;
        unsigned tid;
 
-       tid = vx_rmap_tgid(current->vx_info, name_to_int(dentry));
+       tid = name_to_int(dentry);
        if (tid == ~0U)
                goto out;
-
-/*     handle fakeinit */
+       if (vx_current_initpid(tid))
+               goto out;
 
        read_lock(&tasklist_lock);
        task = find_task_by_pid(tid);
@@ -1655,12 +1659,13 @@ static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry
        if (leader->tgid != task->tgid)
                goto out_drop_task;
 
-       inode = NULL;
-       if (vx_check(vx_task_xid(task), VX_WATCH|VX_IDENT))
-               inode = proc_pid_make_inode(dir->i_sb, task, PROC_TID_INO);
+       if (!vx_check(vx_task_xid(task), VX_WATCH|VX_IDENT))
+               goto out_drop_task;
 
+       inode = proc_pid_make_inode(dir->i_sb, task, PROC_TID_INO);
        if (!inode)
                goto out_drop_task;
+
        inode->i_mode = S_IFDIR|S_IRUGO|S_IXUGO;
        inode->i_op = &proc_tid_base_inode_operations;
        inode->i_fop = &proc_tid_base_operations;
@@ -1696,8 +1701,8 @@ static int get_tgid_list(int index, unsigned long version, unsigned int *tgids)
        read_lock(&tasklist_lock);
        p = NULL;
        if (version) {
-               p = find_task_by_pid(version);
-               if (!thread_group_leader(p))
+               p = find_task_by_real_pid(version);
+               if (p && !thread_group_leader(p))
                        p = NULL;
        }
 
@@ -1715,7 +1720,7 @@ static int get_tgid_list(int index, unsigned long version, unsigned int *tgids)
                        continue;
                if (--index >= 0)
                        continue;
-               tgids[nr_tgids] = vx_map_tgid(current->vx_info, tgid);
+               tgids[nr_tgids] = vx_map_tgid(tgid);
                nr_tgids++;
                if (nr_tgids >= PROC_MAXPIDS)
                        break;
@@ -1749,7 +1754,7 @@ static int get_tid_list(int index, unsigned int *tids, struct inode *dir)
                        continue;
                if (--index >= 0)
                        continue;
-               tids[nr_tids] = vx_map_tgid(current->vx_info, tid);
+               tids[nr_tids] = vx_map_pid(tid);
                nr_tids++;
                if (nr_tids >= PROC_MAXPIDS)
                        break;
@@ -1765,6 +1770,7 @@ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir)
        char buf[PROC_NUMBUF];
        unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY;
        unsigned int nr_tgids, i;
+       int next_tgid;
 
        if (!nr) {
                ino_t ino = fake_ino(0,PROC_TGID_INO);
@@ -1774,26 +1780,45 @@ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir)
                nr++;
        }
 
-       /*
-        * f_version caches the last tgid which was returned from readdir
+       /* f_version caches the tgid value that the last readdir call couldn't
+        * return. lseek aka telldir automagically resets f_version to 0.
         */
-       nr_tgids = get_tgid_list(nr, filp->f_version, tgid_array);
+       next_tgid = filp->f_version;
+       filp->f_version = 0;
+       for (;;) {
+               nr_tgids = get_tgid_list(nr, next_tgid, tgid_array);
+               if (!nr_tgids) {
+                       /* no more entries ! */
+                       break;
+               }
+               next_tgid = 0;
 
-       for (i = 0; i < nr_tgids; i++) {
-               int tgid = tgid_array[i];
-               ino_t ino = fake_ino(tgid,PROC_TGID_INO);
-               unsigned long j = PROC_NUMBUF;
+               /* do not use the last found pid, reserve it for next_tgid */
+               if (nr_tgids == PROC_MAXPIDS) {
+                       nr_tgids--;
+                       next_tgid = tgid_array[nr_tgids];
+               }
 
-               do
-                       buf[--j] = '0' + (tgid % 10);
-               while ((tgid /= 10) != 0);
+               for (i=0;i<nr_tgids;i++) {
+                       int tgid = tgid_array[i];
+                       ino_t ino = fake_ino(tgid,PROC_TGID_INO);
+                       unsigned long j = PROC_NUMBUF;
 
-               if (filldir(dirent, buf+j, PROC_NUMBUF-j, filp->f_pos, ino, DT_DIR) < 0) {
-                       filp->f_version = tgid;
-                       break;
+                       do
+                               buf[--j] = '0' + (tgid % 10);
+                       while ((tgid /= 10) != 0);
+
+                       if (filldir(dirent, buf+j, PROC_NUMBUF-j, filp->f_pos, ino, DT_DIR) < 0) {
+                               /* returning this tgid failed, save it as the first
+                                * pid for the next readir call */
+                               filp->f_version = tgid_array[i];
+                               goto out;
+                       }
+                       filp->f_pos++;
+                       nr++;
                }
-               filp->f_pos++;
        }
+out:
        return 0;
 }