[PATCH] memory hotplug: sysfs and add/remove functions
[pandora-kernel.git] / fs / open.c
index 4ee2dcc..8d06ec9 100644 (file)
--- a/fs/open.c
+++ b/fs/open.c
@@ -24,6 +24,7 @@
 #include <linux/personality.h>
 #include <linux/pagemap.h>
 #include <linux/syscalls.h>
+#include <linux/rcupdate.h>
 
 #include <asm/unistd.h>
 
@@ -737,52 +738,16 @@ asmlinkage long sys_fchown(unsigned int fd, uid_t user, gid_t group)
        return error;
 }
 
-/*
- * Note that while the flag value (low two bits) for sys_open means:
- *     00 - read-only
- *     01 - write-only
- *     10 - read-write
- *     11 - special
- * it is changed into
- *     00 - no permissions needed
- *     01 - read-permission
- *     10 - write-permission
- *     11 - read-write
- * for the internal routines (ie open_namei()/follow_link() etc). 00 is
- * used by symlinks.
- */
-struct file *filp_open(const char * filename, int flags, int mode)
-{
-       int namei_flags, error;
-       struct nameidata nd;
-
-       namei_flags = flags;
-       if ((namei_flags+1) & O_ACCMODE)
-               namei_flags++;
-       if (namei_flags & O_TRUNC)
-               namei_flags |= 2;
-
-       error = open_namei(filename, namei_flags, mode, &nd);
-       if (!error)
-               return dentry_open(nd.dentry, nd.mnt, flags);
-
-       return ERR_PTR(error);
-}
-
-EXPORT_SYMBOL(filp_open);
-
-struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags)
+static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
+                                       int flags, struct file *f,
+                                       int (*open)(struct inode *, struct file *))
 {
-       struct file * f;
        struct inode *inode;
        int error;
 
-       error = -ENFILE;
-       f = get_empty_filp();
-       if (!f)
-               goto cleanup_dentry;
        f->f_flags = flags;
-       f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE;
+       f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK |
+                               FMODE_PREAD | FMODE_PWRITE;
        inode = dentry->d_inode;
        if (f->f_mode & FMODE_WRITE) {
                error = get_write_access(inode);
@@ -797,11 +762,14 @@ struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags)
        f->f_op = fops_get(inode->i_fop);
        file_move(f, &inode->i_sb->s_files);
 
-       if (f->f_op && f->f_op->open) {
-               error = f->f_op->open(inode,f);
+       if (!open && f->f_op)
+               open = f->f_op->open;
+       if (open) {
+               error = open(inode, f);
                if (error)
                        goto cleanup_all;
        }
+
        f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
 
        file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
@@ -827,12 +795,110 @@ cleanup_all:
        f->f_vfsmnt = NULL;
 cleanup_file:
        put_filp(f);
-cleanup_dentry:
        dput(dentry);
        mntput(mnt);
        return ERR_PTR(error);
 }
 
+/*
+ * Note that while the flag value (low two bits) for sys_open means:
+ *     00 - read-only
+ *     01 - write-only
+ *     10 - read-write
+ *     11 - special
+ * it is changed into
+ *     00 - no permissions needed
+ *     01 - read-permission
+ *     10 - write-permission
+ *     11 - read-write
+ * for the internal routines (ie open_namei()/follow_link() etc). 00 is
+ * used by symlinks.
+ */
+struct file *filp_open(const char * filename, int flags, int mode)
+{
+       int namei_flags, error;
+       struct nameidata nd;
+
+       namei_flags = flags;
+       if ((namei_flags+1) & O_ACCMODE)
+               namei_flags++;
+
+       error = open_namei(filename, namei_flags, mode, &nd);
+       if (!error)
+               return nameidata_to_filp(&nd, flags);
+
+       return ERR_PTR(error);
+}
+EXPORT_SYMBOL(filp_open);
+
+/**
+ * lookup_instantiate_filp - instantiates the open intent filp
+ * @nd: pointer to nameidata
+ * @dentry: pointer to dentry
+ * @open: open callback
+ *
+ * Helper for filesystems that want to use lookup open intents and pass back
+ * a fully instantiated struct file to the caller.
+ * This function is meant to be called from within a filesystem's
+ * lookup method.
+ * Note that in case of error, nd->intent.open.file is destroyed, but the
+ * path information remains valid.
+ * If the open callback is set to NULL, then the standard f_op->open()
+ * filesystem callback is substituted.
+ */
+struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry,
+               int (*open)(struct inode *, struct file *))
+{
+       if (IS_ERR(nd->intent.open.file))
+               goto out;
+       if (IS_ERR(dentry))
+               goto out_err;
+       nd->intent.open.file = __dentry_open(dget(dentry), mntget(nd->mnt),
+                                            nd->intent.open.flags - 1,
+                                            nd->intent.open.file,
+                                            open);
+out:
+       return nd->intent.open.file;
+out_err:
+       release_open_intent(nd);
+       nd->intent.open.file = (struct file *)dentry;
+       goto out;
+}
+EXPORT_SYMBOL_GPL(lookup_instantiate_filp);
+
+/**
+ * nameidata_to_filp - convert a nameidata to an open filp.
+ * @nd: pointer to nameidata
+ * @flags: open flags
+ *
+ * Note that this function destroys the original nameidata
+ */
+struct file *nameidata_to_filp(struct nameidata *nd, int flags)
+{
+       struct file *filp;
+
+       /* Pick up the filp from the open intent */
+       filp = nd->intent.open.file;
+       /* Has the filesystem initialised the file for us? */
+       if (filp->f_dentry == NULL)
+               filp = __dentry_open(nd->dentry, nd->mnt, flags, filp, NULL);
+       else
+               path_release(nd);
+       return filp;
+}
+
+struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags)
+{
+       int error;
+       struct file *f;
+
+       error = -ENFILE;
+       f = get_empty_filp();
+       if (f == NULL)
+               return ERR_PTR(error);
+
+       return __dentry_open(dentry, mnt, flags, f, NULL);
+}
 EXPORT_SYMBOL(dentry_open);
 
 /*
@@ -842,14 +908,16 @@ int get_unused_fd(void)
 {
        struct files_struct * files = current->files;
        int fd, error;
+       struct fdtable *fdt;
 
        error = -EMFILE;
        spin_lock(&files->file_lock);
 
 repeat:
-       fd = find_next_zero_bit(files->open_fds->fds_bits, 
-                               files->max_fdset, 
-                               files->next_fd);
+       fdt = files_fdtable(files);
+       fd = find_next_zero_bit(fdt->open_fds->fds_bits,
+                               fdt->max_fdset,
+                               fdt->next_fd);
 
        /*
         * N.B. For clone tasks sharing a files structure, this test
@@ -872,14 +940,14 @@ repeat:
                goto repeat;
        }
 
-       FD_SET(fd, files->open_fds);
-       FD_CLR(fd, files->close_on_exec);
-       files->next_fd = fd + 1;
+       FD_SET(fd, fdt->open_fds);
+       FD_CLR(fd, fdt->close_on_exec);
+       fdt->next_fd = fd + 1;
 #if 1
        /* Sanity check */
-       if (files->fd[fd] != NULL) {
+       if (fdt->fd[fd] != NULL) {
                printk(KERN_WARNING "get_unused_fd: slot %d not NULL!\n", fd);
-               files->fd[fd] = NULL;
+               fdt->fd[fd] = NULL;
        }
 #endif
        error = fd;
@@ -893,9 +961,10 @@ EXPORT_SYMBOL(get_unused_fd);
 
 static inline void __put_unused_fd(struct files_struct *files, unsigned int fd)
 {
-       __FD_CLR(fd, files->open_fds);
-       if (fd < files->next_fd)
-               files->next_fd = fd;
+       struct fdtable *fdt = files_fdtable(files);
+       __FD_CLR(fd, fdt->open_fds);
+       if (fd < fdt->next_fd)
+               fdt->next_fd = fd;
 }
 
 void fastcall put_unused_fd(unsigned int fd)
@@ -924,10 +993,11 @@ EXPORT_SYMBOL(put_unused_fd);
 void fastcall fd_install(unsigned int fd, struct file * file)
 {
        struct files_struct *files = current->files;
+       struct fdtable *fdt;
        spin_lock(&files->file_lock);
-       if (unlikely(files->fd[fd] != NULL))
-               BUG();
-       files->fd[fd] = file;
+       fdt = files_fdtable(files);
+       BUG_ON(fdt->fd[fd] != NULL);
+       rcu_assign_pointer(fdt->fd[fd], file);
        spin_unlock(&files->file_lock);
 }
 
@@ -1010,15 +1080,17 @@ asmlinkage long sys_close(unsigned int fd)
 {
        struct file * filp;
        struct files_struct *files = current->files;
+       struct fdtable *fdt;
 
        spin_lock(&files->file_lock);
-       if (fd >= files->max_fds)
+       fdt = files_fdtable(files);
+       if (fd >= fdt->max_fds)
                goto out_unlock;
-       filp = files->fd[fd];
+       filp = fdt->fd[fd];
        if (!filp)
                goto out_unlock;
-       files->fd[fd] = NULL;
-       FD_CLR(fd, files->close_on_exec);
+       rcu_assign_pointer(fdt->fd[fd], NULL);
+       FD_CLR(fd, fdt->close_on_exec);
        __put_unused_fd(files, fd);
        spin_unlock(&files->file_lock);
        return filp_close(filp, files);