[ARM] 4737/1: Refactor corgi_lcd to improve readability + bugfix
[safe/jmp/linux-2.6] / fs / ext3 / inode.c
index 56665fa..9b162cd 100644 (file)
  *  Copyright (C) 1991, 1992  Linus Torvalds
  *
  *  Goal-directed block allocation by Stephen Tweedie
- *     (sct@redhat.com), 1993, 1998
+ *     (sct@redhat.com), 1993, 1998
  *  Big-endian to little-endian byte-swapping/bitmaps by
  *        David S. Miller (davem@caip.rutgers.edu), 1995
  *  64-bit file support on 64-bit platforms by Jakub Jelinek
- *     (jj@sunsite.ms.mff.cuni.cz)
+ *     (jj@sunsite.ms.mff.cuni.cz)
  *
  *  Assorted race fixes, rewrite of ext3_get_block() by Al Viro, 2000
  */
@@ -27,7 +27,6 @@
 #include <linux/time.h>
 #include <linux/ext3_jbd.h>
 #include <linux/jbd.h>
-#include <linux/smp_lock.h>
 #include <linux/highuid.h>
 #include <linux/pagemap.h>
 #include <linux/quotaops.h>
@@ -36,6 +35,7 @@
 #include <linux/writeback.h>
 #include <linux/mpage.h>
 #include <linux/uio.h>
+#include <linux/bio.h>
 #include "xattr.h"
 #include "acl.h"
 
@@ -470,7 +470,7 @@ static ext3_fsblk_t ext3_find_goal(struct inode *inode, long block,
  *     ext3_blks_to_allocate: Look up the block map and count the number
  *     of direct blocks need to be allocated for the given branch.
  *
- *     @branch: chain of indirect blocks
+ *     @branch: chain of indirect blocks
  *     @k: number of blocks need for indirect blocks
  *     @blks: number of data blocks to be mapped.
  *     @blocks_to_boundary:  the offset in the indirect block
@@ -946,7 +946,7 @@ out:
 static int ext3_get_block(struct inode *inode, sector_t iblock,
                        struct buffer_head *bh_result, int create)
 {
-       handle_t *handle = journal_current_handle();
+       handle_t *handle = ext3_journal_current_handle();
        int ret = 0;
        unsigned max_blocks = bh_result->b_size >> inode->i_blkbits;
 
@@ -1028,7 +1028,7 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode,
                }
                if (buffer_new(&dummy)) {
                        J_ASSERT(create != 0);
-                       J_ASSERT(handle != 0);
+                       J_ASSERT(handle != NULL);
 
                        /*
                         * Now that we do not always journal data, we should
@@ -1073,7 +1073,7 @@ struct buffer_head *ext3_bread(handle_t *handle, struct inode *inode,
                return bh;
        if (buffer_uptodate(bh))
                return bh;
-       ll_rw_block(READ, 1, &bh);
+       ll_rw_block(READ_META, 1, &bh);
        wait_on_buffer(bh);
        if (buffer_uptodate(bh))
                return bh;
@@ -1098,7 +1098,7 @@ static int walk_page_buffers(     handle_t *handle,
 
        for (   bh = head, block_start = 0;
                ret == 0 && (bh != head || !block_start);
-               block_start = block_end, bh = next)
+               block_start = block_end, bh = next)
        {
                next = bh->b_this_page;
                block_end = block_start + blocksize;
@@ -1147,51 +1147,68 @@ static int do_journal_get_write_access(handle_t *handle,
        return ext3_journal_get_write_access(handle, bh);
 }
 
-static int ext3_prepare_write(struct file *file, struct page *page,
-                             unsigned from, unsigned to)
+static int ext3_write_begin(struct file *file, struct address_space *mapping,
+                               loff_t pos, unsigned len, unsigned flags,
+                               struct page **pagep, void **fsdata)
 {
-       struct inode *inode = page->mapping->host;
+       struct inode *inode = mapping->host;
        int ret, needed_blocks = ext3_writepage_trans_blocks(inode);
        handle_t *handle;
        int retries = 0;
+       struct page *page;
+       pgoff_t index;
+       unsigned from, to;
+
+       index = pos >> PAGE_CACHE_SHIFT;
+       from = pos & (PAGE_CACHE_SIZE - 1);
+       to = from + len;
 
 retry:
+       page = __grab_cache_page(mapping, index);
+       if (!page)
+               return -ENOMEM;
+       *pagep = page;
+
        handle = ext3_journal_start(inode, needed_blocks);
        if (IS_ERR(handle)) {
+               unlock_page(page);
+               page_cache_release(page);
                ret = PTR_ERR(handle);
                goto out;
        }
-       if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode))
-               ret = nobh_prepare_write(page, from, to, ext3_get_block);
-       else
-               ret = block_prepare_write(page, from, to, ext3_get_block);
+       ret = block_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+                                                       ext3_get_block);
        if (ret)
-               goto prepare_write_failed;
+               goto write_begin_failed;
 
        if (ext3_should_journal_data(inode)) {
                ret = walk_page_buffers(handle, page_buffers(page),
                                from, to, NULL, do_journal_get_write_access);
        }
-prepare_write_failed:
-       if (ret)
+write_begin_failed:
+       if (ret) {
                ext3_journal_stop(handle);
+               unlock_page(page);
+               page_cache_release(page);
+       }
        if (ret == -ENOSPC && ext3_should_retry_alloc(inode->i_sb, &retries))
                goto retry;
 out:
        return ret;
 }
 
+
 int ext3_journal_dirty_data(handle_t *handle, struct buffer_head *bh)
 {
        int err = journal_dirty_data(handle, bh);
        if (err)
                ext3_journal_abort_handle(__FUNCTION__, __FUNCTION__,
-                                               bh, handle,err);
+                                               bh, handle, err);
        return err;
 }
 
-/* For commit_write() in data=journal mode */
-static int commit_write_fn(handle_t *handle, struct buffer_head *bh)
+/* For write_end() in data=journal mode */
+static int write_end_fn(handle_t *handle, struct buffer_head *bh)
 {
        if (!buffer_mapped(bh) || buffer_freed(bh))
                return 0;
@@ -1200,84 +1217,130 @@ static int commit_write_fn(handle_t *handle, struct buffer_head *bh)
 }
 
 /*
+ * Generic write_end handler for ordered and writeback ext3 journal modes.
+ * We can't use generic_write_end, because that unlocks the page and we need to
+ * unlock the page after ext3_journal_stop, but ext3_journal_stop must run
+ * after block_write_end.
+ */
+static int ext3_generic_write_end(struct file *file,
+                               struct address_space *mapping,
+                               loff_t pos, unsigned len, unsigned copied,
+                               struct page *page, void *fsdata)
+{
+       struct inode *inode = file->f_mapping->host;
+
+       copied = block_write_end(file, mapping, pos, len, copied, page, fsdata);
+
+       if (pos+copied > inode->i_size) {
+               i_size_write(inode, pos+copied);
+               mark_inode_dirty(inode);
+       }
+
+       return copied;
+}
+
+/*
  * We need to pick up the new inode size which generic_commit_write gave us
  * `file' can be NULL - eg, when called from page_symlink().
  *
  * ext3 never places buffers on inode->i_mapping->private_list.  metadata
  * buffers are managed internally.
  */
-static int ext3_ordered_commit_write(struct file *file, struct page *page,
-                            unsigned from, unsigned to)
+static int ext3_ordered_write_end(struct file *file,
+                               struct address_space *mapping,
+                               loff_t pos, unsigned len, unsigned copied,
+                               struct page *page, void *fsdata)
 {
        handle_t *handle = ext3_journal_current_handle();
-       struct inode *inode = page->mapping->host;
+       struct inode *inode = file->f_mapping->host;
+       unsigned from, to;
        int ret = 0, ret2;
 
+       from = pos & (PAGE_CACHE_SIZE - 1);
+       to = from + len;
+
        ret = walk_page_buffers(handle, page_buffers(page),
                from, to, NULL, ext3_journal_dirty_data);
 
        if (ret == 0) {
                /*
-                * generic_commit_write() will run mark_inode_dirty() if i_size
+                * generic_write_end() will run mark_inode_dirty() if i_size
                 * changes.  So let's piggyback the i_disksize mark_inode_dirty
                 * into that.
                 */
                loff_t new_i_size;
 
-               new_i_size = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
+               new_i_size = pos + copied;
                if (new_i_size > EXT3_I(inode)->i_disksize)
                        EXT3_I(inode)->i_disksize = new_i_size;
-               ret = generic_commit_write(file, page, from, to);
+               copied = ext3_generic_write_end(file, mapping, pos, len, copied,
+                                                       page, fsdata);
+               if (copied < 0)
+                       ret = copied;
        }
        ret2 = ext3_journal_stop(handle);
        if (!ret)
                ret = ret2;
-       return ret;
+       unlock_page(page);
+       page_cache_release(page);
+
+       return ret ? ret : copied;
 }
 
-static int ext3_writeback_commit_write(struct file *file, struct page *page,
-                            unsigned from, unsigned to)
+static int ext3_writeback_write_end(struct file *file,
+                               struct address_space *mapping,
+                               loff_t pos, unsigned len, unsigned copied,
+                               struct page *page, void *fsdata)
 {
        handle_t *handle = ext3_journal_current_handle();
-       struct inode *inode = page->mapping->host;
+       struct inode *inode = file->f_mapping->host;
        int ret = 0, ret2;
        loff_t new_i_size;
 
-       new_i_size = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
+       new_i_size = pos + copied;
        if (new_i_size > EXT3_I(inode)->i_disksize)
                EXT3_I(inode)->i_disksize = new_i_size;
 
-       if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode))
-               ret = nobh_commit_write(file, page, from, to);
-       else
-               ret = generic_commit_write(file, page, from, to);
+       copied = ext3_generic_write_end(file, mapping, pos, len, copied,
+                                                       page, fsdata);
+       if (copied < 0)
+               ret = copied;
 
        ret2 = ext3_journal_stop(handle);
        if (!ret)
                ret = ret2;
-       return ret;
+       unlock_page(page);
+       page_cache_release(page);
+
+       return ret ? ret : copied;
 }
 
-static int ext3_journalled_commit_write(struct file *file,
-                       struct page *page, unsigned from, unsigned to)
+static int ext3_journalled_write_end(struct file *file,
+                               struct address_space *mapping,
+                               loff_t pos, unsigned len, unsigned copied,
+                               struct page *page, void *fsdata)
 {
        handle_t *handle = ext3_journal_current_handle();
-       struct inode *inode = page->mapping->host;
+       struct inode *inode = mapping->host;
        int ret = 0, ret2;
        int partial = 0;
-       loff_t pos;
+       unsigned from, to;
 
-       /*
-        * Here we duplicate the generic_commit_write() functionality
-        */
-       pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
+       from = pos & (PAGE_CACHE_SIZE - 1);
+       to = from + len;
+
+       if (copied < len) {
+               if (!PageUptodate(page))
+                       copied = 0;
+               page_zero_new_buffers(page, from+copied, to);
+       }
 
        ret = walk_page_buffers(handle, page_buffers(page), from,
-                               to, &partial, commit_write_fn);
+                               to, &partial, write_end_fn);
        if (!partial)
                SetPageUptodate(page);
-       if (pos > inode->i_size)
-               i_size_write(inode, pos);
+       if (pos+copied > inode->i_size)
+               i_size_write(inode, pos+copied);
        EXT3_I(inode)->i_state |= EXT3_STATE_JDATA;
        if (inode->i_size > EXT3_I(inode)->i_disksize) {
                EXT3_I(inode)->i_disksize = inode->i_size;
@@ -1285,10 +1348,14 @@ static int ext3_journalled_commit_write(struct file *file,
                if (!ret)
                        ret = ret2;
        }
+
        ret2 = ext3_journal_stop(handle);
        if (!ret)
                ret = ret2;
-       return ret;
+       unlock_page(page);
+       page_cache_release(page);
+
+       return ret ? ret : copied;
 }
 
 /*
@@ -1546,7 +1613,7 @@ static int ext3_journalled_writepage(struct page *page,
                        PAGE_CACHE_SIZE, NULL, do_journal_get_write_access);
 
                err = walk_page_buffers(handle, page_buffers(page), 0,
-                               PAGE_CACHE_SIZE, NULL, commit_write_fn);
+                               PAGE_CACHE_SIZE, NULL, write_end_fn);
                if (ret == 0)
                        ret = err;
                EXT3_I(inode)->i_state |= EXT3_STATE_JDATA;
@@ -1651,7 +1718,7 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb,
        /*
         * Reacquire the handle: ext3_get_block() can restart the transaction
         */
-       handle = journal_current_handle();
+       handle = ext3_journal_current_handle();
 
 out_stop:
        if (handle) {
@@ -1706,8 +1773,8 @@ static const struct address_space_operations ext3_ordered_aops = {
        .readpages      = ext3_readpages,
        .writepage      = ext3_ordered_writepage,
        .sync_page      = block_sync_page,
-       .prepare_write  = ext3_prepare_write,
-       .commit_write   = ext3_ordered_commit_write,
+       .write_begin    = ext3_write_begin,
+       .write_end      = ext3_ordered_write_end,
        .bmap           = ext3_bmap,
        .invalidatepage = ext3_invalidatepage,
        .releasepage    = ext3_releasepage,
@@ -1720,8 +1787,8 @@ static const struct address_space_operations ext3_writeback_aops = {
        .readpages      = ext3_readpages,
        .writepage      = ext3_writeback_writepage,
        .sync_page      = block_sync_page,
-       .prepare_write  = ext3_prepare_write,
-       .commit_write   = ext3_writeback_commit_write,
+       .write_begin    = ext3_write_begin,
+       .write_end      = ext3_writeback_write_end,
        .bmap           = ext3_bmap,
        .invalidatepage = ext3_invalidatepage,
        .releasepage    = ext3_releasepage,
@@ -1734,8 +1801,8 @@ static const struct address_space_operations ext3_journalled_aops = {
        .readpages      = ext3_readpages,
        .writepage      = ext3_journalled_writepage,
        .sync_page      = block_sync_page,
-       .prepare_write  = ext3_prepare_write,
-       .commit_write   = ext3_journalled_commit_write,
+       .write_begin    = ext3_write_begin,
+       .write_end      = ext3_journalled_write_end,
        .set_page_dirty = ext3_journalled_set_page_dirty,
        .bmap           = ext3_bmap,
        .invalidatepage = ext3_invalidatepage,
@@ -1767,7 +1834,6 @@ static int ext3_block_truncate_page(handle_t *handle, struct page *page,
        struct inode *inode = mapping->host;
        struct buffer_head *bh;
        int err = 0;
-       void *kaddr;
 
        blocksize = inode->i_sb->s_blocksize;
        length = blocksize - (offset & (blocksize - 1));
@@ -1779,10 +1845,7 @@ static int ext3_block_truncate_page(handle_t *handle, struct page *page,
         */
        if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) &&
             ext3_should_writeback_data(inode) && PageUptodate(page)) {
-               kaddr = kmap_atomic(page, KM_USER0);
-               memset(kaddr + offset, 0, length);
-               flush_dcache_page(page);
-               kunmap_atomic(kaddr, KM_USER0);
+               zero_user_page(page, offset, length, KM_USER0);
                set_page_dirty(page);
                goto unlock;
        }
@@ -1835,11 +1898,7 @@ static int ext3_block_truncate_page(handle_t *handle, struct page *page,
                        goto unlock;
        }
 
-       kaddr = kmap_atomic(page, KM_USER0);
-       memset(kaddr + offset, 0, length);
-       flush_dcache_page(page);
-       kunmap_atomic(kaddr, KM_USER0);
-
+       zero_user_page(page, offset, length, KM_USER0);
        BUFFER_TRACE(bh, "zeroed end of block");
 
        err = 0;
@@ -2540,7 +2599,7 @@ make_io:
                 */
                get_bh(bh);
                bh->b_end_io = end_buffer_read_sync;
-               submit_bh(READ, bh);
+               submit_bh(READ_META, bh);
                wait_on_buffer(bh);
                if (!buffer_uptodate(bh)) {
                        ext3_error(inode->i_sb, "ext3_get_inode_loc",
@@ -2580,6 +2639,25 @@ void ext3_set_inode_flags(struct inode *inode)
                inode->i_flags |= S_DIRSYNC;
 }
 
+/* Propagate flags from i_flags to EXT3_I(inode)->i_flags */
+void ext3_get_inode_flags(struct ext3_inode_info *ei)
+{
+       unsigned int flags = ei->vfs_inode.i_flags;
+
+       ei->i_flags &= ~(EXT3_SYNC_FL|EXT3_APPEND_FL|
+                       EXT3_IMMUTABLE_FL|EXT3_NOATIME_FL|EXT3_DIRSYNC_FL);
+       if (flags & S_SYNC)
+               ei->i_flags |= EXT3_SYNC_FL;
+       if (flags & S_APPEND)
+               ei->i_flags |= EXT3_APPEND_FL;
+       if (flags & S_IMMUTABLE)
+               ei->i_flags |= EXT3_IMMUTABLE_FL;
+       if (flags & S_NOATIME)
+               ei->i_flags |= EXT3_NOATIME_FL;
+       if (flags & S_DIRSYNC)
+               ei->i_flags |= EXT3_DIRSYNC_FL;
+}
+
 void ext3_read_inode(struct inode * inode)
 {
        struct ext3_iloc iloc;
@@ -2607,9 +2685,9 @@ void ext3_read_inode(struct inode * inode)
        }
        inode->i_nlink = le16_to_cpu(raw_inode->i_links_count);
        inode->i_size = le32_to_cpu(raw_inode->i_size);
-       inode->i_atime.tv_sec = le32_to_cpu(raw_inode->i_atime);
-       inode->i_ctime.tv_sec = le32_to_cpu(raw_inode->i_ctime);
-       inode->i_mtime.tv_sec = le32_to_cpu(raw_inode->i_mtime);
+       inode->i_atime.tv_sec = (signed)le32_to_cpu(raw_inode->i_atime);
+       inode->i_ctime.tv_sec = (signed)le32_to_cpu(raw_inode->i_ctime);
+       inode->i_mtime.tv_sec = (signed)le32_to_cpu(raw_inode->i_mtime);
        inode->i_atime.tv_nsec = inode->i_ctime.tv_nsec = inode->i_mtime.tv_nsec = 0;
 
        ei->i_state = 0;
@@ -2632,9 +2710,6 @@ void ext3_read_inode(struct inode * inode)
                 * recovery code: that's fine, we're about to complete
                 * the process of deleting those. */
        }
-       inode->i_blksize = PAGE_SIZE;   /* This is the optimal IO size
-                                        * (for stat), not the fs block
-                                        * size */  
        inode->i_blocks = le32_to_cpu(raw_inode->i_blocks);
        ei->i_flags = le32_to_cpu(raw_inode->i_flags);
 #ifdef EXT3_FRAGMENTS
@@ -2669,8 +2744,10 @@ void ext3_read_inode(struct inode * inode)
                 */
                ei->i_extra_isize = le16_to_cpu(raw_inode->i_extra_isize);
                if (EXT3_GOOD_OLD_INODE_SIZE + ei->i_extra_isize >
-                   EXT3_INODE_SIZE(inode->i_sb))
+                   EXT3_INODE_SIZE(inode->i_sb)) {
+                       brelse (bh);
                        goto bad_inode;
+               }
                if (ei->i_extra_isize == 0) {
                        /* The extra space is currently unused. Use it. */
                        ei->i_extra_isize = sizeof(struct ext3_inode) -
@@ -2738,6 +2815,7 @@ static int ext3_do_update_inode(handle_t *handle,
        if (ei->i_state & EXT3_STATE_NEW)
                memset(raw_inode, 0, EXT3_SB(inode->i_sb)->s_inode_size);
 
+       ext3_get_inode_flags(ei);
        raw_inode->i_mode = cpu_to_le16(inode->i_mode);
        if(!(test_opt(inode->i_sb, NO_UID32))) {
                raw_inode->i_uid_low = cpu_to_le16(low_16_bits(inode->i_uid));
@@ -2876,7 +2954,7 @@ int ext3_write_inode(struct inode *inode, int wait)
                return 0;
 
        if (ext3_journal_current_handle()) {
-               jbd_debug(0, "called recursively, non-PF_MEMALLOC!\n");
+               jbd_debug(1, "called recursively, non-PF_MEMALLOC!\n");
                dump_stack();
                return -EIO;
        }
@@ -3184,7 +3262,7 @@ int ext3_change_inode_journal_flag(struct inode *inode, int val)
         */
 
        journal = EXT3_JOURNAL(inode);
-       if (is_journal_aborted(journal) || IS_RDONLY(inode))
+       if (is_journal_aborted(journal))
                return -EROFS;
 
        journal_lock_updates(journal);