ext4: Reorder fs/Makefile so that ext2 root fs's are mounted using ext2
[safe/jmp/linux-2.6] / fs / ocfs2 / buffer_head_io.c
index ca4ab7c..15c8e6d 100644 (file)
 
 #include "buffer_head_io.h"
 
+/*
+ * Bits on bh->b_state used by ocfs2.
+ *
+ * These MUST be after the JBD2 bits.  Hence, we use BH_JBDPrivateStart.
+ */
+enum ocfs2_state_bits {
+       BH_NeedsValidate = BH_JBDPrivateStart,
+};
+
+/* Expand the magic b_state functions */
+BUFFER_FNS(NeedsValidate, needs_validate);
+
 int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
                      struct inode *inode)
 {
@@ -112,7 +124,7 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block,
                bh = bhs[i];
 
                if (buffer_jbd(bh)) {
-                       mlog(ML_ERROR,
+                       mlog(ML_BH_IO,
                             "trying to sync read a jbd "
                             "managed bh (blocknr = %llu), skipping\n",
                             (unsigned long long)bh->b_blocknr);
@@ -147,15 +159,10 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block,
        for (i = nr; i > 0; i--) {
                bh = bhs[i - 1];
 
-               if (buffer_jbd(bh)) {
-                       mlog(ML_ERROR,
-                            "the journal got the buffer while it was "
-                            "locked for io! (blocknr = %llu)\n",
-                            (unsigned long long)bh->b_blocknr);
-                       BUG();
-               }
+               /* No need to wait on the buffer if it's managed by JBD. */
+               if (!buffer_jbd(bh))
+                       wait_on_buffer(bh);
 
-               wait_on_buffer(bh);
                if (!buffer_uptodate(bh)) {
                        /* Status won't be cleared from here on out,
                         * so we can safely record this and loop back
@@ -170,22 +177,23 @@ bail:
        return status;
 }
 
-int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
+int ocfs2_read_blocks(struct inode *inode, u64 block, int nr,
                      struct buffer_head *bhs[], int flags,
-                     struct inode *inode)
+                     int (*validate)(struct super_block *sb,
+                                     struct buffer_head *bh))
 {
        int status = 0;
-       struct super_block *sb;
        int i, ignore_cache = 0;
        struct buffer_head *bh;
 
-       mlog_entry("(block=(%llu), nr=(%d), flags=%d, inode=%p)\n",
-                  (unsigned long long)block, nr, flags, inode);
+       mlog_entry("(inode=%p, block=(%llu), nr=(%d), flags=%d)\n",
+                  inode, (unsigned long long)block, nr, flags);
 
+       BUG_ON(!inode);
        BUG_ON((flags & OCFS2_BH_READAHEAD) &&
-              (!inode || !(flags & OCFS2_BH_CACHED)));
+              (flags & OCFS2_BH_IGNORE_CACHE));
 
-       if (osb == NULL || osb->sb == NULL || bhs == NULL) {
+       if (bhs == NULL) {
                status = -EINVAL;
                mlog_errno(status);
                goto bail;
@@ -204,26 +212,19 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                goto bail;
        }
 
-       sb = osb->sb;
-
-       if (flags & OCFS2_BH_CACHED && !inode)
-               flags &= ~OCFS2_BH_CACHED;
-
-       if (inode)
-               mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
+       mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
        for (i = 0 ; i < nr ; i++) {
                if (bhs[i] == NULL) {
-                       bhs[i] = sb_getblk(sb, block++);
+                       bhs[i] = sb_getblk(inode->i_sb, block++);
                        if (bhs[i] == NULL) {
-                               if (inode)
-                                       mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
+                               mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
                                status = -EIO;
                                mlog_errno(status);
                                goto bail;
                        }
                }
                bh = bhs[i];
-               ignore_cache = 0;
+               ignore_cache = (flags & OCFS2_BH_IGNORE_CACHE);
 
                /* There are three read-ahead cases here which we need to
                 * be concerned with. All three assume a buffer has
@@ -249,26 +250,25 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                 *    before our is-it-in-flight check.
                 */
 
-               if (flags & OCFS2_BH_CACHED &&
-                   !ocfs2_buffer_uptodate(inode, bh)) {
+               if (!ignore_cache && !ocfs2_buffer_uptodate(inode, bh)) {
                        mlog(ML_UPTODATE,
                             "bh (%llu), inode %llu not uptodate\n",
                             (unsigned long long)bh->b_blocknr,
                             (unsigned long long)OCFS2_I(inode)->ip_blkno);
+                       /* We're using ignore_cache here to say
+                        * "go to disk" */
                        ignore_cache = 1;
                }
 
-               /* XXX: Can we ever get this and *not* have the cached
-                * flag set? */
                if (buffer_jbd(bh)) {
-                       if (!(flags & OCFS2_BH_CACHED) || ignore_cache)
+                       if (ignore_cache)
                                mlog(ML_BH_IO, "trying to sync read a jbd "
                                               "managed bh (blocknr = %llu)\n",
                                     (unsigned long long)bh->b_blocknr);
                        continue;
                }
 
-               if (!(flags & OCFS2_BH_CACHED) || ignore_cache) {
+               if (ignore_cache) {
                        if (buffer_dirty(bh)) {
                                /* This should probably be a BUG, or
                                 * at least return an error. */
@@ -303,7 +303,7 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                         * previously read-ahead buffer may have
                         * completed I/O while we were waiting for the
                         * buffer lock. */
-                       if ((flags & OCFS2_BH_CACHED)
+                       if (!(flags & OCFS2_BH_IGNORE_CACHE)
                            && !(flags & OCFS2_BH_READAHEAD)
                            && ocfs2_buffer_uptodate(inode, bh)) {
                                unlock_buffer(bh);
@@ -312,6 +312,8 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
 
                        clear_buffer_uptodate(bh);
                        get_bh(bh); /* for end_buffer_read_sync() */
+                       if (validate)
+                               set_buffer_needs_validate(bh);
                        bh->b_end_io = end_buffer_read_sync;
                        submit_bh(READ, bh);
                        continue;
@@ -342,20 +344,33 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                                bhs[i] = NULL;
                                continue;
                        }
+
+                       if (buffer_needs_validate(bh)) {
+                               /* We never set NeedsValidate if the
+                                * buffer was held by the journal, so
+                                * that better not have changed */
+                               BUG_ON(buffer_jbd(bh));
+                               clear_buffer_needs_validate(bh);
+                               status = validate(inode->i_sb, bh);
+                               if (status) {
+                                       put_bh(bh);
+                                       bhs[i] = NULL;
+                                       continue;
+                               }
+                       }
                }
 
                /* Always set the buffer in the cache, even if it was
                 * a forced read, or read-ahead which hasn't yet
                 * completed. */
-               if (inode)
-                       ocfs2_set_buffer_uptodate(inode, bh);
+               ocfs2_set_buffer_uptodate(inode, bh);
        }
-       if (inode)
-               mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
+       mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
 
        mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s, flags=0x%x\n", 
             (unsigned long long)block, nr,
-            (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes", flags);
+            ((flags & OCFS2_BH_IGNORE_CACHE) || ignore_cache) ? "no" : "yes",
+            flags);
 
 bail: