[JFFS2] Fix calculation of potential summary marker offset on NOR flash.
authorDavid Woodhouse <dwmw2@infradead.org>
Tue, 30 May 2006 07:59:34 +0000 (08:59 +0100)
committerDavid Woodhouse <dwmw2@infradead.org>
Tue, 30 May 2006 07:59:34 +0000 (08:59 +0100)
Helps if we look _inside_ the buffer, rather than adding jeb->offset to
it. Doh.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
fs/jffs2/scan.c

index 42c1ff21d3521acfa4ec35e668a5df7f2d9047c0..61618080b86f0a9bfe3202e39c8e452f13a21ec5 100644 (file)
@@ -463,7 +463,7 @@ static int jffs2_scan_eraseblock (struct jffs2_sb_info *c, struct jffs2_eraseblo
              
                if (!buf_size) {
                        /* XIP case. Just look, point at the summary if it's there */
-                       sm = (void *)buf + jeb->offset - sizeof(*sm);
+                       sm = (void *)buf + c->sector_size - sizeof(*sm);
                        if (je32_to_cpu(sm->magic) == JFFS2_SUM_MAGIC) {
                                sumptr = buf + je32_to_cpu(sm->offset);
                                sumlen = c->sector_size - je32_to_cpu(sm->offset);