I've tracked down the bug in loop device. The problem was that when loop device created the block
in sparse file it didn't update real_block and so it was writing block to block 0 of the device
(I'm glad that SB is in block 1 ;-). This reminds me if it wouldn't be possible to add a check to
fsck whether has anybody written anything to block 0 and warn about that... The patch I did should
correct the bug (at least now loopback works for me).
Honza.
PS: I wonder how this could ever work ;-)
----------- CUT ME HERE ----------
--- /usr/src/linux/drivers/block/loop.c Wed Jan 13 23:02:21 1999
+++ /usr/src/linux/drivers/block/loop.c Wed Jan 13 23:55:01 1999
@@ -225,6 +225,7 @@
if (!create_missing_block(lo, block, blksize)) {
goto error_out_lock;
}
+ real_block = bmap(lo->lo_dentry->d_inode, block);
}
}
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/