Hi,
On 19/01/18 14:31, Andrew Price wrote:
On 18/01/18 16:04, Andreas Gruenbacher wrote:> diff --git
a/include/uapi/linux/gfs2_ondisk.h b/include/uapi/linux/gfs2_ondisk.h
index 09f0920f07e9..7eb73c32272f 100644
--- a/include/uapi/linux/gfs2_ondisk.h
+++ b/include/uapi/linux/gfs2_ondisk.h
@@ -403,7 +403,15 @@ struct gfs2_ea_header {
* Log header structure
*/
-#define GFS2_LOG_HEAD_UNMOUNT 0x00000001 /* log is clean */
+#define GFS2_LOG_HEAD_UNMOUNT 0x00000001 /* log is clean */
+#define GFS2_LOG_HEAD_FLUSH_NORMAL 0x00000002 /* normal log flush */
+#define GFS2_LOG_HEAD_FLUSH_SYNC 0x00000004 /* Sync log flush */
+#define GFS2_LOG_HEAD_FLUSH_SHUTDOWN 0x00000008 /* Shutdown log
flush */
+#define GFS2_LOG_HEAD_FLUSH_FREEZE 0x00000010 /* Freeze flush */
+#define GFS2_LOG_HEAD_RECOVERY 0x00000020 /* Journal recovery */
+#define GFS2_LOG_HEAD_USERSPACE 0x80000000 /* Written by
gfs2-utils */
+
+#define LH_V1_SIZE (offsetofend(struct gfs2_log_header, lh_hash))
struct gfs2_log_header {
struct gfs2_meta_header lh_header;
@@ -413,6 +421,21 @@ struct gfs2_log_header {
__be32 lh_tail; /* Block number of log tail */
__be32 lh_blkno;
__be32 lh_hash;
+
+ /* Version 2 additional fields start here */
+ __be32 lh_crc; /* crc32 of whole block with this field 0 */
+ __be32 lh_nsec; /* Nano second time stamp */
+ __be64 lh_sec; /* Second based time stamp */
+ __be64 lh_addr; /* Block addr of this log header
(absolute) */
+ __be64 lh_jinode; /* Journal inode number */
+ __be64 lh_statfs_addr; /* Local statfs inode number */
+ __be64 lh_quota_addr; /* Local quota change inode number */
+
+ /* Statfs local changes (i.e. diff from global statfs) */
+ __be64 lh_local_total;
+ __be64 lh_local_free;
+ __be64 lh_local_dinodes;
+ __be32 lh_log_origin; /* The origin of this log header */
};
4 bytes of padding gets added at the end of the struct. Could it be
made explicit with a __pad field? It currently breaks the metadata
description test with "gfs2_log_header: size mismatch between struct
128 and fields 124".
I've pushed a gfs2-utils patch with basic support for the new log
header format to the andyp-lhv2 branch. I need to test it some more
but it should be useful for testing these patches already. I'll
finalise it once the v2 patches are in for-next.
Andy
It could, however nothing follows the structure, so it isn't generally
needed, and I suspect that the test was expecting all the structures to
be of fixed size (which all the others are). Still, it might be the
easiest way to fix that test unless we can find another solution?
Steve.
