[PATCH]Erase block header(revision 1)

[Jörn Engel]

On Mon, 26 September 2005 16:02:50 +0800, zhao forrest wrote:
> 
> >> +struct jffs2_eraseblock_header
> >> +{
> >> +	jint16_t magic;
> >> +	jint16_t nodetype; /* == JFFS2_NODETYPE_ERASEBLOCK_HEADER */
> >> +	jint32_t totlen;
> >> +	jint32_t hdr_crc;
> >> +	uint8_t  fs_version; /* the version of this JFFS2 fs image */
> >
> >Version is completely useless.
> 
> I agree that version field is not used in my patch after compat_fset,
> incompat_fset and rocompat_fset are introduced in my patch.
> But I'm not sure if we should keep this field. 
> Artem,
> What's your opinion about this?

Just give it a useful name like "reserved" or "unused".  We still need
the alignment.

> >> +	uint8_t  compat_fset;
> >> +	uint8_t  incompat_fset;
> >> +	uint8_t  rocompat_fset;

What does "fset" mean?  Why not just "flags_compat" etc.?

> >> +	jint32_t erase_count; /* the erase count of this erase block */
> >> +	jint16_t dsize; /* the size of additional data behind node_crc */
> >
> >Breaks alignment and pretty useless.  Information can be deduced from
> >totlen.  Remove.
> >
> Sometimes we can't deduce the exact additional data size from totlen.
> In particular this is true for NOR ECC. For example, sizeof(struct
> jffs2_eraseblock_header) is 34 bytes; totlen is 40 bytes, dsize is
> 0 byte.

Even then it doesn't matter.  We just copy all fields and ignore the
ones we don't know about.  What we know about is defined in the three
flags fields.

Jörn

-- 
Simplicity is prerequisite for reliability.
-- Edsger W. Dijkstra