svn commit: r198706 - head/sys/sys

[Ed Schouten]

Hi Robert,

* Robert Watson <rwatson at FreeBSD.org> wrote:
> The underlying change seems fine, especially in -CURRENT, but I'm
> confused by the paragraph here on alignment and safety.  What do you
> mean by safe, and does it matter?

With safe I mean that all useful members still begin at the same offset
within the structure. I'm not sure whether it matters at all, because I
can imagine d_devs could be moved around without having any effect, but
I'm not going to make any predictions on that.

> From a casual glance (perhaps mistaken), it looks like this changes
> the KBI, so all modules declaring struct cdevsw will need to be
> rebuilt.

It shouldn't, right?

-	uid_t			d_uid;
-	gid_t			d_gid;
-	mode_t			d_mode;
-	const char		*d_kind;
+
+	int32_t			d_spare0[3];
+	void			*d_spare1[3];

	/* These fields should not be messed with by drivers */
-	LIST_ENTRY(cdevsw)	d_list;

d_uid, d_gid and d_mode is equal to:

	int32_t		d_uid;
	int32_t		d_gid;
	int16_t		d_mode;

Because d_kind is a pointer, there will be a hole in the structure of at
least 2 bytes (maybe even 6), which means we can safely extend d_mode to
32 bits as well. I could have decided to leave it at uint16_t, but that
would only obfuscate it even more when we would try to reclaim some
space there.

d_list is just two pointers, so I merged it with d_kind into an array of
three pointers.

-- 
 Ed Schouten <ed at 80386.nl>
 WWW: http://80386.nl/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/svn-src-all/attachments/20091101/e1ea844e/attachment.pgp