fsid_t

[Poul-Henning Kamp]

In message <xzpvfwf6j4k.fsf at flood.ping.uio.no>, Dag-Erling Smorgrav writes:
>fsid_t is currently defined as a struct containing an array of two
>32-bit ints:
>
>typedef struct fsid { int32_t val[2]; } fsid_t; /* filesystem id type */
>
>which is ridiculous as the only place where this is actually useful is
>when it is initialized (val[0] is set to the udev_t and val[1] to the
>vfs type number)

And we should be very careful with the udev_t thing, since udev_t's
are not very constant across reboots.

>Are there any objections to making fsid_t a uint64_t?

no objection from here.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.