The potential for breakage is there, but I've tested just about everything
I can think of for now.  Ideally there should not be any user visible

In a nutshell:
- several poorly thought out API's for doing some stuff that cam etc needed
has been removed with extreme prejudice (resource_locate() and
resource_query_*() etc) and replaced with resource_find_*() which are a lot
cleaner and more flexible.  Use of these was quite minimal fortunately.
- the backing databases are now the raw string tables.  This means we can
get access to the hints well before malloc() is running.
- resource_get_*() is unchanged, but the implementation is a LOT better, for
example, if you ask for the value of "" as a string, you
will get a string "1" instead of an EINVAL.  If you ask for it as an integer
you will get the integer 1.  This eliminates the evil heuristics to try and
guess what was a string and what was an integer.
- resource_set_*() is currently broken.  For now, use 'set' and 'show'
at the command line to change device wiring instead of userconfig.  Note
that you can override static fallback hints.
This needs a working setenv() for the kernel, should anybody feel energetic.
- because of resource_set_(), this means that old style 'sensitive' isa
drivers cannot set their 'sensitive' flag.  This affects: "rdp"

What could go wrong?  I have had a couple of infinite loops that were
caused by mistakes during the API conversion, that I have since fixed.  I
dont believe there are any more.

"All of this is for nothing if we don't go to the stars" - JMS/B5

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Reply via email to