Hi Wolfram; On Thu, 2008-07-10 at 18:52 +0200, Wolfram Sang wrote: > Hi William, > > On Thu, Jul 10, 2008 at 10:49:23AM -0400, William Case wrote: > > > The tables often have pointers to each other in order to build > > up a full process. Some of the most important tables are; the > > inode, the file descriptor, the socket, the process, the > > interrupt handler, the signal handler and the error handler > > tables. (If I am on the right track, but have forgotten an > > important table please tell me). > > What I would add here are device drivers. They are also collected in a > table, with function pointers to offer their functionality. Still, when > dealing with hardware at this level, I think a reduction to a "table > entry" is not enough. There is quite some logic involved, which is > mostly pure code to handle states of the hardware. It is the end of all > abstractions :) > > In general, your description sounds a bit as if code is > underrepresented, but probably you don't mean it that way. The general > approach should hold, IMHO. > > All the best, > > Wolfram You are absolutely correct; a drivers table has to be added to the list. No, I am not ignoring the importance of code. But, it seems to me, if a programmer has the right table in mind, he can get to it and follow the requisites of the hardware and the code from there. Put another way, tables seem to be the pivotal point for any (most) kernel processes. -- Regards Bill; Fedora 9, Gnome 2.22.3 Evo.2.22.3.1, Emacs 22.2.1 -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
