From: obz@sisd.kodak.com (Orest Zborowski COMP) Subject: Re: changing to another virtual console... Date: Wed, 22 Jul 1992 20:55:54 GMT
davidsen@ariel.crd.GE.COM (william E Davidsen) writes:
>In article <1992Jul22.045943.925@sspiff.ampr.ab.ca>, dje@sspiff.ampr.ab.ca (Doug Evans) writes:
>
>| The SVR4/386 systems and SCO Xenix provide ioctl's to do this.
>| I think Linux should just follow suit and use those. Why add extra
>| pain to people porting applications that use the ones already in existence?
>
> The world is breaking down to BSD and V.4, and BSD is going away...
>please use the SVR4 standard rather than xenix, if they differ. I agree
>that reinventing the wheel is not the way to go.
>--
>bill davidsen, GE Corp. R&D Center; Box 8; Schenectady NY 12345
> It never ceases to amaze me that otherwise rational people, able to
> understand calculus, compound interest, and the income tax form, can
> continue to believe that poker is a game of chance.
i agree. i just read in unix world how svr4.2 is gonna change the face of
unix by being the standard. i dunno about that, but it is certainly true
that most flavors of unix (commercial, that is) are leaning towards svr4
at least in some fashion, and posix is closer to svr4 than bsd.
when i ported the x386 server i was faced with the svr4-isms built in, so
i grabbed the usl svr4 manual describing vt/kd handling and implemented
the bare minimum required by the server. the biggest piece that i omitted
was the vt switching.
under svr4, the kernel switching between vt's with a hot key, as is done
in linux. a process may ask the kernel to be notified if a switch is
pending, something they call process mode as opposed to auto mode. the
kernel is supposed to post a signal to the process, which must then
inform the kernel when it's ok to switch, if at all. the difficulty in
this is the bookeeping involved inside the kernel - which process has
control over what vt, what state are we in, etc, as well as keeping
track when this controlling process terminates. it's not too difficult,
just a bit tricky.
i pretty much copied the header information right from the manpages as
given in this book (the structures were given in pieces), but there wasn't
enough time for me to go back and fill in the details, so our x386 has
this stuff commented out.
please feel free to play with this construct as i don't have access to
a real svr4 machine, so i can't probe the limits of the implementation
(i.e. thats the problem with me fixing the text mode foobar - most svr4
kernels do nifty things which aren't documented too well in my book and
x386 assumes do happen, hence the problems in linux).
zorst
-- zorst (orest zborowski) [obz@raster.Kodak.COM]