vc-switching/scroll back/console driver

Michael Weller (eowmob@exp-math.uni-essen.de)
Fri, 13 Oct 1995 11:24:57 +0100 (MEZ)


I agree that the vc-driver should be enhanced and that a minimal
framebuffer device needs to be in the kernel. That is just because the
console is a shared (multiple Xservers, svgalib applications, text
consoles) device like the harddisk. Someone has to manage that. All
statements in kernel guids that the vga is not shared and doesn't need to
be controlled by the kernel are horribly wrong. It is the most shared and
important device after the harddisk. A real non-shared device is the tty
used by Seyon, so remove that unneeded f* serial tty drivers from the
kernel NOW (just joking).

Other things are that the x86 memory paging and hardware-irq's would be
great to be used by grafix appl. What would really be needed is a
mechanism to reset to textmode on application death, virtualize vga mem
access and paging, irq's driven accelerator queues. vc-switching even back
and forth to dead or currently locked applications. (I've an idea on how
todo that w/o big screen buffers in kernel, with the only disadvantage of
the currently blocked application showing a black screen when switched
back to it until it wakes up again which seems senseable to me.). Device
driver level compatibility to linux-680x0, different rez and FONTS! on
vc's to allow high speed, say APL-3270 or real vt-100 (with correct fonts
that is) or IBM-DOS character set whatever), terminal emulations with own
fonts and the speed-advantages of a textmode screen device.

Well. I didn't continue to work on this for a while, not really due to
lack of time but also because I got tired of pointless religous discussion
about that topic. Thus I won't fight with anyone about above todo's.

What I only would like to ask for if someone intends to work on the vc-driver
to synchronize with me such that hooks for console managing modules as I
started coding might fit in there snuggly and things are done the right way.

Michael.

(eowmob@exp-math.uni-essen.de or eowmob@pollux.exp-math.uni-essen.de
Please do not use my vm or de0hrz1a accounts anymore. In case of real
problems reaching me try mat42b@aixrs1.hrz.uni-essen.de instead.)