Re: Bill Gates slow burns over control-alt-delete -- funny video

On Fri, 25 May 2007, in the Usenet newsgroup comp.os.linux.misc, in article
<135cnpdnu7rj8af@xxxxxxxxxxxxxxxxxx>, Grant Edwards wrote:

On 2007-05-25, AZ Nomad <aznomad.2@xxxxxxxxxxxxxxxxxxx> wrote:

control/alt/del was created by IBM in their design of the original IBM PC.

Correct

The keyboard controller would generate a reset and the
operating system hadn't the slightest clue any of this was
taking place. Or perhaps it was a maskable interrupt -- not
sure about that.

It wasn't a "/RESET" - the pin is on the i8042, but is an input signal
only. The 'control/alt/del' is just another key sequence, and this
only generates the standard "INT 1" interrupt. What this interrupt IF
RECOGNIZED does is totally dependent on the keyboard handler sub-routine.
In DOS and windoze 1.0 to 3.11, this caused the system to jump to a
section of the boot sequence after the RAM and hardware detection tests
(a so-called "warm boot"). Other operating systems may have done other
things on detecting the key-sequence. For one example, you need only
read the inittab(5) man page, where you find

ctrlaltdel
The process will be executed when init receives the
SIGINT signal. This means that someone on the sys-
tem console has pressed the CTRL-ALT-DEL key combi-
nation. Typically one wants to execute some sort of
shutdown either to get into single-user level or to
reboot the machine.

The mystical bullsh!t that the key sequence can't be ignored (and thus
can be used to bypass keystroke loggers, etc.) and is therefore used in
windoze to directly attract the attention of the O/S is demonstrably
false. People really have had systems wedge (because they are ignoring
the interrupts) so hard that the 'control/alt/del' (indeed _any_ key
strokes) does nothing. In older PC hardware, the solution is to press
the front panel 'RESET' switch which is normally hardwired to that
/RESET signal on the motherboard. If that pin is pulled low for more
than a minimum time (tens of CPU clock cycles), then the CPU is reset
to the 'power on reset state'. The /RESET signal also goes to most
peripheral devices (plug in cards, disk drives, etc) including the 8042
or it's modern replacement.

The first I heard of any OS intercepting it was OS/2 shutting
down the file system during such a "reset" around '93.

Then it most certainly isn't a HW reset. There's simply no way
to intercept a HW reset.

The /RESET pin? Correct. Likewise, the system can't bypass the NMI
(Non-Maskable-Interrupt), although what the software handler that
services the NMI may be altered. In the original IBM designs, the NMI
was invoked in the event of a RAM parity error (IBM believing it better
to halt the computer than to continue with possibly corrupt data or
commands - Apple didn't see it that way and didn't use parity RAM).
Some may recall a Linux warning message that reads:

#ifndef CONFIG_IGNORE_NMI
printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n");
printk("You probably have a hardware problem with your RAM chips or a\n");
printk("power saving mode enabled.\n");
#endif

There were probably also TSR programs such as sidekick that
could intercept it, but it sure as hell wasn't MSDOS.

No, the keyboard interrupt handler merely passed the key code to
whatever was highest on the food chain - if that was a TSR, then the TSR
looked at the code to see if it was one of it's combinations. If it was
not, it was passed to the application you were running, which did the
same thing - eventually reaching DOS, which found that the vector for
this code was to start the warm boot sequence.

Hit control/alt/del on an MSDOS system and it was like hitting reset.

As noted above - if the interrupts were masked by a runaway program, or
if the key combination were intercepted by an application, it could or
would be ignored/missed.

Old guy
.

Relevant Pages

  • Re: Softkicking on a DKB Cobra/1240
    ... KillKick that just zeroes the mapped fast ram area and it does kill the ... you don't know what the ROM code of the DKB does. ... don't know what happens with the mapping RESET has been applied. ... JMP -> Direct jump into new ROM ...
    (comp.sys.amiga.programmer)
  • Re: Black box flight recorder for Linux
    ... Some people used to write the oopses into video memory, ... I wouldn't think most BIOSes these days would bother to clear system RAM ... Reset button is like a cold boot and it generally ends up with cleared ...
    (Linux-Kernel)
  • Re: dmesg: PCI interrupts are no longer routed automatically.........
    ... the ISR reads all the INTCSR ... If the PLX had been reset, then the INTCSR bits would all ... available before the chip could generate an interrupt. ... send the line "unsubscribe linux-kernel" in ...
    (Linux-Kernel)
  • Re: R: hc12,can i return from interrupt and jump elsewhere?
    ... then you can use a watchdog reset. ... reception of an interrupt event, ... .I'm trying to implement an USB protocol using a 9s12 microcontroller,that i ... It is raccomended that anytime an USB device receives a bus reset it has to ...
    (comp.arch.embedded)
  • Re: Bug in Synplify?
    ... does your target even support async reset on RAMs? ... I have a code which Synplify detects as Ram in compile step. ... set of pure registers (either due to resource usage, ... The example below should fit in 3 bit counter and 6x8 FF with asynch ...
    (comp.arch.fpga)
Loading