What's the difference with any O/S with manual shutdown and start-upand restart shutdown and start-

K

KCB

Typically, a manual shutdown will turn off the power to the hardware, so you
have to press the power button to restart the machine. This can sometimes
reset some misbehaving part(s). The restart will just restart the computer
without powering off the hardware.
 
P

Paul

KCB said:
Typically, a manual shutdown will turn off the power to the hardware, so
you have to press the power button to restart the machine. This can
sometimes reset some misbehaving part(s). The restart will just restart
the computer without powering off the hardware.
To give an example of this, compare the IDE disk drive and the SATA disk drive.

The IDE disk drive, has a reset signal on the cable. This ensures, that if
you do a reboot, the IDE drive can be returned to a known state. When the
black screen first appears during the reboot, the hardware is reset.

http://pinouts.ru/HD/AtaInternal_pinout.shtml

The SATA drive lacks this feature. The data connection consists of
a TX pair and an RX pair. Unlike the ribbon cable, there isn't a lot
of extra signals to be used for side channel purposes. There
are still signaling means though. Apparently, just not good enough.

http://serialtek.com/sata_protocol_overview.asp

I've had a SATA drive, that would not respond on a reboot. It was
effectively frozen. When I turned off the power, and turned the power
on again, it worked. And it continues to work to this day. So
the techniques in the sata_protocol_overview, apparently aren't
enough to restore a drive to a working state.

The IDE drive would not have done that. And that is because it
has "sanity by design".

There are other parts of the computer that lack "sanity by design".
Certain chipsets cannot be fully reset, during a reboot. And if
there is a problem with them, they require power cycling as well.

Where I used to work, "sanity by design" was a tick box of every
design review. The designer had to demonstrate, how the hardware
could be recovered, without needing to switch the power off and on.
That was done for ASIC designs as well. In the examples above,
the designers would have "failed". And their designs would not
have gone "out the door". And you can imagine where such a
requirement came from - being embarrassed by previous design
errors :)

Our hardware designs could still suck - it's just that when you
pressed the reset button, they'd listen to you :)

Paul
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top