Re: Re[2]: Wakeup from a PCI device

[Charles Curley ETW <ccurley@dwarf.fc.hp.com>]

-> Resent-From: pci-sig-request@znyx.com
-> Resent-Date: Wed, 14 Aug 96 13:21:00 PDT
-> Date: Wed, 14 Aug 96 13:21:00 PDT
-> From: Bruce Young <Bruce_Young@ccm.jf.intel.com>
-> X-Mailing-List: <pci-sig@znyx.com> archive/latest/3467
-> X-Loop: pci-sig@znyx.com
-> Precedence: list
-> Resent-Sender: pci-sig-request@znyx.com
-> 
-> 
-> Text item: 
-> 
-> Yes, the intent is to define a reserved pin for this purpose. It was the add-in 
-> card problem that drove us over the edge to saying we needed a separate pin. The
-> ECR process will address exactly which one. (But of course, no one has used any 
-> of these since that would not be compliant with the spec :-)
-> 
-> BTW, How would a disk controller use this pin? The semantic of the pin is that 
-> it is asserted when the device wants to wake up the system. It is an 
-> asynchronous, active low open drain signal shared between all PCI slots in the 
-> system with the system logic as the receiver of the signal. The system logic 
-> then wakes the system up and once the CPU is running the OS, the OS will then go
-> out and enable the resources it needs such as the disk interfaces.
-> 
-> -Bruce

A disk controller might be designed to *listen* on this line, and wake up
the drives if it is asserted. As disk drive power up can take several
seconds, this could help reduce the latency compared to waiting for the OS
to get around to waking up the drives.

-- 

		-- C^2

		         ___ __ /___ ___ 
			|   /  /_ _ \   |
			|  /  / // / \  |
			|  \ / //_/  /  |
			|___\_ /____/___|
			      /
			  H E W L E T T  
			  P A C K A R D  

Charles Curley is not an employee of Hewlett Packard.

This message does not necessarily represent the opinion(s) of HP.
,
(