Re: Can a PCI Slave generate INTA# ?

["Weng Tianxiang" <wtx@umem.com>]

Hi,
1. If your device wants to generate an Interrupt, you are responsible for
writing an interrupt program to be installed into machine after system
configuration period and before system normal work. In Windows NT, 98, it
may include a sentence in your system.ini file to install your system driver
(another name in software to handle interrupts) at the right time.

2. If you don't write an interrup program while your device generates an
interrupt, the system has a default interrupt program expecting for any
system-unexpected interrupts that would blue your screen. In DOS, it only
skips the interrupt, but in Windows system, it shows you something that is
not expected.

Weng Tianxiang
Micro Memory Inc.
Chatsworth, CA 91311
Phone: 818-998-0070, Fax: 818998-4459


----- Original Message -----
From: Holt, Patrick <pholt@mc.com>
To: 'Ram Y. Gopal' <ramgo@vEngines.com>; <pci-sig@znyx.com>
Sent: Wednesday, August 02, 2000 6:55 AM
Subject: RE: Can a PCI Slave generate INTA# ?


> Morning,
> Since we are the subject of interrupts, I can generate an interrupt
> on all 4 interrupt lines by writting directly to the register of the
device
> on the PCI bus BUT I'm having problems handling the interrupts. They
> always blue screen my PC. Does anyone have any handlers for Win95, WinNT
> for hooking these interrupts.
> Thanks for any input,
> Patrick L. Holt
>
> -----Original Message-----
> From: Ram Y. Gopal [mailto:ramgo@vEngines.com]
> Sent: Wednesday, August 02, 2000 7:06 PM
> To: pci-sig@znyx.com
> Subject: Can a PCI Slave generate INTA# ?
>
>
> Hello:
>
> I would like to know if a PCI Slave device can generate an
> interrupt to the host processor on INTA# line. Are there
> any issues to be addressed?
>
> Please let me know if anyone has done a system design with
> such a requirement.
>
> Thanks in advance...
> --
> Ram Y. Gopal
>
>