Subject: Re: kernel hang on boot/i386, atapi
To: Jason Thorpe <thorpej@nas.nasa.gov>
From: Manuel BOUYER <bouyer@antioche.lip6.fr>
List: current-users
Date: 09/19/1997 18:34:47
On Sep 12, Jason Thorpe wrote
> It seems like the semantics for dealing with ata vs. atapi interrupts are
> wrong. Actually, the return value semantics are wrong.
>
> Interrupt routines are supposed to return:
>
> 1 Interrupt was ours.
> 0 Interrupt was not ours.
> -1 Interrupt might not be ours.
>
[ I've finally got some time to took a this ]
When you says "ours" or "not ours", you mean that the driver was able to
say if his device delivered the interrupt or not ?
In wdc, I should always return 1 after wdc_ata_intr or wdc_atapi_intr (because
i know the driver was waiting an interrupt at this point) and -1 if the driver
didn't wait an interrupt. Rigth ?
--
Manuel Bouyer, LIP6, Universite Paris VI. Manuel.Bouyer@lip6.fr
--