[parisc-linux-cvs] linux-2.6 willy

[Matthew Wilcox]

On Sat, Jan 01, 2005 at 01:24:55AM -0700, Grant Grundler wrote:
> Yeah - I certainly didn't feel like I understood the whole scheme at the time.
> It also wasn't documented anywhere or as well developed as it is today.
> 
> I still think the REVISIT_DESIGN_ISSUE comment was accurate though: one could disable an IRQ
> but not specify which user of a shared IRQ was to be disabled.

Right, you disable an interrupt, not an interrupt handler.  The generic
IRQ code supports nested disable/enable calls, precisely because of this.
Our code does the wrong thing, which is one of the reasons I'm looking
to get rid of it.

> Ditto for mask/unmask calls
> that have been recently removed.

I determined they weren't necessary -- in my latest development tree,
I deleted the enable/disable_irq calls that I replaced them with in the
current tree.  Haven't booted that change yet, but it's worth noting that
dino has *never* masked/unmasked irqs; only wax/lasi/asp do.

> Anyway, I've gotten more than half way through the first diff and clearly
> the more interesting bits are still to come. :^)
> And so far it looks good to me.

Yes, much more to do yet.  I'm trying to commit small pieces, each of
which is an improvement.  That way if other things come up, at least the
next person to tackle this Augean stable has a better place to start from.

Eventually, I hope to be able to create kernel/irq/iosapic.c and
share it between ia64, parisc and i386, but that's a fair way off yet.
Next step is to sort out the gsc_asic interrupt handlers to not require
bit reversal.

-- 
"Next the statesmen will invent cheap lies, putting the blame upon 
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince 
himself that the war is just, and will thank God for the better sleep 
he enjoys after this process of grotesque self-deception." -- Mark Twain