From: Jouni Malinen (jkmaline_at_cc.hut.fi)
Date: 2002-07-20 09:13:30 UTC
On Sat, Jul 20, 2002 at 11:38:50AM +0300, Jouni Malinen wrote:
> I haven't noticed this on any of the systems I'm using (though, I
> haven't really looked for it either). Is there anything special about
> the system that sees this?
This was just because I was not looking.. Indeed there will be an interrupt with no real events after hfa384x_cmd().
> Some minor changes can be done by first reading resp0 and status to
> local variables and then ACK EV_CMD. After this, do res = local_resp0 &
> (bits) and the #ifndef final_version block before
> spin_unlock_irqrestore. However, this will have only a tiny delay
> between ACK and irq restore and might not help at all. How long delay
> you used to get rid of those extra interrupts?
I did some quick tests and this minor change did not have any noticeable efect as expected. I needed to add udelay(4) before irqrestore to get rid of the extra interrupts. However, as I said, this should not be a problem in any other case than possibly with power saving stations. Other RX/TX paths do not use hfa384x_cmd() and do not thus produce these extra interrupts.
-- Jouni Malinen PGP id EFC895FA