netdev
[Top] [All Lists]

Re: [PATCH]: Tigon3 new NAPI locking v2

To: "David S. Miller" <davem@xxxxxxxxxxxxx>
Subject: Re: [PATCH]: Tigon3 new NAPI locking v2
From: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Date: Thu, 16 Jun 2005 21:37:32 +1000
Cc: netdev@xxxxxxxxxxx, mchan@xxxxxxxxxxxx
In-reply-to: <20050603.122558.88474819.davem@davemloft.net>
References: <20050603.122558.88474819.davem@davemloft.net>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.9i
On Fri, Jun 03, 2005 at 07:25:58PM +0000, David S. Miller wrote:
> 
> This version incorporates two bug fixes from Michael.
> 
> 1) Check the mailbox register for 0x1 while polling on the COMPLETE
>    state bit.
> 
> 2) Remove the BUG_ON() check in tg3_restart_ints(), it can legally and
>    harmlessly occur.
> 
> Point #2 may want some refinements, but this patch below is good
> enough for testing.

Nice work Dave.

I was thinking of how we could avoid waiting for the interrupt to
occur after setting SYNC.  Here is one way which is essentially
a hand-coded spin lock.  In fact with a bit of work we could convert
it back to a real spin lock with spin_trylock.

The advantage of this is that we won't have to rely on the interrupt
to occur after setting SYNC.  The disadvantage is that on certain
architectures (sparc64 obviously :) we're now doing the relatively
expensive bit operations on each IRQ.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Attachment: p
Description: Text document

<Prev in Thread] Current Thread [Next in Thread>