netdev
[Top] [All Lists]

Re: Prism54 2.6 kernel update

To: Jeff Garzik <jgarzik@xxxxxxxxx>
Subject: Re: Prism54 2.6 kernel update
From: mcgrof@xxxxxxxxxxxxxxxxxxxx (Luis R. Rodriguez)
Date: Tue, 4 May 2004 16:41:39 -0400
Cc: Margit Schubert-While <msw@xxxxxxxxxxxx>, Margit Schubert-While <margitsw@xxxxxxxxxxx>, Netdev <netdev@xxxxxxxxxxx>
In-reply-to: <4097DCC8.5070301@xxxxxxxxx>
Organization: Rutgers University Student Linux Users Group
References: <5.0.2.1.2.20040429111349.00afb9c0@xxxxxxxxxxx> <4097DCC8.5070301@xxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mutt/1.3.28i
On Tue, May 04, 2004 at 02:11:20PM -0400, Jeff Garzik wrote:
> Margit Schubert-While wrote:
> >2004-04-26      Margit Schubert-While <margitsw@xxxxxxxxxxx>
> >
> >        * islpci_mgt.h : Replace init_wds with a define.
> >        The compiler does not optimize it out (and also generates the
> >        field in the ro section of every module)
> >
> >        * prismcompat(24).h : Include linux/compiler.h
> >        Now we can play with the likely/unlikely macros
> >
> >        * islpci_eth.c, islpci_dev.c : Align skb->data unconditonally
> >        after allocation.  This would appear to improve RX rate.
> >        Do a little bit of likely/unlikely.
> 
> 
> You are violating a standard Linux development rule --
> 
> You MUST send multiple patches, one change per patch.
> 
> "One big patch" is only acceptable the first time, when the driver is 
> initially merged.

Jeff, 

we'll work on sending separate patches. Luckily I tagged things on our
cvs repository since the 2.6 merge <-> our tree matchup so this
shouldn't be that bad.

If I don't get to it, Margit will. 

        Luis
-- 
GnuPG Key fingerprint = 113F B290 C6D2 0251 4D84  A34A 6ADD 4937 E20A 525E

Attachment: pgp4DZUTmREq4.pgp
Description: PGP signature

<Prev in Thread] Current Thread [Next in Thread>
  • Re: Prism54 2.6 kernel update, Luis R. Rodriguez <=