| To: | Mitchell Blank Jr <mitch@xxxxxxxxxx> |
|---|---|
| Subject: | Re: [RFC/PATCH] lockless loopback patch for 2.6 (version 2) |
| From: | Arthur Kepner <akepner@xxxxxxx> |
| Date: | Mon, 14 Jun 2004 12:45:25 -0700 |
| Cc: | netdev@xxxxxxxxxxx |
| In-reply-to: | <20040614182336.GC11280@gaz.sfgoth.com> |
| References: | <Pine.SGI.4.56.0406141000060.479900@neteng.engr.sgi.com> <20040614182336.GC11280@gaz.sfgoth.com> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
On Mon, 14 Jun 2004, Mitchell Blank Jr wrote:
> Arthur Kepner wrote:
> > +#define HARD_TX_LOCK_BH(dev, cpu) { \
> > + if ( dev->features && NETIF_F_LLTX == 0 ) { \
> ^^
>
> Don't you mean '&' instead of '&&' here? It looks like that condition is
> always false, so you've killed the TX locking for all devices with this
> patch.
Ummm, yes. (I believe the appropriate response here is "D'oh!") Thanks.
--
Arthur
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [RFC] ethtool semantics, Marc Herbert |
|---|---|
| Next by Date: | [Prism54] CVS -> bk tree update, Luis R. Rodriguez |
| Previous by Thread: | Re: [RFC/PATCH] lockless loopback patch for 2.6 (version 2), Mitchell Blank Jr |
| Next by Thread: | Re: [RFC/PATCH] lockless loopback patch for 2.6 (version 2), Andi Kleen |
| Indexes: | [Date] [Thread] [Top] [All Lists] |