lkml.org 
[lkml]   [2004]   [Sep]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch] voluntary-preempt-2.6.9-rc1-bk4-R0
From
Date
On Fri, 2004-09-03 at 13:02, K.R. Foley wrote:
> OK. I previously reported about the system hanging trying to boot this
> patch. It may have been a fluke. In any event, it doesn't hang
> consistently. I have rebooted several times now without problems. I do
> however still get some of these:
>
> (ksoftirqd/0/2): new 131 us maximum-latency critical section.
> => started at: <netif_receive_skb+0x82/0x280>
> => ended at: <netif_receive_skb+0x1d7/0x280>

This is actually a good sign if this is the most common one you see.
The netif_receive softirq is basically the longest code path that can't
easily be made preemptible; on my hardware about 150 usecs seems to be
the best we can do. This is really good, I can't imagine an application
that you would use PC hardware for where 150 usecs is not enough, though
I am sure the hard realtime crowd has a few.

Even fixing this one would not help much because it seems there are many
other non-preemtible code paths that are only a little shorter than this
one.

Lee

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 14:05    [from the cache]
©2003-2011 Jasper Spaans