netdev
[Top] [All Lists]

Re: pktgen: causing lots of errors in console log

To: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>
Subject: Re: pktgen: causing lots of errors in console log
From: cliff white <cliffw@xxxxxxxx>
Date: Thu, 10 Mar 2005 10:48:30 -0800
Cc: Stephen Hemminger <shemminger@xxxxxxxx>, Robert Olsson <Robert.Olsson@xxxxxxxxxxx>, netdev@xxxxxxxxxxx
In-reply-to: <16942.55689.924299.906304@xxxxxxxxxxxx>
Organization: OSDL
References: <20050308140826.451435e5@xxxxxxxxxxxxxxxxx> <16942.55689.924299.906304@xxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
On Wed, 9 Mar 2005 12:10:01 +0100
Robert Olsson <Robert.Olsson@xxxxxxxxxxx> wrote:

> 
>  Hello!
> 
> 
> Stephen Hemminger writes:
>  > 1. Errors from IPV4
>  > Freeing alive in_device ffff81007e96a400
> 
> Try:
> 
> --- net/core/pktgen.c.orig    2005-03-09 10:15:01.000000000 +0100
> +++ net/core/pktgen.c 2005-03-09 10:24:19.000000000 +0100
> @@ -151,7 +151,7 @@
>  #include <asm/timex.h>
>  
>  
> -#define VERSION  "pktgen v2.58: Packet Generator for packet performance 
> testing.\n"
> +#define VERSION  "pktgen v2.59: Packet Generator for packet performance 
> testing.\n"
>  
>  /* #define PG_DEBUG(a) a */
>  #define PG_DEBUG(a) 
> @@ -1682,7 +1682,7 @@
>                                       pkt_dev->saddr_min = 
> in_dev->ifa_list->ifa_address;
>                                       pkt_dev->saddr_max = pkt_dev->saddr_min;
>                               }
> -                             in_dev_put(in_dev);     
> +                             __in_dev_put(in_dev);   
>                       }
>                       rcu_read_unlock();
>               }
> 
> 
> 
>  
>  > 2. Cliff is getting this with e1000 and pktgen-conf-1-2
>  >  looks like calling schedule_timeout with lock held.
>  > 
>  
> Need to reproduce this one...

Okay,  I can reproduce it on two machines, no problem.
Please let me know what information you need.

I'm using 2.6.11 from bk, problem also happens on 2.6.10.
.config file is attached. 
I'm using the basic 'send 10 million packets to a second machine' example file. 
lspci for the cards:
0000:04:04.0 Ethernet controller: Intel Corp. 82557/8/9 [Ethernet Pro 100] (rev 
0c)
0000:04:0a.0 Ethernet controller: Intel Corp. 82557/8/9 [Ethernet Pro 100] (rev 
08)

-------------------dmesg ( again )
scheduling while atomic: test-local/0x00000001/3648
 [<c0366f02>] schedule+0x672/0x680
 [<c01891fa>] dnotify_parent+0x3a/0xb0
 [<c0367fa6>] _spin_lock+0x16/0x90
 [<c036843f>] _spin_unlock_irqrestore+0xf/0x30
 [<c01285d9>] __mod_timer+0xe9/0x130
 [<c0367822>] schedule_timeout+0x72/0xd0
 [<c0129190>] process_timeout+0x0/0x10
 [<c0367fa6>] _spin_lock+0x16/0x90
 [<e09926ea>] proc_thread_write+0x20a/0x2f0 [pktgen]
 [<c0367fa6>] _spin_lock+0x16/0x90
 [<c018c727>] proc_file_write+0x37/0x50
 [<c015ca4e>] vfs_write+0xae/0x130
 [<c015cba1>] sys_write+0x51/0x80
 [<c010290f>] syscall_call+0x7/0xb
-----------------------
cliffw

> 
> 
>                                       --ro
> 


-- 
"Ive always gone through periods where I bolt upright at four in the morning; 
now at least theres a reason." -Michael Feldman

Attachment: .config
Description: Text document

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