[tulip] Careful with Conexant LANfinity and 2.4

[Marco Mellia]

On Wednesday 25 July 2001 22:37, you wrote:
> > Donald,
> >
> > 	Any ideas why I'm getting
> > kernel: Warning: kfree_skb on hard IRQ d09df9d0
> > 	?
> > 	This is something to worry about. I just got
> > system lockup and disk corruption. So I do not
> > recommend testing with this on a production system.
>
> Here is the patch to fix the first problem.
> I'm not certain that the second problem is related.
>
> retrieving revision 1.60
> diff -u -r1.60 tulip.c
> --- tulip.c	2001/07/10 02:04:42	1.60
> +++ tulip.c	2001/07/25 20:33:30
> @@ -24,7 +24,7 @@
>
>  /* These identify the driver base version and may not be removed. */
>  static const char version1[] =
> -"tulip.c:v0.92w 7/9/2001  Written by Donald Becker <becker@scyld.com>\n";
> +"tulip.c:v0.92wa 7/11/2001  Written by Donald Becker
> <becker@scyld.com>\n"; static const char version2[] =
>  "  http://www.scyld.com/network/tulip.html\n";
>
> @@ -2740,7 +2740,7 @@
>  				}
>
>  				/* Free the original skb. */
> -				dev_free_skb(tp->tx_skbuff[entry]);
> +				dev_free_skb_irq(tp->tx_skbuff[entry]);
>  				tp->tx_skbuff[entry] = 0;
>  			}

I applied the patch... and the irq warning disappeared... but  now, after a 
given number of pkt sent I get other messages, like

eth0: Out-of-sync dirty pointer, xxxx vs. yyyy, full=1.

where xxxx smaller than yyyy

here is a portion of th dmesg output:
---------------------------------------------
tulip.c:v0.92wa 7/11/2001  Written by Donald Becker <becker@scyld.com>
  http://www.scyld.com/network/tulip.html
eth0: Conexant LANfinity rev 8 at 0xcc894000, 00:50:8B:FA:14:C9, IRQ 9.
eth0:  MII transceiver #1 config 1000 status 782d advertising 01e1.
eth0:  MII transceiver #0 config 1000 status 782d advertising 01e1.
eth0: Out-of-sync dirty pointer, 343 vs. 360, full=1.
eth0: Out-of-sync dirty pointer, 360 vs. 377, full=1.
eth0: Out-of-sync dirty pointer, 377 vs. 414, full=1.
eth0: Out-of-sync dirty pointer, 393 vs. 414, full=1.
eth0: Out-of-sync dirty pointer, 414 vs. 445, full=1.
[...]
eth0: Out-of-sync dirty pointer, 5380 vs. 5411, full=1.
eth0: Out-of-sync dirty pointer, 5414 vs. 5435, full=1.
---------------------------------------------

Here is the output of ifconfig
---------------------------------------------
eth0      Link encap:Ethernet  HWaddr 00:50:8B:FA:14:C9  
          inet addr:130.192.9.154  Bcast:130.192.9.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:11160 errors:0 dropped:0 overruns:0 frame:0
          TX packets:1746 errors:0 dropped:0 overruns:0 carrier:0
          collisions:1043 txqueuelen:100 
          RX bytes:10589425 (10.0 Mb)  TX bytes:121019 (118.1 Kb)
          Interrupt:9 Base address:0x4000 
---------------------------------------------

and here is a dump of the ftp that caused the problem (I had to abort it 
because there were too many packet dropped...)
This is an ftp on the local area network...

---------------------------------------------
ftp> get jre.xpi
local: jre.xpi remote: jre.xpi
200 PORT command successful.
150 Opening BINARY mode data connection for jre.xpi (15514743 bytes).
 
receive aborted
waiting for remote to finish abort
426 Transfer aborted. Data connection closed.
226 Abort successful
9630480 bytes received in 52.9 secs (1.8e+02 Kbytes/sec)
---------------------------------------------


-- 
Ciao,                    /\/\/\rco

+-----------------------------------+  
| Marco Mellia - Assistant Professor|
| Tel: 39-011-5644173               |   /"\  .. . . . . . . . . . . . .
| Politecnico di Torino             |   \ /  . ASCII Ribbon Campaign  .
| Corso Duca degli Abruzzi 24       |    X   .- NO HTML/RTF in e-mail .
| Torino - 10129 - Italy            |   / \  .- NO Word docs in e-mail.
| http://www1.tlc.polito.it/mellia  |        .. . . . . . . . . . . . .
+-----------------------------------+
The box said "Requires Windows 95 or Better." So I installed Linux.