Re: rtl8139 - kernel 2.4.3

[Martin Josefsson]

On Thu, 17 May 2001, Ted Gervais wrote:

> 
> I get the following when ftping from one workstation to another.
> Using kernel 2.4.3 and Redhat7.1:
> 
> Assertion failed! tp->tx_info[entry].skb == NULL,8139too.c,rtl8139_start_xmit,line=1676
> Assertion failed! tp->tx_info[entry].mapping == 0,8139too.c,rtl8139_start_xmit,line=1677
> Assertion failed! tp->tx_info[entry].skb == NULL,8139too.c,rtl8139_start_xmit,line=1676
> Assertion failed! tp->tx_info[entry].mapping == 0,8139too.c,rtl8139_start_xmit,line=1677
> Assertion failed! tp->tx_info[entry].skb == NULL,8139too.c,rtl8139_start_xmit,line=1676
> Assertion failed! tp->tx_info[entry].mapping == 0,8139too.c,rtl8139_start_xmit,line=1677
> eth0: Out-of-sync dirty pointer, 456 vs. 462.
> Assertion failed! tp->tx_info[entry].skb == NULL,8139too.c,rtl8139_start_xmit,line=1676
> Assertion failed! tp->tx_info[entry].mapping == 0,8139too.c,rtl8139_start_xmit,line=1677
> Assertion failed! tp->tx_info[entry].skb == NULL,8139too.c,rtl8139_start_xmit,line=1676
> Assertion failed! tp->tx_info[entry].mapping == 0,8139too.c,rtl8139_start_xmit,line=1677
> 
> 
> Is there a fix for this?  Kernel 2.4.4 is worse. It gives me a 'kernel
> panic'..  doing the same ftp transfer between workstations.

I think I suffer from the same problem. The machine was stable with
2.4.2-ac6 until I started playing with smbfs and hit a bug in that. 
So I upgraded to 2.4.4 and since then the machine has been very unstable.
Maximum uptime so far is 4 days and then it fell over.
And I'm using an rtl8139 card too.

I don't have a monitor attached to the machine but I'm compiling
2.4.4-ac10 (afraid of the LVM changes in -ac11) with the kmsgdump patch so
I'll probably get a stackdump sometime later today.

/Martin