[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: RealTek RTL8129/8139 Fast Ethernet driver for kernel 2.4.14



On 20 Nov 2001, W. Paul Mills wrote:

> neonatus@neonatus.net (Bostjan Muller) writes:
> 
> > * On 20-11-01 at 10:40 Emil Pedersen (emil.pedersen@its.uu.se) wrote:
> > +----Here quoted text begins----+
> > [...]  
> > > I think it's important to be very precise what chip is used;  I've heard
> > > lots of trouble with the chip '8139C' (note the trailing 'C') while I
> > > use three cards based on '8139' (no C) for a long time without any
> > > problem.
> > How does one know which chip he has.. I have allied telesyn and some
> > other noname card, that use the rtl8139 chips.. lspci says this:
> > 00:0d.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
> > 00:14.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
> > 00:11.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
> > (I suppose that is not the rtl8139C class)
> 

With the 8139too driver, I got time-outs.
I compiled & tried the rtl8139 driver, and that one also ends in
time-outs, but what's worse is that it doesn't reset itself. It just keeps
cut off from the network.

About why some people have problems and others not: 
Maybe it depends on the network load? I have an rtl8139 attached to my
cable modem, and I never got problems with that one.

The one in the server gets into trouble a few times a day [maybe during
heavy network traffic?], and if it does it just resets and continues.


The one I don't have problems with: (but isn't attached to a busy network)
  kernel: 2.4.14, 8139too driver
  00:09.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
        Subsystem: Realtek Semiconductor Co., Ltd. RT8139
Index #1: Found a RealTek RTL8139 adapter at 0xde00.
Decoded EEPROM contents:
   PCI IDs -- Vendor 0x10ec, Device 0x8139.
   PCI Subsystem IDs -- Vendor 0x10ec, Device 0x8139.
   PCI timer settings -- minimum grant 32, maximum latency 64.
  General purpose pins --  direction 0xe1  value 0x12.
  Station Address 00:00:1C:D0:FE:A0.
  Configuration register 0/1 -- 0x4c / 0xc2.


//======================================================================
The one that gives problems (both 8139too & rtl8139 driver)
(the card was a genius M.. and something ;-)
 kernel: 2.4.13 
 00:08.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139 (rev 10)
        Subsystem: Realtek Semiconductor Co., Ltd. RT8139

 rtl8139-diag:

 Index #1: Found a RealTek RTL8139 adapter at 0xd800.
Decoded EEPROM contents:
   PCI IDs -- Vendor 0x10ec, Device 0x8139.
   PCI Subsystem IDs -- Vendor 0x10ec, Device 0x8139.
   PCI timer settings -- minimum grant 32, maximum latency 64.
  General purpose pins --  direction 0xe1  value 0x12.
  Station Address 00:C0:DF:0E:F6:38.
  Configuration register 0/1 -- 0x8d / 0xc2.

Errors look like this:
With the rtl8139driver:

rtl8139.c:v1.16a 11/13/2001 Donald Becker, becker@scyld.com.
 http://www.scyld.com/network/rtl8139.html
eth0: RealTek RTL8139C+, 64 bit high performance at 0xd800, IRQ 11,
00:c0:df:0e:f6:38.

eth0: Transmit timeout, status 0d 0000 media 08.
eth0: Tx queue start entry 270360  dirty entry 270356, full.
eth0:  Tx descriptor 0 is 0008203c. (queue head)
eth0:  Tx descriptor 1 is 0008203c.
eth0:  Tx descriptor 2 is 10082436.
eth0:  Tx descriptor 3 is 40084436.
eth0: MII #32 registers are: 1000 782d 0000 0000 01e1 0000 0000 0000.

With the 8139too driver:
8139too Fast Ethernet driver 0.9.20
eth0: RealTek RTL8139 Fast Ethernet at 0xd800, 00:c0:df:0e:f6:38, IRQ 11
eth0:  Identified 8139 chip type 'RTL-8139C'

NETDEV WATCHDOG: eth0: transmit timed out
eth0: Tx queue start entry 2426  dirty entry 2422.
eth0:  Tx descriptor 0 is 00002000.
eth0:  Tx descriptor 1 is 00002000.
eth0:  Tx descriptor 2 is 00002000. (queue head)
eth0:  Tx descriptor 3 is 00002000.
eth0: Setting half-duplex based on auto-negotiated partner ability 0000.


  In any case, both cards work. I can live with about one time-out /
60min... The strange thing is that the new driver says it's a C+ (haven't
tried the C+ driver from 2.4.x yet)

Hope this was of some use to somebody ;-)

Dries Kimpe




Reply to: