On 20 Nov 2001, W. Paul Mills wrote:
> [EMAIL PROTECTED] (Bostjan Muller) writes:
>
> > * On 20-11-01 at 10:40 Emil Pedersen ([EMAIL PROTECTED]) 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, [EMAIL PROTECTED]
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
