rl(4) is broken in STABLE

[Marco Trentini]

Bill Paul wrote:
>>Bill Paul wrote:
>>
>>
>>>Aw come on guys! Give me a hint here! I can't even find this
>>>mysterious PR that was allegedly filed! Since you don't know the number,
>>
>>I've finally got the number of this PR.
>>Look: http://www.freebsd.org/cgi/query-pr.cgi?pr=55727
>>
>>Eugene Grosbein
>>
> 
> 
> You didn't provide enough information. Unfortunately, the only way
> to diagnose this properly is with tcpdump:
> 
> # tcpdump -n -e -i rl0 (or rl1, or whatever)
> 
> Look at the incoming packets, see if they appear sane. If there are
> no packets at all, then you need to report this as well. Also, you
> should show me what "ifconfig -a" says. Note: don't just tell me
> that it "looks normal." I need to see the raw data to interpret it
> myself.
> 
> You should also check netstat -in to see if the input/output packet
> and error counts are increasing.
> 
> Checking the transmit side requires running tcopdump on a remote host.
> (Running tcpdump locally only shows you packets going into the TX
> routine of the driver. You can't actually tell that the frames have
> made it onto the wire unless you scan from another machine.)
> 
> The 8139D isn't treated in any special way. Other people have reported
> that 'normal' 8139 NICs work again. (The commit on August 18th fixed
> the problem I introduced on August 17th.)
> 
> I don't have one of these cards, so I'm relying on you people to provide
> detailed information. All of the other 8139 cards I've tested seem to
> work, however I only have cards with the original 8139 and 8139A chips.
> The code path for the 8139D should be the same as for all the others.
> The only ones that are different are the 8139C+ and the 8169. There's
> a slim chance the D-Link card is being mis-identified, but to know
> that for sure, it NEED you to show me the output of "ifconfig -a."

I have checked two impeached cards on one STABLE (RELENG_4) of 
the 21 Aug. The cards are: RTL8139D and D-Link DFE-528TX 
(recognized as one D-Link DFE-530TX+ 10/100BaseTX).
Result: the cards work *correctly* (at least on my system).

uname -v
FreeBSD 4.8-STABLE #0: Thu Aug 21 17:29:19 CEST 2003 
root at einstein.lab:/usr/obj/usr/src/sys/EINSTEIN

dmesg.boot
....
rl0: <D-Link DFE-530TX+ 10/100BaseTX, rev. 8139D/8100B/8100C> 
port 0xd800-0xd8ff mem 0xd4000000-0xd40000ff irq 11 at device 
13.0 on pci0
rl0: Ethernet address: 00:40:05:0d:46:99
miibus0: <MII bus> on rl0
rlphy0: <RealTek internal media interface> on miibus0
rlphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
...

and (RTL8139D)

...
rl0: <RealTek 8139 10/100BaseTX, rev. 8139D/8100B/8100C> port 
0xd800-0xd8ff mem 0xd4000000-0xd40000ff irq 11 at device 13.0 on pci0
rl0: Ethernet address: 00:08:a1:2a:b4:30
miibus0: <MII bus> on rl0
rlphy0: <RealTek internal media interface> on miibus0
rlphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
...

ifconfig -a
rl0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
	inet 192.168.168.2 netmask 0xffffff00 broadcast 192.168.168.255
	ether 00:40:05:0d:46:99
	media: Ethernet autoselect (100baseTX <full-duplex>)
	status: active
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 16384
	inet 127.0.0.1 netmask 0xff000000

and (RTL8139D)

rl0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
	inet 192.168.168.2 netmask 0xffffff00 broadcast 192.168.168.255
	ether 00:08:a1:2a:b4:30
	media: Ethernet autoselect (100baseTX <full-duplex>)
	status: active
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 16384
	inet 127.0.0.1 netmask 0xff000000

netstat -in
Name  Mtu   Network       Address            Ipkts Ierrs    Opkts 
Oerrs  Coll
rl0   1500  <Link#1>    00:40:05:0d:46:99      671     0      757 
     0     0
rl0   1500  192.168.168   192.168.168.2        670     -      751 
     -     -
lo0   16384 <Link#2>                             7     0        7 
     0     0
lo0   16384 127           127.0.0.1              7     -        7 
     -     -

and (RTL8139D)

Name  Mtu   Network       Address            Ipkts Ierrs    Opkts 
Oerrs  Coll
rl0   1500  <Link#1>    00:08:a1:2a:b4:30        5     0        6 
     0     0
rl0   1500  192.168.168   192.168.168.2          4     -        4 
     -     -
lo0   16384 <Link#2>                             2     0        2 
     0     0
lo0   16384 127           127.0.0.1              2     -        2 
     -     -

Most likely in the last check I don't use the last revision of 
the if_rl.c.



-- 
Marco Trentini                mark at remotelab.org
http://www.remotelab.org/