Re: [ECOS] PCI Ethernet card

Re: [ECOS] PCI Ethernet card

[Raghu BK]

Hi Mike/Eric,

I've a custom board based on XScale 255, PCI Bridge IT8152 and RTL8139. I'm 
in the process of porting RedBoot to this platform. I would really 
appreciate if you could share the RTL8139 controller driver with me.

Mike: Since you already ported Eric's driver to Redboot, could you please 
share the driver code you ported with me? This would cut short my 
development time and your driver would be validated on one more platform. 
I'll keep you folks updated on the status.

Thanks in Advance,
Raghu


>From: Eric Doenges <Eric.Doenges@DynaPel.com>
>To: Michael Anburaj <embeddedeng@hotmail.com>
>CC: Eric.Doenges@DynaPel.de, 
>gary@mlbassoc.com,ecos-discuss@sources.redhat.com
>Subject: Re: [ECOS] PCI Ethernet card
>Date: Wed, 20 Aug 2003 09:49:02 +0200
>
>Michael Anburaj wrote:
>
>>I found the problem & fixed it.
>
>Great !
>
>>
>>Problem:
>>-----------
>>
>>I displayed the value of ?header? each time after the following line of 
>>code at rltk8139_deliver():
>>
>>header = HAL_LE32TOC(*((cyg_uint32 *)&rltk8139_info->rx_ring[rx_pos]));
>>
>>And it was most of the time = 0x0.
>>I initially guessed that probably the Rx_Ring_buffer was not setup right. 
>>But, the actual problem was with the caches < cache coherency -same issue 
>>that Gary Thomas addressed in one of the emails sent to ecos-patches >. 
>>So, I pulled revision 3 of your driver & also defined 
>>CYGPKG_DEVS_ETH_RLTK_8139_SOFTWARE_CACHE_COHERENCY in my platform-specific 
>>.inl file. It worked well & got the IP parameters in 1 shot from the DHCP 
>>server. Right now it?s configured as eth0 interface, I will try making it 
>>eth1. Thanks a lot to you and all who participated in this discussion.
>>
>>FYI: I exchanged my PCI 2.2 compliant <with WOL> NIC with a PCI 2.1 one, 
>>but having the same chip - RTL8139D.
>
>It's nice to hear that my code actually works 8^). What I find so
>interesting is that carefull analysis of the MMU configuration of my
>hardware platform leads me to believe that cache coherency must be
>handled in software (there is no special uncached address space or bus
>snooping logic), yet the driver works well without it. I guess this just
>proves how beneficial it is to make code open source so it can be
>tested on many different platforms by many different people ...
>
>Ome more thing that you might have already done yourself - since my
>third revision of the driver I've found a simple way to ensure that the 
>buffers used are actually placed on cache line boundaries. In your
>platform-specific .inl-File, add the following definition:
>
>#define CACHE_ALIGNED __attribute__ ((aligned (HAL_DCACHE_LINE_SIZE)))
>
>Then, use this attribute on the transmit and receive buffer
>definitions:
>
>static cyg_uint8 rltk8139_eth0_rx_ring[RX_BUF_TOT_LEN] CACHE_ALIGNED;
>
>Note that this might fail for long cache lines. Gcc documentation
>unfortunately does not specify what the maximum allowed value for
>the 'aligned' attribute is, and it also seems to be linker specific.
>(This information will go into the README for the fourth revision,
>planned for release when either a bug is fixed, a feature added, or
>the acknowledgement of the copyright assignment to RedHat arrives,
>in which case I plan to hand the driver over to the eCos maintainers).
>--
>--------------------------------------------------------------------
>|     Eric Doenges              |     DynaPel Laboratories GmbH    |
>|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
>|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
>--------------------------------------------------------------------
>
>
>--
>Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
>and search the list archive: http://sources.redhat.com/ml/ecos-discuss
>

_________________________________________________________________
<b>Get MSN 8</b> and enjoy automatic e-mail virus protection.  
http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Raghu,


>From: "Raghu BK" <raghu_bk@hotmail.com>
>To: embeddedeng@hotmail.com, Eric.Doenges@DynaPel.com
>CC: ecos-discuss@sources.redhat.com
>Subject: Re: [ECOS] PCI Ethernet card
>Date: Thu, 21 Aug 2003 02:15:50 +0000
>
>Hi Eric/Mike,
>
>The driver file (rltk8139.tgz) that I have is of size 17,680 bytes. Is it 
>the revision 3? How do I know the revision?

RLT8139 driver rev 3 from ecos-patches:
---------------------------
http://sources.redhat.com/ml/ecos-patches/2003-08/msg00040/rltk8139.tgz

>
>Mike: Thanks for the summary. Did you had to make any change for the driver 
>to work in Polled Mode since it is redboot?

The driver has functions for polling mode also which gets called by Redboot. 
Try this & let me know how it goes.

-Mike.

_________________________________________________________________
<b>Get MSN 8</b> and enjoy automatic e-mail virus protection.   
http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Raghu BK]

Hi Eric/Mike,

The driver file (rltk8139.tgz) that I have is of size 17,680 bytes. Is it 
the revision 3? How do I know the revision?

Mike: Thanks for the summary. Did you had to make any change for the driver 
to work in Polled Mode since it is redboot?

Thanks,
Raghu



>From: "Michael Anburaj" <embeddedeng@hotmail.com>
>To: raghu_bk@hotmail.com, Eric.Doenges@DynaPel.com
>CC: ecos-discuss@sources.redhat.com
>Subject: Re: [ECOS] PCI Ethernet card
>Date: Thu, 21 Aug 2003 01:01:56 +0000
>
>Hi Raghu,
>
>>From: "Raghu BK" <raghu_bk@hotmail.com>
>>To: Eric.Doenges@DynaPel.com, embeddedeng@hotmail.com
>>CC: ecos-discuss@sources.redhat.com
>>Subject: Re: [ECOS] PCI Ethernet card
>>Date: Wed, 20 Aug 2003 23:49:52 +0000
>>
>>Hi Mike/Eric,
>>
>>I've a custom board based on XScale 255, PCI Bridge IT8152 and RTL8139. 
>>I'm in the process of porting RedBoot to this platform. I would really 
>>appreciate if you could share the RTL8139 controller driver with me.
>>
>>Mike: Since you already ported Eric's driver to Redboot, could you please 
>>share the driver code you ported with me? This would cut short my 
>>development time and your driver would be validated on one more platform. 
>>I'll keep you folks updated on the status.
>>
>
>EricÂ’s driver code works on redboot without any change. I did not modify 
>his code.
>
>Did the following:
>1. Extracted the rltk8139.tgz <revision 3>
>2. Copied the files under devs/eth/i386/pc/rltk8139/ into 
>devs/eth/mips/atlas/rltk8139/
>3. Changed all reference made to i386 to atlas <my MIPS platform name>
>4. Included the following #define in devs_eth_atlas_pc_rltk8139.inl to 
>solve the cache coherency issue:
>#define CYGPKG_DEVS_ETH_RLTK_8139_SOFTWARE_CACHE_COHERENCY
>5. Included this package in ecos.db & created a new redboot_rltk.ecm <that 
>included this package instead of the existing onboard one>.
>
>I rebuilt redboot after this & it worked fine.
>
>Cheers,
>-Mike.
>

_________________________________________________________________
<b>Get MSN 8</b> and help protect your children with advanced parental 
controls.  http://join.msn.com/?page=features/parental


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Raghu,

>From: "Raghu BK" <raghu_bk@hotmail.com>
>To: Eric.Doenges@DynaPel.com, embeddedeng@hotmail.com
>CC: ecos-discuss@sources.redhat.com
>Subject: Re: [ECOS] PCI Ethernet card
>Date: Wed, 20 Aug 2003 23:49:52 +0000
>
>Hi Mike/Eric,
>
>I've a custom board based on XScale 255, PCI Bridge IT8152 and RTL8139. I'm 
>in the process of porting RedBoot to this platform. I would really 
>appreciate if you could share the RTL8139 controller driver with me.
>
>Mike: Since you already ported Eric's driver to Redboot, could you please 
>share the driver code you ported with me? This would cut short my 
>development time and your driver would be validated on one more platform. 
>I'll keep you folks updated on the status.
>

EricÂ’s driver code works on redboot without any change. I did not modify his 
code.

Did the following:
1. Extracted the rltk8139.tgz <revision 3>
2. Copied the files under devs/eth/i386/pc/rltk8139/ into 
devs/eth/mips/atlas/rltk8139/
3. Changed all reference made to i386 to atlas <my MIPS platform name>
4. Included the following #define in devs_eth_atlas_pc_rltk8139.inl to solve 
the cache coherency issue:
#define CYGPKG_DEVS_ETH_RLTK_8139_SOFTWARE_CACHE_COHERENCY
5. Included this package in ecos.db & created a new redboot_rltk.ecm <that 
included this package instead of the existing onboard one>.

I rebuilt redboot after this & it worked fine.

Cheers,
-Mike.

_________________________________________________________________
<b>MSN 8:</b> Get 6 months for $9.95/month. 
http://join.msn.com/?page=dept/dialup


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Eric,

I found the problem & fixed it.

Problem:
-----------

I displayed the value of ‘header’ each time after the following line of code 
at rltk8139_deliver():

header = HAL_LE32TOC(*((cyg_uint32 *)&rltk8139_info->rx_ring[rx_pos]));

And it was most of the time = 0x0.
I initially guessed that probably the Rx_Ring_buffer was not setup right. 
But, the actual problem was with the caches < cache coherency -same issue 
that Gary Thomas addressed in one of the emails sent to ecos-patches >. So, 
I pulled revision 3 of your driver & also defined 
CYGPKG_DEVS_ETH_RLTK_8139_SOFTWARE_CACHE_COHERENCY in my platform-specific 
.inl file. It worked well & got the IP parameters in 1 shot from the DHCP 
server. Right now itÂ’s configured as eth0 interface, I will try making it 
eth1. Thanks a lot to you and all who participated in this discussion.

FYI: I exchanged my PCI 2.2 compliant <with WOL> NIC with a PCI 2.1 one, but 
having the same chip - RTL8139D.

Thanks & Cheers,
-Mike.

_________________________________________________________________
<b>Get MSN 8</b> and help protect your children with advanced parental 
controls.  http://join.msn.com/?page=features/parental


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:

> I found the problem & fixed it.

Great !

> 
> Problem:
> -----------
> 
> I displayed the value of ?header? each time after the following line of 
> code at rltk8139_deliver():
> 
> header = HAL_LE32TOC(*((cyg_uint32 *)&rltk8139_info->rx_ring[rx_pos]));
> 
> And it was most of the time = 0x0.
> I initially guessed that probably the Rx_Ring_buffer was not setup 
> right. But, the actual problem was with the caches < cache coherency 
> -same issue that Gary Thomas addressed in one of the emails sent to 
> ecos-patches >. So, I pulled revision 3 of your driver & also defined 
> CYGPKG_DEVS_ETH_RLTK_8139_SOFTWARE_CACHE_COHERENCY in my 
> platform-specific .inl file. It worked well & got the IP parameters in 1 
> shot from the DHCP server. Right now it?s configured as eth0 interface, 
> I will try making it eth1. Thanks a lot to you and all who participated 
> in this discussion.
> 
> FYI: I exchanged my PCI 2.2 compliant <with WOL> NIC with a PCI 2.1 one, 
> but having the same chip - RTL8139D.

It's nice to hear that my code actually works 8^). What I find so
interesting is that carefull analysis of the MMU configuration of my
hardware platform leads me to believe that cache coherency must be
handled in software (there is no special uncached address space or bus
snooping logic), yet the driver works well without it. I guess this just
proves how beneficial it is to make code open source so it can be
tested on many different platforms by many different people ...

Ome more thing that you might have already done yourself - since my
third revision of the driver I've found a simple way to ensure that the 
buffers used are actually placed on cache line boundaries. In your
platform-specific .inl-File, add the following definition:

#define CACHE_ALIGNED __attribute__ ((aligned (HAL_DCACHE_LINE_SIZE)))

Then, use this attribute on the transmit and receive buffer
definitions:

static cyg_uint8 rltk8139_eth0_rx_ring[RX_BUF_TOT_LEN] CACHE_ALIGNED;

Note that this might fail for long cache lines. Gcc documentation
unfortunately does not specify what the maximum allowed value for
the 'aligned' attribute is, and it also seems to be linker specific.
(This information will go into the README for the fourth revision,
planned for release when either a bug is fixed, a feature added, or
the acknowledgement of the copyright assignment to RedHat arrives,
in which case I plan to hand the driver over to the eCos maintainers).
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Eric,

I have passed on incorrect info. sorry

The values I sent earlier were sampled at the 1st call to rlt8139_poll() 
<probably the Auto-neg was not done yet at that point>. Now I sample at 
every change in ISR (status),

BMCR = 0x1100
BMSR = 0x782d
ANAR = 0x01e1
ANLPAR = 0x45e1
ANER = 0x0001
DIS = 0x0000
FCSC = 0x0000

And one more thing, even if the DHCP server is not connected, I see the Link 
status change interrupt (status) bit going high after every packet sent. Is 
there a register to look at for the cause to this link status change <The 
above registers don't change at all>? I would probably display all register 
& try comparing them :)

Thanks,
-Mike.

_________________________________________________________________
Help STOP SPAM with the new MSN 8 and get 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:

> The values I sent earlier were sampled at the 1st call to rlt8139_poll() 
> <probably the Auto-neg was not done yet at that point>. Now I sample at 
> every change in ISR (status),
> 
> BMCR = 0x1100
> BMSR = 0x782d
> ANAR = 0x01e1
> ANLPAR = 0x45e1
> ANER = 0x0001
> DIS = 0x0000
> FCSC = 0x0000
> 
> And one more thing, even if the DHCP server is not connected, I see the 
> Link status change interrupt (status) bit going high after every packet 
> sent. Is there a register to look at for the cause to this link status 
> change <The above registers don't change at all>? I would probably 
> display all register & try comparing them :)

Well, there is the 'media status register' (0x58; bit #2 contains the
inverse of the link status), and the 'NWay Test Register' (0x70/0x71;
however I'm not sure if this would help any).

I've noticed two things in your register dump:

1. In BMCR, it seems duplex mode is set to full duplex. If you are
connected to a hub, this is incorrect. However, the 8139 documentation
sort of implies this bit is ignored together with bit 13 when bit 12
(enable autonegotiation) is enabled.

2. In ANAR, 'ACK' (bit #14) is not set. IMO this indicates that
auto-negotiation is failing for some reason.

Have you tried setting speed and duplex mode manually, i.e.
BMCR = 0x0000 (this would set 10MBit/s, half-duplex). In order to
change these bits, you must set bits 6 and 7 in in the '93C46 Command
Register' (0x50). Don't forget to clear those bits after changing
BMCR.
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Eric,

>a. Check the value of the basic mode status register (0x64), the
>auto-negotiation registers (0x66, 0x68, 0x6A), and the disconnect (0x6c)
>and false carrier sense (0x6e) counters.
>

BMCR = 0x1000            - enable auto-negotiation
BMSR = 0x7809 (default)
ANAR = 0x01e1 (default)
ANLPAR = 0x0000
ANER = 0x0000
DIS = 0x0000
FCSC = 0x0000

>b. Force the 8139 to use a particular speed and duplex mode by setting
>bits #13 and #8 in the basic mode control register (0x62) explicitely.
>Don't forget to disable autonegotiation by clearing bit #12. (I'm
>assuming that the EEPROM on your card selects autonegotiation by
>default).

Yes you are right. To keep things simple, I will start out by doing this. 
But, is it not ok to set this bit (auto-negotiation)?

I am still reading the documentation & trying to understand it better.

Questions:
1. Is the auto-negotiation (sending & receiving negotiation packets) under 
software (device driver)  control?
       a. If yes, does the rltk8139 driver Ver 2 support this feature? Or 
does it assume fixed link parameters?
      b.If no, then is there a problem with the Hub (the link partner on the 
other side of this link)? < Presumably, the both sides understand 
“auto-negotiation”, should negotiate the link parameters & settle down for 
actual transfers (with the sable connection - Link Status bit of BMSR set), 
am I right? >

Thanks,
-Mike.

_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:
> Hi Eric,
> 
>> a. Check the value of the basic mode status register (0x64), the
>> auto-negotiation registers (0x66, 0x68, 0x6A), and the disconnect (0x6c)
>> and false carrier sense (0x6e) counters.
>>
> 
> BMCR = 0x1000            - enable auto-negotiation
> BMSR = 0x7809 (default)
> ANAR = 0x01e1 (default)
> ANLPAR = 0x0000
> ANER = 0x0000
> DIS = 0x0000
> FCSC = 0x0000

I think auto-negotiation is not working in your setup - ANLPAR shouldn't
be all 0s (at least bit #14, ACK should be set IMO). To me it looks like
the two parties never get to negotiating.

> Yes you are right. To keep things simple, I will start out by doing 
> this. But, is it not ok to set this bit (auto-negotiation)?

If auto-negotiation is enabled, the 8139 ignores bits 13 and 8
('Spd_Set' and 'Duplex Mode').

> I am still reading the documentation & trying to understand it better.

Welcome to the club =8^)

> Questions:
> 1. Is the auto-negotiation (sending & receiving negotiation packets) 
> under software (device driver)  control?

No, this is done by the hardware. You can enable or disable
auto-negotiation, and tell the 8139 which link options to accept in
software, but the negotiation itself is done by the hardware (if it is
enabled).

>       a. If yes, does the rltk8139 driver Ver 2 support this feature? Or 
> does it assume fixed link parameters?

The driver assumes the serial EEPROM will have set the correct
parameters after reset.

>      b.If no, then is there a problem with the Hub (the link partner on 
> the other side of this link)? < Presumably, the both sides understand 
> 'auto-negotiation', should negotiate the link parameters & settle down 
> for actual transfers (with the sable connection - Link Status bit of 
> BMSR set), am I right?

I think you are, but I am no expert on Ethernet devices in general or
the 8139 in particular, so I may be wrong. That being said, old
networking equipment may not implement auto-negotiation correctly; I've
had problems in the past between 3Com switches and DEC 'tulip', as well
as some SMC NICs (this was back in 1998 or so). With modern equipment,
auto-negotiation seems to works just fine.

If you can, try hooking your board up to a different hub/switch or
directly to the DHCP server; if that works then your hub is probably to
blame.
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi,

> > >  * Turn on the debug messages in the RLTK driver (..._CHATTER)

I did this. Interrupt Status Register's bit 5 ( PUN/LinkChg- Packet 
Underrun/Link Change ) was getting set each time after a packet was sent 
through the RealTek 8139 NIC.

The status printed showed this (0x20 or 0x24),
diag_printf("rltk8139_deliver(%s): %04x\n", sc->dev_name, status);

Please let me know why this is happening? Is this a common thing or an error 
condition. Apart from the spec.Â’s (which gives very little details about 
these registers & the 8139 chip operation), is there a better document (or a 
detailed document for some other related chip)? If so, please send me the 
URL for the same.

Thanks,
-Mike.

_________________________________________________________________
Protect your PC - get McAfee.com VirusScan Online  
http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:

> I did this. Interrupt Status Register's bit 5 ( PUN/LinkChg- Packet 
> Underrun/Link Change ) was getting set each time after a packet was sent 
> through the RealTek 8139 NIC.
> 
> The status printed showed this (0x20 or 0x24),
> diag_printf("rltk8139_deliver(%s): %04x\n", sc->dev_name, status);
> 
> Please let me know why this is happening? Is this a common thing or an 
> error condition.

I'd say that is an error condition. The way I understand the
documentation, PUN/LinkChg is set when either:

1. CAPR is written by the driver software even though the chip thinks
    it's receive buffer is empty.

2. The link status changed.

Since the driver only touches CAPR when there is in fact a packet in the
receive buffer, I think you're getting frequent link status changes,
which is an error and explains why communication is not working. In a
previous life as a network administrator, we had problems with
auto-negotiation between ethernet NICs and 3Com switches that resulted
in frequent link up/link down events. With more modern equipment, these
problems went away (mostly). Some suggestions:

a. Check the value of the basic mode status register (0x64), the
auto-negotiation registers (0x66, 0x68, 0x6A), and the disconnect (0x6c)
and false carrier sense (0x6e) counters.

b. Force the 8139 to use a particular speed and duplex mode by setting
bits #13 and #8 in the basic mode control register (0x62) explicitely.
Don't forget to disable autonegotiation by clearing bit #12. (I'm
assuming that the EEPROM on your card selects autonegotiation by
default).

> Apart from the specs (which gives very little details 
> about these registers & the 8139 chip operation), is there a better 
> document (or a detailed document for some other related chip)? If so, 
> please send me the URL for the same.

Well, if you manage to find better documentation than what RealTek
provides on their website, be sure to let me know ! The best
documentation I found for how the chip actually works was the original
OpenBSD driver (sorry, don't remember the URL) -- but that doesn't
handle that condition.
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Gary,

>  * Turn on the debug messages in the RLTK driver (..._CHATTER)

I will do this.

>  * Run a network snooper ('ethereal' is what I use) to watch the network
>    traffic.

I tried installing ethereal-base-0.9.11-1.7.2.i386.rpm, but it needs 
ucd-snmp. My RH 9 has net-snmp & is not allowing me to install ucd-snmp (I 
guess, ucd-snmp is older version of net-snmp). Your comments on this. Where 
can I find ethereal binaries that will work with packages from RH 9 without 
issues ( I mean, with net-snmp)?

>  Does the request even make it to your DHCP server machine?
>    Is there a reply?

I guess there were packet exchanges between the 2, but it did not get to 
completion.

>>DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
>>DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
>>
>>< The same (DHCPDISCOVER-DHCPOFFER) pair repeated for a couple of times>

Please point me to ethereal binaries...

Thanks,
-Mike.

_________________________________________________________________
MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*.  
http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Gary Thomas]

On Sat, 2003-08-09 at 14:22, Michael Anburaj wrote:
> Hi Gary,
> 
> >  * Turn on the debug messages in the RLTK driver (..._CHATTER)
> 
> I will do this.
> 
> >  * Run a network snooper ('ethereal' is what I use) to watch the network
> >    traffic.
> 
> I tried installing ethereal-base-0.9.11-1.7.2.i386.rpm, but it needs 
> ucd-snmp. My RH 9 has net-snmp & is not allowing me to install ucd-snmp (I 
> guess, ucd-snmp is older version of net-snmp). Your comments on this. Where 
> can I find ethereal binaries that will work with packages from RH 9 without 
> issues ( I mean, with net-snmp)?
> 
> >  Does the request even make it to your DHCP server machine?
> >    Is there a reply?
> 
> I guess there were packet exchanges between the 2, but it did not get to 
> completion.
> 
> >>DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
> >>DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
> >>
> >>< The same (DHCPDISCOVER-DHCPOFFER) pair repeated for a couple of times>
> 
> Please point me to ethereal binaries...
> 

I'm running just what ships with Red Hat 9:
        [root@hermes log]# rpm -qa | grep snmp
        net-snmp-5.0.6-17
        net-snmp-utils-5.0.6-17
        net-snmp-devel-5.0.6-17
        [root@hermes log]# rpm -qa | grep ethereal
        ethereal-gnome-0.9.8-6
        ethereal-0.9.8-6

-- 
Gary Thomas <gary@mlbassoc.com>
MLB Associates


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Chuck McManis]

At 03:08 PM 8/9/2003 -0600, Gary Thomas wrote:
>I'm running just what ships with Red Hat 9:

Gary the problem is on the ECOS side. Your DHCP server accurately printed 
out that your machine requested an address, and it offered it a valid 
address. For what ever reason the BOOTP/DHCP code on the ECOS side did not 
act on that offer.

I'd look at the IP code to see if the return packet even made it past the 
ethernet driver. The _CHATTER flag will help here. Its possible that even 
though you have BOOTP enabled the DHCP server responded with a DHCP ack. 
That is what ethereal will tell you.

--Chuck



-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Gary Thomas]

On Sat, 2003-08-09 at 14:22, Michael Anburaj wrote:
> Hi Gary,
> 
> >  * Turn on the debug messages in the RLTK driver (..._CHATTER)
> 
> I will do this.
> 
> >  * Run a network snooper ('ethereal' is what I use) to watch the network
> >    traffic.
> 
> I tried installing ethereal-base-0.9.11-1.7.2.i386.rpm, but it needs 
> ucd-snmp. My RH 9 has net-snmp & is not allowing me to install ucd-snmp (I 
> guess, ucd-snmp is older version of net-snmp). Your comments on this. Where 
> can I find ethereal binaries that will work with packages from RH 9 without 
> issues ( I mean, with net-snmp)?
> 
> >  Does the request even make it to your DHCP server machine?
> >    Is there a reply?
> 
> I guess there were packet exchanges between the 2, but it did not get to 
> completion.
> 
> >>DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
> >>DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
> >>
> >>< The same (DHCPDISCOVER-DHCPOFFER) pair repeated for a couple of times>
> 
> Please point me to ethereal binaries...
> 

One more thing - make sure that your RedBoot sources are up to date.
There was a bug in the BOOTP/DHCP code from June 24 through July 15,
so make sure your sources are at least that new.

-- 
Gary Thomas <gary@mlbassoc.com>
MLB Associates


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi,

I configured the RLTK eth driver to be eth0 (to keep things simple for now, 
later I will try having both eth drivers).

Redboot's console dump:
---------------------------------
RedBoot> ++... waiting for BOOTP information
Ethernet eth0: MAC address 00:50:fc:9b:95:ec
Can't get BOOTP info for device!


dhcpd dump:
-----------------
[root@localhost bin]# /usr/sbin/dhcpd -f -d
Internet Software Consortium DHCP Server V3.0pl1
Copyright 1995-2001 Internet Software Consortium.
All rights reserved.
For info, please visit http://www.isc.org/products/DHCP
Wrote 0 deleted host decls to leases file.
Wrote 0 new dynamic host decls to leases file.
Wrote 0 leases to leases file.
Listening on LPF/eth0/00:e0:98:9e:84:96/4.35.2.0/24
Sending on   LPF/eth0/00:e0:98:9e:84:96/4.35.2.0/24
Sending on   Socket/fallback/fallback-net
DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0

< The same (DHCPDISCOVER-DHCPOFFER) pair repeated for a couple of times>


My dhcpd.cong:
---------------------

allow bootp;
ddns-update-style none; #ad-hoc;
subnet 4.35.2.0 netmask 255.255.255.0 {
    host atlas {
        hardware ethernet 00:d0:a0:00:00:ed;
        fixed-address 4.35.2.3;
    }

    host atlasRLTK {
        hardware ethernet 00:50:fc:9b:95:ec;
        fixed-address 4.35.2.4;
    }
}


RH 9 PC info:
------------------
Ethernet port on the PC is set wirh the same netmask (255.255.255.0) & IP = 
4.35.2.2. The PC & the Atlas board are connected using a HUB. This is the 
1st time I am working with DHCP,/BootP please let me know where the problem 
could be.

Thanks,
-Mike.

_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Gary Thomas]

On Sat, 2003-08-09 at 04:50, Michael Anburaj wrote:
> Hi,
> 
> I configured the RLTK eth driver to be eth0 (to keep things simple for now, 
> later I will try having both eth drivers).
> 
> Redboot's console dump:
> ---------------------------------
> RedBoot> ++... waiting for BOOTP information
> Ethernet eth0: MAC address 00:50:fc:9b:95:ec
> Can't get BOOTP info for device!
> 
> 
> dhcpd dump:
> -----------------
> [root@localhost bin]# /usr/sbin/dhcpd -f -d
> Internet Software Consortium DHCP Server V3.0pl1
> Copyright 1995-2001 Internet Software Consortium.
> All rights reserved.
> For info, please visit http://www.isc.org/products/DHCP
> Wrote 0 deleted host decls to leases file.
> Wrote 0 new dynamic host decls to leases file.
> Wrote 0 leases to leases file.
> Listening on LPF/eth0/00:e0:98:9e:84:96/4.35.2.0/24
> Sending on   LPF/eth0/00:e0:98:9e:84:96/4.35.2.0/24
> Sending on   Socket/fallback/fallback-net
> DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
> DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
> DHCPDISCOVER from 00:50:fc:9b:95:ec via eth0
> DHCPOFFER on 4.35.2.4 to 00:50:fc:9b:95:ec via eth0
> 
> < The same (DHCPDISCOVER-DHCPOFFER) pair repeated for a couple of times>
> 
> 
> My dhcpd.cong:
> ---------------------
> 
> allow bootp;
> ddns-update-style none; #ad-hoc;
> subnet 4.35.2.0 netmask 255.255.255.0 {
>     host atlas {
>         hardware ethernet 00:d0:a0:00:00:ed;
>         fixed-address 4.35.2.3;
>     }
> 
>     host atlasRLTK {
>         hardware ethernet 00:50:fc:9b:95:ec;
>         fixed-address 4.35.2.4;
>     }
> }
> 
> 
> RH 9 PC info:
> ------------------
> Ethernet port on the PC is set wirh the same netmask (255.255.255.0) & IP = 
> 4.35.2.2. The PC & the Atlas board are connected using a HUB. This is the 
> 1st time I am working with DHCP,/BootP please let me know where the problem 
> could be.

A couple of things to try:
 * Turn on the debug messages in the RLTK driver (..._CHATTER)
 * Run a network snooper ('ethereal' is what I use) to watch the network
   traffic.  Does the request even make it to your DHCP server machine?
   Is there a reply?

-- 
Gary Thomas <gary@mlbassoc.com>
MLB Associates


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi,

I took the code by "Eric Doenges" for RLTK 8139 & added it to my Atlas 
Redboot config. Now I want to switch to this ethernet 'eth1' from the 
on-board ethernet port 'eth0' for Redboot.

Does this, 'CYGNUM_REDBOOT_DEFAULT_NETWORK_DEVICE ' govern the selection of 
eth device for redboot? This option shows in gray & not able to edit the 
value. And also 'CYGHWR_NET_DRIVERS' is 1 here. Should this be 2 if I need 
to have 2 options (eth0 & eth1)? If so, How can it be changed to 2? please 
shar your ideas on this.

Thanks,
-Mike.

_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Nick Garnett]

"Michael Anburaj" <embeddedeng@hotmail.com> writes:

> Hi,
> 
> I took the code by "Eric Doenges" for RLTK 8139 & added it to my Atlas
> Redboot config. Now I want to switch to this ethernet 'eth1' from the
> on-board ethernet port 'eth0' for Redboot.
> 
> Does this, 'CYGNUM_REDBOOT_DEFAULT_NETWORK_DEVICE ' govern the
> selection of eth device for redboot? This option shows in gray & not
> able to edit the value. And also 'CYGHWR_NET_DRIVERS' is 1
> here. Should this be 2 if I need to have 2 options (eth0 & eth1)? If
> so, How can it be changed to 2? please shar your ideas on this.

At least initially, it probably make sense to work only with the RLTK
driver. Remove the default ethernet device entirely. Only when that is
working 100% should you then try running with both.

Note that in the past the only boards we have had with more than one
ethernet interface have always had identical devices. While the
software is structured to allow multiple different devices, it has
never been execised in this way, so there may be lurking bugs.


-- 
Nick Garnett                    eCos Kernel Architect
http://www.ecoscentric.com      The eCos and RedBoot experts


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

RE: [ECOS] PCI Ethernet card

[Andy Dyer]

> Sounds scary! I have only 1 MIPS board & if it gets busted, that’s the
end 
> of it. Cannot afford another MIPS board (Hobby development at home). I
guess 
> the Atlas board has a 5v signaling PCI slot (the key ridge on the slot
looks 
> like it). The COMPUSA NIC ($17 – but looks very much like the 
> http://www.memorywizards.com/pd_edimax_en9130tx.cfm - The component
layout 
> look very much like this, except it has more capacitors) has notches
at both 
> ends. So, I guess I am safe (because the PCI on-board is 5v), Am I
right? Is 
> there anything else I should take care before I insert this NIC on the
slot?
> 
> After all, the COMPUSA NIC could suffer the same issue. If so, why pay
3 
> times more for the same stuff with a different name. Any comments?

If your MIPS board has 5V PCI I wouldn't expect any problems.  Most PC
h/w
vendors are going to test on a few 5V PCI motherboards and ship it.
3V PCI is still a little weird and mostly seen in servers (66MHz
PCI is 3.3V).  Anybody who puts a $10 realtek NIC into a server deserves
what they get :-)

5V PCI has the 'keyway' away from the "back plate" end of the board,
3.3V PCI has the 'keyway' towards the "back plate".



--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

RE: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Andy,

>As for PCI cards and slots beware!  I blew up my 3.3V PCI target board
>last
>week by plugging in a cheap "Edimax" realtek NIC, which was laid out as
>a
>universal PCI card (had both notches), but had connections between Vio
>(3.3V
>in my system) and +5V dedicated pins.

Sounds scary! I have only 1 MIPS board & if it gets busted, thatÂ’s the end 
of it. Cannot afford another MIPS board (Hobby development at home). I guess 
the Atlas board has a 5v signaling PCI slot (the key ridge on the slot looks 
like it). The COMPUSA NIC ($17 – but looks very much like the 
http://www.memorywizards.com/pd_edimax_en9130tx.cfm - The component layout 
look very much like this, except it has more capacitors) has notches at both 
ends. So, I guess I am safe (because the PCI on-board is 5v), Am I right? Is 
there anything else I should take care before I insert this NIC on the slot?

After all, the COMPUSA NIC could suffer the same issue. If so, why pay 3 
times more for the same stuff with a different name. Any comments?

-Mike.

_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

RE: [ECOS] PCI Ethernet card

[Andy Dyer]

> If 'WOL' stands for 'wake on lan' I think you can just ignore 
> it if the
> card physically fits into the slot (fits without using any force, that
> is). Unless I'm very mistaken, the only thing you must do is avoid
> sticking 3.3V only cards into 5V only slots (or vice versa), but the
> physical encoding of the connectors should prevent this.

I believe the WOL signal is an output from the card to the motherboard,
and the other two wires are the PC standby 3.3V and ground.

As for PCI cards and slots beware!  I blew up my 3.3V PCI target board
last
week by plugging in a cheap "Edimax" realtek NIC, which was laid out as
a
universal PCI card (had both notches), but had connections between Vio
(3.3V 
in my system) and +5V dedicated pins.

--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi Eric,

<I don't know much about PCI bus yet. So.. >

I bought a RTL8139 based ethernet card from COMPUSA (for PCI 2.2). But, the 
thing is it needs WOL (3 pin connector) for PCI 2.1 boards.

The MIPS Atlas board supports only PCI 2.1 & does not have WOL (3 pin 
connector). Is there a possible, simple workaround? or Are there Cards 
(Based on RTL8139) out there that can work with PCI 2.1 without needing the 
WOL?

Thanks,
-Mike.

_________________________________________________________________
MSN 8 with e-mail virus protection service: 2 months FREE*  
http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Nikolay V. Pyatkov]

On Sunday August 3 2003 06:13 pm, Michael Anburaj wrote:
> Hi Eric,
>
> <I don't know much about PCI bus yet. So.. >
>
> I bought a RTL8139 based ethernet card from COMPUSA (for PCI 2.2). But, the
> thing is it needs WOL (3 pin connector) for PCI 2.1 boards.
>
> The MIPS Atlas board supports only PCI 2.1 & does not have WOL (3 pin
> connector). Is there a possible, simple workaround? or Are there Cards
> (Based on RTL8139) out there that can work with PCI 2.1 without needing the
> WOL?

If you do not need it do not use it
Regards,
Nikolay

>
> Thanks,
> -Mike.
>
> _________________________________________________________________
> MSN 8 with e-mail virus protection service: 2 months FREE*
> http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:

> <I don't know much about PCI bus yet. So.. >

Neither do I =8^)

> I bought a RTL8139 based ethernet card from COMPUSA (for PCI 2.2). But, 
> the thing is it needs WOL (3 pin connector) for PCI 2.1 boards.
> 
> The MIPS Atlas board supports only PCI 2.1 & does not have WOL (3 pin 
> connector). Is there a possible, simple workaround? or Are there Cards 
> (Based on RTL8139) out there that can work with PCI 2.1 without needing 
> the WOL?

If 'WOL' stands for 'wake on lan' I think you can just ignore it if the
card physically fits into the slot (fits without using any force, that
is). Unless I'm very mistaken, the only thing you must do is avoid
sticking 3.3V only cards into 5V only slots (or vice versa), but the
physical encoding of the connectors should prevent this.
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

RE: [ECOS] PCI Ethernet card

[Michael Anburaj]

Hi All,

Thanks all for sharing info.

>for one of the commodity PCI ethernet chipsets. That would be a useful

Can you point me to a particular commodity PCI Ethernet chipset, whichÂ’s 
good & well documented & cards based on them affordable :-) (or I could go 
for a i82559 based card). Then I would certainly work on it & contribute it 
back.

Thanks,
-Mike.

_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE*  
http://join.msn.com/?page=features/junkmail


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Nick Garnett]

"Michael Anburaj" <embeddedeng@hotmail.com> writes:

> Hi All,
> 
> Thanks all for sharing info.
> 
> >for one of the commodity PCI ethernet chipsets. That would be a useful
> 
> Can you point me to a particular commodity PCI Ethernet chipset,
> which’s good & well documented & cards based on them affordable :-)

Not offhand. I'm not sure what gets put into cheap ethernet cards
these days. I know someone is already doing a RealTek driver. 

> (or I could go for a i82559 based card). Then I would certainly work
> on it & contribute it back.

Given you appear to be able to pick up 82559 cards cheaply in the US,
then that may be the best way to go. 

-- 
Nick Garnett                    eCos Kernel Architect
http://www.ecoscentric.com      The eCos and RedBoot experts


--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Nick Garnett wrote:

> "Michael Anburaj" <embeddedeng@hotmail.com> writes:
> 
> 
>>Hi All,
>>
>>Thanks all for sharing info.
>>
>>
>>>for one of the commodity PCI ethernet chipsets. That would be a useful
>>
>>Can you point me to a particular commodity PCI Ethernet chipset,
>>whichÂ’s good & well documented & cards based on them affordable :-)
> 
> 
> Not offhand. I'm not sure what gets put into cheap ethernet cards
> these days. I know someone is already doing a RealTek driver. 
[ ... ]

According to RealTek, they held over 60% of the total market for
Ethernet and Fast Ethernet in 2002, so chances are good most low-end
100 MBit cards will use one of the RealTek 8139 variants. If Michael
wants to try his hand at improving my driver for it he is of course
more than welcome to do so (the latest code was posted to ecos-patches
on Friday); however the RealTek 8139 definitely does not fall into the
'well documented' category (you can get a data sheet without signing
anything, but it's not very helpful since it is missing a few important
points and contains a number of mistakes).
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

RE: [ECOS] PCI Ethernet card

[Andy Dyer]

 

	-----Original Message----- 
	From: ecos-discuss-owner@sources.redhat.com on behalf of Nick
Garnett 
	Sent: Fri 8/1/2003 8:26 AM 
	To: Michael Anburaj 
	Cc: ecos-discuss@sources.redhat.com 
	Subject: Re: [ECOS] PCI Ethernet card
	
	
	
	The most common PCI card we support is the Intel i82559 based
	Ether-Pro cards. But these are relatively expensive.
	

	I just picked up an 82559 card for $4 at gearxs.com (no
affiliation).  I haven't tried it yet, and I probably won't try it with
ecos because the card is 5V pci and my target is 3.3V.

[ECOS] PCI Ethernet card

[Michael Anburaj]

Hi,

Can somebody suggest be a ethernet card, that can go with my Atlas board 
(has a 33MHz, PCI slot). And for which eCOS has a driver. - Available out of 
Fry's or circuit city kinda shores.


Dumb Questions:
1. Can any Ethernet card available in the market be used with my board?
2. What specs should I look for when considering one?
3. And about the config registers & all other registers viewable over PCI & 
their details for the particular card; will that be provided by the 
manufacturer? So, if there is an Ethernet card that eCOS supports, please 
let me know.

This is just for learning purposes.

Thanks,
-Mike.

_________________________________________________________________
MSN 8 with e-mail virus protection service: 2 months FREE*  
http://join.msn.com/?page=features/virus


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Eric Doenges]

Michael Anburaj wrote:

> 1. Can any Ethernet card available in the market be used with my board?

No, unless you plan on writing a driver for it if one doesn't already
exist.

> 2. What specs should I look for when considering one?

You should look at which chips have driver support.

> 3. And about the config registers & all other registers viewable over 
> PCI & their details for the particular card; will that be provided by 
> the manufacturer?

In theory it should be; in practice some manufacturers will supply you
with very inferior documentation. Fortunately, you can always look at
Linux driver code to help figure out how a given chip works (if a
Linux driver exists).

> So, if there is an Ethernet card that eCOS supports, 
> please let me know.

Check what drivers eCos has for your platform. My source repository
lists a single driver for the Atlas, the SAA9730 (which seems to be 
onboard). However, you should be able to adapt an existing driver to
work with your board with minimal to low effort. Take a look at the
Intel 82559 driver and how it is used for the i386/pc platform in
eCos. The chip itself is found on Intel EtherExpress Pro/100 cards,
and should be availible just about everywhere.

A cheaper card would be any RealTek 8139-based card, but I'm still
developing the driver for that chip, so unless you want to be on the
bleeding edge and help me debug the driver, I would pass this one by
for now.
-- 
--------------------------------------------------------------------
|     Eric Doenges              |     DynaPel Laboratories GmbH    |
|     Tel: +49 89 962428 23     |     Fraunhoferstrasse 9/2        |
|     Fax: +49 89 962428 90     |     D - 85737 Ismaning, Germany  |
--------------------------------------------------------------------


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Jonathan Larmour]

Eric Doenges wrote:
> 
>> So, if there is an Ethernet card that eCOS supports, please let me know.
> 
> 
> Check what drivers eCos has for your platform. My source repository
> lists a single driver for the Atlas, the SAA9730 (which seems to be 
> onboard). However, you should be able to adapt an existing driver to
> work with your board with minimal to low effort. Take a look at the
> Intel 82559 driver and how it is used for the i386/pc platform in
> eCos. The chip itself is found on Intel EtherExpress Pro/100 cards,
> and should be availible just about everywhere.

http://ecos.sourceware.org/ecos/hardware.html#ethernet has a list of some 
of the chipsets (although there are more).

> A cheaper card would be any RealTek 8139-based card, but I'm still
> developing the driver for that chip, so unless you want to be on the
> bleeding edge and help me debug the driver, I would pass this one by
> for now.

Search the ecos-patches mailing list archive. Also there's an NE2000 
driver at http://sources.redhat.com/ecos/contrib.html, but that won't be 
able to be included in the eCos sources for licence reasons. Similarly 
<http://ecos.sourceware.org/ml/ecos-patches/2002-07/msg00002.html> / 
<http://www.vsr.si/ecos/patches/rtl8019.html>

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] PCI Ethernet card

[Nick Garnett]

"Michael Anburaj" <embeddedeng@hotmail.com> writes:

> Hi,
> 
> Can somebody suggest be a ethernet card, that can go with my Atlas
> board (has a 33MHz, PCI slot). And for which eCOS has a driver. -
> Available out of Fry's or circuit city kinda shores.
> 
> 
> Dumb Questions:
> 1. Can any Ethernet card available in the market be used with my
> board?

No, only those for which we have drivers -- we don't have the luxury
of having the manufacturers do it for us, like Windows, or an army
of hackers beavering away in their bedrooms like Linux.

The most common PCI card we support is the Intel i82559 based
Ether-Pro cards. But these are relatively expensive. We also have
support for AMD PCNET based cards. Take a look at the ethernet device
driver packages in ecos.db for an idea of what is supported.

One of the problems with a $10 commodity card is that it can be
difficult -- until you've bought it -- to work out exactly what
chipset it uses. And manufacturers sometimes change the design
entirely without changing the model number.

> 2. What specs should I look for when considering one?
> 3. And about the config registers & all other registers viewable over
> PCI & their details for the particular card; will that be provided by
> the manufacturer?

You may be lucky and find the specs on the chipset manufacturer's
website. But these are often under NDA, even for quite old chipsets.
Of course, there is always the driver sources in Linux to look at --
but never copy!

> So, if there is an Ethernet card that eCOS supports,
> please let me know.
> 
> This is just for learning purposes.

If you really want to learn you could look at writing a device driver
for one of the commodity PCI ethernet chipsets. That would be a useful
contribution.

-- 
Nick Garnett                    eCos Kernel Architect
http://www.ecoscentric.com      The eCos and RedBoot experts


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss