linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
* EMAC2 in ibm_newemac driver
@ 2008-10-02  4:38 vb
  2008-10-02  5:51 ` Stefan Roese
  0 siblings, 1 reply; 4+ messages in thread
From: vb @ 2008-10-02  4:38 UTC (permalink / raw)
  To: linuxppc-embedded

I have a 460GT based system with three ethernet interfaces, and I
don't seem to be able to bring up EMAC2 in RGMII mode. The PHY works
just fine, the counterpart on the other side of the serdes link
autonegotiates successfully, etc. When I attempt to send something
from the interface though, nothing shows up on the RGMII lines (I use
a scope to probe it). The received traffic is passed from the  PHY to
the MAC, but again, nothing is reported by the rx driver.

I am using the latest released code from 2.6.26.5 (I started with
2.6.25 release, and in that case just EMAC0 worked, updgrading to to
2.6.26.5 made the second interface work, but the third one seems to be
stuck. I see that the Denx tree has a lot of changes in the driver,
but it is not as cleanly portable, so advancing  there is not an
option.

I am trying to debug it, but  am wondering if anyone had more than two
interfaces going on this architecture and if so - what are the
gotchas.

Any hints will be highly appreciated,

cheers,
VadimB

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: EMAC2 in ibm_newemac driver
  2008-10-02  4:38 EMAC2 in ibm_newemac driver vb
@ 2008-10-02  5:51 ` Stefan Roese
  2008-10-03  4:57   ` vb
  0 siblings, 1 reply; 4+ messages in thread
From: Stefan Roese @ 2008-10-02  5:51 UTC (permalink / raw)
  To: linuxppc-embedded; +Cc: vb

On Thursday 02 October 2008, vb wrote:
> I have a 460GT based system with three ethernet interfaces, and I
> don't seem to be able to bring up EMAC2 in RGMII mode. The PHY works
> just fine, the counterpart on the other side of the serdes link
> autonegotiates successfully, etc. When I attempt to send something
> from the interface though, nothing shows up on the RGMII lines (I use
> a scope to probe it). The received traffic is passed from the  PHY to
> the MAC, but again, nothing is reported by the rx driver.
>
> I am using the latest released code from 2.6.26.5 (I started with
> 2.6.25 release, and in that case just EMAC0 worked, updgrading to to
> 2.6.26.5 made the second interface work, but the third one seems to be
> stuck. I see that the Denx tree has a lot of changes in the driver,
> but it is not as cleanly portable, so advancing  there is not an
> option.

The ibm_newemac driver in the DENX repository has absolutely no changes to the 
kernel.org version.

> I am trying to debug it, but  am wondering if anyone had more than two
> interfaces going on this architecture and if so - what are the
> gotchas.

The AMCC Glacier 460GT eval board uses all 4 EMAC's with RGMII. IIRC, then you 
need to configure the GPIO multiplexing differently (compared to Canyonlands 
460EX) to fully support all 4 EMAC's on 460GT. This is done in U-Boot. I 
suggest that you check this configuration first.

Best regards,
Stefan

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: EMAC2 in ibm_newemac driver
  2008-10-02  5:51 ` Stefan Roese
@ 2008-10-03  4:57   ` vb
  2008-10-03 14:37     ` Stefan Roese
  0 siblings, 1 reply; 4+ messages in thread
From: vb @ 2008-10-03  4:57 UTC (permalink / raw)
  To: Stefan Roese; +Cc: linuxppc-embedded

On Wed, Oct 1, 2008 at 10:51 PM, Stefan Roese <sr@denx.de> wrote:
> On Thursday 02 October 2008, vb wrote:
>>
>
> The ibm_newemac driver in the DENX repository has absolutely no changes to the
> kernel.org version.
>

I am comparing between  2.6.26.5 and your tree, these are the
differences I referred to.

>> I am trying to debug it, but  am wondering if anyone had more than two
>> interfaces going on this architecture and if so - what are the
>> gotchas.
>
> The AMCC Glacier 460GT eval board uses all 4 EMAC's with RGMII. IIRC, then you
> need to configure the GPIO multiplexing differently (compared to Canyonlands
> 460EX) to fully support all 4 EMAC's on 460GT. This is done in U-Boot. I
> suggest that you check this configuration first.
>

Stefan, you are a gentleman and a scholar: this is exactly where my
problem was. Thank you for saving me from wasting lots of time trying
to figure out what's 'wrong' with the driver.

Now, comparing glacier GPIO settings with the 460Gx datasheet (rev
1.07 of May 29) it seems that the datasheet is wrong and suggests
using of ALT2 mode for the RGMII2 gpio pins(does not work) as opposed
ALT1 used in glacier (and actually working apart from a problem I
still am seeing with the TX). Do you remember anything about this?

TIA,
vadim

> Best regards,
> Stefan
>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: EMAC2 in ibm_newemac driver
  2008-10-03  4:57   ` vb
@ 2008-10-03 14:37     ` Stefan Roese
  0 siblings, 0 replies; 4+ messages in thread
From: Stefan Roese @ 2008-10-03 14:37 UTC (permalink / raw)
  To: vb; +Cc: linuxppc-embedded

Vadim,

On Friday 03 October 2008, vb wrote:
> > The AMCC Glacier 460GT eval board uses all 4 EMAC's with RGMII. IIRC,
> > then you need to configure the GPIO multiplexing differently (compared to
> > Canyonlands 460EX) to fully support all 4 EMAC's on 460GT. This is done
> > in U-Boot. I suggest that you check this configuration first.
>
> Stefan, you are a gentleman and a scholar: this is exactly where my
> problem was. Thank you for saving me from wasting lots of time trying
> to figure out what's 'wrong' with the driver.
>
> Now, comparing glacier GPIO settings with the 460Gx datasheet (rev
> 1.07 of May 29)

That's an old version. I suggest you download the latest users manual. I have 
rev. 1.14 (Sept. 2) here.

> it seems that the datasheet is wrong and suggests 
> using of ALT2 mode for the RGMII2 gpio pins(does not work) as opposed
> ALT1 used in glacier (and actually working apart from a problem I
> still am seeing with the TX). Do you remember anything about this?

I remember that the early manual had some problems. GPIO being only one of 
them. Not sure if this is really fixed in the current manual. But the U-Boot 
code should show what's really correct. Code can't lie! ;)

Best regards,
Stefan

=====================================================================
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-0 Fax: +49-8142-66989-80  Email: office@denx.de
=====================================================================

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2008-10-03 14:37 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-10-02  4:38 EMAC2 in ibm_newemac driver vb
2008-10-02  5:51 ` Stefan Roese
2008-10-03  4:57   ` vb
2008-10-03 14:37     ` Stefan Roese

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).