[U-Boot] smc911x not functional on top of tree

[U-Boot] smc911x not functional on top of tree

[Fabio Estevam]

Hi,

When using U-boot from top of tree I am not able to get networking to
work on a mx31pdk:

uboot> dhcp
smc911x: detected LAN9217 controller
smc911x: phy initialized
smc911x: MAC 00:04:9f:00:8d:6e
BOOTP broadcast 1
BOOTP broadcast 2
DHCP client bound to address 10.29.244.21
BOOTP broadcast 3
DHCP client bound to address 10.29.244.21
BOOTP broadcast 4
DHCP client bound to address 10.29.244.21
BOOTP broadcast 5
DHCP client bound to address 10.29.244.21

Retry count exceeded; starting again

I haven't started debugging this, but would like to check first if
this is a known issue.

Running v2011.06 makes networking to work fine.

I have not enabled the caches on this board yet and I do see a warning
in boot saying
that caches are not enabled.

Does anyone have smc911x working with the latest U-boot?

Thanks,

Fabio Estevam

[U-Boot] smc911x not functional on top of tree

[stefano babic]

Am 15/09/2011 21:39, schrieb Fabio Estevam:
> Hi,
>
Hi Fabio,
> When using U-boot from top of tree I am not able to get networking to
> work on a mx31pdk:
>
> uboot> dhcp
> smc911x: detected LAN9217 controller
> smc911x: phy initialized
> smc911x: MAC 00:04:9f:00:8d:6e
> BOOTP broadcast 1
> BOOTP broadcast 2
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 3
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 4
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 5
> DHCP client bound to address 10.29.244.21
>
> Retry count exceeded; starting again
I see...
> I haven't started debugging this, but would like to check first if
> this is a known issue.
I have not heard about this issue.
> Running v2011.06 makes networking to work fine.
>
> I have not enabled the caches on this board yet and I do see a warning
> in boot saying
> that caches are not enabled.
I know the commit, the same warning appears on a lot of other boards. It
is only a remind that cache is not enabled because the drivers do not
work if cache is enabled (as th FEC driver, for example).
> Does anyone have smc911x working with the latest U-boot?

I have now tested on a mx35pdk with Debug board. It has a SMC911x. It works:

U-Boot 2011.09-rc1-00020-g56fa45d (Sep 15 2011 - 22:08:55)

CPU:   Freescale i.MX35 at 532 MHz
Board: MX35 PDK 2.0 i.MX35 2.0 [WDT]
I2C:   ready
DRAM:  256 MiB
WARNING: Caches not enabled
Flash: 64 MiB
NAND:  2048 MiB
In:    serial
Out:   serial
Err:   serial
Net:   smc911x-0, FECWarning: failed to set MAC address

Hit any key to stop autoboot:  0
MX35 U-Boot >

MX35 U-Boot > run load
smc911x: detected LAN9217 controller
smc911x: phy initialized
smc911x: MAC 00:04:9f:00:f4:99
Using smc911x-0 device
TFTP from server 192.168.2.14; our IP address is 192.168.2.97
Filename 'mx35pdk/u-boot.bin'.
Load address: 0x80800000
Loading: #################
done
Bytes transferred = 248892 (3cc3c hex)

Best  regards,
Stefano Babic

-- 
=====================================================================
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 at denx.de
=====================================================================

[U-Boot] smc911x not functional on top of tree

[Marek Vasut]

On Thursday, September 15, 2011 10:21:04 PM stefano babic wrote:
> Am 15/09/2011 21:39, schrieb Fabio Estevam:
> > Hi,
> 
> Hi Fabio,
> 
> > When using U-boot from top of tree I am not able to get networking to
> > work on a mx31pdk:
> > 
> > uboot> dhcp
> > smc911x: detected LAN9217 controller
> > smc911x: phy initialized
> > smc911x: MAC 00:04:9f:00:8d:6e
> > BOOTP broadcast 1
> > BOOTP broadcast 2
> > DHCP client bound to address 10.29.244.21
> > BOOTP broadcast 3
> > DHCP client bound to address 10.29.244.21
> > BOOTP broadcast 4
> > DHCP client bound to address 10.29.244.21
> > BOOTP broadcast 5
> > DHCP client bound to address 10.29.244.21
> > 
> > Retry count exceeded; starting again
> 
> I see...
> 
> > I haven't started debugging this, but would like to check first if
> > this is a known issue.
> 
> I have not heard about this issue.
> 
> > Running v2011.06 makes networking to work fine.
> > 
> > I have not enabled the caches on this board yet and I do see a warning
> > in boot saying
> > that caches are not enabled.
> 
> I know the commit, the same warning appears on a lot of other boards. It
> is only a remind that cache is not enabled because the drivers do not
> work if cache is enabled (as th FEC driver, for example).

I'm working on the FEC patch that allows it to work with caches actually. I have 
it ready, but it needs some polishing.

> 
> > Does anyone have smc911x working with the latest U-boot?
> 
> I have now tested on a mx35pdk with Debug board. It has a SMC911x. It
> works:
> 
> U-Boot 2011.09-rc1-00020-g56fa45d (Sep 15 2011 - 22:08:55)
> 
> CPU:   Freescale i.MX35 at 532 MHz
> Board: MX35 PDK 2.0 i.MX35 2.0 [WDT]
> I2C:   ready
> DRAM:  256 MiB
> WARNING: Caches not enabled
> Flash: 64 MiB
> NAND:  2048 MiB
> In:    serial
> Out:   serial
> Err:   serial
> Net:   smc911x-0, FECWarning: failed to set MAC address
> 
> Hit any key to stop autoboot:  0
> MX35 U-Boot >
> 
> MX35 U-Boot > run load
> smc911x: detected LAN9217 controller
> smc911x: phy initialized
> smc911x: MAC 00:04:9f:00:f4:99
> Using smc911x-0 device
> TFTP from server 192.168.2.14; our IP address is 192.168.2.97
> Filename 'mx35pdk/u-boot.bin'.
> Load address: 0x80800000
> Loading: #################
> done
> Bytes transferred = 248892 (3cc3c hex)
> 
> Best  regards,
> Stefano Babic

[U-Boot] smc911x not functional on top of tree

[Fabio Estevam]

Hi Stefano,

On Thu, Sep 15, 2011 at 5:21 PM, stefano babic <sbabic@denx.de> wrote:
..
>> Does anyone have smc911x working with the latest U-boot?
>
> I have now tested on a mx35pdk with Debug board. It has a SMC911x. It works:

Thanks for testing! I tried it at home and it is working fine here.

Maybe it was an issue at the company's network.

Thanks,

Fabio Estevam

[U-Boot] smc911x not functional on top of tree

[Stefano Babic]

On 09/16/2011 01:13 AM, Fabio Estevam wrote:
> Hi Stefano,
> 
> On Thu, Sep 15, 2011 at 5:21 PM, stefano babic <sbabic@denx.de> wrote:
> ..
>>> Does anyone have smc911x working with the latest U-boot?
>>
>> I have now tested on a mx35pdk with Debug board. It has a SMC911x. It works:
> 
> Thanks for testing! I tried it at home and it is working fine here.
> 
> Maybe it was an issue at the company's network.

Your network works better as your company's network.

Best regards,
Stefano Babic


-- 
=====================================================================
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 at denx.de
=====================================================================

[U-Boot] smc911x not functional on top of tree

[Helmut Raiger]

On 09/15/2011 09:39 PM, Fabio Estevam wrote:
> Hi,
>
> When using U-boot from top of tree I am not able to get networking to
> work on a mx31pdk:
>
> uboot>  dhcp
> smc911x: detected LAN9217 controller
> smc911x: phy initialized
> smc911x: MAC 00:04:9f:00:8d:6e
> BOOTP broadcast 1
> BOOTP broadcast 2
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 3
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 4
> DHCP client bound to address 10.29.244.21
> BOOTP broadcast 5
> DHCP client bound to address 10.29.244.21
>
> Retry count exceeded; starting again
>

I can confirm this issue, same with our i.MX31 board. Interestingly the 
tftpboot command works, so its probably a DHCP issue.

Helmut


--
Scanned by MailScanner.

[U-Boot] smc911x not functional on top of tree

[Stefano Babic]

On 09/19/2011 11:20 AM, Helmut Raiger wrote:
> On 09/15/2011 09:39 PM, Fabio Estevam wrote:
>> Hi,
>>
>> When using U-boot from top of tree I am not able to get networking to
>> work on a mx31pdk:
>>
>> uboot>  dhcp
>> smc911x: detected LAN9217 controller
>> smc911x: phy initialized
>> smc911x: MAC 00:04:9f:00:8d:6e
>> BOOTP broadcast 1
>> BOOTP broadcast 2
>> DHCP client bound to address 10.29.244.21
>> BOOTP broadcast 3
>> DHCP client bound to address 10.29.244.21
>> BOOTP broadcast 4
>> DHCP client bound to address 10.29.244.21
>> BOOTP broadcast 5
>> DHCP client bound to address 10.29.244.21
>>
>> Retry count exceeded; starting again
>>
> 
> I can confirm this issue, same with our i.MX31 board. Interestingly the 
> tftpboot command works, so its probably a DHCP issue.

Without checking in code, it seems the timout for DHCP elapses before
the answer arrives. This confirms Fabio's tests, that is something
related to the network. Some boards have set CONFIG_ARP_TIMEOUT to a
higher value as default (200 instead of 5 mSec). Maybe it is related,
maybe not...you need to sniff your network ;-)

Best regards,
Stefano Babic

-- 
=====================================================================
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 at denx.de
=====================================================================

[U-Boot] smc911x not functional on top of tree

[Helmut Raiger]

On 09/19/2011 11:20 AM, Helmut Raiger wrote:
> On
> I can confirm this issue, same with our i.MX31 board. Interestingly the
> tftpboot command works, so its probably a DHCP issue.
>
> Helmut

Bisecting the problem, got me:

093498669e77597635a24f326f11efeab213d394 is the first bad commit
commit 093498669e77597635a24f326f11efeab213d394
Author: Simon Glass <sjg@chromium.org>
Date:   Mon Jun 13 16:13:12 2011 -0700

     Put common autoload code into auto_load() function

     This is a small clean-up patch.

     Signed-off-by: Simon Glass <sjg@chromium.org>
     Tested-by: Eric B?nard <eric@eukrea.com>

:040000 040000 cbc4fde5e7708a24f3e5ceb16946996e2b8048a5 
cc66136f66e0fb1e5e1f7e3aef0787f57072a4b1 M    net

Still analyzing ...
Helmut


--
Scanned by MailScanner.

[U-Boot] [PATCH] net: fix a regression in bootp.c

[Helmut Raiger]

This fixes a bug introduced by 093498669e77597635a24f326f11efeab213d394.
TftpStart() was not called unless the 'autoload' environment variable was set.

Signed-off-by: Helmut Raiger <helmut.raiger@hale.at>
---
 net/bootp.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/net/bootp.c b/net/bootp.c
index 3db08ea..3157548 100644
--- a/net/bootp.c
+++ b/net/bootp.c
@@ -164,8 +164,9 @@ static void auto_load(void)
 			return;
 		}
 #endif
-	TftpStart();
 	}
+
+	TftpStart();
 }
 
 #if !defined(CONFIG_CMD_DHCP)
-- 
1.7.4.4



--
Scanned by MailScanner.

[U-Boot] smc911x not functional on top of tree

[Peter Korsgaard]

>>>>> "Helmut" == Helmut Raiger <helmut.raiger@hale.at> writes:

 Helmut> On 09/19/2011 11:20 AM, Helmut Raiger wrote:
 >> On
 >> I can confirm this issue, same with our i.MX31 board. Interestingly the
 >> tftpboot command works, so its probably a DHCP issue.
 >> 
 >> Helmut

 Helmut> Bisecting the problem, got me:

 Helmut> 093498669e77597635a24f326f11efeab213d394 is the first bad commit
 Helmut> commit 093498669e77597635a24f326f11efeab213d394
 Helmut> Author: Simon Glass <sjg@chromium.org>
 Helmut> Date:   Mon Jun 13 16:13:12 2011 -0700

 Helmut>      Put common autoload code into auto_load() function

 Helmut>      This is a small clean-up patch.

 Helmut>      Signed-off-by: Simon Glass <sjg@chromium.org>
 Helmut>      Tested-by: Eric B?nard <eric@eukrea.com>

 Helmut> :040000 040000 cbc4fde5e7708a24f3e5ceb16946996e2b8048a5 
 Helmut> cc66136f66e0fb1e5e1f7e3aef0787f57072a4b1 M    net

 Helmut> Still analyzing ...

Yes, I sent a patch that earlier today:

http://lists.denx.de/pipermail/u-boot/2011-September/101692.html

-- 
Bye, Peter Korsgaard

[U-Boot] smc911x not functional on top of tree

[Helmut Raiger]

On 09/19/2011 03:01 PM, Peter Korsgaard wrote:
>>>>>> "Helmut" == Helmut Raiger<helmut.raiger@hale.at>  writes:
>   Helmut>  On 09/19/2011 11:20 AM, Helmut Raiger wrote:
>   >>  On
>   >>  I can confirm this issue, same with our i.MX31 board. Interestingly the
>   >>  tftpboot command works, so its probably a DHCP issue.
>   >>
>   >>  Helmut
>
>   Helmut>  Bisecting the problem, got me:
>
>   Helmut>  093498669e77597635a24f326f11efeab213d394 is the first bad commit
>   Helmut>  commit 093498669e77597635a24f326f11efeab213d394
>   Helmut>  Author: Simon Glass<sjg@chromium.org>
>   Helmut>  Date:   Mon Jun 13 16:13:12 2011 -0700
>
>   Helmut>       Put common autoload code into auto_load() function
>
>   Helmut>       This is a small clean-up patch.
>
>   Helmut>       Signed-off-by: Simon Glass<sjg@chromium.org>
>   Helmut>       Tested-by: Eric B?nard<eric@eukrea.com>
>
>   Helmut>  :040000 040000 cbc4fde5e7708a24f3e5ceb16946996e2b8048a5
>   Helmut>  cc66136f66e0fb1e5e1f7e3aef0787f57072a4b1 M    net
>
>   Helmut>  Still analyzing ...
>
> Yes, I sent a patch that earlier today:
>
> http://lists.denx.de/pipermail/u-boot/2011-September/101692.html
>
Yes I saw it 10 minutes after sending the patch.
It's already applied, perfect.

Helmut



--
Scanned by MailScanner.