Quantcast

Overo STORM USB Failures

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Overo STORM USB Failures

Hapgood, Ryan (STRX)
We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).

It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:

https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc

and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
Anyone else having these problems?
Is anyone else using these changes?

We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.

Regards,
Ryan Hapgood
Digital Development Manager
       



PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Niek Kooij-2
Hi Ryan,

We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?

Greetings Niek

-----Original Message-----
From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
Sent: maandag 5 augustus 2013 6:18
To: [hidden email]
Subject: [Gumstix-users] Overo STORM USB Failures

We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).

It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:

https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc

and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
Anyone else having these problems?
Is anyone else using these changes?

We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.

Regards,
Ryan Hapgood
Digital Development Manager
       



PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Ash Charles
Hi Ryan,

I've pulled in the kernel patch you cited as well as the corresponding
u-boot patch (http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/165309).
 Can you give these a try to see if they fix the problem?

U-boot
=====
Source: https://github.com/gumstix/u-boot/tree/v2013-overo-usb
Binaries:
* https://s3.amazonaws.com/gumstix-misc/usb/MLO
* https://s3.amazonaws.com/gumstix-misc/usb/u-boot.img

Linux
=====
Source: https://github.com/gumstix/linux/tree/omap-3.5-usb
Binaries:
* https://s3.amazonaws.com/gumstix-misc/usb/uImage
* https://s3.amazonaws.com/gumstix-misc/usb/modules.tar.gz

Checksums: https://s3.amazonaws.com/gumstix-misc/usb/md5sums.txt

--Ash

On Mon, Aug 5, 2013 at 12:50 AM, Niek Kooij <[hidden email]> wrote:

> Hi Ryan,
>
> We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?
>
> Greetings Niek
>
> -----Original Message-----
> From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
> Sent: maandag 5 augustus 2013 6:18
> To: [hidden email]
> Subject: [Gumstix-users] Overo STORM USB Failures
>
> We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).
>
> It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:
>
> https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc
>
> and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
> Anyone else having these problems?
> Is anyone else using these changes?
>
> We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.
>
> Regards,
> Ryan Hapgood
> Digital Development Manager
>
>
>
>
> PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.
>
> ------------------------------------------------------------------------------
> Get your SQL database under version control now!
> Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
> _______________________________________________
> gumstix-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>
> ------------------------------------------------------------------------------
> Get your SQL database under version control now!
> Version control is standard for application code, but databases havent
> caught up. So what steps can you take to put your SQL databases under
> version control? Why should you start doing it? Read more to find out.
> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
> _______________________________________________
> gumstix-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Hapgood, Ryan (STRX)
In reply to this post by Niek Kooij-2
Niek,

I manually applied the changes to a 3.4 Sakoman kernel easily. I would suggest you could do the same to 3.5 with no problems.

>From what we have experienced I would definitely give it a try.

Regards,
Ryan.

-----Original Message-----
From: Niek Kooij [mailto:[hidden email]]
Sent: Monday, 5 August 2013 5:51 PM
To: General mailing list for gumstix users.
Subject: Re: [Gumstix-users] Overo STORM USB Failures

Hi Ryan,

We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?

Greetings Niek

-----Original Message-----
From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
Sent: maandag 5 augustus 2013 6:18
To: [hidden email]
Subject: [Gumstix-users] Overo STORM USB Failures

We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).

It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:

https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc

and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
Anyone else having these problems?
Is anyone else using these changes?

We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.

Regards,
Ryan Hapgood
Digital Development Manager
       



PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

______________________________________________________________________
CAUTION: This message was sent via the Public Internet and its authenticity cannot be guaranteed.

PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Niek Kooij-2
Yes, i also manually applied the patch and that was pretty straightforward. Never had any problems during boot so I leave the uboot patch as suggested by Ash out. Ill keep you updated if we see the problems again.

Greetings Niek

-----Original Message-----
From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
Sent: maandag 5 augustus 2013 23:55
To: General mailing list for gumstix users.
Subject: Re: [Gumstix-users] Overo STORM USB Failures

Niek,

I manually applied the changes to a 3.4 Sakoman kernel easily. I would suggest you could do the same to 3.5 with no problems.

>From what we have experienced I would definitely give it a try.

Regards,
Ryan.

-----Original Message-----
From: Niek Kooij [mailto:[hidden email]]
Sent: Monday, 5 August 2013 5:51 PM
To: General mailing list for gumstix users.
Subject: Re: [Gumstix-users] Overo STORM USB Failures

Hi Ryan,

We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?

Greetings Niek

-----Original Message-----
From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
Sent: maandag 5 augustus 2013 6:18
To: [hidden email]
Subject: [Gumstix-users] Overo STORM USB Failures

We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).

It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:

https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc

and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
Anyone else having these problems?
Is anyone else using these changes?

We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.

Regards,
Ryan Hapgood
Digital Development Manager
       



PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

______________________________________________________________________
CAUTION: This message was sent via the Public Internet and its authenticity cannot be guaranteed.

PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Peter A. Bigot
In reply to this post by Ash Charles
I've been seeing USB problems where the device list reloads every 60
seconds, but only for non-graphical environments: once the display
modules are loaded it starts working.  I do not get an EMI-related
warning, just repeated disconnects and reconnects.

The u-boot patch fixes this problem for me.

The Linux patch does not fix my problem.

They both "do the same thing", but u-boot writes 0x1bb05 to
CM_CLKSEL4_PLL while Linux writes 0x1bc05.

On closer examination, the Linux patch is wrong: it improperly adds 1 to
the multiplier value, while it should be doing this only for the n
value.  This is confirmed by reviewing
https://github.com/RobertCNelson/armv7-multiplatform/pull/3 which does
not adjust m.

So while the linux may be fixing the EMI problem, it's not doing it
quite correctly.

I'll add both the u-boot and a corrected Linux patch to my queue for
gumstix dylan.  Assuming it gets merged I'd appreciate it if people
would try the new kernel just to make sure it's right. (Alternatively,
pre-check it by using the existing patch but with 433 instead of 433+1
as the m values.)

Peter


On 08/05/2013 02:21 PM, Ash Charles wrote:

> Hi Ryan,
>
> I've pulled in the kernel patch you cited as well as the corresponding
> u-boot patch (http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/165309).
>   Can you give these a try to see if they fix the problem?
>
> U-boot
> =====
> Source: https://github.com/gumstix/u-boot/tree/v2013-overo-usb
> Binaries:
> * https://s3.amazonaws.com/gumstix-misc/usb/MLO
> * https://s3.amazonaws.com/gumstix-misc/usb/u-boot.img
>
> Linux
> =====
> Source: https://github.com/gumstix/linux/tree/omap-3.5-usb
> Binaries:
> * https://s3.amazonaws.com/gumstix-misc/usb/uImage
> * https://s3.amazonaws.com/gumstix-misc/usb/modules.tar.gz
>
> Checksums: https://s3.amazonaws.com/gumstix-misc/usb/md5sums.txt
>
> --Ash
>
> On Mon, Aug 5, 2013 at 12:50 AM, Niek Kooij <[hidden email]> wrote:
>> Hi Ryan,
>>
>> We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?
>>
>> Greetings Niek
>>
>> -----Original Message-----
>> From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
>> Sent: maandag 5 augustus 2013 6:18
>> To: [hidden email]
>> Subject: [Gumstix-users] Overo STORM USB Failures
>>
>> We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).
>>
>> It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:
>>
>> https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc
>>
>> and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
>> Anyone else having these problems?
>> Is anyone else using these changes?
>>
>> We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.
>>
>> Regards,
>> Ryan Hapgood
>> Digital Development Manager
>>
>>
>>
>>
>> PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.
>>
>> ------------------------------------------------------------------------------
>> Get your SQL database under version control now!
>> Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>> _______________________________________________
>> gumstix-users mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>>
>> ------------------------------------------------------------------------------
>> Get your SQL database under version control now!
>> Version control is standard for application code, but databases havent
>> caught up. So what steps can you take to put your SQL databases under
>> version control? Why should you start doing it? Read more to find out.
>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>> _______________________________________________
>> gumstix-users mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
> ------------------------------------------------------------------------------
> Get your SQL database under version control now!
> Version control is standard for application code, but databases havent
> caught up. So what steps can you take to put your SQL databases under
> version control? Why should you start doing it? Read more to find out.
> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
> _______________________________________________
> gumstix-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gumstix-users


------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Peter A. Bigot
A correction for the record:

The u-boot patch does not solve my problem; it eliminates the periodic
disconnects by completely breaking USB, and will do so whenever it
configures a 960MHz dpll5 with a divide-by-eight and then passes control
to a Linux that sets the default dpll5 120MHz and leaves the
divide-by-eight untouched under the assumption it's still at its
power-up value of divide-by-one.  This is what happens with Linux 3.5
and probably most other releases.

The u-boot patch should not be applied unless you need usb to function
in u-boot.  If so, you'll also need to either reconfigure CM_CLKSEL5_PLL
back to 1 before handing off control to Linux, or only boot a Linux that
also configures dpll5's divider.

Peter

On 08/14/2013 05:05 PM, Peter A. Bigot wrote:

> I've been seeing USB problems where the device list reloads every 60
> seconds, but only for non-graphical environments: once the display
> modules are loaded it starts working.  I do not get an EMI-related
> warning, just repeated disconnects and reconnects.
>
> The u-boot patch fixes this problem for me.
>
> The Linux patch does not fix my problem.
>
> They both "do the same thing", but u-boot writes 0x1bb05 to
> CM_CLKSEL4_PLL while Linux writes 0x1bc05.
>
> On closer examination, the Linux patch is wrong: it improperly adds 1 to
> the multiplier value, while it should be doing this only for the n
> value.  This is confirmed by reviewing
> https://github.com/RobertCNelson/armv7-multiplatform/pull/3 which does
> not adjust m.
>
> So while the linux may be fixing the EMI problem, it's not doing it
> quite correctly.
>
> I'll add both the u-boot and a corrected Linux patch to my queue for
> gumstix dylan.  Assuming it gets merged I'd appreciate it if people
> would try the new kernel just to make sure it's right. (Alternatively,
> pre-check it by using the existing patch but with 433 instead of 433+1
> as the m values.)
>
> Peter
>
>
> On 08/05/2013 02:21 PM, Ash Charles wrote:
>> Hi Ryan,
>>
>> I've pulled in the kernel patch you cited as well as the corresponding
>> u-boot patch (http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/165309).
>>    Can you give these a try to see if they fix the problem?
>>
>> U-boot
>> =====
>> Source: https://github.com/gumstix/u-boot/tree/v2013-overo-usb
>> Binaries:
>> * https://s3.amazonaws.com/gumstix-misc/usb/MLO
>> * https://s3.amazonaws.com/gumstix-misc/usb/u-boot.img
>>
>> Linux
>> =====
>> Source: https://github.com/gumstix/linux/tree/omap-3.5-usb
>> Binaries:
>> * https://s3.amazonaws.com/gumstix-misc/usb/uImage
>> * https://s3.amazonaws.com/gumstix-misc/usb/modules.tar.gz
>>
>> Checksums: https://s3.amazonaws.com/gumstix-misc/usb/md5sums.txt
>>
>> --Ash
>>
>> On Mon, Aug 5, 2013 at 12:50 AM, Niek Kooij <[hidden email]> wrote:
>>> Hi Ryan,
>>>
>>> We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?
>>>
>>> Greetings Niek
>>>
>>> -----Original Message-----
>>> From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
>>> Sent: maandag 5 augustus 2013 6:18
>>> To: [hidden email]
>>> Subject: [Gumstix-users] Overo STORM USB Failures
>>>
>>> We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).
>>>
>>> It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:
>>>
>>> https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc
>>>
>>> and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
>>> Anyone else having these problems?
>>> Is anyone else using these changes?
>>>
>>> We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.
>>>
>>> Regards,
>>> Ryan Hapgood
>>> Digital Development Manager
>>>
>>>
>>>
>>>
>>> PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.
>>>
>>> ------------------------------------------------------------------------------
>>> Get your SQL database under version control now!
>>> Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
>>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>>> _______________________________________________
>>> gumstix-users mailing list
>>> [hidden email]
>>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>>>
>>> ------------------------------------------------------------------------------
>>> Get your SQL database under version control now!
>>> Version control is standard for application code, but databases havent
>>> caught up. So what steps can you take to put your SQL databases under
>>> version control? Why should you start doing it? Read more to find out.
>>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>>> _______________________________________________
>>> gumstix-users mailing list
>>> [hidden email]
>>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>> ------------------------------------------------------------------------------
>> Get your SQL database under version control now!
>> Version control is standard for application code, but databases havent
>> caught up. So what steps can you take to put your SQL databases under
>> version control? Why should you start doing it? Read more to find out.
>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>> _______________________________________________
>> gumstix-users mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>
> ------------------------------------------------------------------------------
> Get 100% visibility into Java/.NET code with AppDynamics Lite!
> It's a free troubleshooting tool designed for production.
> Get down to code-level detail for bottlenecks, with <2% overhead.
> Download for free and get started troubleshooting in minutes.
> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> _______________________________________________
> gumstix-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gumstix-users


------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Steve Modica
In reply to this post by Peter A. Bigot
Can someone point me to something that talks about the "EMI problems" mentioned below with Storm boards?

I have a USB device that works reliably on non-storm boards and fails reliably on storm boards.  While it always fails, sometimes it enumerates, sometimes it sees a little receive data, sometimes it gets tons of errors and disappears. My device puts out a lot of EMI too.

Steve

On Aug 14, 2013, at 5:05 PM, Peter A. Bigot <[hidden email]> wrote:

So while the linux may be fixing the EMI problem, it's not doing it 
quite correctly.

--
Steve Modica
CTO -  Small Tree Communications
phone: 651-209-6509 ext 301
mobile: 651-261-3201







------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Overo STORM USB Failures

Ash Charles
In reply to this post by Peter A. Bigot
Hi,

The patch in linux seems to be working so I've pushed the usb host fix
to the default overo branch
(https://github.com/gumstix/linux/tree/omap-3.5) and deleted the
omap-3.5-usb branch.  I will not add the USB patch to u-boot because
of the problems Peter noted.

--Ash

On Wed, Aug 14, 2013 at 8:07 PM, Peter A. Bigot <[hidden email]> wrote:

> A correction for the record:
>
> The u-boot patch does not solve my problem; it eliminates the periodic
> disconnects by completely breaking USB, and will do so whenever it
> configures a 960MHz dpll5 with a divide-by-eight and then passes control
> to a Linux that sets the default dpll5 120MHz and leaves the
> divide-by-eight untouched under the assumption it's still at its
> power-up value of divide-by-one.  This is what happens with Linux 3.5
> and probably most other releases.
>
> The u-boot patch should not be applied unless you need usb to function
> in u-boot.  If so, you'll also need to either reconfigure CM_CLKSEL5_PLL
> back to 1 before handing off control to Linux, or only boot a Linux that
> also configures dpll5's divider.
>
> Peter
>
> On 08/14/2013 05:05 PM, Peter A. Bigot wrote:
>> I've been seeing USB problems where the device list reloads every 60
>> seconds, but only for non-graphical environments: once the display
>> modules are loaded it starts working.  I do not get an EMI-related
>> warning, just repeated disconnects and reconnects.
>>
>> The u-boot patch fixes this problem for me.
>>
>> The Linux patch does not fix my problem.
>>
>> They both "do the same thing", but u-boot writes 0x1bb05 to
>> CM_CLKSEL4_PLL while Linux writes 0x1bc05.
>>
>> On closer examination, the Linux patch is wrong: it improperly adds 1 to
>> the multiplier value, while it should be doing this only for the n
>> value.  This is confirmed by reviewing
>> https://github.com/RobertCNelson/armv7-multiplatform/pull/3 which does
>> not adjust m.
>>
>> So while the linux may be fixing the EMI problem, it's not doing it
>> quite correctly.
>>
>> I'll add both the u-boot and a corrected Linux patch to my queue for
>> gumstix dylan.  Assuming it gets merged I'd appreciate it if people
>> would try the new kernel just to make sure it's right. (Alternatively,
>> pre-check it by using the existing patch but with 433 instead of 433+1
>> as the m values.)
>>
>> Peter
>>
>>
>> On 08/05/2013 02:21 PM, Ash Charles wrote:
>>> Hi Ryan,
>>>
>>> I've pulled in the kernel patch you cited as well as the corresponding
>>> u-boot patch (http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/165309).
>>>    Can you give these a try to see if they fix the problem?
>>>
>>> U-boot
>>> =====
>>> Source: https://github.com/gumstix/u-boot/tree/v2013-overo-usb
>>> Binaries:
>>> * https://s3.amazonaws.com/gumstix-misc/usb/MLO
>>> * https://s3.amazonaws.com/gumstix-misc/usb/u-boot.img
>>>
>>> Linux
>>> =====
>>> Source: https://github.com/gumstix/linux/tree/omap-3.5-usb
>>> Binaries:
>>> * https://s3.amazonaws.com/gumstix-misc/usb/uImage
>>> * https://s3.amazonaws.com/gumstix-misc/usb/modules.tar.gz
>>>
>>> Checksums: https://s3.amazonaws.com/gumstix-misc/usb/md5sums.txt
>>>
>>> --Ash
>>>
>>> On Mon, Aug 5, 2013 at 12:50 AM, Niek Kooij <[hidden email]> wrote:
>>>> Hi Ryan,
>>>>
>>>> We also have been seeing USB problems (sometimes ending with the "disabled by hub (EMI?)" message). We are using the FireSTORM and the IronSTORM with the 3.5 Sakoman kernel. We attach a lot of USB devices to our system (via a powered hub) so we were blaming that part for the problems :).  This patch is for 3.5 Sakoman?
>>>>
>>>> Greetings Niek
>>>>
>>>> -----Original Message-----
>>>> From: Hapgood, Ryan (STRX) [mailto:[hidden email]]
>>>> Sent: maandag 5 augustus 2013 6:18
>>>> To: [hidden email]
>>>> Subject: [Gumstix-users] Overo STORM USB Failures
>>>>
>>>> We are experiencing USB failures ("disabled by hub (EMI?)") on a significant number of WaterSTORM COMs (~10% in manufacturing) and it is starting to look like it's caused by the clock issues described in the DM3730 & DM3725 Silicon Errata (SPRZ319E) (P112).
>>>>
>>>> It seems that some COMs are less susceptible than others but according to the errata a 26MHz system clock the problem cannot be truly eliminated, just reduced. To reduce the problem I have tried the following kernel changes:
>>>>
>>>> https://bitbucket.org/sr105/linux-sakoman/commits/01cfa0831c55496c27ee3fc6a1141ad73f4020fc
>>>>
>>>> and things are looking promising but I have noticed that these changes are not in Steve Sakoman's kernel by default.
>>>> Anyone else having these problems?
>>>> Is anyone else using these changes?
>>>>
>>>> We are dealing with Gumstix on this matter also, but I thought others may benefit from the heads up.
>>>>
>>>> Regards,
>>>> Ryan Hapgood
>>>> Digital Development Manager
>>>>
>>>>
>>>>
>>>>
>>>> PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged.  It is intended for the recipient only.  If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail.  If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Get your SQL database under version control now!
>>>> Version control is standard for application code, but databases havent caught up. So what steps can you take to put your SQL databases under version control? Why should you start doing it? Read more to find out.
>>>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>>>> _______________________________________________
>>>> gumstix-users mailing list
>>>> [hidden email]
>>>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Get your SQL database under version control now!
>>>> Version control is standard for application code, but databases havent
>>>> caught up. So what steps can you take to put your SQL databases under
>>>> version control? Why should you start doing it? Read more to find out.
>>>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>>>> _______________________________________________
>>>> gumstix-users mailing list
>>>> [hidden email]
>>>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>>> ------------------------------------------------------------------------------
>>> Get your SQL database under version control now!
>>> Version control is standard for application code, but databases havent
>>> caught up. So what steps can you take to put your SQL databases under
>>> version control? Why should you start doing it? Read more to find out.
>>> http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
>>> _______________________________________________
>>> gumstix-users mailing list
>>> [hidden email]
>>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>>
>> ------------------------------------------------------------------------------
>> Get 100% visibility into Java/.NET code with AppDynamics Lite!
>> It's a free troubleshooting tool designed for production.
>> Get down to code-level detail for bottlenecks, with <2% overhead.
>> Download for free and get started troubleshooting in minutes.
>> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
>> _______________________________________________
>> gumstix-users mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gumstix-users
>
>
> ------------------------------------------------------------------------------
> Get 100% visibility into Java/.NET code with AppDynamics Lite!
> It's a free troubleshooting tool designed for production.
> Get down to code-level detail for bottlenecks, with <2% overhead.
> Download for free and get started troubleshooting in minutes.
> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
> _______________________________________________
> gumstix-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gumstix-users

------------------------------------------------------------------------------
Introducing Performance Central, a new site from SourceForge and
AppDynamics. Performance Central is your source for news, insights,
analysis and resources for efficient Application Performance Management.
Visit us today!
http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
_______________________________________________
gumstix-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gumstix-users
Loading...