[U-Boot-Users] MPC8360 Rev 2 patch?

Hello--
I read the recent post regarding the MPC8360 rev 2 patch that is currently in development for u-boot 1.2.0. How soon will that patch be available (or the code submitted to the tree)? I have a custom board arriving at the end of the week that will have an 8360 rev2....and it has DDR2 memory so I'll need DDR2 support. Is there somewhere that I can download the latest bits to get started?
Thanks, Steve

On Mon, 05 Feb 2007 17:15:21 -0600 Steven Hein ssh@sgi.com wrote:
I read the recent post regarding the MPC8360 rev 2 patch that is currently in development for u-boot 1.2.0. How soon will that patch be available (or the code submitted to the tree)? I have a custom board arriving at the end of the week that will have an 8360 rev2....and it has DDR2 memory so I'll need DDR2 support. Is there somewhere that I can download the latest bits to get started?
it'll be in the tree before you get your board :)
Kim

Kim Phillips wrote:
On Mon, 05 Feb 2007 17:15:21 -0600 Steven Hein ssh@sgi.com wrote:
I read the recent post regarding the MPC8360 rev 2 patch that is currently in development for u-boot 1.2.0. How soon will that patch be available (or the code submitted to the tree)? I have a custom board arriving at the end of the week that will have an 8360 rev2....and it has DDR2 memory so I'll need DDR2 support. Is there somewhere that I can download the latest bits to get started?
it'll be in the tree before you get your board :)
Kim
Hi Kim,
I just received word that my board will be in my hands at tomorrow (Friday) morning. Will the 8360 rev.2 support be available?
Thanks, Steve

On Thu, 08 Feb 2007 10:28:09 -0600 Steven Hein ssh@sgi.com wrote:
I just received word that my board will be in my hands at tomorrow (Friday) morning. Will the 8360 rev.2 support be available?
it's already in the 'mpc83xx' branch on the opensource.freescale.com u-boot-83xx.git tree. DDR2 seems to be working fine.
I hesitated to broadcast this info since I haven't finished testing it on our boards, plus the patches haven't been posted here yet.
For future reference, the stuff considered stable on mpc83xx will be pulled into the 'upstream' branch. Once Rev.2 support tests ok, I'll ask WD to pull from the 'upstream' branch.
Kim

Kim Phillips wrote:
On Thu, 08 Feb 2007 10:28:09 -0600 Steven Hein ssh@sgi.com wrote:
I just received word that my board will be in my hands at tomorrow (Friday) morning. Will the 8360 rev.2 support be available?
it's already in the 'mpc83xx' branch on the opensource.freescale.com u-boot-83xx.git tree. DDR2 seems to be working fine.
I hesitated to broadcast this info since I haven't finished testing it on our boards, plus the patches haven't been posted here yet.
For future reference, the stuff considered stable on mpc83xx will be pulled into the 'upstream' branch. Once Rev.2 support tests ok, I'll ask WD to pull from the 'upstream' branch.
Kim
Thanks Kim! After a little bit of "git for dummies", I figured out how to grab the mpc83xx branch. But before your mail, I was just looking in WD's git tree for the changes--I didn't know that freescale had its own git trees for u-boot.
One question about the 8360 version. When I compile it straight up for the MPC8360EMDS_HOST_66 configuration, I never see any output at the startup of U-boot. But if I comment out the open firmware defines in the include/configs/MPC8360EMDS.h config file, then all works. Why doesn't this work for me straight up on the board it was compiled for? (This has been true for all U-boot versions I've tried for the 8360).
Thanks, Steve

On Thu, 08 Feb 2007 16:50:50 -0600 Steven Hein ssh@sgi.com wrote:
Thanks Kim! After a little bit of "git for dummies", I figured out how to grab the mpc83xx branch. But before your mail, I was just looking in WD's git tree for the changes--I didn't know that freescale had its own git trees for u-boot.
it was set up primarily for WD's convenience, so he could just pull from it. It's also useful to have avant garde mpc83xx users such as yourself get 'early access.' Ultimately, though, an mpc83xx tree should be hosted on denx.de.
One question about the 8360 version. When I compile it straight up for the MPC8360EMDS_HOST_66 configuration, I never see any output at the startup of U-boot. But if I comment out the open firmware defines in the include/configs/MPC8360EMDS.h config file, then all works. Why doesn't this work for me straight up on the board it was compiled for? (This has been true for all U-boot versions I've tried for the 8360).
that's odd - MPC8360EMDS_HOST_66 works fine here and there's nothing obvious I can see in the code that would warrant that sort of behaviour. Can you provide details?
Kim

Kim Phillips wrote:
One question about the 8360 version. When I compile it straight up for the MPC8360EMDS_HOST_66 configuration, I never see any output at the startup of U-boot. But if I comment out the open firmware defines in the include/configs/MPC8360EMDS.h config file, then all works. Why doesn't this work for me straight up on the board it was compiled for? (This has been true for all U-boot versions I've tried for the 8360).
that's odd - MPC8360EMDS_HOST_66 works fine here and there's nothing obvious I can see in the code that would warrant that sort of behaviour. Can you provide details?
I think the problem is in his device tree. Either it's missing or incorrect.

Timur Tabi wrote:
Kim Phillips wrote:
One question about the 8360 version. When I compile it straight up for the MPC8360EMDS_HOST_66 configuration, I never see any output at the startup of U-boot. But if I comment out the open firmware defines in the include/configs/MPC8360EMDS.h config file, then all works. Why doesn't this work for me straight up on the board it was compiled for? (This has been true for all U-boot versions I've tried for the 8360).
that's odd - MPC8360EMDS_HOST_66 works fine here and there's nothing obvious I can see in the code that would warrant that sort of behaviour. Can you provide details?
I think the problem is in his device tree. Either it's missing or incorrect.
I'm sure that you are right, Timur. Every version of U-boot I've tried to use, all the way back to 1.3.x, for the 8360 as well as the 8349EMDS, has had the same problem. And as soon as I comment out the OF stuff in the config file, all is well. (I haven't looked into it in detail, so I can't provide a lot of info at the moment. ......and my new custom MPC8358 board arrived just about an your ago! Friday afternoon, right on time! ;-) ).
At this point, I know *nothing* about device trees or OF...... Is this something that lives outside of u-boot?
I guess someone probably needs to point me to some "device trees for dummies" info........

Hi Steve,
Try Documentation/powerpc/booting-without-of.txt in the Linux tree. You can find some device trees in arch/powerpc/boot/dts. You can get/clone the dtc compiler from git://www.jdl.com/software/dtc.git.
Cheers, Emil.
This e-mail, and any associated attachments have been classified as: ==================================================================== [x] Public [ ] Freescale Semiconductor Internal Use Only [ ] Freescale Semiconductor Confidential Proprietary
-----Original Message----- From: u-boot-users-bounces@lists.sourceforge.net [mailto:u-boot-users-bounces@lists.sourceforge.net] On Behalf Of Steven Hein Sent: Friday, February 09, 2007 3:44 PM To: Tabi Timur-B04825 Cc: u-boot-users@lists.sourceforge.net; Phillips Kim-R1AAHA Subject: Re: [U-Boot-Users] MPC8360 Rev 2 patch?
Timur Tabi wrote:
Kim Phillips wrote:
One question about the 8360 version. When I compile it straight up for the MPC8360EMDS_HOST_66 configuration, I never see any output at the startup of U-boot. But if I comment out the open firmware defines in the include/configs/MPC8360EMDS.h config file, then all works. Why doesn't this work for me straight up on the board it was compiled for? (This has been true for all U-boot versions I've tried for the
8360).
that's odd - MPC8360EMDS_HOST_66 works fine here and there's nothing obvious I can see in the code that would warrant that sort of behaviour. Can you provide details?
I think the problem is in his device tree. Either it's missing or incorrect.
I'm sure that you are right, Timur. Every version of U-boot I've tried to use, all the way back to 1.3.x, for the 8360 as well as the 8349EMDS, has had the same problem. And as soon as I comment out the OF stuff in the config file, all is well. (I haven't looked into it in detail, so I can't provide a lot of info at the moment. ......and my new custom MPC8358 board arrived just about an your ago! Friday afternoon, right on time! ;-) ).
At this point, I know *nothing* about device trees or OF...... Is this something that lives outside of u-boot?
I guess someone probably needs to point me to some "device trees for dummies" info........

Steven Hein wrote:
At this point, I know *nothing* about device trees or OF...... Is this something that lives outside of u-boot?
Yes. The device tree source is in arch/powerpc/boot/dts in the kernel. You compile that with dtc into a dtb, transfer the dtb to U-Boot via tftp, and then boot the kernel, passing the dtb as the third parameter.
I guess someone probably needs to point me to some "device trees for dummies" info........
I don't know where such a document exists.

Steven,
On Fri, 2007-02-09 at 15:53 -0600, Timur Tabi wrote:
Yes. The device tree source is in arch/powerpc/boot/dts in the kernel. You compile that with dtc into a dtb, transfer the dtb to U-Boot via tftp, and then boot the kernel, passing the dtb as the third parameter.
If you want your board to be self-contained, put the .dtb file in a flash sector. I expect this is how most people will use it 'in the field'. As Timur alludes to, the boot command is:
bootm <kernel address> <initrd address> <dtb address>
If you don't have an initrd, use '-'
If you're still confused I'd be happy to help with the details.
This whole U-boot/device tree/linux process is sorely lacking in documentation. The 'booting-without-of.txt' file is nice and all, but it's jibberish to people who are just starting out.
Having figured this out the hard way, maybe I'll put together a 'for dummies' document.
regards, Ben
participants (5)
-
Ben Warren
-
Kim Phillips
-
Medve Emilian-EMMEDVE1
-
Steven Hein
-
Timur Tabi