Does the Latest BeagleBoard-X15 Image Run Out of the Box on The Sitara am571x IDK?

Jeff_Andich1 · June 6, 2017, 7:15pm

Hi,

Does the BeagleBoard-X15 image run, out of the box on the 5718 Industrial Development Kit (IDK) in addition to the 572x EVM or are configuration changes needed to the image to make it compatible with the 5718 IDK?

Has anyone tried this specific configuration with the BB-X15 image? We’re running with the BeagleBoard-X15 image on the 572x EVM, literally out of the box. Debian on the BB-X15 image facilitates testing by our software team.

We’re at a crucial decision point on whether to design-in the am5718 vs the am5728 chip on a board.

Any information is greatly appreciated!!

Thanks!

RobertCNelson · June 6, 2017, 7:19pm

Hi Jeff,

Hi,

Does the BeagleBoard-X15 image run, out of the box on the 5718 Industrial
Development Kit (IDK) in addition to the 572x EVM or are configuration
changes needed to the image to make it compatible with the 5718 IDK?

Has anyone tried this specific configuration with the BB-X15 image? We're
running with the BeagleBoard-X15 image on the 572x EVM, literally out of the
box. Debian on the BB-X15 image facilitates testing by our software team.

We're at a crucial decision point on whether to design-in the am5718 vs the
am5728 chip on a board.

Any information is greatly appreciated!!

Give the 2017-06-04 snapshot a shot:

https://rcn-ee.net/rootfs/bb.org/testing/2017-06-04/

There's one am571x based device off the x15 that i know it personally
boots just fine on..

the 5718 idk 'should' work, but please test it and give us a serial
boot log if there is is issues.

Regards,

Jeff_Andich1 · June 6, 2017, 7:43pm

Thanks a lot Robert!!

Will update what happens here… Regards, Jeff

Jeff_Andich1 · June 9, 2017, 8:25pm

Hi Robert,

Here’s what I’m seeing so far, what I’ve tried, and what’s coming to mind to try to get the debian 8.8 image going on the 571x IDK:

*When our 5718 IDK, is booting the bbx15-debian-8.8-lxqt-4gb-armhf-2017-06-04-4gb.img image from SD Card, the console output dies shortly after the SPL startup string (see below).

*The debian 8.8 LXQT image boots fine on out 572x EVM (console output of 2017.01 u-boot shown below).

*The latest TI SDK image, am57xx-evm-03.03.00.04 boots ok on both the 5718 IDK and 5728 EVM (boot console printed below).

After thinking the problem could be related to pinmux issues, I built the 2017.01 u-boot , but substituted the TI SDK’s pad config. file, mux_data.h for the mux_data.h in 2017.01 u-boot. I then “dd-ed” just the MLO and u-boot.img to an SD Card ( per https://eewiki.net/display/linuxonarm/BeagleBoard-X15#BeagleBoard-X15-SetupmicroSDcard), and booted the 5718 IDK, but I apparently get the same result (console dies after SPL).

*Am not sure what to try next, but am wondering if getting u-boot to load the am571x-idk.dtb instead of am57xx-evm-reva3.dtb would help. I can’t see whether the TI SDK image is loading a different dtb for the 571x vs the 572x as it doesn’t appear to advertise this during bootup.

Also wondering if it has anything to do with the different DDR3 configuration on the 572x EVM (4 DDR3 chips on 2 EMIFs) vs. 571X IDK (2 DDR3 chips on 1 EMIF).

Please let me know what to try next…

Thanks a lot in advance!! jeff

Following are the different boot console outputs of interest:

Jeff_Andich1 · June 11, 2017, 4:54pm

in the 2017.01 u-boot, I attempted to change the default device tree that u-boot is using to see if that would allow u-boot to boot on the 5718 IDK, but the console output still stops just after the SPL prints its output (please see below). I get the same result regardless of whether I use the 2017.01 vs TI SDK’s board/ti/am57xx/ IO pad configuration in mux_data.h. I believe that for all of the builds I have tried, 2017.01 u-boot seems to boot on the 572x EVM.

To attempt to change the device tree which u-boot uses:

In configs/am57xx_evm_defconfig:

changed CONFIG_DEFAULT_DEVICE_TREE=“am572x-idk” to

CONFIG_DEFAULT_DEVICE_TREE=“am57xx-beagle-x15” /* I did this because TI SDK’s u-boot assigns the am57xx-beagle-x15 device tree name */

Then re-built the 2017.01 u-boot.

Note: When I re-built 2017.01 u-boot, I re-used the am57xx-beagle-x15 dts file within the 2017.01 u-boot file tree. I did not copy
the am57xx-beagle-x15.dts or #included dtsi files from the TI SDK file tree. The dts files with the same name are different in the two trees.

Am I barking up the right tree?
Robert, when you ran the BeagleBoard-X15 image on the 571x IDK, what u-boot defconfig, u-boot device tree, and pad mux did you use?
Does this console output mean that SPL has jumped to u-boot, but u-boot has not started up correctly?

Thanks a lot in advance!!!

Console Output:

RobertCNelson · June 11, 2017, 5:07pm

in the 2017.01 u-boot, I attempted to change the default device tree that
u-boot is using to see if that would allow u-boot to boot on the 5718 IDK,
but the console output still stops just after the SPL prints its output
(please see below). I get the same result regardless of whether I use the
2017.01 vs TI SDK's board/ti/am57xx/ IO pad configuration in mux_data.h. I
believe that for all of the builds I have tried, 2017.01 u-boot seems to
boot on the 572x EVM.

To attempt to change the device tree which u-boot uses:

In configs/am57xx_evm_defconfig:

changed CONFIG_DEFAULT_DEVICE_TREE="am572x-idk" to

CONFIG_DEFAULT_DEVICE_TREE="am57xx-beagle-x15" /* I did this because TI
SDK's u-boot assigns the am57xx-beagle-x15 device tree name */

(just noticed before hitting reply) Oh fudge, did they fubar the eeprom?

Then re-built the 2017.01 u-boot.

Note: When I re-built 2017.01 u-boot, I re-used the am57xx-beagle-x15 dts
file within the 2017.01 u-boot file tree. I did not copy
the am57xx-beagle-x15.dts or #included dtsi files from the TI SDK file tree.
The dts files with the same name are different in the two trees.

1) Am I barking up the right tree?

i'd expect it to just work with ti's tree...

2) Robert, when you ran the BeagleBoard-X15 image on the 571x IDK, what
u-boot defconfig, u-boot device tree, and pad mux did you use?

It was a am571x "based" board, not the IDK (i don't have the idk)..
It boots the exact same X15 image..

3) Does this console output mean that SPL has jumped to u-boot, but u-boot
has not started up correctly?

Thanks a lot in advance!!!!

Console Output:
**************************

U-Boot SPL 2017.01-00319-g9fb6b936b8 (May 16 2017 - 10:31:53)
DRA722-GP ES2.0
Trying to boot from MMC1

** Unable to use mmc 0:1 for loading the env **
Using defau environment

This means you only got into the MLO file (u-boot spl), it never made
the jump to u-boot.img

i think your stuck here:

http://git.ti.com/gitweb/?p=ti-u-boot/ti-u-boot.git;a=blob;f=board/ti/am57xx/board.c;h=b8798d6f2a24c58a24ace604939182fa49d2f883;hb=refs/heads/ti-u-boot-2017.01#l1050

change the -1 to a 0, it should catch the

1046 } else if (board_is_am571x_idk() && !strcmp(name, "am571x-idk")) {
1047 return 0;
1048 }

but it's not..

Regards,

Jeff_Andich1 · June 11, 2017, 6:03pm

Thanks Robert!!!

The information/education we’re getting from this site is really helpful!!

Changed board.c::board_fit_config_name_match(…) to always return ‘0’, but the MLO/SPL appears to be stuck at the same place…

Since the MLO/SPL, is printing out its startup string on console, tomorrow I will try to print out the MLO/SPL’s board ID to the console or some other debug interface, and then update this thread if successful.

Also, I may be dd’ing the MLO and u-boot image to the SD card incorrectly, as I get

Created a new DOS disklabel with disk identifier 0xad1707b9.

Start sector 8192 out of range.

Failed to add partition: Numerical result out of range

After the sudo sfdisk step…

Regards,

jeff

Jeff_Andich1 · June 12, 2017, 8:45pm

Hi Robert,

Yesterday, I built u-boot wrong after making your recommended change to board/ti/am57xx/board.c. The build date on my u-boot from yesterday was showing May 16, 2017.

When I made your recommended change in another 2017.01 u-boot tree on another VM, the change allowed my 571x/5718 IDK to get to u-boot (see below).

I’m sorry, I should have noticed the build date issue yesterday!!

My console output thus far follows. Please note that this build of u-boot pulls in the default device tree for 2017.01 u-boot CONFIG_DEFAULT_DEVICE_TREE=“am572x-idk” and the TI SDK’s mux_data.h file

!!! Thanks a lot for your help yesterday !!!

I think our board engineer is designing in the 5728 at this point, but I’ll relay these results to the manager and team.

**** Console Output after making change.

U-Boot SPL 2017.01-00318-g9fd6070-dirty (Jun 12 2017 - 15:31:55)
DRA722-GP ES2.0
Trying to boot from MMC1
** First descriptor is NOT a primary desc on 0:1 **
** No partion table - mmc 0 **
Using default environment

U-Boot 2017.01-00318-g9fd6070-dirty (Jun 12 2017 - 15:31:5-0500)

CPU : DRA722-GP ES2.0
Model: TI AM5728 BeagleBoard15
Board: AM571x IDK REV 1.3B
DRAM: 1 GiB
MMC: OMAP SD/MMC: 0, OMAP SD/MMC: 1
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** rst descriptor is NOT a primary desc on 0:1 **
** No partition table - mmc 0 **
Usindefault environment

setup_board_eeprom_env: am571x_idk
am5_idk_lcd_detect: Touch screen detect failed: 1!
SCSI: SATA link 0 timeout.
AHCI 0001.0300 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
fgs: 64bit ncq stag pm led clo only pmp pio slum part ccc apst scanning bus for devices…
Found 0 device(s).
Net: not set. Validating first E-fuse MAC
cpsw
Press SCE to abort autoboot in 2 seconds
usb_boot is currently disabled
scsi_boot is currently disabled
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** rst descriptor is NOT a primary desc on 0:1 **
switch to partions #0, OK
mmc0 is current device
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a priry desc on 0:1 **
Scanning mmc device 0
Checking for: /uEnv.t …
** No partition table - mmc 0 **
Checking for: /boot/uEnv.txt …
** No partition table - mmc 0 **
** partition table - mmc 0 **

No partition table - mmc 0 ***
** No partition table - mmc 0 **
** No partition table - mmc 0 **
** No partition tle - mmc 0 **
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a primary desc on 1:1 **
switch toartitions #0, OK
mmc1(part 0) is current device
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a pmary desc on 1:1 **
Scanning mmc device 1
Checking for: /uEntxt …
** No partition table - mmc 1 **
Checking for: /boot/uEnv.txt …
** No partition table - mmc 1 **
** No ptition table - mmc 1 **
** No partition table - mmc 1 **
** partition table - mmc 1 **
** No partition table - mmc 1 **
** No partition table - mmc 1 **
** No partition tabl- mmc 1 **

U-Boot SPL 2017.01-00318-g9fd6070-dirty (Jun 12 2017 - 15:31:55)
DRA722-GP ES2.0
Trying to boot from MMC1
** First descriptor is NOT a primary desc on 0:1 **
** No partion table - mmc 0 **
Using default environment

U-Boot 2017.01-00318-g9fd6070-dirty (Jun 12 2017 - 15:31:5-0500)

CPU : DRA722-GP ES2.0
Model: TI AM5728 BeagleBoard15
Board: AM571x IDK REV 1.3B
DRAM: 1 GiB
MMC: OMAP SD/MMC: 0, OMAP SD/MMC: 1
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** rst descriptor is NOT a primary desc on 0:1 **
** No partition table - mmc 0 **
Usindefault environment

setup_board_eeprom_env: am571x_idk
am5_idk_lcd_detect: Touch screen detect failed: 1!
SCSI: SATA link 0 timeout.
AHCI 0001.0300 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
fgs: 64bit ncq stag pm led clo only pmp pio slum part ccc apst scanning bus for devices…
Found 0 device(s).
Net: not set. Validating first E-fuse MAC
cpsw
Press SCE to abort autoboot in 2 seconds
usb_boot is currently disabled
scsi_boot is currently disabled
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** rst descriptor is NOT a primary desc on 0:1 **
switch to partions #0, OK
mmc0 is current device
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a priry desc on 0:1 **
Scanning mmc device 0
Checking for: /uEnv.t …
** No partition table - mmc 0 **
Checking for: /boot/uEnv.txt …
** No partition table - mmc 0 **
** partition table - mmc 0 **

No partition table - mmc 0 ***
** No partition table - mmc 0 **
** No partition table - mmc 0 **
** No partition tle - mmc 0 **
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a primary desc on 1:1 **
switch toartitions #0, OK
mmc1(part 0) is current device
CACHE: Misaligned operation at range [bffbb290, bffbba90]
** First descriptor is NOT a pmary desc on 1:1 **
Scanning mmc device 1
Checking for: /uEntxt …
** No partition table - mmc 1 **
Checking for: /boot/uEnv.txt …
** No partition table - mmc 1 **
** No ptition table - mmc 1 **
** No partition table - mmc 1 **
** partition table - mmc 1 **
** No partition table - mmc 1 **
** No partition table - mmc 1 **
** No partition tabl- mmc 1 **

Jeff_Andich1 · June 13, 2017, 10:25pm

Well right now we’re moving forward with the BeagleBoard-X15 image on the 5728 part, but I’m still taking a look at booting the BB-X15 image on the 571x IDK - at least through today…

Today I re-imaged a 16 GB SD Card with the entire BB-X15 image file and then dd’ed just the re-built MLO and u-boot (with RCN’s recommended fix for board.c) to the SD Card.

On the 571x IDK, the kernel boots, but when it mounts the FS, but it ends up in initramfs. The same SD card, when inserted in the am572x EVM boots and comes up with the debian login prompt.

I also tried a mixture of device trees (am57x-idk.dtb) am57xx-beagle-x15.dtb from the BeagleBoard-X15 image and TI SDK image, but each time, I’m still ending up in initramfs.

Here are excerpts from the boot console log on the 571x IDK. At the end of the console printout, there are some questions.

Let me know what else you would like to see.

Regards, jeff

Jeff_Andich1 · June 14, 2017, 10:29pm

Update on BeagleBoard Debian on the 571x IDK:

I just tried imaging a smaller SD Card than what I was attempting before. This time I used a 4 GB SD card, whereas previously used a 32 GB card.

***Am able to boot into Debian (BeagleBoard.org Debian Image 2017-06-04) and am able to login! ***

Getting a few of these error messages after the FS is mounted (for different sectors):

[ 29.628082] mmcblk0: error -110 transferring data, sector 4154243 (et. al). nr 165, cmd response 0x900card status 0x0

Previously, when running with an image on a 32 GB SD card, I kept getting the following when ?? booting the kernel??

[ 5.802193] omap_hsmmc 4809c000.mmc: card busy

[ 5.849414] mmc0: error -110 whilst initialising SD card

I’ve seen other posts on TI E2E related to card busy and error -110. One TI employee suggested trying another SD Card.
I’m not sure why the 4 GB SD Card would make things better?? Not sure if it’s a timing issue, or because I’m improperly formatting the card (??leaving a bunch of junk after the ext4 partition??).
I didn’t experience this problem on the 572x EVM - just on the 571x IDK…
Found this patch ( https://py3.patchwork.dja.id.au/series/1807/mbox), ?? for omap_hsmmc “Add-ADMA-support and …” but not sure if it has anything to do with this issue…

Again, please let me know if you want me to fetch anything else from this test configuration and thanks again for your help on Sunday Robert!!

rgds, jeff