Skip to contentSkip to conversation

AM335x : failure to boot SPL from NAND

U-Boot Mark Jackson 2 participants 6 messages Feb 14, 2013 – Feb 21, 2013

lore
Share & tools
lore
#0 lore raw

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Following the TI wiki
(http://processors.wiki.ti.com/index.php/AM335x_U-Boot_User%27s_Guide#Flashing_images_to_NAND_in_SD_boot)
I:-

(1) boot from MMC
(2) erase the nand
(3) copy MLO from MMC into NAND
(4) verified it copied correctly (using crc32)

When I reboot the board in NAND mode, I get nothing on UART0.

If I reboot in MMC mode, SPL and U-Boot load correctly.

Can anyone give me some pointers on the boot sequence, and where I might look to help debug the problem ?

Regards
Mark JACKSON

#1 lore raw
Quoted 13 lines · Mark Jackson

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Following the TI wiki
(http://processors.wiki.ti.com/index.php/AM335x_U-Boot_User%27s_Guide#Flashing_images_to_NAND_in_SD_boot)
I:-

(1) boot from MMC
(2) erase the nand
(3) copy MLO from MMC into NAND
(4) verified it copied correctly (using crc32)

Quoted 6 lines

When I reboot the board in NAND mode, I get nothing on UART0.

If I reboot in MMC mode, SPL and U-Boot load correctly.

Can anyone give me some pointers on the boot sequence, and where I might look to help debug the problem ?

Assuming you're using mainline U-Boot and can rule out "vendor problems", if you can access the SYSBOOT pins, set it up for a mode that does NAND and UART. If you never see the 'CCC' on console (or only see it the first time if you do UART then NAND), then you are starting SPL and dying in there. If you just see a stream of C then your file is not written to NAND correctly.

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20130215/93e765d7/attachment.pgp>
#2 lore raw
Quoted 22 lines · Tom Rini
Quoted 13 lines · Mark Jackson

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Following the TI wiki
(http://processors.wiki.ti.com/index.php/AM335x_U-Boot_User%27s_Guide#Flashing_images_to_NAND_in_SD_boot)
I:-

(1) boot from MMC
(2) erase the nand
(3) copy MLO from MMC into NAND
(4) verified it copied correctly (using crc32)

Quoted 6 lines

When I reboot the board in NAND mode, I get nothing on UART0.

If I reboot in MMC mode, SPL and U-Boot load correctly.

Can anyone give me some pointers on the boot sequence, and where I might look to help debug the problem ?

Quoted 6 lines

Assuming you're using mainline U-Boot and can rule out "vendor problems", if you can access the SYSBOOT pins, set it up for a mode that does NAND and UART. If you never see the 'CCC' on console (or only see it the first time if you do UART then NAND), then you are starting SPL and dying in there. If you just see a stream of C then your file is not written to NAND correctly.

Interesting ... I don't get any 'CCC' on the console.

However, I then tested this by booting via MMC, erasing the NAND chip and then trying to reboot via NAND again.

I still get no 'CCC' on the console !?!

This is using boot mode 10011 (NAND, NANDI2C, MMC0, UART0), so I would expect to either boot via MMC (if I left it in) or get some 'CCC' output on the console.

I can see that the NAND signals always have a burst of activity every 10ms. That must be a timeout of some sort ... do you know if that's a hardware or software timeout ?

Mark J.

#3 lore raw
Quoted 30 lines · Mark Jackson
Quoted 22 lines · Tom Rini
Quoted 13 lines · Mark Jackson

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Following the TI wiki
(http://processors.wiki.ti.com/index.php/AM335x_U-Boot_User%27s_Guide#Flashing_images_to_NAND_in_SD_boot)
I:-

(1) boot from MMC
(2) erase the nand
(3) copy MLO from MMC into NAND
(4) verified it copied correctly (using crc32)

Quoted 6 lines

When I reboot the board in NAND mode, I get nothing on UART0.

If I reboot in MMC mode, SPL and U-Boot load correctly.

Can anyone give me some pointers on the boot sequence, and where I might look to help debug the problem ?

Quoted 6 lines

Assuming you're using mainline U-Boot and can rule out "vendor problems", if you can access the SYSBOOT pins, set it up for a mode that does NAND and UART. If you never see the 'CCC' on console (or only see it the first time if you do UART then NAND), then you are starting SPL and dying in there. If you just see a stream of C then your file is not written to NAND correctly.

Quoted 6 lines

Interesting ... I don't get any 'CCC' on the console.

However, I then tested this by booting via MMC, erasing the NAND chip and then trying to reboot via NAND again.

I still get no 'CCC' on the console !?!

How did you wire the SYSBOOT pins, and which UART is connected to console? If you have UART0 as the one hooked up to console and have SYSBOOT[4:0] as 00100, the order is UART0,XIP (NOR),MMC0,NAND. So you should see 'CCCCCCCC' (UART0 trying to initiate X-Modem), then it will try MMC0 and then NAND..

Quoted 3 lines

This is using boot mode 10011 (NAND, NANDI2C, MMC0, UART0), so I would expect to either boot via MMC (if I left it in) or get some 'CCC' output on the console.

Correct. If you have an empty NAND and nothing inserted in MMC, it should be an endless cycle of 'C'.

Quoted 3 lines

I can see that the NAND signals always have a burst of activity every 10ms. That must be a timeout of some sort ... do you know if that's a hardware or software timeout ?

Moving beyond what I can easily help with, sorry. Hit up the TI forums at http://e2e.ti.com/ as this is getting a bit off-topic for the U-Boot list as well :) Thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20130218/01c6683e/attachment.pgp>
#4 lore raw
Quoted 7 lines · Tom Rini
Quoted 6 lines · Mark Jackson
Quoted 5 lines · Tom Rini
Quoted 3 lines · Mark Jackson

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Just a quick update (in case anyone else is in the same boat) ...

Looks like the upper data bus (AD[15:8]) needs to have pullups on it, even though you're booting from an 8bit device.

I guess the AM335x internal boot code assumes everything is 16bit (even though SYSBOOT[8] is used to configure 8/16bit device).

Mark J.

#5 lore raw
Quoted 5 lines · Mark Jackson
Quoted 4 lines · Tom Rini
Quoted 3 lines · Mark Jackson
Quoted 2 lines · Tom Rini

On Thu, Feb 14, 2013 at 03:54:23PM +0000, Mark Jackson wrote:

Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

I'm trying to diagnose why our AM335x based CPU board (based on the AM335x Starter Kit) can boot SPL and U-Boot from an MMC card, but is unable to boot from NAND (connected to CS0).

Just a quick update (in case anyone else is in the same boat) ...

Looks like the upper data bus (AD[15:8]) needs to have pullups on it, even though you're booting from an 8bit device.

Just to update this ... *pulldowns* (not pullups) were needed to get it to boot reliably.