Skip to contentSkip to conversation

Attached DTB no longer supported u-boot.imx?

U-Boot Bill Pringlemeir 2 participants 5 messages Nov 28, 2014 – Dec 3, 2014

lore
Share & tools
lore
#0 lore raw

I tried to update the new u-boot I get,

Bytes transferred = 3413041 (341431 hex) Kernel image @ 0x86000000 [ 0x000000 - 0x33e150 ] Could not find a valid device tree

The 'bootcmd' is,

tftp ${loadaddr} zImage.dtb; bootz ${loadaddr}

has the syntax changed and I am not up to date? I think I saw something about an attached DTB not supported.

u-boot-imx$ git describe
v2009.08-18413-g5c84206

err, that is no good. I have 2a82ec77d27ef5f860a107c4b764643a655dceeb as 'Prepare v2015.01-rc2' and I had to apply 5c84206 to compile with a 4.9 hard-float ARM compiler.

I pulled the main 'u-boot.git' and this does not happen.

u-boot$ git describe
v2015.01-rc2-87-g0f89149

It is '38cd8c4253013ccdd4052ee021f6066fe9a52551', with the same patch as 'pwclient git-am 415336'

Is the imx just using newer patches to be merged to the main tree and I must switch to a detached DT? Or is there something else?

Thanks,
Bill.

#1 lore raw
Quoted 6 lines · bpringlemeir at nbsps.com

I tried to update the new u-boot I get,

Bytes transferred = 3413041 (341431 hex)
Kernel image @ 0x86000000 [ 0x000000 - 0x33e150 ]
Could not find a valid device tree

[snip]

Quoted 2 lines

Is the imx just using newer patches to be merged to the main tree and I must switch to a detached DT? Or is there something else?

Sorry, both trees behave the same. I will switch to separate binaries.

#2 lore raw
Quoted 9 lines · Bill Pringlemeir
Quoted 6 lines · bpringlemeir at nbsps.com

I tried to update the new u-boot I get,

Bytes transferred = 3413041 (341431 hex)
Kernel image @ 0x86000000 [ 0x000000 - 0x33e150 ]
Could not find a valid device tree

Quoted 6 lines

[snip]

Quoted 2 lines

Is the imx just using newer patches to be merged to the main tree and I must switch to a detached DT? Or is there something else?

Sorry, both trees behave the same. I will switch to separate binaries.

This is not intentional. I suspect I know what commit did this. Can you please do a git bisect from v2015.01-rc1 to -rc2 and see what it says the bad commit is? Thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20141128/7f8186a4/attachment.pgp>
#3 lore raw
Quoted 5 lines · Bill Pringlemeir
Quoted 3 lines

Sorry, both trees behave the same. I will switch to separate binaries.

Quoted 3 lines · trini at ti.com

This is not intentional. I suspect I know what commit did this. Can you please do a git bisect from v2015.01-rc1 to -rc2 and see what it says the bad commit is? Thanks!

I have the 'ext4-div' patch and I must reduce the vf610 image size in order to boot (why I have 'git reset --hard').

git reset --hard; git bisect bad
HEAD is now at c6150aa image-fdt: boot_get_fdt() return value when no DTB exists c6150aaf2f2745141a7c2ceded58d7efbfeace7d is the first bad commit commit c6150aaf2f2745141a7c2ceded58d7efbfeace7d

Author: Noam Camus <[email protected]>
Date:   Wed Oct 22 17:17:49 2014 +0300

image-fdt: boot_get_fdt() return value when no DTB exists

I believe that when no DTB is around we should return 1. This why I fixed such scenarious to not return zero anymore. Else kernel might get NULL pointer to DTB which doesn't exists.

:040000 040000 f343f802d1491dc54545271694669c21f5efaa65 1120435d7aee7a9c12a6c8e2c0b19a364cbaf185 M common

If I take rc2, revert c6150aaf2 and rebuild then the attached DTB image seems fine for booting.

Regards,
Bill Pringlemeir.

#4 lore raw
Quoted 28 lines · Bill Pringlemeir
Quoted 5 lines · Bill Pringlemeir
Quoted 3 lines

Sorry, both trees behave the same. I will switch to separate binaries.

Quoted 3 lines · trini at ti.com

This is not intentional. I suspect I know what commit did this. Can you please do a git bisect from v2015.01-rc1 to -rc2 and see what it says the bad commit is? Thanks!

I have the 'ext4-div' patch and I must reduce the vf610 image size in order to boot (why I have 'git reset --hard').

git reset --hard; git bisect bad
HEAD is now at c6150aa image-fdt: boot_get_fdt() return value when no DTB exists c6150aaf2f2745141a7c2ceded58d7efbfeace7d is the first bad commit commit c6150aaf2f2745141a7c2ceded58d7efbfeace7d

Author: Noam Camus <[email protected]>
Date:   Wed Oct 22 17:17:49 2014 +0300

image-fdt: boot_get_fdt() return value when no DTB exists

I believe that when no DTB is around we should return 1. This why I fixed such scenarious to not return zero anymore. Else kernel might get NULL pointer to DTB which doesn't exists.

Quoted 6 lines

:040000 040000 f343f802d1491dc54545271694669c21f5efaa65 1120435d7aee7a9c12a6c8e2c0b19a364cbaf185 M common

If I take rc2, revert c6150aaf2 and rebuild then the attached DTB image seems fine for booting.

Great, thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20141203/2b28b95e/attachment.pgp>