arm-boot-config-0.62-1 causes boot to fail on BeagleBone (white)

[Clive Messer]

Peter,

Just a friendly heads-up.... 

"Set bootm_size for dtb passing issues, use better load addresses for
TI"

After updating arm-boot-config package yesterday, regenerating boot.cmd
with the new version causes boot to fail on BeagleBone (white). (Starts
loading the initramfs. Hangs there.)

I've attached a diff of boot.cmd generated with 59, against that
generated with 62.

config_am33xx 59:
u_k_addr=0x80300000; u_ramfs_addr=0x82000000; u_dtb_addr=0x81600000;

config_am33xx 62:
setenv bootm_size 0x20000000
u_k_addr=0x88080000; u_ramfs_addr=0x82000000; u_dtb_addr=0x88000000;

Regards

Clive
-- 
Clive Messer <clive.m.messer(a)gmail.com&gt;

[Peter Robinson]

Hi Clive,

>  Just a friendly heads-up.... 
Thanks.

>  "Set bootm_size for dtb passing issues, use better load
> addresses for
>  TI"
>
>  After updating arm-boot-config package yesterday, regenerating boot.cmd
>  with the new version causes boot to fail on BeagleBone (white). (Starts
>  loading the initramfs. Hangs there.)
>
>  I've attached a diff of boot.cmd generated with 59, against that
>  generated with 62.
>
>  config_am33xx 59:
>  u_k_addr=0x80300000; u_ramfs_addr=0x82000000; u_dtb_addr=0x81600000;
>
>  config_am33xx 62:
>  setenv bootm_size 0x20000000
>  u_k_addr=0x88080000; u_ramfs_addr=0x82000000; u_dtb_addr=0x88000000; 
Can we get this reported into bugzilla please so it can be easily
tracked. From memory the BBW has 256Mb RAM? The link below will open
you up into the right spot. Can you also add details of the kernel
uboot versions to the bug and just reply here with the BZ number.

Thanks,
Peter

https://bugzilla.redhat.com/enter_bug.cgi?product=Fedora&version=rawh...

[Dennis Gilmore]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 28 Mar 2014 09:55:16 +0000
Peter Robinson <pbrobinson(a)gmail.com&gt; wrote:

>  Hi Clive,
>  
>  > Just a friendly heads-up....
>  
>  Thanks.
>  
>  > "Set bootm_size for dtb passing issues, use better load addresses
>  > for TI"
>  >
>  > After updating arm-boot-config package yesterday, regenerating
>  > boot.cmd with the new version causes boot to fail on BeagleBone
>  > (white). (Starts loading the initramfs. Hangs there.)
>  >
>  > I've attached a diff of boot.cmd generated with 59, against that
>  > generated with 62.
>  >
>  > config_am33xx 59:
>  > u_k_addr=0x80300000; u_ramfs_addr=0x82000000; u_dtb_addr=0x81600000;
>  >
>  > config_am33xx 62:
>  > setenv bootm_size 0x20000000
>  > u_k_addr=0x88080000; u_ramfs_addr=0x82000000; u_dtb_addr=0x88000000;
>  
>  Can we get this reported into bugzilla please so it can be easily
>  tracked. From memory the BBW has 256Mb RAM? The link below will open
>  you up into the right spot. Can you also add details of the kernel
>  uboot versions to the bug and just reply here with the BZ number.
>  
I noticed this getting rawhide working yesterday we need to adjust
bootm_size to 0x10000000 since there is less memory 

Dennis
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.22 (GNU/Linux)
>
> iQIcBAEBAgAGBQJTNYcHAAoJEH7ltONmPFDR+jAQAKDSLiW8KiOHQEUuHUW0kaLg
> lDmclxQAXLqt+wQqG1Ylbf4H51HxayGnEY4voJjhHIHlX9tv6Xm1aF97FgpUh2WA
> hi77r+KGnYvpHgvy6Di1EEsd6ySehaTrRE936SN9WA2w4sqWR7RoJ0GHjrU/QT7j
> 5hr4bQbIpEogNNQVIo3yGJq/ypYeO6nIUYjpvVPqdDfyIN85k+NnA9amtyjkZfs0
> rnLACre3PO/Whip6+mtWs3r8JR1x91B8LPMi5qWsx5leyLEnXYKxv6pKABzJq95J
> FRxMdVvHdkj79xHJlsYSzd7V37FM6Juj0KpfxAhPIHH/2y102kxvOPX/QpKdGTkJ
> UHLq6BhSPEpdmrv8q6ey4ym80lI3b1W50XNHgfWaan8fMNMKC2dk3PVzTzwz9gHv
> /A3vwjSW8ounz17PN3HmabokQQ8LiwgvaFYH1DCH/s6/v2Tk9Pxsgu3ypDtPiaHW
> c2FOY+nydD02+19XpfPMzQwsZnJzPj0RRnyHSc88jLUlP4Tbz5D4QC5VmJOKlunN
> Exd9wdzCUIT78SUNrno+GEklGwu9pvmy4W752RHgdHEtlXhHgFYoaJ3/5SLVKOh0
> 3XaSqSd/clglzWVhtJMwGePcOXNkxSsA77WWWyGINEggs5nHPJ6eHxl2bxXnFf1L
> o4CJZQD1lWDSOeV5JW9Q
> =pNOO
> -----END PGP SIGNATURE-----
>

[Brendan Conoboy]

On 03/28/2014 07:28 AM, Dennis Gilmore wrote:
>  I noticed this getting rawhide working yesterday we need to adjust
>  bootm_size to 0x10000000 since there is less memory 
Is that 0x10000000 for all devices or just low-memory devices like BBW?

-- 
Brendan Conoboy / Red Hat, Inc. / blc(a)redhat.com

[Dennis Gilmore]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 28 Mar 2014 07:34:36 -0700
Brendan Conoboy <blc(a)redhat.com&gt; wrote:

>  On 03/28/2014 07:28 AM, Dennis Gilmore wrote:
>  > I noticed this getting rawhide working yesterday we need to adjust
>  > bootm_size to 0x10000000 since there is less memory
>  
>  Is that 0x10000000 for all devices or just low-memory devices like
>  BBW?
>  
>  
0x10000000 is 256Mb i've reported the issue upstream to u-boot. seems
its a bug in u-boot it should only relocate to the end of memory even
if bootm_size is bigger than physical memory. 

a device that only has 128Mb like my beagleboard needs 0x08000000 

ultimately this will need to be fixed upstream in u-boot

Dennis
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.22 (GNU/Linux)
>
> iQIcBAEBAgAGBQJTNYotAAoJEH7ltONmPFDRtvsP/j2L1W5xTdZDrVcVaHtebySI
> 6rGkIE16gyg6txWYxf/RiUZ1D+cHC78Pu3DiDw79S9LuTUidTHQUSFwUUqA4zq+c
> nqk2vSlZ/0DIjYCRiitq+/M8bQNZnQGUqCLNxpKRzdTo2oMNcQkd5Ooe5ahATBUy
> 24Go9ezawCKck6jA0/eMAK7lpv+wFtcavWdNHF5GstYs23BQyzaosOAWQv6dmZZa
> 1tmlD+0IEvfUa3z79x/YvnKGDcLF0FACZdA8zAw5eYfIlxye+/Ca4iGEoRIVudfd
> +v7+7uJpLUy/dDVVGryD29zmzZbBANkjGj5+EoCOHPS+9Qwv3osV44uZ/LfUrSxe
> SZnYG0SscqvamfQavu6oYNIQhmnVragX+1RLSlddRzyx3Limk9VV0/cUQTd/AT1o
> 9AihEVzthEgq3y3y9iHW7z0nyZ6hgIVI8HWy9HxS4qyFJ62/2G6fb3exttW38rSd
> WhBQpo92fEZFMZ4Q9R5Q9Rwm2C6kLNte21iSsKOGwwhBdnu9KJdd5C+stViaLkP7
> EQj8L+du6G3Slkg5LBrxN9TeL9B3DcmvR8YlEk5sd2kIb+WCRj723wWvW90TMA8h
> xcdeuIhvfzCtweYFXM+YIlTWkp5CHQV8mx74LTiQaV2J2FdZipWZmvrrUI3WSosB
> kpzsSf1ZabG/09MXcknM
> =7UaI
> -----END PGP SIGNATURE-----
>