[U-Boot] [PATCH] da850evm: fix linux bootparam address
This patch fixes the LINUX_BOOT_PARAM_ADDR define to be based off of PHYS_SDRAM_1 instead of CONFIG_SYS_MEMTEST_START. On da830 they are the same thing but on da850 the CONFIG_SYS_MEMSTART define is offset from the PHYS_SDRAM_1 start.
Without this patch it is not possible to boot linux on da850 -- bootm hangs at "Uncompressing Linux... done, booting the kernel."
Signed-off-by: Ben Gardiner bengardiner@nanometrics.ca CC: Sudhakar Rajashekhara sudhakar.raj@ti.com
---
I'm not sure if this is a bugfix or not -- there is currently no way to load a uImage from anywhere in the da850evm config on master. But if you did modify the configuration enough to load one then your bootm command would hang at "Uncompressing Linux... done, booting the kernel."
I'm submitting it for inclusion into 2010.09 since the definiion of LINUX_BOOT_PARAM_ADDR is wrong for da850. But it's OK to push off to next since you couldn't load a uImage with the default config anyways.
--- include/configs/da850evm.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/include/configs/da850evm.h b/include/configs/da850evm.h index 1b6fe72..21db022 100644 --- a/include/configs/da850evm.h +++ b/include/configs/da850evm.h @@ -139,7 +139,7 @@ /* * Linux Information */ -#define LINUX_BOOT_PARAM_ADDR (CONFIG_SYS_MEMTEST_START + 0x100) +#define LINUX_BOOT_PARAM_ADDR (PHYS_SDRAM_1 + 0x100) #define CONFIG_CMDLINE_TAG #define CONFIG_SETUP_MEMORY_TAGS #define CONFIG_BOOTARGS \
Dear Sudhakar,
On Wed, Sep 22, 2010 at 12:45 PM, Ben Gardiner bengardiner@nanometrics.ca wrote:
This patch fixes the LINUX_BOOT_PARAM_ADDR define to be based off of PHYS_SDRAM_1 instead of CONFIG_SYS_MEMTEST_START. On da830 they are the same thing but on da850 the CONFIG_SYS_MEMSTART define is offset from the PHYS_SDRAM_1 start.
Without this patch it is not possible to boot linux on da850 -- bootm hangs at "Uncompressing Linux... done, booting the kernel."
Signed-off-by: Ben Gardiner bengardiner@nanometrics.ca CC: Sudhakar Rajashekhara sudhakar.raj@ti.com
I just noticed today that there is a patch committed [1] in the u-boot-davinci tree (Sandeep Paulraj, the maintainer, has been added to the cc) which does the same as this patch and also predates it. I am sorry: I was unaware of the da850 changes in that tree when I worked on this and other patches. In my defense, the description of that repo is "DaVinci DM3xx/DM6446 U-Boot integration/staging tree." Whereas I had been porting changes from the u-boot-omapl1 tree [2] which is described as the "DA8xx/OMAP-L1xx U-Boot integration/staging tree."
Also relevant are the following patches where you have made changes to the u-boot-davinci tree which predate the changes I have submitted for 2010.12: "da850: Enable Davinci EMAC driver" [3] which predates my "[PATCH v6 2/2][NEXT] da850evm: basic MII EMAC support." [4] The changes here are very similar except for the differences resulting from my implementing the davinci_eth_set_mac_addr extraction as requested by Ben Warren (on the CC since I am singling him out).
I apologize for being oblivious to the integration efforts in that tree. I would very much like to see more complete da850 support in 2010.12 -- how would you like to proceed? I am happy to withdraw the patches and review/test any that you post in their place.
Best Regards, Ben Gardiner
[1] http://arago-project.org/git/people/?p=sandeep/u-boot-davinci.git;a=commit;h... [2] http://arago-project.org/git/projects/?p=u-boot-omapl1.git;a=summary [3] http://arago-project.org/git/people/?p=sandeep/u-boot-davinci.git;a=commit;h... [4] http://www.mail-archive.com/u-boot@lists.denx.de/msg38461.html
--- Nanometrics Inc. http://www.nanometrics.ca
participants (1)
-
Ben Gardiner