[PATCH v4 06/22] fdt: Redo devicetree setup

[Jonas Karlman]

Hi Simon,

On 2025-05-10 15:42, Simon Glass wrote:
> The existing logic for handling receiving a devicetree from a previous
> phase is quite complicated. Simplify it in preparation for introducing
> standard passage.
> 
> Add a Kconfig called OF_PASSAGE which indicates that the devicetree
> must come from standard passage.
> 
> Signed-off-by: Simon Glass <sjg at chromium.org>
> ---
> 
> Changes in v4:
> - Drop now-unused label
> 
> Changes in v3:
> - Add new patch to redo how a devicetree is set up
> 
>  dts/Kconfig  |  8 ++++++++
>  lib/fdtdec.c | 35 ++++++-----------------------------
>  2 files changed, 14 insertions(+), 29 deletions(-)
> 
> diff --git a/dts/Kconfig b/dts/Kconfig
> index c544f948a15..57101bddabb 100644
> --- a/dts/Kconfig
> +++ b/dts/Kconfig
> @@ -159,6 +159,14 @@ config OF_EMBED
>  	  and development only and is not recommended for production devices.
>  	  Boards in the mainline U-Boot tree should not use it.
>  
> +config OF_PASSAGE
> +	bool "DTB is provided by a standard passage, in a bloblist"
> +	help
> +	  Select this to read the devicetree using the standard passage
> +	  mechanism, i.e. from a previous phase in bloblist using arch-specific
> +	  register conventions. If enabled, the previous phase must provide
> +	  the devicetree, or U-Boot will not function correctly.
> +
>  endchoice
>  
>  config OF_INITIAL_DTB_READONLY
> diff --git a/lib/fdtdec.c b/lib/fdtdec.c
> index c38738b48c7..8559616873b 100644
> --- a/lib/fdtdec.c
> +++ b/lib/fdtdec.c
> @@ -1689,35 +1689,13 @@ void fdtdec_setup_embed(void)
>  
>  int fdtdec_setup(void)
>  {
> -	int ret = -ENOENT;
> -
> -	/*
> -	 * If allowing a bloblist, check that first. There was discussion about
> -	 * adding an OF_BLOBLIST Kconfig, but this was rejected.
> -	 *
> -	 * The necessary test is whether the previous phase passed a bloblist,
> -	 * not whether this phase creates one.
> -	 */
> -	if (CONFIG_IS_ENABLED(BLOBLIST) &&
> -	    (xpl_prev_phase() != PHASE_TPL ||
> -	     IS_ENABLED(CONFIG_TPL_BLOBLIST))) {
> -		ret = bloblist_maybe_init();
> -		if (!ret) {
> -			gd->fdt_blob = bloblist_find(BLOBLISTT_CONTROL_FDT, 0);
> -			if (gd->fdt_blob) {
> -				gd->fdt_src = FDTSRC_BLOBLIST;
> -				log_debug("Devicetree is in bloblist at %p\n",
> -					  gd->fdt_blob);
> -				goto setup_fdt;
> -			} else {
> -				log_debug("No FDT found in bloblist\n");
> -				ret = -ENOENT;
> -			}
> -		}
> -	}
> +	int ret;
>  
> -	/* Otherwise, the devicetree is typically appended to U-Boot */
> -	if (ret) {
> +	if (CONFIG_IS_ENABLED(OF_BLOBLIST)) {

This looks wrong, should possible be OF_PASSAGE ?

Also looks like there is more patches in need of a correction to use
OF_PASSAGE instead of OF_BLOBLIST in commit message and/or patch content
in this series.

Regards,
Jonas

> +		printf("Previous phase failed to provide standard passage\n");
> +		return -ENOENT;
> +	} else {
> +		/* The devicetree is typically appended to U-Boot */
>  		if (IS_ENABLED(CONFIG_OF_SEPARATE)) {
>  			gd->fdt_blob = fdt_find_separate();
>  			gd->fdt_src = FDTSRC_SEPARATE;
> @@ -1752,7 +1730,6 @@ int fdtdec_setup(void)
>  		}
>  	}
>  
> -setup_fdt:
>  	if (CONFIG_IS_ENABLED(MULTI_DTB_FIT))
>  		setup_multi_dtb_fit();
>