Subject: Re: [PATCH 0/3] of/fdt: Rework early FDT scanning functions
To:     Rob Herring <robh@kernel.org>,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        John Crispin <john@phrozen.org>,
        Michael Ellerman <mpe@ellerman.id.au>,
        Paul Mackerras <paulus@samba.org>,
        Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc:     devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-mips@vger.kernel.org, linuxppc-dev@lists.ozlabs.org
References: <20211118181213.1433346-1-robh@kernel.org>
From:   Frank Rowand <frowand.list@gmail.com>
Message-ID: <7c3286d4-2da4-b083-b3a4-88d32e475617@gmail.com>
Date:   Wed, 24 Nov 2021 12:01:33 -0500
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.13.0
MIME-Version: 1.0
In-Reply-To: <20211118181213.1433346-1-robh@kernel.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 11/18/21 1:12 PM, Rob Herring wrote:
> The early FDT scanning functions use of_scan_flat_dt() which implements 
> its own node walking method. This function predates libfdt and is an 
> unnecessary indirection. This series reworks 
> early_init_dt_scan_chosen(), early_init_dt_scan_root(), and 
> early_init_dt_scan_memory() to be called directly and use libfdt calls.
> 
> Ultimately, I want to remove of_scan_flat_dt(). Most of the remaining 
> of_scan_flat_dt() users are in powerpc.
> 
> Rob
> 
> 
> Rob Herring (3):
>   of/fdt: Rework early_init_dt_scan_chosen() to call directly
>   of/fdt: Rework early_init_dt_scan_root() to call directly
>   of/fdt: Rework early_init_dt_scan_memory() to call directly
> 
>  arch/mips/ralink/of.c                |  16 +---
>  arch/powerpc/kernel/prom.c           |  22 ++---
>  arch/powerpc/mm/nohash/kaslr_booke.c |   4 +-
>  drivers/of/fdt.c                     | 121 ++++++++++++++-------------
>  include/linux/of_fdt.h               |   9 +-
>  5 files changed, 79 insertions(+), 93 deletions(-)
> 


"checkpatch --strict" reports some "CHECK" issues, but review of the patches
for correctness becomes much more difficult if they are addressed, so they
should be ignored for this series.

-Frank