Re: [PATCH 2/2] arm: dts: k3-am625: add template for tispl.bin

On Tue Jul 28, 2026 at 4:46 PM IST, Simon Glass wrote:
> On 2026-07-17T10:42:51, Anshul Dalal <[email protected]> wrote:
>> arm: dts: k3-am625: add template for tispl.bin
>>
>> This patch addresses the duplication of the binman nodes responsible for
>> generating tispl.bin and tispl.bin_unsigned for TI's AM625 SoC based
>> boards.
>>
>> The common nodes are factored out into a template in the SoC level
>> k3-am625-binman.dtsi that each board in-turn includes.
>>
>> No functional change is intended from the patch asides from the addition
>> of tifsstub-hs and tifsstub-fs nodes in beagleplay's tispl.bin which
>> should have no consequence asides from growing the binary size by few
>> kilo bytes.
>>
>> Signed-off-by: Anshul Dalal <[email protected]>
>>
>> arch/arm/dts/k3-am625-beagleplay-u-boot.dtsi      |  92 +------
>>  arch/arm/dts/k3-am625-binman.dtsi                 | 240 ++++++++++++++++++
>>  arch/arm/dts/k3-am625-phycore-som-binman.dtsi     | 241 +-----------------
>>  arch/arm/dts/k3-am625-sk-binman.dtsi              | 288 ++--------------------
>>  arch/arm/dts/k3-am625-verdin-wifi-dev-binman.dtsi | 185 +-------------
>>  5 files changed, 270 insertions(+), 776 deletions(-)
>
>> diff --git a/arch/arm/dts/k3-am625-beagleplay-u-boot.dtsi b/arch/arm/dts/k3-am625-beagleplay-u-boot.dtsi
>> @@ -63,99 +63,11 @@
>>       ti-spl_unsigned {
>> -             filename = "tispl.bin_unsigned";
>> +             insert-template = <&ti_spl_unsigned_am625_template>;
>>               symlink = "tispl.bin";
>> -             pad-byte = <0xff>;
>> -
>>               fit {
>> -                     description = "Configuration to load ATF and SPL";
>> -                     #address-cells = <1>;
>> -
>>                       images {
>
> Just to check: the FIT for beagleplay now also contains tifsstub-hs /
> tifsstub-fs entries even though conf-0's loadables only references
> tifsstub-gp. Have you confirmed that the resulting tispl.bin still
> parses cleanly and SPL does not try to load the unused entries? Please
> note the beagleplay boot test in the commit message.
>
> I was unable to try this in my lab:
>
>        arm:  +   am62x_beagleplay_r5
> +binman: Node '/binman/tiboot3-am62x-gp-evm.bin/ti-secure-rom': Cannot
> find node for phandle 100
> +make[1]: *** [Makefile:1410: .binman_stamp] Error 1
> +make: *** [Makefile:189: __sub-make] Error 2

This is due to this[1] missing fix which is mentioned as a dependency in
the cover-letter. I have verified the patch series on the beagleplay
with the delta to tispl being the addition of tifsstub-hs and
tifsstub-fs which add a total of ~60KiB to the totalling 1MiB file.

The runtime correctly only loads the tifsstub-gp binary.

[1]: https://lore.kernel.org/u-boot/20260709-binman_template_phandle_copy_fix-v2-1-6c4017eec04d@ti.com/

>
>> diff --git a/arch/arm/dts/k3-am625-binman.dtsi b/arch/arm/dts/k3-am625-binman.dtsi
>> @@ -121,3 +121,243 @@
>> +#ifdef CONFIG_ARM64
>> +
>> +&binman {
>> +     tifsstub-hs {
>
> The three tifsstub-{hs,fs,gp} nodes are unconditional within the
> CONFIG_ARM64 block, so every A53 board including this dtsi will
> produce tifsstub.bin_hs / _fs / _gp regardless of variant. That is
> fine since the blobs are optional, but please confirm this is intended
> and does not regress packaging scripts that consume these artefacts.
>

It was only beagleplay that did not have these binaries included and the
build/boot works just fine with the addition.

I will make the requested changes in the v2, thanks for the review.
Anshul