[ . . . ] > > diff --git a/tools/binman/etype/qcom_appsbl.py > > b/tools/binman/etype/qcom_appsbl.py > > @@ -0,0 +1,23 @@ > > + Properties / Entry arguments: > > + - filename: Filename of U-Boot MBN (default 'u-boot.mbn') > > + > > + This is the U-Boot MBN image. > > + """ > > + def __init__(self, section, etype, node): > > + super().__init__(section, etype, node, 'qcom-appsbl') > > Not quite. blob_named_by_arg looks up an entry-arg called > <blob_fname>-path, so the actual property is qcom-appsbl-path > (matching what your doc patch passes with -a qcom-appsbl-path=...). > There is no filename property and no default. Please update the > docstring to match the other two new etypes.
Ok. > > diff --git a/tools/binman/etype/qcom_appsbl.py > > b/tools/binman/etype/qcom_appsbl.py > > @@ -0,0 +1,23 @@ > > + def __init__(self, section, etype, node): > > + super().__init__(section, etype, node, 'qcom-appsbl') > > qcom_config and qcom_lib set self.external = True, but this one > doesn't, even though u-boot.mbn is supplied via -a qcom-appsbl-path=. > Should this also be external? Otherwise a missing blob produces a hard > error rather than the usual missing-blob warning. Ok. > > diff --git a/tools/binman/etype/qcom_config.py > > b/tools/binman/etype/qcom_config.py > > @@ -0,0 +1,26 @@ > > +from binman.entry import Entry > > +from binman.etype.blob_named_by_arg import Entry_blob_named_by_arg > > The Entry import is unused - same in qcom_lib.py. Please drop it. Ok. > > diff --git a/tools/binman/etype/qcom_appsbl.py > > b/tools/binman/etype/qcom_appsbl.py > > @@ -0,0 +1,23 @@ > > +# SPDX-License-Identifier: GPL-2.0 > > The rest of tools/binman/etype/ uses GPL-2.0+. Please align (all three > new files). Our legal doesn't allow GPL-2.0+. If that is not acceptable, let me know will talk to them and see what can be done. > > diff --git a/tools/binman/etype/qcom_appsbl.py > > b/tools/binman/etype/qcom_appsbl.py > > @@ -0,0 +1,23 @@ > > + MBN is a Qualcomm‑specific signed firmware container that packages > > + executable images (typically ELF) along with metadata such as hash, > > + signature, and certificate chain, enabling the secure boot chain and > > Root > > + of Trust (RoT) to load and authenticate images. > > Just to clarify - the doc patch shows this entry is fed u-boot.mbn > produced by running elftombn.py over u-boot.elf, but the binman node > in the dtsi treats it as an ELF (fit,operation = 'split-elf', os = > 'U-Boot'). Is the input actually a raw MBN container or the wrap-elf? > The docstring should describe whichever binman consumes. Input is the MBN container. wrap-elf is for the SPL. Will update the docstring. > > > diff --git a/tools/binman/etype/qcom_appsbl.py > > b/tools/binman/etype/qcom_appsbl.py > > @@ -0,0 +1,23 @@ > > +class Entry_qcom_appsbl(Entry_blob_named_by_arg): > > + """U-Boot mbn image > > New binman etypes normally come with an ftest entry and a small test > data file. Please add coverage for these three. Ok. Thanks -Varada
