This area of binman can be a bit confusing. Add some more comments to help.
Signed-off-by: Simon Glass sjg@chromium.org ---
(no changes since v3)
Changes in v3: - Add new patch with more detail on how ObtainContents() works
tools/binman/entry.py | 3 +++ tools/binman/etype/section.py | 32 +++++++++++++++++++++++++++++++- 2 files changed, 34 insertions(+), 1 deletion(-)
diff --git a/tools/binman/entry.py b/tools/binman/entry.py index 328b5bc568a9..f20f32213a9b 100644 --- a/tools/binman/entry.py +++ b/tools/binman/entry.py @@ -474,6 +474,9 @@ class Entry(object): def ObtainContents(self, skip_entry=None, fake_size=0): """Figure out the contents of an entry.
+ For missing blobs (where allow-missing is enabled), the contents are set + to b'' and self.missing is set to True. + Args: skip_entry (Entry): Entry to skip when obtaining section contents fake_size (int): Size of fake file to create if needed diff --git a/tools/binman/etype/section.py b/tools/binman/etype/section.py index d56cc11d1023..d9b9e428024a 100644 --- a/tools/binman/etype/section.py +++ b/tools/binman/etype/section.py @@ -316,12 +316,15 @@ class Entry_section(Entry): This should be overridden by subclasses which want to build their own data structure for the section.
+ Missing entries will have be given empty (or fake) data, so are + processed normally here. + Args: required: True if the data must be present, False if it is OK to return None
Returns: - Contents of the section (bytes) + Contents of the section (bytes), None if not available """ section_data = bytearray()
@@ -711,6 +714,33 @@ class Entry_section(Entry): def GetEntryContents(self, skip_entry=None): """Call ObtainContents() for each entry in the section
+ The overall goal of this function is to read in any available data in + this entry and any subentries. This includes reading in blobs, setting + up objects which have predefined contents, etc. + + Since entry types which contain entries call ObtainContents() on all + those entries too, the result is that ObtainContents() is called + recursively for the whole tree below this one. + + Entries with subentries are generally not *themselves& processed here, + i.e. their ObtainContents() implementation simply obtains contents of + their subentries, skipping their own contents. For example, the + implementation here (for entry_Section) does not attempt to pack the + entries into a final result. That is handled later. + + Generally, calling this results in SetContents() being called for each + entry, so that the 'data' and 'contents_size; properties are set, and + subsequent calls to GetData() will return value data. + + Where 'allow_missing' is set, this can result in the 'missing' property + being set to True if there is no data. This is handled by setting the + data to b''. This function will still return success. Future calls to + GetData() for this entry will return b'', or in the case where the data + is faked, GetData() will return that fake data. + + Args: + skip_entry: (single) Entry to skip, or None to process all entries + Note that this may set entry.absent to True if the entry is not actually needed """