Ehm, maybe we have a terminology problem here. When I talk about signature
header immutable region, I mean HEADER_SIGNATURES region. That's not used for
digests or signatures and when adding/removing signatures the signature header
changes more fundamentally than just padding adjustments. So yes if you look at
it in terms of what might change, only the build-time added digests and sizes
etc should go into the region and actual signatures etc that might be added
over time should go outside it, but rpm has "forever" reloaded the signature
into a region after each signature add/remove and all. Padding only changes
when signatures are added or removed, so technically it can just as well be in
the region (or not). So yes the region doesn't mean a whole lot in the
signature header...
As for the bug allegation, the details are right there in my previous posts:
the code (copy-pasted from rpm5 as you must've noticed) calls headerReload()
after adding the padding, which I read as "make it a single (immutable)
region". And despite that, the padding ends up being outside the region. I'm
not arguing whether it should be in the region or not, just observing that the
result doesn't match what was indicated by the code, which to me seems like a
possible bug someplace, which I'm pointing out as in "you might want to look at
that", no ill-will / FUD / whatever intended. If I'm mistaken then by golly
apologies for speculating the possibility of lurking bugs in a piece of
software.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/270#issuecomment-347774166
_______________________________________________
Rpm-maint mailing list
[email protected]
http://lists.rpm.org/mailman/listinfo/rpm-maint
