Re: [bitbake-devel][PATCH] fetch: Upgrade shown checksum to SHA-512

Joshua Watt <[email protected]>
Newsgroups org.openembedded.lists.bitbake-devel
Message-ID <CAJdd5GbsTRCyFh=5=yhETHZ_rqPzGm_bhqeun0=mEvFTvuY4MQ@mail.gmail.com>
On Wed, May 13, 2026 at 2:21 PM Alexander Kanavin
<[email protected]> wrote:
>
> On Wed, 13 May 2026 at 16:46, Joshua Watt via lists.openembedded.org
> <[email protected]> wrote:
> >  CHECKSUM_LIST = [ "goh1", "md5", "sha256", "sha1", "sha384", "sha512" ]
> > -SHOWN_CHECKSUM_LIST = ["sha256"]
> > +SHOWN_CHECKSUM_LIST = ["sha256", "sha512"]
>
> This raises so many questions :)
> - why suggest adding/fixing both sha256 and sha512 to users, and not
> just sha512? One checksum is enough, and causes less visual clutter in
> recipes

I suspect we'd want to make SHA-512 the only option by the next LTS. I
don't have a strong opinion if we do that now or in a future release.

> - if sha512 is suggested, should devtool be tweaked to add that in
> version upgrades and newly added recipes? (something similar was
> previously done to assist md5 to sha256 transition)

Probably, but only if we make SHA 512 the only option, so see above.

> - last but not least, should we look into adding support for better,
> newer ways to identify commits in git, replacing the classic,
> not-that-secure sha1 ?

I've not seen any guidance about what to do with git SHAs in e.g. BSI
TR-03183, so I'm operating under the assumption that SHA 1 is fine for
now. If you know of a better way to do it, and/or see some
specification that demands we do something different, we can discuss
that. Either way, I would consider changing the way git works as out
of scope for this change.

>
> Alex
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.