RedHat and Monolithic Kernel Patches

Posted May 31, 2018 5:41 UTC (Thu) by Garak (guest, #99377)
In reply to: The Software Freedom Conservancy on Tesla's GPL compliance by hkario
Parent article: The Software Freedom Conservancy on Tesla's GPL compliance

The work Red Hat does is actually part of upstream kernel: https://lwn.net/Articles/742672/

All of it? Since that 2011 LWN article? Even if so, what goal (whose interests) is served by not providing the non-monolithic version? It always sounded like a corner-case gaming of the system to further their bottom line ($) to me. Maybe I was wrong. The more recent controversy with pax/grs(?) I viewed as a similar tactic with a similar presumed motive (though different, perhaps significantly so in the details).

and the changes being distributed to customers the way they are, not as just recompiled upstream branch, is because of backporting: https://access.redhat.com/security/updates/backporting

Are you saying that Redhat's backporting methodology precludes their ability to behave as they did prior to 2011, i.e. providing a more granular patch series instead of a monolithic one? I don't grok that. The 2011 article/event seemed to demonstrate a financially motivated information control tactic. Yes, I am paranoid, but do please clarify if I'm mistaken. I.e. when Corbet wrote in 2011:

Distribution in this form should satisfy the GPL, but it makes life hard for anybody else wanting to see what has been done with this kernel. Hopefully it is simply a mistake which will be corrected soon.

was it a prelude to a story I didn't see/don't remember where the "mistake" was "corrected soon"?

RedHat and Monolithic Kernel Patches

Posted Jun 1, 2018 2:48 UTC (Fri) by lsl (subscriber, #86508) [Link]

> All of it? Since that 2011 LWN article?

I think so, yes. RHEL kernels contain lots of bugfix and feature backports but it's all code originally developed for later upstream kernels. In some cases, the work has been proposed for upstream inclusion but was rejected in its current form. A recent example for the latter is the Secure-Boot-related stuff.

The particular patches on top of the ancient 3.10 kernel are of no interest to the upstream Linux community. They *are* useful to other folks trying to support these ancient kernels.