AMD K8 Support Stripped Out Of Coreboot

[schmidtbag]

There absolutely was an Athlon X2 line, and looking it up, it turns out there are K8 and K10h based models- so I was also a little wrong. Brisbane got used for the K8 ones and Kuma, which was a chip harvest of Agena, the Phenom X4's core, got used for the K10h ones.

You are getting waaay too literal here and it isn't helping your point. The "Athlon X2" line is still from the Athlon 64 family (K8). The names are used interchangeably (even the wikipedia article uses both names when referencing the same product lineups). I guess I should've been more explicit in saying "there's no such thing as a different Athlon X2 vs an Athlon 64 X2; they're both the same thing".

Keep in mind, there are also the socket FM2 Athlon X2s, which also don't have the "64" in the name, despite being 64 bit. But unlike the K8 Athlon 64 vs Athlon X2, the FM2 models are a totally different architecture (K12?). I'm guessing these were not the ones you were referring to.

[Adarion]

Hopefully those old boards won't need any new coreboot work, as the old code (like the old boards) still exists. Presumably the coreboot devs will be keeping it around in the usual way git and other version control systems work and/or as release old tarballs. Still, given how many new laptops are getting shit like Intel boot-guard and all the new blobs on newer procs, it seems to me that the Coreboot project should not be dropping older boards as well. Maybe the libreboot project will revert the offending commits on their branches, at least so long as that remains possible without a new rewrite.

I dearly hope so!
For once the new HW is crappy when it comes to Coreboot due to lack of manpower to support it and of course due to DRM-style crap.
On the other hand there is support for old boards coming in every now and then, but often those boards are no longer available (unless used). So on one hand they publish support for old boards, but then hamper people from having other old things supported.
I know they do lack some manpower to really take care of all the code, and I wish I was a coder so I could support the cause. But narrowing coreboot support still hurts a lot.

And a lot of people are so fed up with all that UEFI crap, there should really be a welcoming audience for Coreboot/LibreCore/Libreboot.

[lilos]

Never buy new bord with UFEI crap if not possible coreboot to be intaled !
I think Linux kernel must run in some virtual hyperviser and in this way we can escape spectre v xxx hack !

[AndyChow]

Maybe eventually coreboot will stop supporting all cpus. That would decrease the code complexity quite a bit. I mean, right now that means coreboot is chromebook, purism, and a few old lenovos.

[avph]

Coreboot has removed target support from the master on numerous occasions in the past. The reason is always that the code is not up to standards and hinders further development. Often the code goes untested for years, partly due to the hardware becoming unavailable, with in the end very little hope of the code in master actually working. Due to the lack of developers and automatic hardware testing, coreboot developers chose to drop support over having that code hanging around in a degraded state. This is the best for developers but also for users since they will likely have better hardware support with older revisions. This is certainly the case with amdk8. It should ofc be noted that if one cleans up the code and brings it up to standards, nothing prevents it from being merged again.