[coreboot] AMD's binary-only AGESA libraries

David Hubbard david.c.hubbard+coreboot at gmail.com
Wed Nov 5 19:36:41 CET 2014

> What do you think … ?

I am happy Kyösti has written such a good analysis of the situation. Thank
you! I only have one question and one business-level observation.

First, the observation: AMD can't abandon the open-source AGESA. It *does*
affect business decisions to go with AMD over Intel: Intel's last binary
holdout is FSP. AMD's binary blob is the GPU firmware. We might see an AMD
Chromebook with open-source AGESA, but we can abandon all hope of that if
AGESA is a binary blob and AMD GPUs are a binary blob too. More
specifically, I'll probably be contributing to a community-maintained fork
of AGESA in the event AMD finally really does abandon it.

Second, the question: AMD Graphics cards have announced a project to create
a unified driver that links with both the open-source radeon and closed
source "fglrx" currently available in linux. AMD claims this approach will
save them effort as the valuable bits in fglrx are separated from all the
API overhead. Why not try a similar approach here?

SAGE can have their chimera-coreboot (with binaryPI), getting to market
faster. See Kyösti's question about the GPL license.

AMD can't abandon the open-source AGESA releases but now that AMD has at
least announced a rationale for their actions, the community better
understands where their priorities are. See Kyösti's question of when AMD
will release scrubbed AGESA code.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.coreboot.org/pipermail/coreboot/attachments/20141105/6b693691/attachment.html>

More information about the coreboot mailing list