Thanks for everyone who has submitted status reports and now proved that they work.
BTW the boards devices support ASPM so you can remove the pcie_aspm=off in your kernel command line - I like to force it on myself with pcie_aspm=force.
On 04/06/2018 09:54 PM, David Hendricks wrote:
On Fri, Apr 6, 2018 at 3:40 PM, Taiidan@gmx.com Taiidan@gmx.com wrote:
Like I have said before these types of policies are eventually going to result in coreboot only having unobtainable development boards in the tree (that are of course not owner controlled)
It simply isn't right.
Indeed, this isn't right (as in correct) so don't spread this FUD. The boards are still in the tree, you just need to check out whatever coreboot version is known to actually work with the board. For example if a board was last reported working in coreboot-4.6, then `git checkout 4.6` or checkout a specific hash reported on the board_status repo.
I swear I heard "removed from the tree" somewhere - irregardless only unobtainable closed-source development boards will benefit from new coreboot features.
Eventually the old versions of coreboot will become non-functional for whatever reason so it won't be that simple (ie: don't checkout master) if one doesn't require the new features - already for example there are various old libs that coreboot requires which are only available *unsigned* from a single site.
It does no good for users to have hundreds of boards in master that fail to boot, and no good for developers who need to maintain and refactor code for boards that nobody tests and have been abandoned.
I am not complaining because some random boards from 2005 that no one uses are being removed - this is because the last owner controlled x86_64 boards will eventually be functionally removed (and would have been if no one had submitted status for the D8).
This isn't hundreds of boards that fail to boot - it is a few boards that do boot - and people know that they do (as everyone can see now that status have been submitted :D)
As an example I don't mind the removal of the H8SCM because AGESA doesn't support IOMMU and the boards are now quite expensive (and not worth it for the money).
There's obviously a few people on this list using the Asus boards mentioned which is great. The issue we need to solve is getting more people to submit test results so that this isn't a problem in the future.
Even if I didn't use mine for something important I am unable to submit results because I refuse to provide my "real" name and am too honest to use a fake name.