Sam Kuper wrote:
GNUtoo in 2013: GNUtoo in 2013: phcoder in 2013: phcoder in 2014: Kl3 in 2015
So, it would seem that during 2013-2015, native graphics init became operational within Coreboot for the X201.
TL;DR: That code was always poor, don't expect it to work reliably.
Don't assume that any merged code is general and robust, only because it might have worked for some developers in some case.
If you need reliable graphics init then please use the VGA BIOS until you can [sponsor someone who can] analyze and address (potentially also design) problems in the native graphics init C code.
If you do go on into development, then please do not waste time on the native graphics init C code, but instead please work on making use of libgfxinit on the platform you are interested in. It is the way to go forward.
Thanks
//Peter
On 24/03/2017, Peter Stuge peter@stuge.se wrote:
Sam Kuper wrote:
GNUtoo in 2013: GNUtoo in 2013: phcoder in 2013: phcoder in 2014: Kl3 in 2015
So, it would seem that during 2013-2015, native graphics init became operational within Coreboot for the X201.
TL;DR: That code was always poor, don't expect it to work reliably.
Don't assume that any merged code is general and robust, only because it might have worked for some developers in some case.
If you need reliable graphics init then please use the VGA BIOS until you can [sponsor someone who can] analyze and address (potentially also design) problems in the native graphics init C code.
If you do go on into development, then please do not waste time on the native graphics init C code, but instead please work on making use of libgfxinit on the platform you are interested in. It is the way to go forward.
Thanks Peter. I appreciate your work on Coreboot, and your insight above.
I would still be interested in working .config files for the X201, though ;) Why? Not only to see if your (or someone else's) .config would succeed in building Coreboot for my X201 with native graphics init, but also because more generally, from the perspective of reproducible builds and reproducible research, it would be great to see Coreboot users and developers publishing, for builds that they have made:
- the .config ,
- the Git revision,
- whether me_cleaner was used,
- a hash (e.g. SHA512) of the resulting coreboot.rom , if the build succeeded,
- the success or otherwise of booting the build, and
- maybe the coreboot.rom file itself.
But perhaps I should air that general wish in another thread some time, instead of hijacking this one! I should also research whether anyone has aired similar wishes already :)
Best regards.
-
Peter Stuge -
Sam Kuper