Am 23.07.2015 um 14:38 schrieb Alexander Graf: >> Am 23.07.2015 um 14:22 schrieb Guillaume Gardet <[email protected]>: >> Le 23/07/2015 14:16, Andreas Färber a écrit : >>>> Am 20.07.2015 um 14:28 schrieb Guillaume Gardet: >>>> Le 19/07/2015 18:15, Misha Komarovskiy a écrit : >>>>> Also lcd screen stays black after kernel start, probably some drm >>>>> module still missing. >>>> Upstream kernel is broken. :( >>> How did you test? Actually there's a known issue that the Exynos IOMMU >>> support breaks graphics on the Chromebooks because they enable a >>> framebuffer early in bootloaders, before Linux initializes the IOMMU. >>> Since some other platforms do need the IOMMU, I fear that we're going to >>> need a new kernel flavor... :/ Not sure if that can be scripted as lpae >>> minus certain options? >> >> You mean that if we disable IOMMU in kernel, graphics are working fine? >> CONFIG_DRM_EXYNOS_IOMMU is enabled. Should we disable it for chromebooks? > > Considering that chromebooks are the only Exynos devices with an actual user > base, whatever we need to do to enable those wins over other random Exynos > garbage. So if all out takes is to disable the Exynos IOMMU I'm all for it.
Ehrm, given the amount of breakage and lateness of reports, I disagree that Chromebooks have any significant userbase on Tumbleweed. And depending on which options we need to disable, it may affect non-Exynos users as well, which I would strongly object to. Finding a way to blacklist (or patch?) things specifically for the Chromebooks, as Guillaume suggests, sounds much more appealing to me. Regards, Andreas -- SUSE Linux GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Dilip Upmanyu, Graham Norton; HRB 21284 (AG Nürnberg) -- To unsubscribe, e-mail: [email protected] To contact the owner, e-mail: [email protected]
