Igor

That nobody knows anything about your situation: https://www.armbian.com/bugs/

Exactly. I am crying that we need your help to be able to help you, but you are keep seeking shortcuts. You are almost certainly not the only one with the problem. Research is expense and attention is expensive - did you search forum and Jira if this was already reported, solved or dismissed. Nobody like doubled bug report especially from "customers" that are actually contributing nothing. I always check who is asking. If this is first post, this is abusing forum as technical support service ... This is forum. Make friends first, help, then ask for help. Don't get me wrong and don't take it personal. I am just trying to explain you what "sales" are not telling you in order to keep you happy. What if there are 1000 x more people that are seeking out for time and attention? Majority is trying on abusing minority? Reality is harsh and for problem solving, reality check is needed. Read FAQ for start https://docs.armbian.com/User-Guide_FAQ/ Since you are coming from a world where supporting another hardware means 30 minutes of work, it is difficult not to offend you. We can easily waste years for the same, before its good enough that they put their label on. And we maintain it daily. FAQ will reveal how much that costs project maintainers. I am pretty sure Hardkernel minority - their paid engineer - will not support this problem since this HW is EOL, they also don't have endless resources. If you finance this bug, we can check it once next year. When we will deal with others that wait ...

This is forum which we are maintaining for people that are using unmodified releases. Supporting known Linux is hard and expensive, generate big hole in our pocket, while any small and unknown change significantly raise expenses of support. Not to mention moral aspects. Even long term Armbian users, contributors and supporters does not have access to technical support service you might picture in your head. Project can't afford to cover expenses, so support is best effort in any case. Do you understand what you are asking? When you will finance us for dealing with your problems, history on this forum (posts, subscriptions, crowdfunding) will be very relevant. Start making it.

You mean adjusting HW features? Because there are no overlays and nobody developed them. No, we have not plans nor budget for that. We are also waiting and waiting that you realise that you need to support R&D in order to have some small rights to complain This graph below shows how many days is needed before reported issues is resolved. Terrible, right? https://docs.armbian.com/User-Guide_FAQ/#development-time Remember that if you don't contribute on the project, you can't add bugs / tasks to the system at all. With another words - we have no resources not even for write bugs into the system and if we would listen to your problems, situation for the project that helps you, but you don't help it, would only be much worse. If all people that complain and report bugs to themselves would have Angel https://forum.armbian.com/subscriptions/ we would be able to perhaps hire full time person for you that would record bugs and solve them one by one. Since this is not the case and you expect the impossible, graph can only get steeper.

This bug is here for a while. https://github.com/armbian/config/issues/33 You can enable this manually too. You might need to downgrade kernel to 5.10.y

With a previous release cycle we have added support for Raspberry Pi4. Currently we are producing daily images only (rolling release): https://www.armbian.com/rpi4b/ When someone steps as a maintainer, it can become officially supported: https://docs.armbian.com/User-Guide_Board-Support-Rules/ We only support Ubuntu based builds for now and kernel source is attached to official RaspberryPi sources. Which is better then mainline.

@Heisath In light of those changes - do we need separate tags for Clearfog pro and base model?

Yes, it can go into release. Wrote on mobile

@Contributor/Maintainer If something has to be squeezed into the code ... I guess merging armbian-next has to be postponed. After Sunday, bugs bugs bugs. Solving, not generating new ones

This is open source. What prevents anyone to study the code? And understand this world a bit more before starting (too?) difficult path. Kernel you are asking defconfig about and kernel Armbian is providing can be / are very different. Perhaps thousands of hours diff. Kernel 5.10.y from Orange has many dirty hacks which nobody will not even look into ... and if you want analysis of that diff, Armbian will certainly not provide that. We have no time and we had it, the answer would be the same. if you are curious, this forum and the code has many answers ... search forums, peek into the code, tell others what you have find out. People that understand it are busy and rare. I highly recommend you to build the kernel the recommended way. If not, good luck, you will need it. In order to not waste too much time answering the same over and over again https://docs.armbian.com/User-Guide_FAQ/ And. This is forum where many people are solving difficult questions. Forum. Make friends, help community helping you. And thank you for supporting Armbian. Welcome!

If you start on Jammy, https://github.com/armbian/build/blob/armbian-next/lib/functions/host/prepare-host.sh#L46 is no more. Also office unblocked at kernel.org

In case you are not planning to develop or sponsor this adventure I suggest you to stick to stock kernel. Functions in modern kernels are never complete, especially this is critical when comes to external devices such as displays. Hardkernel never supported modern kernels nor our work on this device while end user donations are only enough to address one medium complex bug in whole project. Per year. This here could pull some serious R&D ... but someone perhaps knows something. Hard to judge without investigation, but what I am pretty confident it won't be as simple as this idea.

Ban is still in place I am writing an email right now. In general it would be better to use Google mirror primarily. Or we need to use kernel.org Git for bundles? You mean like having our own full kernel mirror? I would not push on moving everything to distro compilers at once. What works, works and we do this gradually. Going on half now, removing them all in one year. Something like that. Even today, sometimes we will need to deal with a special compiler, so support has to remain. And a few of them if there is too much work with adopting the code. Thanks for update. Will do another cycle of review when possible.

Best way is to start here: https://forum.armbian.com/staffapplications/ Welcome!

It is. Try here: https://armbian.systemonachip.net/archive/ but also get a bit more familiar with a build system. Official fresh images when someone takes responsibility for https://docs.armbian.com/Release_Board-Maintainers/ those boards. Until then ...

Search Odroid forums for similar problems. Then search Raspbery Pi forums for similar problems ... then ask storage experts on some storage oriented forums if they recommend USB connected drives and why not. USB protocol / drives were not designed for 24/7 usage and Armbian can't fix what is fundamentally broken. Disabling UASP might help.

@private/contributor Unifying TAGS as much as possible - as low count / universal as possible has been enforced to GitHub merge requests and issues. From now on, we have only 12 labels:

Release is not just freezing the code and generating images. Its a time when we pull together stronger, agree and deal with common goals and ideas. Things we are dealing with are complicated. Linux is complex, our framework is getting there too and only together we can deal with all this. I started on freezing the kernel version. There is also some breakage, some related to our patches, some not. But since we have to freeze it sooner or later, perhaps now its the right time. This brings a few good things at once: temporally release people that are working close with maintaining kernel code @going@jock@balbes150@amazingfate myself ... this can bring more resources on trying to match with / merge armbian-next. @rpardiniworked for moths on this. Lets review the code and help sorting out bugs. Again. IMHO it is possible to merge, but if code remain unstable by the code freeze date (2022-05-14), we are merging after new images are generated. asking @Contributor/Maintainer board maintainers to engage in images building and testing (currently still from master branch) and bug reporting as earlier as possible (once PR 3736 gets merged) reproducible build framework in release branch. Which is most important long term consequence. Additionally, we would like to finish u-boot upgrade on Rockchip and release major forum refactoring @TonyMac32 @piter75 @balbes150 Happy Workers' Day!

Read this https://docs.armbian.com/User-Guide_FAQ/ and act responsibly.

Bug was introduced with last upstream kernel changes and was reported by several people, including you. With free support, this is it. With paid support - its holiday.

@jimmij If you find more inspiration: https://github.com/armbian/wallpapers

Don't expect everything to work. That is close to mission impossible in this world. Armbian maintainers generate huge loss by dealing with their "customers", so this is not a problem of Armbian in any way. https://docs.armbian.com/User-Guide_FAQ/#why-does-hardware-feature-xy-work-in-old-kernel-but-not-in-more-recent-one This might shed some light. Support on modern kernel is written / ported basically from scratch and its mainly maintained by amateurs. Until they find fun / need in that. You build and boot image from sources that works is already a nice success. A lot of work is needed in the backstage that this looks so simple.

We generate different images when images can't run without issues on both devices or when its too expensive creating support for both in one binary. Diff goes between not booting at all and lack of some functions. If this data is not marked / written somewhere, a research is needed to answer this questions. Try and report to M4V2 board maintainer if you discover something useful ...

Costs from 2 USD on and its a very valued tool in this world: https://linux-sunxi.org/UART

"Delete My Account" https://forum.armbian.com/terms @Moderators