But, this is the best approach for developers and people who want to dive deeper in the OS. For plain users, and we have observed that happening, the whole situation seems frustrating and they leave the community. There is a need to keep a healthy and productive environment for people to feel that things are moving forward. I just hope that the number of people who would care about the survival and expandability of the platform will increase every year.

..Waiting for new SDK release, tnx Walkero!

Our goal for clib4 is not to bring confusion to the developers, but provide a better option and free them from any barriers and blocks. Instead of bitching about having to set a gcc argument for a different libc, we all should be grateful that a handful of people are doing what is needed for the development environments for our platform. They develop, they update, they test, and they benchmark as much as possible.

This thread is a great example of the barriers to entry for Amiga development.

Opensource OS4 newlib could be a solution, at least clib4 could benefit from some OS4 specific code and resolve I/O speed issues and maybe other tips too.

George is working on a native add-on for the SDK that will include clib4 and other components, so it can be used if needed. It should be released soon for those who require it.

While I'm not advocating for anything specific, in my opinion, building modern projects on OS4 with an "out-of-the-box in minutes" experience, like on major platforms, is not feasible and likely never will be.

You don't need a static libdl.a to load an additional library.
The program can be compiled completely statically but by dlopen through libdl.so an additional .so library for your needs.
I'm not quite sure if that's what you meant in what you wrote.

There is no libdl.a because under AmigaOS you cannot use the libdl functions if you are not using dynamic linking.

Well, because it's not just an Amiga blog. Ko-fi is a system for creators that lets you - apart from presenting content - build a network of supporters. It makes it easy to inform them about what you have done/written, and allows them to support your efforts financially, even on a regular basis via custom-defined support tiers. As payments are involved, the technology needs to be up-to-date and secure. I couldn't possibly achieve all that with my blogsite tweaked and scaled down for Amiga compatibility.

these drivers have not been officially released.

Client can be not that hard to do , but question is if there any advantage in compare with just ftp or samba then ? As 9p is network based protocol, mean same working network, same sharing as we already can do with samba ? Or there is advantage of usage p9 instead ?

Maybe the manufacturer is expecting a driver module in the Linux kernel? Mmm, hadn't considered that.

I figured the architecture is similar enough that they might eventually stumble on the answer. That is, if they perform Linux DMA tests comparing the Merari with the X5000 and find both are similar, then maybe it's a limitation of the SoC ?

Oh, I hadn't realized that! Then kudos to them both for undertaking this project!

Using a kickstart module text file for config:
If you have access to the AmigaOS 4.x sources check for example my diskboot.kmod sources (Hyperion has no permission to use it anymore, but you can use the part of my sources parsing the config for your drivers).
IIRC it's something like...

@geennams's nearly 2 GB/s nvme.device benchmark result on the X5000 with PCIe x4 v2.0 is only a theoretical result, transferring 128 MB of data in a single DMA transfer is next to never done with real world DOS/filesystem operations.
Using 16 KB it's only 222 MB/s, and with the max. transfer size NGFS uses (128 KB) it's 450 MB/s.
I don't know anything about 3D GPUs, but I guess most GPU texture, etc. DMA transfers are quite small as well and not several MB at once either.

On real hardware sure but I meant for QEMU with vfio-pci. I don't even know if it would work currently so maybe it's not top priority over other possible improvements. But adding a config for it helps testing it in any case.

They say optimists are oblivious pessimists But it does look like from my perspective that your work is just going into a black hole...