Robert

At least on aarch64, this generates a large amount of warnings in every single file that includes the generated header. This buries any other warning. It would be nice to...

> added a new config option to the `syscall-shim` submenu to suppress the warnings. The config option works great for me, thank you for adding it. I experienced no problems...

> Hi @kubanrob, I just noticed that this one cause a compiler warning, as the `off` variable is no longer used: Hey @michpappas, the commit renames the `off` variable to...

@skuenzer > I assume having spaces in file paths did work before. Can we confirm this before merging this PR? Since there is no quoting (that I am aware of)...

@skuenzer It fails really early. If I am missing something obvious to make it work with spaces in paths, please let me know. ``` ~/tmp/build/app-hello world$ make menuconfig make[1]: ***...

I added the `FIXME` commit. I also tried retake the measurement. I am not sure what changed since Unikraft 0.10.0 (or in my setup), but _I can not reproduce the...

> Is there no speed difference anymore or is it just smaller? I did not see any speed difference, at least in this order of magnitude. I remember that was...

> @kubanrob, so you are unable the reproduce even on the older 0.10 version? I did not try with the 0.10 version as the musl support was not there yet,...

> However, when the parent unit address is missing from the DTB, we can assume that the parent node's #address-size property is equal to 0, as per Devicetree specification v0.3...

> Our test application has some loops that busy wait on a shared variable, which obviously deadlocks when the writer is waiting on the same hardware thread and never gets...