Indan Zupancic

> The intention of the change seems right to me, I don't know enough about the hardware details to judge all implementation aspects. I double checked with the ARM documentation...

@heshamelmatary, do you remember why it was done the way it was done?

> Let's not merge this before we have found out. Not knowing is not a good enough reason to change something. Whatever it is, it warrants a comment. Only reason...

> [seL4/rust-sel4#167](https://github.com/seL4/rust-sel4/pull/167) adds support for resettable protection domains. The best way that I've come up with to achieve this requires manipulating the program image _after_ linking. If you copied the...

> If this does not work (i.e. this creates collisions during linking) It should work, statically declared functions don't generate global symbols.

Same for RISCV and x86, they all count on an invalid VSpace cap and falling back to the global page table. I think it would be cleaner if we introduced...

I've only seen this when running Linux in a VM, not when only running normal seL4 tasks yet.

> It would be really useful if the IRQ number could be printed out as well. What platform? That's what I thought too when I had it, but on our...

> I _believe_ this issue has been resolved by [#1183](https://github.com/seL4/seL4/pull/1183). At least, from the tests we ran ([#1183 (comment)](https://github.com/seL4/seL4/pull/1183#issuecomment-2677276986)), no longer occur[1](#user-content-fn-1-93eab00f62a96e5807b3cd066c46cd95). Most likely yes, especially if it has only...

> I've seen this issue happen on GICv2 and GICv3 platforms It would really help if people would share information like this earlier and more often, especially subtle hardware specific...