Axel Heider

Results 136 issues of Axel Heider

Follow up from https://github.com/riscv/riscv-isa-manual/pull/691 The PLIC chapter has been removed from the privileged specification and will be maintained at https://github.com/riscv/riscv-plic-spec now.

In the section 'Interrupt Completion' is says "_... gateway will only forward additional interrupts to the PLIC core after receiving the completion message._". But what is the behavior between the...

The 'K' constraint is for 5-bit unsigned integer immediate operands only in both GCC and LLVM. It does not apply to the value written. See manuals: - https://llvm.org/docs/LangRef.html#supported-constraint-code-list - https://gcc.gnu.org/onlinedocs/gcc/Machine-Constraints.html...

cleanup

The thread context structure layout is explicitly designed to have the register context first, as this simplifies saving the user context on entry in assembly code. On the exit path...

cleanup

Remove the return value from `configure_sched_context()`, because it never fails. As a consequence, `create_idle_thread()` also never fails and does not need a return value.

cleanup
proof-test

Use the thread passed as parameter instead of making assumption that this is in sync with the global state. Using `NODE_STATE(ksCurThread)` feels semantically wrong when `cur_thread` is passed to the...

cleanup

From reviewing the code, I think the refactoring of commit https://github.com/seL4/seL4/commit/4b491dcf0c2c19860c5c9b503539bf0e9b0d1667 picked the wrong define name here. In `include/arch/arm/arch/machine/debug_conf.h` the define `ARM_CP14_SAVE_AND_RESTORE_NATIVE_THREADS` is created. This change aligns the code with...

cleanup

- doc: improve comments - use types time_t and tickts_t properly, currently the semantics are incorrect - print period and budget with units The PR in seL4_libs is not necessary,...

docs
cleanup

Having a bit of documentation for the HTIF would be good. Currently this still feels a bit like a undocumented useful hack that has been there of quite some time...

Calling `make clean` requires `PLATFORM` to be valid. Is this intended or should cleaning always work because all that really matters for the cleaning process is the parameter `O=some/output/path`? I'm...