alwin-joshy

Results 11 issues of alwin-joshy

This is basically a combination of [Shane's signal fastpath](https://github.com/seL4/seL4/pull/322/commits) from last year and [Anna's signal/IRQ fastpath](https://github.com/pingerino/seL4/blob/phd/src/fastpath/fastpath.c#L421). Shane's fastpath was able to handle signalling to both lower and higher prio threads,...

MCS
proof-test
hw-test

Implementation of the VM fault fastpath with some stubs implemented for other types of faults. Up to 62% faster for faulter -> handler direction and up to 67% faster for...

This pull request relates to the porting of the hardware debug API to aarch64. The API exposes hardware debugging features to userspace and enables the development of debugging tools on...

hw-test
hw-bench

Last year, I published a [pre-rfc](https://sel4.discourse.group/t/pre-rfc-protectn/478) regarding new mapping operations which allowed you to batch unmaps and access right changes in order to make operations on a process' virtual address...

verification

The existing page mapping benchmarks only appear to be implemented for x86 platforms. In this PR, I have extended the benchmarks to work on AARCH64 and fixed a bug with...

New benchmarks for testing VM fault performance - converted some of the tests from seL4test to benchmarks (can later be extended for other types of benchmarks) as well as added...

hw-test

The implementation of TCB_SetSchedParams did not follow the API reference, as it would fail if: a. The TCB already had a scheduling context, even if this was the same as...

verification
MCS

For the seL4_TCB_SetSchedParams (MCS) invocation, the API reference states the following about the `sched_context` parameter > If the scheduling context is already bound to a notification or TCB that is...

Instead of hard-coding invocation numbers, we now get them from a file generated by the seL4 kernel build. Depends on: [bd1c39240944207ef71237f98663acb06cc971f6](https://github.com/seL4/seL4/commit/bd1c39240944207ef71237f98663acb06cc971f6)