| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2024-11-02 | check spawn exhaustion | Kimplul | |
| 2024-11-02 | add spawn benchmark | Kimplul | |
| 2024-11-02 | intial working sys_spawn | Kimplul | |
| 2024-11-02 | initial exec support | Kimplul | |
| 2024-11-01 | use correct sfence.vma | Kimplul | |
| 2024-11-01 | simplify enter_rpc, remove finalize_rpc | Kimplul | |
| + The simulated benchmarks shows very slightly more than 2M requests per second, new record! | |||
| 2024-11-01 | calculate rpc stack top slightly differently | Kimplul | |
| + Saves a multiplication, very minor | |||
| 2024-11-01 | mark some hot path functions inline | Kimplul | |
| + Gives a hint to the LTO optimizer to more agressively inline the functions, seems to bring a slight performance increase | |||
| 2024-11-01 | remove benchmarks autogenerated file | Kimplul | |
| 2024-11-01 | flush only current tlb during page fault | Kimplul | |
| 2024-11-01 | copy rpc stacks directly | Kimplul | |
| + Speeds up fork a little bit | |||
| 2024-11-01 | move stack modifications to be earlier in rpc | Kimplul | |
| 2024-11-01 | speed up creating threads quite a bit | Kimplul | |
| 2024-11-01 | test all ipc variants more-or-less properly | Kimplul | |
| 2024-11-01 | formatting and docs updates | Kimplul | |
| 2024-11-01 | update some docs | Kimplul | |
| 2024-11-01 | add note about running on real hardware | Kimplul | |
| 2024-11-01 | smp on visionfive2 'works' | Kimplul | |
| 2024-11-01 | remember to set regs to some valid address | Kimplul | |
| + Worked in qemu, failed in visionfive2, fixed. | |||
| 2024-11-01 | rpc actually marks and unmarks stack regions | Kimplul | |
| + Processes won't be able to read previous stack frames etc | |||
| 2024-10-30 | clarify visionfive2 docs | Kimplul | |
| 2024-10-30 | bookkeeping | Kimplul | |
| 2024-10-30 | tests pass | Kimplul | |
| 2024-10-30 | most of the way to passing tests | Kimplul | |
| 2024-10-29 | start moving towards threads always being in rpc | Kimplul | |
| 2024-08-30 | improvements to shared memory handling | Kimplul | |
| + Now each shared region is reference counter (technically each region is refernce counted, private regions just have a count of 1). Also, syscalls that touch the TLB get flushed, but should probably look into where else this flushing might be needed. | |||
| 2024-08-30 | add some basic benchmarks | Kimplul | |
| 2024-08-30 | fix a couple memory issues | Kimplul | |
| + get_mem_node() clears out the memory so we don't accidentally read stale flags + Thread creation simplified a bit, each thread references itself and data freeing is based on when the reference count reaches zero. | |||
| 2024-08-30 | specify mabi and march | Kimplul | |
| 2024-08-30 | speed up ipc_resp a bit | Kimplul | |
| + Skip unwinding stack and just jump directly to userspace | |||
| 2024-08-27 | convert addr to page in find_closest_used_region | Kimplul | |
| 2024-08-27 | add some missing documentation | Kimplul | |
| 2024-08-27 | add optimizations for memcpy and memset | Kimplul | |
| + Apparently primary bottlenecks for sys_fork() and sys_create(), speeds things up quite a bit | |||
| 2024-08-26 | add magic assert | Kimplul | |
| 2024-08-26 | fix memory leaks destroying threads | Kimplul | |
| 2024-08-26 | make sys_create() better + test | Kimplul | |
| 2024-08-26 | check for running out of memory during fork | Kimplul | |
| 2024-08-26 | improve killing processes/threads, test | Kimplul | |
| + Remove sys_kill() as we should be using a two-stage process where a thread is first orphaned by sys_detach(), and then the thread itself calls sys_exit() after it has done all necessary cleanup in init. | |||
| 2024-08-26 | don't free the physical address of page tables | Kimplul | |
| 2024-08-26 | simplify tests | Kimplul | |
| 2024-08-22 | add fork test | Kimplul | |
| 2024-08-22 | add ipc_notify testcase, remove ipc_ghost | Kimplul | |
| + A bit easier to implement just ignoring the ipc_resp() arguments than try to enforce that a notification is exited from by ipc_ghost(), especially in cases where the root process has been killed. | |||
| 2024-08-22 | add note about tests | Kimplul | |
| 2024-08-22 | change sys_ret handling to always include id | Kimplul | |
| 2024-08-22 | add basic ipc-req test | Kimplul | |
| 2024-08-21 | ensure our memory is actually used up fully | Kimplul | |
| 2024-08-21 | add malloc test | Kimplul | |
| 2024-08-18 | small fixes to issues reported by analyzer | Kimplul | |
| + Currently analyzer has to be run manually with something like make CFLAGS='-fanalyzer -Wno-analyzer-infinite-loop' I use an infinite loop as an assert, I know it's not great/technically UB but it's just a fallback. + Analyzer is still somewhat limited, I could add in more attributes about different functions, such as memory allocation sizes etc. + If I ever set up a CI pipeline, remember to use analyzer? | |||
| 2024-07-18 | mark some ipc functions inline | Kimplul | |
| + Seems to improve code quality a little bit | |||
| 2024-07-18 | add ipc_tail() | Kimplul | |
| + ipc_kick() does a forward and a tail at the same time The idea with ipc_tail() is to allow 'trusted' calls, so for example a process is not allowed to willy-nilly open a file, as it has to go via init(), making the eid 1. This way the receiver can know that the request has gone through init() and can open up a new connection (file, whatever) after which requests from that pid/tid are allowed without init() intervention | |||
 |
index : kmi | |
| ACTIVE - Kernel with migrating threads as IPC | kimi.h.kuparinen+kmi@gmail.com |
| aboutsummaryrefslogtreecommitdiff |