| Rev |
Age |
Author |
Path |
Log message |
Diff |
| 2446 |
6362 d 9 h |
jermar |
/trunk/kernel/generic/ |
More efficient and simpler task termination.
Based on the assumption, that after its creation, only the task itself can create more threads for itself,
the last thread with userspace context to execute thread_exit() will perform futex and IPC cleanup. When
the task has no threads, it is destroyed. Both the cleanup and destruction is controlled by reference
counting.
As for userspace threads, even though there could be a global garbage collector for joining threads, it is
much simpler if the uinit thread detaches itself before switching to userspace.
task_kill() is now an idempotent operation. It just instructs the threads within a task to exit.
Change in the name of a thread state: Undead -> JoinMe. |
|
| 2445 |
6362 d 10 h |
decky |
/trunk/ |
initial merge of branches/fs
(not finished, huge cleanup is needed) |
|
| 2444 |
6363 d 1 h |
jermar |
/trunk/kernel/ |
First fixes for suncc support.
It is going to be a long way... |
|
| 2441 |
6363 d 3 h |
decky |
/trunk/kernel/ |
fix ICC compilation |
|
| 2440 |
6363 d 3 h |
jermar |
/trunk/kernel/generic/ |
Fix the problem with sys_thread_create() by splitting the create and attach functionality of
thread_create(). Now it is possible to specify a flag that will cause thread_create() to only allocate and
initialize the thread structure. A call to thread_attach() will make the thread visible to the system.
This arrangement makes it easier to undo creation of a thread in case of a failure in sys_thread_create(). |
|
| 2310 |
6387 d 6 h |
jermar |
/trunk/kernel/generic/ |
Cleanup the waitq_wakeup() interface.
Replace numeric constants (i.e. 0)
and boolean constants (i.e. false) with
more readable WAKEUP_FIRST. Also change the
type of the second argument of waitq_wakeup()
to a newly introduced type wakeup_mode_t.
Fix behaviour of waitq_wakeup() in case
that WAKEUP_ALL semantics is required
but no threads are sleeping in the wait
queue. This is a similar fix to that of
Jan Hudecek committed in the RCU branch,
but, IMHO, is more straightforward and
also doesn't eat up previous missed
wakeups. |
|
| 2275 |
6404 d 6 h |
decky |
/trunk/kernel/generic/ |
add fancy uptime kconsole command |
|
| 2268 |
6406 d 5 h |
jermar |
/trunk/kernel/generic/ |
Document and assert proper usage of THREAD_FLAG_WIRED.
Original patch by Jan Hudecek. |
|
| 2227 |
6415 d 18 h |
decky |
/trunk/kernel/ |
start shutdown infrastructure |
|
| 2218 |
6417 d 1 h |
decky |
/trunk/kernel/ |
support the possibility to send EOI or Interrupt Acknowledgement
prior to processing the interrupt
(this is essential on some architectures to prevent preemption deadlock) |
|
| 2216 |
6417 d 4 h |
decky |
/trunk/ |
make thread ID 64 bit (task ID is 64 bit already)
cleanup thread syscalls |
|
| 2208 |
6417 d 7 h |
decky |
/trunk/kernel/generic/ |
use spinlock only on console output, not other print functions
cleanup |
|
| 2187 |
6418 d 12 h |
decky |
/trunk/kernel/generic/ |
add syscall for getting current thread ID |
|
| 2183 |
6419 d 8 h |
jermar |
/trunk/kernel/generic/ |
Continue to de-oversynchronize the kernel.
- replace as->refcount with an atomic counter; accesses to this
reference counter are not to be done when the as->lock mutex is held;
this gets us rid of mutex_lock_active();
Remove the possibility of a deadlock between TLB shootdown and asidlock.
- get rid of mutex_lock_active() on as->lock
- when locking the asidlock spinlock, always do it conditionally and with
preemption disabled; in the unsuccessful case, enable interrupts and try again
- there should be no deadlock between TLB shootdown and the as->lock mutexes
- PLEASE REVIEW !!!
Add DEADLOCK_PROBE's to places where we have spinlock_trylock() loops. |
|
| 2170 |
6424 d 2 h |
jermar |
/trunk/kernel/ |
Simplify synchronization in as_switch().
The function was oversynchronized, which
was causing deadlocks on the address
space mutex.
Now, address spaces can only be switched
when the asidlock is held. This also protects
stealing of ASIDs. No other synchronization
is necessary. |
|
| 2142 |
6427 d 9 h |
jermar |
/trunk/kernel/generic/include/mm/ |
Remove out of date comment. |
|
| 2128 |
6471 d 5 h |
jermar |
/trunk/ |
Add arm32 architecture. The 32 suffix is used to specify that 16-bit Thumb
instructions are not used. The arm32 code is mostly composed of placeholders
that need to be replaced by real implementation. So far, the arm32 tree
only compiles. If run under GXEmul simulator, an infinit loop at the
kernel entry point will be entered. |
|
| 2126 |
6472 d 2 h |
decky |
/trunk/kernel/ |
small cleanup, more work coming |
|
| 2125 |
6472 d 4 h |
decky |
/trunk/ |
experimental support for Objective C
(disabled by default) |
|
| 2124 |
6472 d 6 h |
decky |
/trunk/kernel/generic/ |
implement simple realloc() |
|