Sorry, but I thought that I uploaded the code!
I think it is in the function "scheduler_enqueue_smp", because if I comment out the code for taking the spinlock of the thread then something get printed and I get a pagefault. With the code I get a deadlock.
sync on a smp system (again)
Re:sync on a smp system (again)
Ok, you can?t find a deadlock, because there isn?t anyone ::) I forgot to change my multitask init function, which worked with the old code and there I set the spinlock acquired for the idle thread and so the scheduler can?t get this spinlock and will loop forever
Now I get a page fault and I will look if I can find the failure.
Now I get a page fault and I will look if I can find the failure.
Re:sync on a smp system (again)
Yeah, I got it finally working I implemented thread status and my scheduler_add_scheduler function waits now till the THREAD_RUN flag is cleared! It is working and I saved a lot of spinlocking the thread struc.