RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-05-31 05:38:11 +00:00
Code Issues Activity Actions

Kernel: Various context switch fixes

Browse source 
These changes solve a number of problems with the software
context swithcing:

* The scheduler lock really should be held throughout context switches
* Transitioning from the initial (idle) thread to another needs to
  hold the scheduler lock
* Transitioning from a dying thread to another also needs to hold
  the scheduler lock
* Dying threads cannot necessarily be finalized if they haven't
  switched out of it yet, so flag them as active while a processor
  is running it (the Running state may be switched to Dying while
  it still is actually running)
This commit is contained in:
Tom 2020-07-05 14:32:07 -06:00 • committed by Andreas Kling
parent 49f5069b76
commit 2a82a25fec
9 changed files with 235 additions and 89 deletions
Download patch file Download diff file Expand all files Collapse all files

14
Kernel/Lock.cpp
View file

@ -73,8 +73,13 @@ void Lock::lock(Mode mode)
// switch will happen, so yield.
// The assumption is that if we call this from a critical section
// that we DO want to temporarily leave it
TemporaryChange change(Processor::current().in_critical(), 0u);
u32 prev_flags;
u32 prev_crit = Processor::current().clear_critical(prev_flags, !Processor::current().in_irq());
Scheduler::yield();
// Note, we may now be on a different CPU!
Processor::current().restore_critical(prev_crit, prev_flags);
}
}
}
@ -105,8 +110,13 @@ void Lock::unlock()
// I don't know *who* is using "m_lock", so just yield.
// The assumption is that if we call this from a critical section
// that we DO want to temporarily leave it
TemporaryChange change(Processor::current().in_critical(), 0u);
u32 prev_flags;
u32 prev_crit = Processor::current().clear_critical(prev_flags, false);
Scheduler::yield();
// Note, we may now be on a different CPU!
Processor::current().restore_critical(prev_crit, prev_flags);
}
}