Index of /archives/linux/kernel.org/kernel/people/rml/sched/for_alan
Name Last modified Size Description
Parent Directory -
110-remove-wake-up-sync.patch 2002-04-21 08:40 5.3K
120-need_resched-abstraction.patch 2002-04-21 08:40 2.4K
130-frozen-lock.patch 2002-04-21 08:40 2.3K
140-sched_yield.patch 2002-04-21 08:40 1.4K
145-more-sched_yield.patch 2002-05-08 03:31 4.0K
150-need_resched-check.patch 2002-04-21 08:40 715
160-maxrtprio-1.patch 2002-04-21 08:40 2.1K
165-maxrtprio.patch 2002-05-08 03:31 2.1K
166-maxrtprio.patch 2002-05-31 08:37 3.2K
170-migration_thread.patch 2002-04-21 08:40 12K
175-updated-migration_init.patch 2002-05-08 03:31 5.1K
180-misc-stuff.patch 2002-04-21 08:40 5.4K
185-more-misc-stuff.patch 2002-05-31 01:26 3.8K
190-documentation.patch 2002-05-31 01:26 13K
200-sched-yield.patch 2002-09-12 07:02 3.5K
210-sched-comments.patch 2002-09-12 07:02 12K
220-task_cpu.patch 2002-09-12 07:02 5.2K
230-sched-misc.patch 2002-09-12 07:02 4.9K
README 2002-09-12 07:09 4.0K
sha256sums.asc 2023-04-26 06:12 2.5K
Against 2.4.19-pre7-ac2
145, 165, and 175 are against 2.4.19-pre7-ac4.
166, 185, and 190 are against 2.4.19-pre8-ac5
200, 210, 220, and 230 are against 2.4.20-pre5-ac4
110-remove-wake-up-sync.patch
We do not need sync wakeups anymore, as the load balancer handles
the case fine. Remove wake_up_sync and friends and the sync flag
in the __wake_up method.
120-need_resched-abstraction.patch
Abstract away access to need_resched into set_need_resched, etc.
130-frozen-lock.patch
Fix scheduler deadlock on some platforms. I'll let DaveM (the author)
explain:
Some platforms need to grab mm->page_table_lock during switch_mm().
On the other hand code like swap_out() in mm/vmscan.c needs to hold
mm->page_table_lock during wakeups which needs to grab the runqueue
lock. This creates a conflict and the resolution chosen here is to
not hold the runqueue lock during context_switch().
The implementation is specifically a "frozen" state implemented as a
spinlock, which is held around the context_switch() call. This allows
the runqueue lock to be dropped during this time yet prevent another cpu
from running the "not switched away from yet" task.
140-sched_yield.patch
Optimize sched_yield.
145-more-sched_yield.patch
More abstractions to yield()
150-need_resched-check.patch
A new task can become runnable during schedule(). We always want to
return from scheduler with the highest priority task running, so we
should check need_resched before returning to see if we should rerun
ourselves through schedule. This used to be in the scheduler but was
removed and then readded.
160-maxrtprio-1.patch
Cleanup assumptions over what the value of MAX_RT_PRIO. No change to
object code; just replace magic numbers with defines.
165-maxrtprio.patch
Separate notion of "maximum real-time priority" from "maximum
user-space real-time priority" via MAX_RT_PRIO vs MAX_USER_RT_PRIO
defines.
166-maxrtprio.patch
Further cleanup the code and move the defines to sched.h
170-migration_thread.patch
Backport of the migration_thread migration code from 2.5. This
includes my interrupt-off bugfix and wli's new migration_init code.
The migration_thread code allows arch-independent task migration
via set_cpus_allowed and allows the creation of things like task cpu
affinity interfaces.
175-updated_migration_init.patch
Rewrite of migration_init using Erich Focht's simpler method of using
the initial migration_thread to migrate any future threads. Also
includes a fix for arches where logical != physical CPU mapping.
180-misc-stuff.patch
Lots of misc stuff, almost entiirely invariant and trivial cleanups.
Specifically:
- rename lock_task_rq -> task_rq_lock
- rename unlock_task_rq -> task_rq_lock
- cleanup lock_task_rq
- list_del_init -> list_del fix in dequeue_task
- comment cleanups and additions
- load_balance fixes and cleanups
- simple optimization (rt_task -> policy!=SCHED_OTHER)
185-more-misc-stuff.patch
More misc. cleanups and improvements:
- move sched_find_first_bit from mmu_context.h to bitops.h
as in 2.5. Why it was ever in mmu_context.h is beyond me.
- remove the RUN_CHILD_FIRST cruft from kernel/fork.c.
Pretty clear this works great; we do not need the ifdefs.
- Add comments to top of kernel/sched.c to briefly explain
new scheduler design, give credit, and update copyright.
- set_cpus_allowed optimization from Mike Kravetz: we do not
need to invoke the migration_thread's if the task is not
running; just update task->cpu.
190-documentation.patch
Add sched-coding.txt, a description of scheduler methods and
locking rules, and sched-design.txt, Ingo's original lkml
email detailing the goals, design, and implementation of the
scheduler.
200-sched-yield.patch
Fix sched_yield for good. Seriously.
210-sched-comments.patch
Glorious comments everywhere.
220-task_cpu.patch
"set_task_cpu()" and "task_cpu()" abstraction.
230-sched-misc.patch
misc. and trivial cleanups