linux_dsm_epyc7002/arch/x86_64/kernel
Nick Piggin 64c7c8f885 [PATCH] sched: resched and cpu_idle rework
Make some changes to the NEED_RESCHED and POLLING_NRFLAG to reduce
confusion, and make their semantics rigid.  Improves efficiency of
resched_task and some cpu_idle routines.

* In resched_task:
- TIF_NEED_RESCHED is only cleared with the task's runqueue lock held,
  and as we hold it during resched_task, then there is no need for an
  atomic test and set there. The only other time this should be set is
  when the task's quantum expires, in the timer interrupt - this is
  protected against because the rq lock is irq-safe.

- If TIF_NEED_RESCHED is set, then we don't need to do anything. It
  won't get unset until the task get's schedule()d off.

- If we are running on the same CPU as the task we resched, then set
  TIF_NEED_RESCHED and no further action is required.

- If we are running on another CPU, and TIF_POLLING_NRFLAG is *not* set
  after TIF_NEED_RESCHED has been set, then we need to send an IPI.

Using these rules, we are able to remove the test and set operation in
resched_task, and make clear the previously vague semantics of
POLLING_NRFLAG.

* In idle routines:
- Enter cpu_idle with preempt disabled. When the need_resched() condition
  becomes true, explicitly call schedule(). This makes things a bit clearer
  (IMO), but haven't updated all architectures yet.

- Many do a test and clear of TIF_NEED_RESCHED for some reason. According
  to the resched_task rules, this isn't needed (and actually breaks the
  assumption that TIF_NEED_RESCHED is only cleared with the runqueue lock
  held). So remove that. Generally one less locked memory op when switching
  to the idle thread.

- Many idle routines clear TIF_POLLING_NRFLAG, and only set it in the inner
  most polling idle loops. The above resched_task semantics allow it to be
  set until before the last time need_resched() is checked before going into
  a halt requiring interrupt wakeup.

  Many idle routines simply never enter such a halt, and so POLLING_NRFLAG
  can be always left set, completely eliminating resched IPIs when rescheduling
  the idle task.

  POLLING_NRFLAG width can be increased, to reduce the chance of resched IPIs.

Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Con Kolivas <kernel@kolivas.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-11-09 07:56:33 -08:00
..
acpi [PATCH] x86-64: more gratitious linux/irq.h includes 2005-09-12 10:50:58 -07:00
cpufreq Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
aperture.c [PATCH] x86-64: Don't allocate aperture when swiotlb is enabled and no AGP found 2005-09-12 10:49:56 -07:00
apic.c [PATCH] x86-64: more gratitious linux/irq.h includes 2005-09-12 10:50:58 -07:00
asm-offsets.c [PATCH] x86-64: Some cleanup and optimization to the processor data area. 2005-09-12 10:49:58 -07:00
crash.c [PATCH] x86-64: more gratitious linux/irq.h includes 2005-09-12 10:50:58 -07:00
e820.c [PATCH] x86_64: e820.c needs module.h 2005-09-17 11:50:00 -07:00
early_printk.c [PATCH] x86-64: Add simnow console 2005-09-12 10:49:58 -07:00
entry.S [PATCH] x86-64: Fix CFI information 2005-09-12 10:50:56 -07:00
genapic_cluster.c [PATCH] x86-64: Remove obsolete APIC "write around" bug workaround 2005-09-12 10:49:55 -07:00
genapic_flat.c [PATCH] x86-64: Delivery mode should be APIC_DM_FIXED when using physical mode. 2005-09-12 10:50:56 -07:00
genapic.c [PATCH] x86-64: Use physflat on Intel for < 8 CPUs with CPU hotplug 2005-09-12 10:50:56 -07:00
head64.c [PATCH] reorganize x86-64 NUMA and DISCONTIGMEM config options 2005-06-23 09:45:06 -07:00
head.S [PATCH] x86_64: Drop global bit from early low mappings 2005-10-04 15:56:52 -07:00
i387.c [PATCH] x86_64: Change init sections for CPU hotplug support 2005-06-25 16:24:30 -07:00
i8259.c [PATCH] introduce setup_timer() helper 2005-10-30 17:37:17 -08:00
init_task.c [PATCH] x86-64: Set the stack pointer correctly in init_thread and init_tss 2005-09-12 10:50:57 -07:00
io_apic.c Partially revert "Fix time going twice as fast problem on ATI Xpress chipsets" 2005-09-14 15:56:27 -07:00
ioport.c Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
irq.c [PATCH] Remove unnecessary BUG_ON in irq.c 2005-09-12 10:50:55 -07:00
kprobes.c [PATCH] Kprobes: preempt_disable/enable() simplification 2005-11-07 07:53:46 -08:00
ldt.c Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
machine_kexec.c [PATCH] x86_64 machine_kexec: Use standard pagetable helpers 2005-07-29 13:12:49 -07:00
Makefile [PATCH] swiotlb: move from arch/ia64/lib/ to lib/ 2005-09-29 14:42:42 -07:00
mce_intel.c [PATCH] x86_64: Change init sections for CPU hotplug support 2005-06-25 16:24:30 -07:00
mce.c [PATCH] x86_64: Fix mce_log 2005-09-29 15:41:42 -07:00
module.c [PATCH] uml: kludgy compilation fixes for x86-64 subarch modules support 2005-05-05 16:36:33 -07:00
mpparse.c [PATCH] x86-64: more gratitious linux/irq.h includes 2005-09-12 10:50:58 -07:00
nmi.c [PATCH] x86_64: NMI watchdog frequency calculation adjustments 2005-09-13 08:22:33 -07:00
pci-dma.c Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
pci-gart.c [PATCH] gfp_t: dma-mapping (amd64) 2005-10-28 08:16:48 -07:00
pci-nommu.c [PATCH] gfp_t: dma-mapping (amd64) 2005-10-28 08:16:48 -07:00
pmtimer.c [PATCH] x86_64: Add pmtimer support 2005-05-17 07:59:15 -07:00
process.c [PATCH] sched: resched and cpu_idle rework 2005-11-09 07:56:33 -08:00
ptrace.c [PATCH] consolidate sys_ptrace() 2005-11-07 07:53:42 -08:00
reboot.c [PATCH] x86_64 sync machine_power_off with i386 2005-07-26 14:35:45 -07:00
relocate_kernel.S [PATCH] kexec: x86_64 kexec implementation 2005-06-25 16:24:50 -07:00
setup64.c [PATCH] x86_64: Allocate cpu local data for all possible CPUs 2005-10-10 16:33:25 -07:00
setup.c [PATCH] x86: vmx cpu feature detection 2005-10-30 17:37:13 -08:00
signal.c [PATCH] convert signal handling of NODEFER to act like other Unix boxes. 2005-08-29 10:03:11 -07:00
smp.c [PATCH] x86-64: more gratitious linux/irq.h includes 2005-09-12 10:50:58 -07:00
smpboot.c [PATCH] sched: disable preempt in idle tasks 2005-11-09 07:56:33 -08:00
suspend_asm.S [PATCH] x86_64: Set up safe page tables during resume 2005-10-10 08:36:46 -07:00
suspend.c [PATCH] swsusp: rework memory freeing on resume 2005-10-30 17:37:14 -08:00
sys_x86_64.c [PATCH] Avoiding mmap fragmentation 2005-06-21 18:46:16 -07:00
syscall.c Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
time.c [PATCH] hpet-RTC: cache the comparator register 2005-10-30 17:37:30 -08:00
trampoline.S Linux-2.6.12-rc2 2005-04-16 15:20:36 -07:00
traps.c [PATCH] x86-64: reduce x86-64 bug frame by 4 bytes 2005-09-12 10:50:58 -07:00
vmlinux.lds.S [PATCH] x86_64 linker script cleanups for debug sections 2005-09-10 12:00:17 -07:00
vsyscall.c [PATCH] x86-64: Prevent gcc 4 from optimizing away vsyscalls 2005-09-12 10:50:56 -07:00
x8664_ksyms.c [LIB]: Consolidate _atomic_dec_and_lock() 2005-09-14 21:47:01 -07:00