linux-toradex.git/drivers/base/cpu.c, branch v3.10.78

numa, cpu hotplug: change links of CPU and node when changing node number by onlining CPU

2013-04-29T22:54:39+00:00

When booting x86 system contains memoryless node, node numbers of CPUs
on memoryless node were changed to nearest online node number by
init_cpu_to_node() because the node is not online.

In my system, node numbers of cpu#30-44 and 75-89 were changed from 2 to
0 as follows:

  $ numactl --hardware
  available: 2 nodes (0-1)
  node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 30 31 32 33 34 35 36 37 38 39 40
  41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 75 76 77 78 79 80 81 82
  83 84 85 86 87 88 89
  node 0 size: 32394 MB
  node 0 free: 27898 MB
  node 1 cpus: 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 60 61 62 63 64 65 66
  67 68 69 70 71 72 73 74
  node 1 size: 32768 MB
  node 1 free: 30335 MB

If we hot add memory to memoryless node and offine/online all CPUs on
the node, node numbers of these CPUs are changed to correct node numbers
by srat_detect_node() because the node become online.

In this case, node numbers of cpu#30-44 and 75-89 were changed from 0 to
2 in my system as follows:

  $ numactl --hardware
  available: 3 nodes (0-2)
  node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 45 46 47 48 49 50 51 52 53 54 55
  56 57 58 59
  node 0 size: 32394 MB
  node 0 free: 27218 MB
  node 1 cpus: 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 60 61 62 63 64 65 66
  67 68 69 70 71 72 73 74
  node 1 size: 32768 MB
  node 1 free: 30014 MB
  node 2 cpus: 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 75 76 77 78 79 80 81
  82 83 84 85 86 87 88 89
  node 2 size: 16384 MB
  node 2 free: 16384 MB

But "cpu to node" and "node to cpu" links were not changed as follows:

  $ ls /sys/devices/system/cpu/cpu30/|grep node
  node0
  $ ls /sys/devices/system/node/node0/|grep cpu30
  cpu30

"numactl --hardware" shows that cpu30 belongs to node 2.  But sysfs
links does not change.

This patch changes "cpu to node" and "node to cpu" links when node
number changed by onlining CPU.

Signed-off-by: Yasuaki Ishimatsu 
Cc: KOSAKI Motohiro 
Cc: Ingo Molnar 
Cc: "H. Peter Anvin" 
Cc: "Srivatsa S. Bhat" 
Cc: Greg KH 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

sysfs: fix crash_notes_size build warning

2013-04-03T18:09:03+00:00

commit eca4549f57 "sysfs: Add crash_notes_size to export percpu
note size" adds a printk that outputs a size_t value as %lu
when it should be %zu, resulting in this warning.

drivers/base/cpu.c: In function 'show_crash_notes_size':
drivers/base/cpu.c:142:2: warning: format '%lu' expects argument of type 'long unsigned int', but argument 3 has type 'unsigned int' [-Wformat=]

Reported-by: kbuild test robot 
Signed-off-by: Arnd Bergmann 
Cc: Zhang Yanfei 
Cc: Simon Horman 
Signed-off-by: Greg Kroah-Hartman

sysfs: Add crash_notes_size to export percpu note size

2013-03-29T16:17:22+00:00

For percpu notes, we are exporting only address and not size. So
the userspace tool kexec-tools is putting an upper limit of 1024
and putting the value in p_memsz and p_filesz fields. So the patch
add the new sysfile crash_notes_size to export the exact percpu
note size and let the kexec-tools parse it intead of using 1024.

The idea came from Vivek Goyal. And a later patch will be sent to
kexec-tools to let it parse the size.

Cc: "Eric W. Biederman" 
Cc: Vivek Goyal 
Signed-off-by: Zhang Yanfei 
Acked-by: Simon Horman 
Signed-off-by: Greg Kroah-Hartman

drivers/base/cpu.c: Fix typo in comment

2013-01-16T20:34:34+00:00

[ We should make fun of people who can't speel too, but then we'd have
  no time for any real work at all  - Linus ]

Signed-off-by: Linus Torvalds

sched: Remove stale power aware scheduling remnants and dysfunctional knobs

2012-05-17T11:48:56+00:00

It's been broken forever (i.e. it's not scheduling in a power
aware fashion), as reported by Suresh and others sending
patches, and nobody cares enough to fix it properly ...
so remove it to make space free for something better.

There's various problems with the code as it stands today, first
and foremost the user interface which is bound to topology
levels and has multiple values per level. This results in a
state explosion which the administrator or distro needs to
master and almost nobody does.

Furthermore large configuration state spaces aren't good, it
means the thing doesn't just work right because it's either
under so many impossibe to meet constraints, or even if
there's an achievable state workloads have to be aware of
it precisely and can never meet it for dynamic workloads.

So pushing this kind of decision to user-space was a bad idea
even with a single knob - it's exponentially worse with knobs
on every node of the topology.

There is a proposal to replace the user interface with a single
3 state knob:

 sched_balance_policy := { performance, power, auto }

where 'auto' would be the preferred default which looks at things
like Battery/AC mode and possible cpufreq state or whatever the hw
exposes to show us power use expectations - but there's been no
progress on it in the past many months.

Aside from that, the actual implementation of the various knobs
is known to be broken. There have been sporadic attempts at
fixing things but these always stop short of reaching a mergable
state.

Therefore this wholesale removal with the hopes of spurring
people who care to come forward once again and work on a
coherent replacement.

Signed-off-by: Peter Zijlstra 
Cc: Suresh Siddha 
Cc: Arjan van de Ven 
Cc: Vincent Guittot 
Cc: Vaidyanathan Srinivasan 
Cc: Linus Torvalds 
Cc: Andrew Morton 
Link: http://lkml.kernel.org/r/1326104915.2442.53.camel@twins
Signed-off-by: Ingo Molnar

Merge 3.3-rc6 into driver-core-next

2012-03-09T20:35:53+00:00

This was done to resolve a conflict in the drivers/base/cpu.c file.

Signed-off-by: Greg Kroah-Hartman

driver-core: cpu: fix kobject warning when hotplugging a cpu

2012-02-08T23:11:17+00:00

Due to the sysdev conversion to struct device, the cpu objects get
reused when adding a cpu after offlining it, which causes a big warning
that the kobject portion is not properly initialized.

So clear out the object before we register it again, so all is quiet.

Reported-by: Konrad Rzeszutek Wilk 
Tested-by: Konrad Rzeszutek Wilk 
Signed-off-by: Greg Kroah-Hartman

driver core: cpu: remove kernel warning when removing a cpu

2012-02-02T18:43:16+00:00

With the movement of the cpu sysdev code to be real stuct devices, now
when we remove a cpu from the system, the driver core rightfully
complains that there is not a release method for this device.

For now, paper over this issue by quieting the driver core, but comment
this in detail.  This will be resolved in future kernels to be solved
properly.

Reported-by: Konrad Rzeszutek Wilk 
Tested-by: Konrad Rzeszutek Wilk 
Signed-off-by: Greg Kroah-Hartman

CPU: Introduce ARCH_HAS_CPU_AUTOPROBE and X86 parts

2012-01-27T00:49:08+00:00

This patch is based on Andi Kleen's work:
Implement autoprobing/loading of modules serving CPU
specific features (x86cpu autoloading).

And Kay Siever's work to get rid of sysdev cpu structures
and making use of struct device instead.

Before, the cpuid driver had to be loaded to get the x86cpu
autoloading feature. With this patch autoloading works through
the /sys/devices/system/cpu object

Cc: Kay Sievers 
Cc: Dave Jones 
Cc: Jens Axboe 
Cc: Herbert Xu 
Cc: Huang Ying 
Cc: Len Brown 
Acked-by: Andi Kleen 
Signed-off-by: Thomas Renninger 
Acked-by: H. Peter Anvin 
Signed-off-by: Greg Kroah-Hartman

cpu: Register a generic CPU device on architectures that currently do not

2012-01-11T23:50:11+00:00

frv, h8300, m68k, microblaze, openrisc, score, um and xtensa currently
do not register a CPU device.  Add the config option GENERIC_CPU_DEVICES
which causes a generic CPU device to be registered for each present CPU,
and make all these architectures select it.

Richard Weinberger  covered UML and suggested using
per_cpu.

Signed-off-by: Ben Hutchings 
Signed-off-by: Linus Torvalds