Skip to content

cloud-kernel
cloud-kernel

openanolis / cloud-kernel
1 年多 前同步成功

通知 160
Star 36
Fork 7
cloud-kernel
cloud-kernel
切换分支/标签
  1. 12 4月, 2013 1 次提交
  3. 24 3月, 2013 1 次提交
  5. 12 3月, 2013 15 次提交
  7. 04 3月, 2013 5 次提交
  9. 03 3月, 2013 1 次提交
    • Y
      x86, ACPI, mm: Revert movablemem_map support · 20e6926d
      由 Yinghai Lu 提交于 
      Tim found:

  WARNING: at arch/x86/kernel/smpboot.c:324 topology_sane.isra.2+0x6f/0x80()
  Hardware name: S2600CP
  sched: CPU #1's llc-sibling CPU #0 is not on the same node! [node: 1 != 0]. Ignoring dependency.
  smpboot: Booting Node   1, Processors  #1
  Modules linked in:
  Pid: 0, comm: swapper/1 Not tainted 3.9.0-0-generic #1
  Call Trace:
    set_cpu_sibling_map+0x279/0x449
    start_secondary+0x11d/0x1e5

Don Morris reproduced on a HP z620 workstation, and bisected it to
commit e8d19552 ("acpi, memory-hotplug: parse SRAT before memblock
is ready")

It turns out movable_map has some problems, and it breaks several things

1. numa_init is called several times, NOT just for srat. so those
	nodes_clear(numa_nodes_parsed)
	memset(&numa_meminfo, 0, sizeof(numa_meminfo))
   can not be just removed.  Need to consider sequence is: numaq, srat, amd, dummy.
   and make fall back path working.

2. simply split acpi_numa_init to early_parse_srat.
   a. that early_parse_srat is NOT called for ia64, so you break ia64.
   b.  for (i = 0; i < MAX_LOCAL_APIC; i++)
	     set_apicid_to_node(i, NUMA_NO_NODE)
     still left in numa_init. So it will just clear result from early_parse_srat.
     it should be moved before that....
   c.  it breaks ACPI_TABLE_OVERIDE...as the acpi table scan is moved
       early before override from INITRD is settled.

3. that patch TITLE is total misleading, there is NO x86 in the title,
   but it changes critical x86 code. It caused x86 guys did not
   pay attention to find the problem early. Those patches really should
   be routed via tip/x86/mm.

4. after that commit, following range can not use movable ram:
  a. real_mode code.... well..funny, legacy Node0 [0,1M) could be hot-removed?
  b. initrd... it will be freed after booting, so it could be on movable...
  c. crashkernel for kdump...: looks like we can not put kdump kernel above 4G
	anymore.
  d. init_mem_mapping: can not put page table high anymore.
  e. initmem_init: vmemmap can not be high local node anymore. That is
     not good.

If node is hotplugable, the mem related range like page table and
vmemmap could be on the that node without problem and should be on that
node.

We have workaround patch that could fix some problems, but some can not
be fixed.

So just remove that offending commit and related ones including:

 f7210e6c ("mm/memblock.c: use CONFIG_HAVE_MEMBLOCK_NODE_MAP to
    protect movablecore_map in memblock_overlaps_region().")

 01a178a9 ("acpi, memory-hotplug: support getting hotplug info from
    SRAT")

 27168d38 ("acpi, memory-hotplug: extend movablemem_map ranges to
    the end of node")

 e8d19552 ("acpi, memory-hotplug: parse SRAT before memblock is
    ready")

 fb06bc8e ("page_alloc: bootmem limit with movablecore_map")

 42f47e27 ("page_alloc: make movablemem_map have higher priority")

 6981ec31 ("page_alloc: introduce zone_movable_limit[] to keep
    movable limit for nodes")

 34b71f1e ("page_alloc: add movable_memmap kernel parameter")

 4d59a751 ("x86: get pg_data_t's memory from other node")

Later we should have patches that will make sure kernel put page table
and vmemmap on local node ram instead of push them down to node0.  Also
need to find way to put other kernel used ram to local node ram.
Reported-by: NTim Gardner <tim.gardner@canonical.com>
Reported-by: NDon Morris <don.morris@hp.com>
Bisected-by: NDon Morris <don.morris@hp.com>
Tested-by: NDon Morris <don.morris@hp.com>
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Thomas Renninger <trenn@suse.de>
Cc: Tejun Heo <tj@kernel.org>
Cc: Tang Chen <tangchen@cn.fujitsu.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      20e6926d
  11. 26 2月, 2013 2 次提交
  13. 24 2月, 2013 4 次提交
    • T
      acpi, memory-hotplug: parse SRAT before memblock is ready · e8d19552
      由 Tang Chen 提交于 
      On linux, the pages used by kernel could not be migrated.  As a result,
if a memory range is used by kernel, it cannot be hot-removed.  So if we
want to hot-remove memory, we should prevent kernel from using it.

The way now used to prevent this is specify a memory range by
movablemem_map boot option and set it as ZONE_MOVABLE.

But when the system is booting, memblock will allocate memory, and
reserve the memory for kernel.  And before we parse SRAT, and know the
node memory ranges, memblock is working.  And it may allocate memory in
ranges to be set as ZONE_MOVABLE.  This memory can be used by kernel,
and never be freed.

So, let's parse SRAT before memblock is called first.  And it is early
enough.

The first call of memblock_find_in_range_node() is in:

  setup_arch()
    |-->setup_real_mode()

so, this patch add a function early_parse_srat() to parse SRAT, and call
it before setup_real_mode() is called.

NOTE:

1) early_parse_srat() is called before numa_init(), and has initialized
   numa_meminfo.  So DO NOT clear numa_nodes_parsed in numa_init() and DO
   NOT zero numa_meminfo in numa_init(), otherwise we will lose memory
   numa info.

2) I don't know why using count of memory affinities parsed from SRAT
   as a return value in original acpi_numa_init().  So I add a static
   variable srat_mem_cnt to remember this count and use it as the return
   value of the new acpi_numa_init()

[mhocko@suse.cz: parse SRAT before memblock is ready fix]
Signed-off-by: NTang Chen <tangchen@cn.fujitsu.com>
Reviewed-by: NWen Congyang <wency@cn.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Jiang Liu <jiang.liu@huawei.com>
Cc: Jianguo Wu <wujianguo@huawei.com>
Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Wu Jianguo <wujianguo@huawei.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Len Brown <lenb@kernel.org>
Cc: "Brown, Len" <len.brown@intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      e8d19552
    • W
      cpu-hotplug, memory-hotplug: try offlining the node when hotremoving a cpu · 76bba142
      由 Wen Congyang 提交于 
      The node will be offlined when all memory/cpu on the node is hotremoved.
So we should try offline the node when hotremoving a cpu on the node.
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NTang Chen <tangchen@cn.fujitsu.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Jiang Liu <liuj97@gmail.com>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Len Brown <lenb@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      76bba142
    • T
      memory-hotplug: remove sysfs file of node · 60a5a19e
      由 Tang Chen 提交于 
      Introduce a new function try_offline_node() to remove sysfs file of node
when all memory sections of this node are removed.  If some memory
sections of this node are not removed, this function does nothing.
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>
Signed-off-by: NTang Chen <tangchen@cn.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Jiang Liu <jiang.liu@huawei.com>
Cc: Jianguo Wu <wujianguo@huawei.com>
Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: Wu Jianguo <wujianguo@huawei.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      60a5a19e
    • R
      ACPI / PM: Take unusual configurations of power resources into account · b5d667eb
      由 Rafael J. Wysocki 提交于 
      Commit d2e5f0c1 (ACPI / PCI: Rework the setup and cleanup of device
wakeup) moved the initial disabling of system wakeup for PCI devices
into a place where it can actually work and that exposed a hidden old
issue with crap^Wunusual system designs where the same power
resources are used for both wakeup power and device power control at
run time.

Namely, say there is one power resource such that the ACPI power
state D0 of a PCI device depends on that power resource (i.e. the
device is in D0 when that power resource is "on") and it is used
as a wakeup power resource for the same device.  Then, calling
acpi_pci_sleep_wake(pci_dev, false) for the device in question will
cause the reference counter of that power resource to drop to 0,
which in turn will cause it to be turned off.  As a result, the
device will go into D3cold at that point, although it should have
stayed in D0.

As it turns out, that happens to USB controllers on some laptops
and USB becomes unusable on those machines as a result, which is
a major regression from v3.8.

To fix this problem, (1) increment the reference counters of wakup
power resources during their initialization if they are "on"
initially, (2) prevent acpi_disable_wakeup_device_power() from
decrementing the reference counters of wakeup power resources that
were not enabled for wakeup power previously, and (3) prevent
acpi_enable_wakeup_device_power() from incrementing the reference
counters of wakeup power resources that already are enabled for
wakeup power.

In addition to that, if it is impossible to determine the initial
states of wakeup power resources, avoid enabling wakeup for devices
whose wakeup power depends on those power resources.
Reported-by: NDave Jones <davej@redhat.com>
Reported-by: NFabio Baltieri <fabio.baltieri@linaro.org>
Tested-by: NFabio Baltieri <fabio.baltieri@linaro.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
      b5d667eb
  15. 23 2月, 2013 1 次提交
  17. 22 2月, 2013 1 次提交
  19. 17 2月, 2013 1 次提交
  21. 16 2月, 2013 1 次提交
  23. 13 2月, 2013 7 次提交
    • R
      ACPI / hotplug: Fix concurrency issues and memory leaks · 3757b948
      由 Rafael J. Wysocki 提交于 
      This changeset is aimed at fixing a few different but related
problems in the ACPI hotplug infrastructure.

First of all, since notify handlers may be run in parallel with
acpi_bus_scan(), acpi_bus_trim() and acpi_bus_hot_remove_device()
and some of them are installed for ACPI handles that have no struct
acpi_device objects attached (i.e. before those objects are created),
those notify handlers have to take acpi_scan_lock to prevent races
from taking place (e.g. a struct acpi_device is found to be present
for the given ACPI handle, but right after that it is removed by
acpi_bus_trim() running in parallel to the given notify handler).
Moreover, since some of them call acpi_bus_scan() and
acpi_bus_trim(), this leads to the conclusion that acpi_scan_lock
should be acquired by the callers of these two funtions rather by
these functions themselves.

For these reasons, make all notify handlers that can handle device
addition and eject events take acpi_scan_lock and remove the
acpi_scan_lock locking from acpi_bus_scan() and acpi_bus_trim().
Accordingly, update all of their users to make sure that they
are always called under acpi_scan_lock.

Furthermore, since eject operations are carried out asynchronously
with respect to the notify events that trigger them, with the help
of acpi_bus_hot_remove_device(), even if notify handlers take the
ACPI scan lock, it still is possible that, for example,
acpi_bus_trim() will run between acpi_bus_hot_remove_device() and
the notify handler that scheduled its execution and that
acpi_bus_trim() will remove the device node passed to
acpi_bus_hot_remove_device() for ejection.  In that case, the struct
acpi_device object obtained by acpi_bus_hot_remove_device() will be
invalid and not-so-funny things will ensue.  To protect agaist that,
make the users of acpi_bus_hot_remove_device() run get_device() on
ACPI device node objects that are about to be passed to it and make
acpi_bus_hot_remove_device() run put_device() on them and check if
their ACPI handles are not NULL (make acpi_device_unregister() clear
the device nodes' ACPI handles for that check to work).

Finally, observe that acpi_os_hotplug_execute() actually can fail,
in which case its caller ought to free memory allocated for the
context object to prevent leaks from happening.  It also needs to
run put_device() on the device node that it ran get_device() on
previously in that case.  Modify the code accordingly.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
      3757b948
    • R
      ACPI / scan: Full transition to D3cold in acpi_device_unregister() · 0aa120a0
      由 Rafael J. Wysocki 提交于 
      In order to drop reference counts of all power resources used by an
ACPI device node being removed, acpi_device_unregister() calls
acpi_power_transition(device, ACPI_STATE_D3_COLD), which effectively
transitions the device node into D3cold if it uses any power
resources.  However, for some device nodes it may not be appropriate
to remove power from them entirely before putting them into D3hot
before.  On the other hand, executing _PS3 for devices that don't
use power resources before removing them shouldn't really hurt.
In fact, that is done by acpi_bus_hot_remove_device(), but this is
not the right place to do it, because the bus trimming may have
caused power to be removed from the device node in question already
before.

For these reasons, make acpi_device_unregister() carry out full
power-off transition for all device nodes supporting that and remove
the direct evaluation of _PS3 from acpi_bus_hot_remove_device().
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
      0aa120a0
    • R
      ACPI / scan: Make acpi_bus_hot_remove_device() acquire the scan lock · f058cdf4
      由 Rafael J. Wysocki 提交于 
      The ACPI scan lock has been introduced to prevent acpi_bus_scan()
and acpi_bus_trim() from running in parallel with each other for
overlapping ACPI namespace scopes.  However, it is not sufficient
to do that, because if acpi_bus_scan() is run (for an overlapping
namespace scope) right after the acpi_bus_trim() in
acpi_bus_hot_remove_device(), the subsequent eject will remove
devices without removing the corresponding struct acpi_device
objects (and possibly companion "physical" device objects).
Therefore acpi_bus_hot_remove_device() has to acquire the scan
lock before carrying out the bus trimming and hold it through
the evaluation of _EJ0, so make that happen.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NToshi Kani <toshi.kani@hp.com>
      f058cdf4
    • R
      ACPI: Drop the container.h header file · 87d4a4da
      由 Rafael J. Wysocki 提交于 
      The include/acpi/container.h only contains a definition of a
structure that is not used any more, so drop it entirely.

Similar change was proposed earlier by Toshi Kani.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NToshi Kani <toshi.kani@hp.com>
      87d4a4da
    • R
      ACPI / scan: Make container driver use struct acpi_scan_handler · 737f1a9f
      由 Rafael J. Wysocki 提交于 
      Make the ACPI container driver use struct acpi_scan_handler for
representing the object used to initialize ACPI containers and remove
the ACPI driver structure used previously and the data structures
created by it, since in fact they were not used for any purpose.

This simplifies the code and reduces the kernel's memory footprint by
avoiding the registration of a struct device_driver object with the
driver core and creation of its sysfs directory which is unnecessary.

In addition to that, make the namespace walk callback used for
installing the notify handlers for ACPI containers more
straightforward.

This change includes fixes from Toshi Kani.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NYinghai Lu <yinghai@kernel.org>
Acked-by: NYasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Tested-by: NYasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Reviewed-by: NToshi Kani <toshi.kani@hp.com>
Tested-by: NToshi Kani <toshi.kani@hp.com>
      737f1a9f
    • R
      ACPI / scan: Remove useless #ifndef from acpi_eject_store() · 38475b3b
      由 Rafael J. Wysocki 提交于 
      Since the FORCE_EJECT symbol is never defined, the
#ifndef FORCE_EJECT in acpi_eject_store() is always true, so drop it.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NYasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Tested-by: NYasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Reviewed-by: NToshi Kani <toshi.kani@hp.com>
Tested-by: NToshi Kani <toshi.kani@hp.com>
      38475b3b
    • T
      ACPI: Unbind ACPI drv when probe failed · 5f27ee8e
      由 Toshi Kani 提交于 
      When acpi_device_install_notify_handler() failed in acpi_device_probe(),
it calls acpi_drv->ops.remove() and fails the probe.  However, the ACPI
driver is left bound to the acpi_device.  Fix it by clearing the driver
and driver_data fields.
Signed-off-by: NToshi Kani <toshi.kani@hp.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
      5f27ee8e