Skip to content

L
linux
Switch branch/tag
  1. 14 Aug, 2009 21 commits
    • Tejun Heo's avatar
      x86,percpu: use embedding for 64bit NUMA and page for 32bit NUMA · 4518e6a0
      Tejun Heo authored 
      Embedding percpu first chunk allocator can now handle very sparse unit
mapping.  Use embedding allocator instead of lpage for 64bit NUMA.
This removes extra TLB pressure and the need to do complex and fragile
dancing when changing page attributes.

For 32bit, using very sparse unit mapping isn't a good idea because
the vmalloc space is very constrained.  32bit NUMA machines aren't
exactly the focus of optimization and it isn't very clear whether
lpage performs better than page.  Use page first chunk allocator for
32bit NUMAs.

As this leaves setup_pcpu_*() functions pretty much empty, fold them
into setup_per_cpu_areas().
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Andi Kleen <andi@firstfloor.org>
      4518e6a0
    • Tejun Heo's avatar
      percpu: update embedding first chunk allocator to handle sparse units · c8826dd5
      Tejun Heo authored 
      Now that percpu core can handle very sparse units, given that vmalloc
space is large enough, embedding first chunk allocator can use any
memory to build the first chunk.  This patch teaches
pcpu_embed_first_chunk() about distances between cpus and to use
alloc/free callbacks to allocate node specific areas for each group
and use them for the first chunk.

This brings the benefits of embedding allocator to NUMA configurations
- no extra TLB pressure with the flexibility of unified dynamic
allocator and no need to restructure arch code to build memory layout
suitable for percpu.  With units put into atom_size aligned groups
according to cpu distances, using large page for dynamic chunks is
also easily possible with falling back to reuglar pages if large
allocation fails.

Embedding allocator users are converted to specify NULL
cpu_distance_fn, so this patch doesn't cause any visible behavior
difference.  Following patches will convert them.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      c8826dd5
    • Tejun Heo's avatar
      percpu: use group information to allocate vmap areas sparsely · 6563297c
      Tejun Heo authored 
      ai->groups[] contains which units need to be put consecutively and at
what offset from the chunk base address.  Compile this information
into pcpu_group_offsets[] and pcpu_group_sizes[] in
pcpu_setup_first_chunk() and use them to allocate sparse vm areas
using pcpu_get_vm_areas().

This will be used to allow directly using sparse NUMA memories as
percpu areas.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: Nick Piggin <npiggin@suse.de>
      6563297c
    • Tejun Heo's avatar
      vmalloc: implement pcpu_get_vm_areas() · ca23e405
      Tejun Heo authored 
      To directly use spread NUMA memories for percpu units, percpu
allocator will be updated to allow sparsely mapping units in a chunk.
As the distances between units can be very large, this makes
allocating single vmap area for each chunk undesirable.  This patch
implements pcpu_get_vm_areas() and pcpu_free_vm_areas() which
allocates and frees sparse congruent vmap areas.

pcpu_get_vm_areas() take @offsets and @sizes array which define
distances and sizes of vmap areas.  It scans down from the top of
vmalloc area looking for the top-most address which can accomodate all
the areas.  The top-down scan is to avoid interacting with regular
vmallocs which can push up these congruent areas up little by little
ending up wasting address space and page table.

To speed up top-down scan, the highest possible address hint is
maintained.  Although the scan is linear from the hint, given the
usual large holes between memory addresses between NUMA nodes, the
scanning is highly likely to finish after finding the first hole for
the last unit which is scanned first.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: Nick Piggin <npiggin@suse.de>
      ca23e405
    • Tejun Heo's avatar
      vmalloc: separate out insert_vmalloc_vm() · cf88c790
      Tejun Heo authored 
      Separate out insert_vmalloc_vm() from __get_vm_area_node().
insert_vmalloc_vm() initializes vm_struct from vmap_area and inserts
it into vmlist.  insert_vmalloc_vm() only initializes fields which can
be determined from @vm, @flags and @caller The rest should be
initialized by the caller.  For __get_vm_area_node(), all other fields
just need to be cleared and this is done by using kzalloc instead of
kmalloc.

This will be used to implement pcpu_get_vm_areas().
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: Nick Piggin <npiggin@suse.de>
      cf88c790
    • Tejun Heo's avatar
      percpu: add chunk->base_addr · bba174f5
      Tejun Heo authored 
      The only thing percpu allocator wants to know about a vmalloc area is
the base address.  Instead of requiring chunk->vm, add
chunk->base_addr which contains the necessary value.  This simplifies
the code a bit and makes the dummy first_vm unnecessary.  This change
will ease allowing a chunk to be mapped by multiple vms.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      bba174f5
    • Tejun Heo's avatar
      percpu: add pcpu_unit_offsets[] · fb435d52
      Tejun Heo authored 
      Currently units are mapped sequentially into address space.  This
patch adds pcpu_unit_offsets[] which allows units to be mapped to
arbitrary offsets from the chunk base address.  This is necessary to
allow sparse embedding which might would need to allocate address
ranges and memory areas which aren't aligned to unit size but
allocation atom size (page or large page size).  This also simplifies
things a bit by removing the need to calculate offset from unit
number.

With this change, there's no need for the arch code to know
pcpu_unit_size.  Update pcpu_setup_first_chunk() and first chunk
allocators to return regular 0 or -errno return code instead of unit
size or -errno.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: David S. Miller <davem@davemloft.net>
      fb435d52
    • Tejun Heo's avatar
      percpu: introduce pcpu_alloc_info and pcpu_group_info · fd1e8a1f
      Tejun Heo authored 
      Till now, non-linear cpu->unit map was expressed using an integer
array which maps each cpu to a unit and used only by lpage allocator.
Although how many units have been placed in a single contiguos area
(group) is known while building unit_map, the information is lost when
the result is recorded into the unit_map array.  For lpage allocator,
as all allocations are done by lpages and whether two adjacent lpages
are in the same group or not is irrelevant, this didn't cause any
problem.  Non-linear cpu->unit mapping will be used for sparse
embedding and this grouping information is necessary for that.

This patch introduces pcpu_alloc_info which contains all the
information necessary for initializing percpu allocator.
pcpu_alloc_info contains array of pcpu_group_info which describes how
units are grouped and mapped to cpus.  pcpu_group_info also has
base_offset field to specify its offset from the chunk's base address.
pcpu_build_alloc_info() initializes this field as if all groups are
allocated back-to-back as is currently done but this will be used to
sparsely place groups.

pcpu_alloc_info is a rather complex data structure which contains a
flexible array which in turn points to nested cpu_map arrays.

* pcpu_alloc_alloc_info() and pcpu_free_alloc_info() are provided to
  help dealing with pcpu_alloc_info.

* pcpu_lpage_build_unit_map() is updated to build pcpu_alloc_info,
  generalized and renamed to pcpu_build_alloc_info().
  @cpu_distance_fn may be NULL indicating that all cpus are of
  LOCAL_DISTANCE.

* pcpul_lpage_dump_cfg() is updated to process pcpu_alloc_info,
  generalized and renamed to pcpu_dump_alloc_info().  It now also
  prints which group each alloc unit belongs to.

* pcpu_setup_first_chunk() now takes pcpu_alloc_info instead of the
  separate parameters.  All first chunk allocators are updated to use
  pcpu_build_alloc_info() to build alloc_info and call
  pcpu_setup_first_chunk() with it.  This has the side effect of
  packing units for sparse possible cpus.  ie. if cpus 0, 2 and 4 are
  possible, they'll be assigned unit 0, 1 and 2 instead of 0, 2 and 4.

* x86 setup_pcpu_lpage() is updated to deal with alloc_info.

* sparc64 setup_per_cpu_areas() is updated to build alloc_info.

Although the changes made by this patch are pretty pervasive, it
doesn't cause any behavior difference other than packing of sparse
cpus.  It mostly changes how information is passed among
initialization functions and makes room for more flexibility.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: David Miller <davem@davemloft.net>
      fd1e8a1f
    • Tejun Heo's avatar
      percpu: move pcpu_lpage_build_unit_map() and pcpul_lpage_dump_cfg() upward · 033e48fb
      Tejun Heo authored 
      Unit map handling will be generalized and extended and used for
embedding sparse first chunk and other purposes.  Relocate two
unit_map related functions upward in preparation.  This patch just
moves the code without any actual change.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      033e48fb
    • Tejun Heo's avatar
      percpu: add @align to pcpu_fc_alloc_fn_t · 3cbc8565
      Tejun Heo authored 
      pcpu_fc_alloc_fn_t is about to see more interesting usage, add @align
parameter.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      3cbc8565
    • Tejun Heo's avatar
      percpu: make @dyn_size mandatory for pcpu_setup_first_chunk() · 1d9d3257
      Tejun Heo authored 
      Now that all actual first chunk allocation and copying happen in the
first chunk allocators and helpers, there's no reason for
pcpu_setup_first_chunk() to try to determine @dyn_size automatically.
The only left user is page first chunk allocator.  Make it determine
dyn_size like other allocators and make @dyn_size mandatory for
pcpu_setup_first_chunk().
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      1d9d3257
    • Tejun Heo's avatar
      percpu: drop @static_size from first chunk allocators · 9a773769
      Tejun Heo authored 
      First chunk allocators assume percpu areas have been linked using one
of PERCPU_*() macros and depend on __per_cpu_load symbol defined by
those macros, so there isn't much point in passing in static area size
explicitly when it can be easily calculated from __per_cpu_start and
__per_cpu_end.  Drop @static_size from all percpu first chunk
allocators and helpers.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      9a773769
    • Tejun Heo's avatar
      percpu: generalize first chunk allocator selection · f58dc01b
      Tejun Heo authored 
      Now that all first chunk allocators are in mm/percpu.c, it makes sense
to make generalize percpu_alloc kernel parameter.  Define PCPU_FC_*
and set pcpu_chosen_fc using early_param() in mm/percpu.c.  Arch code
can use the set value to determine which first chunk allocator to use.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      f58dc01b
    • Tejun Heo's avatar
      percpu: build first chunk allocators selectively · 08fc4580
      Tejun Heo authored 
      There's no need to build unused first chunk allocators in.  Define
CONFIG_NEED_PER_CPU_*_FIRST_CHUNK and let archs enable them
selectively.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      08fc4580
    • Tejun Heo's avatar
      percpu: rename 4k first chunk allocator to page · 00ae4064
      Tejun Heo authored 
      Page size isn't always 4k depending on arch and configuration.  Rename
4k first chunk allocator to page.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Cc: David Howells <dhowells@redhat.com>
      00ae4064
    • Tejun Heo's avatar
      percpu: improve boot messages · 004018e2
      Tejun Heo authored 
      Improve percpu boot messages such that they're uniform and contain
more information.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Reviewed-by: default avatarChristoph Lameter <cl@linux-foundation.org>
      004018e2
    • Tejun Heo's avatar
      percpu: fix pcpu_reclaim() locking · 971f3918
      Tejun Heo authored 
      pcpu_reclaim() calls pcpu_depopulate_chunk() which makes use of pages
array and bitmap returned by pcpu_get_pages_and_bitmap() and thus
should be called under pcpu_alloc_mutex.  pcpu_reclaim() released the
mutex before calling depopulate leading to double free and other
strange problems caused by the unexpected concurrent usages of pages
array and bitmap.  Fix it.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Reviewed-by: default avatarChristoph Lameter <cl@linux-foundation.org>
      971f3918
    • Tejun Heo's avatar
      Merge branch 'percpu-for-linus' into percpu-for-next · 384be2b1
      Tejun Heo authored 
      Conflicts:
	arch/sparc/kernel/smp_64.c
	arch/x86/kernel/cpu/perf_counter.c
	arch/x86/kernel/setup_percpu.c
	drivers/cpufreq/cpufreq_ondemand.c
	mm/percpu.c

Conflicts in core and arch percpu codes are mostly from commit
ed78e1e078dd44249f88b1dd8c76dafb39567161 which substituted many
num_possible_cpus() with nr_cpu_ids.  As for-next branch has moved all
the first chunk allocators into mm/percpu.c, the changes are moved
from arch code to mm/percpu.c.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      384be2b1
    • Amerigo Wang's avatar
      percpu: use the right flag for get_vm_area() · 142d44b0
      Amerigo Wang authored 
      get_vm_area() only accepts VM_* flags, not GFP_*.

And according to the doc of get_vm_area(), here should be
VM_ALLOC.
Signed-off-by: default avatarWANG Cong <amwang@redhat.com>
Acked-by: default avatarTejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
      142d44b0
    • Tejun Heo's avatar
      percpu, sparc64: fix sparse possible cpu map handling · 74d46d6b
      Tejun Heo authored 
      percpu code has been assuming num_possible_cpus() == nr_cpu_ids which
is incorrect if cpu_possible_map contains holes.  This causes percpu
code to access beyond allocated memories and vmalloc areas.  On a
sparc64 machine with cpus 0 and 2 (u60), this triggers the following
warning or fails boot.

 WARNING: at /devel/tj/os/work/mm/vmalloc.c:106 vmap_page_range_noflush+0x1f0/0x240()
 Modules linked in:
 Call Trace:
  [00000000004b17d0] vmap_page_range_noflush+0x1f0/0x240
  [00000000004b1840] map_vm_area+0x20/0x60
  [00000000004b1950] __vmalloc_area_node+0xd0/0x160
  [0000000000593434] deflate_init+0x14/0xe0
  [0000000000583b94] __crypto_alloc_tfm+0xd4/0x1e0
  [00000000005844f0] crypto_alloc_base+0x50/0xa0
  [000000000058b898] alg_test_comp+0x18/0x80
  [000000000058dad4] alg_test+0x54/0x180
  [000000000058af00] cryptomgr_test+0x40/0x60
  [0000000000473098] kthread+0x58/0x80
  [000000000042b590] kernel_thread+0x30/0x60
  [0000000000472fd0] kthreadd+0xf0/0x160
 ---[ end trace 429b268a213317ba ]---

This patch fixes generic percpu functions and sparc64
setup_per_cpu_areas() so that they handle sparse cpu_possible_map
properly.

Please note that on x86, cpu_possible_map() doesn't contain holes and
thus num_possible_cpus() == nr_cpu_ids and this patch doesn't cause
any behavior difference.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
Acked-by: default avatarDavid S. Miller <davem@davemloft.net>
Cc: Ingo Molnar <mingo@elte.hu>
      74d46d6b
    • Tejun Heo's avatar
      init: set nr_cpu_ids before setup_per_cpu_areas() · d6647bdf
      Tejun Heo authored 
      nr_cpu_ids is dependent only on cpu_possible_map and
setup_per_cpu_areas() already depends on cpu_possible_map and will use
nr_cpu_ids.  Initialize nr_cpu_ids before setting up percpu areas.
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      d6647bdf
  3. 13 Aug, 2009 19 commits
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7