Git next: Second stage of cpu_alloc patches
From: Christoph Lameter
Date: Tue Nov 04 2008 - 21:57:25 EST
The second stage of the cpu_alloc patchset can be pulled from
kernel.org/pub/scm/linux/kernel/git/christoph/work.git cpu_alloc_stage2
Stage 2 includes the conversion of the page allocator, vm statistics
and slub allocator to the use of the cpu allocator and it includes the
core of the atomic vs. interrupt cpu ops.
commit c9112914d224fbead980438877212c0f003c624e
Author: Christoph Lameter <clameter@xxxxxxx>
Date: Tue Nov 6 11:33:51 2007 -0800
cpu alloc: page allocator conversion
Use the new cpu_alloc functionality to avoid per cpu arrays in struct zone.
This drastically reduces the size of struct zone for systems with a large
amounts of processors and allows placement of critical variables of struct
zone in one cacheline even on very large systems.
Another effect is that the pagesets of one processor are placed near one
another. If multiple pagesets from different zones fit into one cacheline
then additional cacheline fetches can be avoided on the hot paths when
allocating memory from multiple zones.
Surprisingly this clears up much of the painful NUMA bringup. Bootstrap
becomes simpler if we use the same scheme for UP, SMP, NUMA. #ifdefs are
reduced and we can drop the zone_pcp macro.
Hotplug handling is also simplified since cpu alloc can bring up and
shut down cpu areas for a specific cpu as a whole. So there is no need to
allocate or free individual pagesets.
Signed-off-by: Christoph Lameter <clameter@xxxxxxx>
commit 8a6784bb0da03ea87bab414e9f8de51dcd399490
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:39 2008 -0600
x86_64: Support for cpu ops
Support fast cpu ops in x86_64 by providing a series of functions that
generate the proper instructions.
Define CONFIG_HAVE_CPU_OPS so that core code
can exploit the availability of fast per cpu operations.
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
commit e872b7e4dc22f746df8435f1b4b7208b56a68fd6
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:39 2008 -0600
VM statistics: Use CPU ops
The use of CPU ops here avoids the offset calculations that we used to have
to do with per cpu operations. The result of this patch is that event counters
are coded with a single instruction the following way:
incq %gs:offset(%rip)
Without these patches this was:
mov %gs:0x8,%rdx
mov %eax,0x38(%rsp)
mov xxx(%rip),%eax
mov %eax,0x48(%rsp)
mov varoffset,%rax
incq 0x110(%rax,%rdx,1)
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
commit a81123720c0786d2847f4eaa145da61348f369ab
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:38 2008 -0600
cpu ops: Core piece for generic atomic per cpu operations
Currently the per cpu subsystem is not able to use the atomic capabilities
that are provided by many of the available processors.
This patch adds new functionality that allows the optimizing of per cpu
variable handling. In particular it provides a simple way to exploit
atomic operations in order to avoid having to disable interrupts or
performing address calculation to access per cpu data.
F.e. Using our current methods we may do
unsigned long flags;
struct stat_struct *p;
local_irq_save(flags);
/* Calculate address of per processor area */
p = CPU_PTR(stat, smp_processor_id());
p->counter++;
local_irq_restore(flags);
The segment can be replaced by a single atomic CPU operation:
CPU_INC(stat->counter);
Most processors have instructions to perform the increment using a
a single atomic instruction. Processors may have segment registers,
global registers or per cpu mappings of per cpu areas that can be used
to generate atomic instructions that combine the following in a single
operation:
1. Adding of an offset / register to a base address
2. Read modify write operation on the address calculated by
the instruction.
If 1+2 are combined in an instruction then the instruction is atomic
vs interrupts. This means that percpu atomic operations do not need
to disable interrupts to increments counters etc.
The existing methods in use in the kernel cannot utilize the power of
these atomic instructions. local_t is not really addressing the issue
since the offset calculation performed before the atomic operation. The
operation is therefor not atomic. Disabling interrupt or preemption is
required in order to use local_t.
local_t is also very specific to the x86 processor. The solution here can
utilize other methods than just those provided by the x86 instruction set.
On x86 the above CPU_INC translated into a single instruction:
inc %%gs:(&stat->counter)
This instruction is interrupt safe since it can either be completed
or not. Both adding of the offset and the read modify write are combined
in one instruction.
The determination of the correct per cpu area for the current processor
does not require access to smp_processor_id() (expensive...). The gs
register is used to provide a processor specific offset to the respective
per cpu area where the per cpu variable resides.
Note that the counter offset into the struct was added *before* the segment
selector was added. This is necessary to avoid calculations. In the past
we first determine the address of the stats structure on the respective
processor and then added the field offset. However, the offset may as
well be added earlier. The adding of the per cpu offset (here through the
gs register) must be done by the instruction used for atomic per cpu
access.
If "stat" was declared via DECLARE_PER_CPU then this patchset is capable of
convincing the linker to provide the proper base address. In that case
no calculations are necessary.
Should the stat structure be reachable via a register then the address
calculation capabilities can be leveraged to avoid calculations.
On IA64 we can get the same combination of operations in a single instruction
by using the virtual address that always maps to the local per cpu area:
fetchadd &stat->counter + (VCPU_BASE - __per_cpu_start)
The access is forced into the per cpu address reachable via the virtualized
address. IA64 allows the embedding of an offset into the instruction. So the
fetchadd can perform both the relocation of the pointer into the per cpu
area as well as the atomic read modify write cycle.
In order to be able to exploit the atomicity of these instructions we
introduce a series of new functions that take either:
1. A per cpu pointer as returned by cpu_alloc() or CPU_ALLOC().
2. A per cpu variable address as returned by per_cpu_var(<percpuvarname>).
CPU_READ()
CPU_WRITE()
CPU_INC
CPU_DEC
CPU_ADD
CPU_SUB
CPU_XCHG
CPU_CMPXCHG
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
commit bf49f4db6d6582675dbcb38189fc943e2dc8ba8d
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:38 2008 -0600
cpu alloc: Remove slub fields
Remove the fields in kmem_cache_cpu that were used to cache data from
kmem_cache when they were in different cachelines. The cacheline that holds
the per cpu array pointer now also holds these values. We can cut down the
struct kmem_cache_cpu size to almost half.
The get_freepointer() and set_freepointer() functions that used to be only
intended for the slow path now are also useful for the hot path since access
to the field does not require accessing an additional cacheline anymore. This
results in consistent use of setting the freepointer for objects throughout
SLUB.
Also we initialize all possible kmem_cache_cpu structures when a slab is
created. No need to initialize them when a processor or node comes online.
And all fields are set to zero. So just use __GFP_ZERO on cpu alloc.
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
commit e88a3b4cf66e40d81e8fcda870260a237a58a920
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:38 2008 -0600
cpu alloc: Use in slub
Using cpu alloc removes the needs for the per cpu arrays in the kmem_cache struct.
These could get quite big if we have to support system of up to thousands of cpus.
The use of cpu_alloc means that:
1. The size of kmem_cache for SMP configuration shrinks since we will only
need 1 pointer instead of NR_CPUS. The same pointer can be used by all
processors. Reduces cache footprint of the allocator.
2. We can dynamically size kmem_cache according to the actual nodes in the
system meaning less memory overhead for configurations that may potentially
support up to 1k NUMA nodes / 4k cpus.
3. We can remove the diddle widdle with allocating and releasing of
kmem_cache_cpu structures when bringing up and shutting down cpus. The cpu
alloc logic will do it all for us. Removes some portions of the cpu hotplug
functionality.
4. Fastpath performance increases.
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
commit 240314584fcc19042c6d016954f3b9971d94bf2f
Author: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
Date: Tue Nov 4 16:32:38 2008 -0600
Increase default reserve percpu area
SLUB now requires a portion of the per cpu reserve. There are on average
about 70 real slabs on a system (aliases do not count) and each needs 12 bytes
of per cpu space. Thats 840 bytes. In debug mode all slabs will be real slabs
which will make us end up with 150 -> 1800.
Things work fine without this patch but then slub will reduce the percpu reserve
for modules.
Percpu data must be available regardless if modules are in use or not. So get
rid of the #ifdef CONFIG_MODULES.
Make the size of the percpu area dependant on the size of a machine word. That
way we have larger sizes for 64 bit machines. 64 bit machines need more percpu
memory since the pointer and counters may have double the size. Plus there is
lots of memory available on 64 bit.
Signed-off-by: Christoph Lameter <cl@xxxxxxxxxxxxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/