linux/arch
Davidlohr Bueso 615d6e8756 mm: per-thread vma caching
This patch is a continuation of efforts trying to optimize find_vma(),
avoiding potentially expensive rbtree walks to locate a vma upon faults.
The original approach (https://lkml.org/lkml/2013/11/1/410), where the
largest vma was also cached, ended up being too specific and random,
thus further comparison with other approaches were needed.  There are
two things to consider when dealing with this, the cache hit rate and
the latency of find_vma().  Improving the hit-rate does not necessarily
translate in finding the vma any faster, as the overhead of any fancy
caching schemes can be too high to consider.

We currently cache the last used vma for the whole address space, which
provides a nice optimization, reducing the total cycles in find_vma() by
up to 250%, for workloads with good locality.  On the other hand, this
simple scheme is pretty much useless for workloads with poor locality.
Analyzing ebizzy runs shows that, no matter how many threads are
running, the mmap_cache hit rate is less than 2%, and in many situations
below 1%.

The proposed approach is to replace this scheme with a small per-thread
cache, maximizing hit rates at a very low maintenance cost.
Invalidations are performed by simply bumping up a 32-bit sequence
number.  The only expensive operation is in the rare case of a seq
number overflow, where all caches that share the same address space are
flushed.  Upon a miss, the proposed replacement policy is based on the
page number that contains the virtual address in question.  Concretely,
the following results are seen on an 80 core, 8 socket x86-64 box:

1) System bootup: Most programs are single threaded, so the per-thread
   scheme does improve ~50% hit rate by just adding a few more slots to
   the cache.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 50.61%   | 19.90            |
| patched        | 73.45%   | 13.58            |
+----------------+----------+------------------+

2) Kernel build: This one is already pretty good with the current
   approach as we're dealing with good locality.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 75.28%   | 11.03            |
| patched        | 88.09%   | 9.31             |
+----------------+----------+------------------+

3) Oracle 11g Data Mining (4k pages): Similar to the kernel build workload.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 70.66%   | 17.14            |
| patched        | 91.15%   | 12.57            |
+----------------+----------+------------------+

4) Ebizzy: There's a fair amount of variation from run to run, but this
   approach always shows nearly perfect hit rates, while baseline is just
   about non-existent.  The amounts of cycles can fluctuate between
   anywhere from ~60 to ~116 for the baseline scheme, but this approach
   reduces it considerably.  For instance, with 80 threads:

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 1.06%    | 91.54            |
| patched        | 99.97%   | 14.18            |
+----------------+----------+------------------+

[akpm@linux-foundation.org: fix nommu build, per Davidlohr]
[akpm@linux-foundation.org: document vmacache_valid() logic]
[akpm@linux-foundation.org: attempt to untangle header files]
[akpm@linux-foundation.org: add vmacache_find() BUG_ON]
[hughd@google.com: add vmacache_valid_mm() (from Oleg)]
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: adjust and enhance comments]
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Michel Lespinasse <walken@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Tested-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-07 16:35:53 -07:00
..
alpha PCI changes for the v3.15 merge window: 2014-04-01 15:14:04 -07:00
arc Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial 2014-04-02 16:23:38 -07:00
arm IOMMU Upates for Linux v3.15 2014-04-05 18:46:26 -07:00
arm64 Merge tag 'kvm-3.15-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm 2014-04-02 14:50:10 -07:00
avr32 Char/Misc driver patches for 3.15-rc1 2014-04-01 16:13:21 -07:00
blackfin Most of the changes were largely clean ups, and some documentation. 2014-04-03 10:26:31 -07:00
c6x Merge branch 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2014-03-31 10:59:39 -07:00
cris Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial 2014-04-02 16:23:38 -07:00
frv PCI changes for the v3.15 merge window: 2014-04-01 15:14:04 -07:00
hexagon Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rkuo/linux-hexagon-kernel 2014-04-05 13:19:32 -07:00
ia64 Most of the changes were largely clean ups, and some documentation. 2014-04-03 10:26:31 -07:00
m32r arch: Remove stub cputime.h headers 2014-03-13 16:09:30 +01:00
m68k Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gerg/m68knommu 2014-04-05 13:18:52 -07:00
metag Most of the changes were largely clean ups, and some documentation. 2014-04-03 10:26:31 -07:00
microblaze Most of the changes were largely clean ups, and some documentation. 2014-04-03 10:26:31 -07:00
mips Most of the changes were largely clean ups, and some documentation. 2014-04-03 10:26:31 -07:00
mn10300 Merge branch 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2014-04-01 11:22:57 -07:00
openrisc locking/mcs: Allow architecture specific asm files to be used for contended case 2014-02-09 21:18:52 +01:00
parisc Merge branch 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2014-04-01 11:22:57 -07:00
powerpc A number of cleanups plus support for the RDSEED instruction, which 2014-04-04 21:25:28 -07:00
s390 mm: revert "thp: make MADV_HUGEPAGE check for mm->def_flags" 2014-04-07 16:35:51 -07:00
score score: remove unused CPU_SCORE7 Kconfig parameter 2014-04-03 16:20:52 -07:00
sh Merge branch 'akpm' (incoming from Andrew) 2014-04-03 16:22:16 -07:00
sparc Merge branch 'for-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup 2014-04-03 13:05:42 -07:00
tile Merge git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile 2014-04-06 08:11:57 -07:00
um Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2014-04-02 20:53:45 -07:00
unicore32 mm: per-thread vma caching 2014-04-07 16:35:53 -07:00
x86 A number of cleanups plus support for the RDSEED instruction, which 2014-04-04 21:25:28 -07:00
xtensa Merge branch 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2014-04-01 11:22:57 -07:00
.gitignore
Kconfig uprobes: Kconfig dependency fix 2014-03-18 16:39:33 -04:00