diff options
| author | Nathan Zimmer <nzimmer@sgi.com> | 2016-01-15 00:18:36 +0100 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2016-01-15 01:00:49 +0100 |
| commit | 4a8c7bb59ac85b038c29adf6d32ff56e11fbb267 (patch) | |
| tree | 76ded28807d90dab68c27f1d33412209be004d27 /fs/hugetlbfs | |
| parent | mm: add PHYS_PFN, use it in __phys_to_pfn() (diff) | |
| download | linux-4a8c7bb59ac85b038c29adf6d32ff56e11fbb267.tar.xz linux-4a8c7bb59ac85b038c29adf6d32ff56e11fbb267.zip | |
mm/mempolicy.c: convert the shared_policy lock to a rwlock
When running the SPECint_rate gcc on some very large boxes it was
noticed that the system was spending lots of time in
mpol_shared_policy_lookup(). The gamess benchmark can also show it and
is what I mostly used to chase down the issue since the setup for that I
found to be easier.
To be clear the binaries were on tmpfs because of disk I/O requirements.
We then used text replication to avoid icache misses and having all the
copies banging on the memory where the instruction code resides. This
results in us hitting a bottleneck in mpol_shared_policy_lookup() since
lookup is serialised by the shared_policy lock.
I have only reproduced this on very large (3k+ cores) boxes. The
problem starts showing up at just a few hundred ranks getting worse
until it threatens to livelock once it gets large enough. For example
on the gamess benchmark at 128 ranks this area consumes only ~1% of
time, at 512 ranks it consumes nearly 13%, and at 2k ranks it is over
90%.
To alleviate the contention in this area I converted the spinlock to an
rwlock. This allows a large number of lookups to happen simultaneously.
The results were quite good reducing this consumtion at max ranks to
around 2%.
[akpm@linux-foundation.org: tidy up code comments]
Signed-off-by: Nathan Zimmer <nzimmer@sgi.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Nadia Yvette Chambers <nyc@holomorphy.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'fs/hugetlbfs')
| -rw-r--r-- | fs/hugetlbfs/inode.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index f6820ecf0a11..a1cb8fd2289b 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -738,7 +738,7 @@ static struct inode *hugetlbfs_get_inode(struct super_block *sb, /* * The policy is initialized here even if we are creating a * private inode because initialization simply creates an - * an empty rb tree and calls spin_lock_init(), later when we + * an empty rb tree and calls rwlock_init(), later when we * call mpol_free_shared_policy() it will just return because * the rb tree will still be empty. */ |