This is a patch for coalescing access to res_counter at charging by percpu
caching. At charge, memcg charges 64pages and remember it in percpu
cache. Because it's cache, drain/flush if necessary.
This version uses public percpu area.
2 benefits for using public percpu area.
1. Sum of stocked charge in the system is limited to # of cpus
not to the number of memcg. This shows better synchonization.
2. drain code for flush/cpuhotplug is very easy (and quick)
The most important point of this patch is that we never touch res_counter
in fast path. The res_counter is system-wide shared counter which is modified
very frequently. We shouldn't touch it as far as we can for avoiding
false sharing.
On x86-64 8cpu server, I tested overheads of memcg at page fault by
running a program which does map/fault/unmap in a loop. Running
a task per a cpu by taskset and see sum of the number of page faults
in 60secs.
[without memcg config]
40156968 page-faults # 0.085 M/sec ( +- 0.046% )
27.67 cache-miss/faults
[root cgroup]
36659599 page-faults # 0.077 M/sec ( +- 0.247% )
31.58 cache miss/faults
[in a child cgroup]
18444157 page-faults # 0.039 M/sec ( +- 0.133% )
69.96 cache miss/faults
[ + coalescing uncharge patch]
27133719 page-faults # 0.057 M/sec ( +- 0.155% )
47.16 cache miss/faults
[ + coalescing uncharge patch + this patch ]
34224709 page-faults # 0.072 M/sec ( +- 0.173% )
34.69 cache miss/faults
Changelog (since Oct/2):
- updated comments
- replaced get_cpu_var() with __get_cpu_var() if possible.
- removed mutex for system-wide drain. adds a counter instead of it.
- removed CONFIG_HOTPLUG_CPU
Changelog (old):
- rebased onto the latest mmotm
- moved charge size check before __GFP_WAIT check for avoiding unnecesary
- added asynchronous flush routine.
- fixed bugs pointed out by Nishimura-san.
[akpm@linux-foundation.org: tweak comments]
[nishimura@mxp.nes.nec.co.jp: don't do INIT_WORK() repeatedly against the same work_struct]
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In massive parallel enviroment, res_counter can be a performance
bottleneck. One strong techinque to reduce lock contention is reducing
calls by coalescing some amount of calls into one.
Considering charge/uncharge chatacteristic,
- charge is done one by one via demand-paging.
- uncharge is done by
- in chunk at munmap, truncate, exit, execve...
- one by one via vmscan/paging.
It seems we have a chance to coalesce uncharges for improving scalability
at unmap/truncation.
This patch is a for coalescing uncharge. For avoiding scattering memcg's
structure to functions under /mm, this patch adds memcg batch uncharge
information to the task. A reason for per-task batching is for making use
of caller's context information. We do batched uncharge (deleyed
uncharge) when truncation/unmap occurs but do direct uncharge when
uncharge is called by memory reclaim (vmscan.c).
The degree of coalescing depends on callers
- at invalidate/trucate... pagevec size
- at unmap ....ZAP_BLOCK_SIZE
(memory itself will be freed in this degree.)
Then, we'll not coalescing too much.
On x86-64 8cpu server, I tested overheads of memcg at page fault by
running a program which does map/fault/unmap in a loop. Running
a task per a cpu by taskset and see sum of the number of page faults
in 60secs.
[without memcg config]
40156968 page-faults # 0.085 M/sec ( +- 0.046% )
27.67 cache-miss/faults
[root cgroup]
36659599 page-faults # 0.077 M/sec ( +- 0.247% )
31.58 miss/faults
[in a child cgroup]
18444157 page-faults # 0.039 M/sec ( +- 0.133% )
69.96 miss/faults
[child with this patch]
27133719 page-faults # 0.057 M/sec ( +- 0.155% )
47.16 miss/faults
We can see some amounts of improvement.
(root cgroup doesn't affected by this patch)
Another patch for "charge" will follow this and above will be improved more.
Changelog(since 2009/10/02):
- renamed filed of memcg_batch (as pages to bytes, memsw to memsw_bytes)
- some clean up and commentary/description updates.
- added initialize code to copy_process(). (possible bug fix)
Changelog(old):
- fixed !CONFIG_MEM_CGROUP case.
- rebased onto the latest mmotm + softlimit fix patches.
- unified patch for callers
- added commetns.
- make ->do_batch as bool.
- removed css_get() at el. We don't need it.
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A memory cgroup has a memory.memsw.usage_in_bytes file. It shows the sum
of the usage of pages and swapents in the cgroup. Presently the root
cgroup's memsw.usage_in_bytes shows the wrong value - the number of
swapents are not added.
So take MEM_CGROUP_STAT_SWAPOUT into account.
Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Reviewed-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Example is outdated, it still uses old ->read_proc interfaces and "fb"
example is plain racy. There are better examples all over the tree.
Docbook itself says almost nothing about /proc and contain quite a number
of simply wrong facts, e.g. device nodes support. What it does is
describing at great length interface which are going to be removed.
There are Documentation/filesystems/seq_file.txt in exchange.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Erik Mouw <mouw@nl.linux.org>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
/proc/fs/reiserfs/version is on the way of removing ->read_proc interface.
It's empty however, so simply remove it instead of doing dummy
conversion. It's hard to see what information userspace can extract from
empty file.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When an IO error happens while writing metadata buffers, we should better
report it and call ext2_error since the filesystem is probably no longer
consistent. Sometimes such IO errors happen while flushing thread does
background writeback, the buffer gets later evicted from memory, and thus
the only trace of the error remains as AS_EIO bit set in blockdevice's
mapping. So we check this bit in ext2_fsync and report the error although
we cannot be really sure which buffer we failed to write.
Signed-off-by: Jan Kara <jack@suse.cz>
Cc: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Convert code away from ->read_proc/->write_proc interfaces. Switch to
proc_create()/proc_create_data() which make addition of proc entries
reliable wrt NULL ->proc_fops, NULL ->data and so on.
Problem with ->read_proc et al is described here commit
786d7e1612 "Fix rmmod/read/write races in
/proc entries"
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Adam Belay <abelay@mit.edu>
Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The intelfb driver sets color map depending on currently active pipe.
However, if an LVDS display is attached (like in laptop) the active pipe
variable is never set. The default value is PIPE_A and can be wrong. Set
up the pipe variable during driver initialization after hardware state was
read.
Also, the detection of the active display (and hence the pipe) is wrong.
The pipes are assigned to so called planes. Both pipes are always enabled
on my laptop but only one plane is enabled (the plane A for the CRT or the
plane B for the LVDS). Change active pipe detection code to take into
account a status of the plane assigned to each pipe.
The problem is visible in the 8 bpp mode if colors above 15 are used. The
first 16 color entries are displayed correctly.
The graphics chip description is here (G45 vol. 3):
http://intellinuxgraphics.org/documentation.html
Addresses http://bugzilla.kernel.org/show_bug.cgi?id=13285
Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Cc: Michal Suchanek <hramrach@centrum.cz>
Cc: Dean Menezes <samanddeanus@yahoo.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add a helper function to enable raster. Also add one member in the
private data structure to track the current blank status, another function
pointer which takes in the platform specific callback function to control
panel power.
These updates will help in adding suspend/resume and frame buffer blank
operation features.
Signed-off-by: Chaithrika U S <chaithrika@ti.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch provides the acceleration entry points for the SM501
framebuffer driver.
This patch provides the sync, copyarea and fillrect entry points, using
the SM501's 2D acceleration engine to perform the operations in-chip
rather than across the bus.
Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Simtec Linux Team <linux@simtec.co.uk>
Signed-off-by: Vincent Sanders <vince@simtec.co.uk>
Cc: Krzysztof Helt <krzysztof.h1@poczta.fm>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There are several places in the SM501 fb driver that could do with using
resource_size() to calculate the size of a resource.
Also fix a bug where request_mem_region() is being passed one too few
bytes when requesting the register memory region, which was causing the
following in /proc/iomem:
13e80000-13e8ffff : sm501-fb.0
13e80000-13e8fffe : sm501-fb
fixed, this reads:
13e80000-13e8ffff : sm501-fb.0
13e80000-13e8ffff : sm501-fb
Signed-off-by: Ben Dooks <ben@simtec.co.uk>
Signed-off-by: Simtec Linux Team <linux@simtec.co.uk>
Cc: Vincent Sanders <vince@simtec.co.uk>
Cc: Krzysztof Helt <krzysztof.h1@poczta.fm>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The clock divider value can be calculated from the pixel clock value for
the panel. This gives more flexiblity to the driver to change the divider
value on the fly as in the case of cpufreq feature- support for which will
be added shortly.
Signed-off-by: Chaithrika U S <chaithrika@ti.com>
Cc: Sudhakar Rajashekhara <sudhakar.raj@ti.com>
Cc: Steve Chen <schen@mvista.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add check if palette register number is in correct range for few drivers
which miss it. The regno value comes indirectly from user space.
Two drivers has converted check from BUG_ON() macro to just return an
error (non-zero value).
Signed-off-by: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Drivers may use gpiolib sysfs as part of their public user space
interface. The GPIO number and polarity might change from board to
board. The gpio_export_link() call can be used to hide the GPIO number
from user space. Add support for also hiding the GPIO line polarity
changes from user space.
Signed-off-by: Jani Nikula <ext-jani.1.nikula@nokia.com>
Cc: David Brownell <david-b@pacbell.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
