arts的共享空间

请参考照片栏中以下的pic：

DSC00168_DT18-70mm/F10/-0.3EV/1|160s：

采用DT 18-70mm普焦镜头，在强太阳光环境下，采用风景模式以柔和整体色调，光圈值为F10，同时曝光补偿低到－0.3EV，镜头稍微提到20mm将这个轮廓凸现，提高快门到1|160S，将ISO降到100增强图片饱满度，白平衡打到日光模式下降低因强烈阳光带来的刺激。

结果：图像曝光全面，刺激感强，整个图像结构明显、分层清晰、错落有致，降低曝光补偿，凸现阴影感，造成整个图像的阶梯感明显。缺点是：背景蓝天色素处理不好，应该增加图像对比度，更好凸现鸟巢的层次感和逼真度。

Note：怎么看鸟巢怎么像唐僧的僧帽！～～～

如果使用长焦，估计效果会更好些。

 

。。。。。。

 

 

 

 

请分别参考本博客照片中的：

DSC00209_DT 75-300:DT75-300 mm->230mm/F5.6/+2.0EV/ISO1600/1|60s/无反光灯/色调柔和  &&

DSC00214_DT 18-70:DT18-70mm->70mm/F5.6/+2.0EV/ISO1600/1|30S/无闪光灯/色调柔和。

 

DSC00209_DT 75-300:DT75-300 mm->230mm/F5.6/+2.0EV/ISO1600/1|60s/无反光灯/色调柔和：

采用长焦，在日光灯环境下，镜头拉倒230mm，光圈挂到5.6，加上2个曝光补偿，提高快门速度到1|60S，同时调到ISO1600再次提高快门速度，没有打反光灯，色调采用风景背景凸现柔和色调，着重渲染中心点位置

结果：非常清晰，栩栩如生般，效果更佳

 

DSC00214_DT 18-70:DT18-70mm->70mm/F5.6/+2.0EV/ISO1600/1|30S/无闪光灯/色调柔和：

采用普焦，在日光灯环境下，镜头拉倒最大70mm，光圈挂到5.6，加上2个曝光补偿，提高快门速度到1|30S，同时调到ISO1600再次提高快门速度，没有打反光灯，色调采用风景背景凸现柔和色调，着重渲染中心点位置

结果：色调突兀，色素偏硬，模糊，图像有失真

 

此种情况下，长焦(DT75-300mm)更显优势！

 

 

 

入门级普通单反：Sony A200W(带双头：DT18-70mm和DT75-300mm)

教材：美国纽约摄影学院摄影教材(网上n多下载，Note:上下两册，pdf86M)

        a200用户手册(Special for A200 Users)

最重要的：Interest、Money、Time

座右铭：最好的照片是PS出来的，即使有充分的内容。

所以，玩单反要学会photoshope.

Note: 长焦镜头非常非常舒服，有普焦所不能替代的功能.(有关这点将在下面的文章中以图解释)

目标：一年后蔡司超声波马达镜头！

建议：如果有筒子们想买入门级单反的话，推荐Sony A300W，此款功能已经是最全面的入门级了。

 

接下来开始曝Pic(简称曝P)

 

。。。。。。

 

 

 

 

    =============================
    NO-MMU MEMORY MAPPING SUPPORT
    =============================

The kernel has limited support for memory mapping under no-MMU conditions, such
as are used in uClinux environments. From the userspace point of view, memory
mapping is made use of in conjunction with the mmap() system call, the shmat()
call and the execve() system call. From the kernel's point of view, execve()
mapping is actually performed by the binfmt drivers, which call back into the
mmap() routines to do the actual work.

Memory mapping behaviour also involves the way fork(), vfork(), clone() and
ptrace() work. Under uClinux there is no fork(), and clone() must be supplied
the CLONE_VM flag.

The behaviour is similar between the MMU and no-MMU cases, but not identical;
and it's also much more restricted in the latter case:

 (*) Anonymous mapping, MAP_PRIVATE

 In the MMU case: VM regions backed by arbitrary pages; copy-on-write
 across fork.

 In the no-MMU case: VM regions backed by arbitrary contiguous runs of
 pages.

 (*) Anonymous mapping, MAP_SHARED

 These behave very much like private mappings, except that they're
 shared across fork() or clone() without CLONE_VM in the MMU case. Since
 the no-MMU case doesn't support these, behaviour is identical to
 MAP_PRIVATE there.

 (*) File, MAP_PRIVATE, PROT_READ / PROT_EXEC, !PROT_WRITE

 In the MMU case: VM regions backed by pages read from file; changes to
 the underlying file are reflected in the mapping; copied across fork.

 In the no-MMU case:

         - If one exists, the kernel will re-use an existing mapping to the
           same segment of the same file if that has compatible permissions,
           even if this was created by another process.

         - If possible, the file mapping will be directly on the backing device
           if the backing device has the BDI_CAP_MAP_DIRECT capability and
           appropriate mapping protection capabilities. Ramfs, romfs, cramfs
           and mtd might all permit this.

  - If the backing device device can't or won't permit direct sharing,
           but does have the BDI_CAP_MAP_COPY capability, then a copy of the
           appropriate bit of the file will be read into a contiguous bit of
           memory and any extraneous space beyond the EOF will be cleared

  - Writes to the file do not affect the mapping; writes to the mapping
    are visible in other processes (no MMU protection), but should not
    happen.

 (*) File, MAP_PRIVATE, PROT_READ / PROT_EXEC, PROT_WRITE

 In the MMU case: like the non-PROT_WRITE case, except that the pages in
 question get copied before the write actually happens. From that point
 on writes to the file underneath that page no longer get reflected into
 the mapping's backing pages. The page is then backed by swap instead.

 In the no-MMU case: works much like the non-PROT_WRITE case, except
 that a copy is always taken and never shared.

 (*) Regular file / blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

 In the MMU case: VM regions backed by pages read from file; changes to
 pages written back to file; writes to file reflected into pages backing
 mapping; shared across fork.

 In the no-MMU case: not supported.

 (*) Memory backed regular file, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

 In the MMU case: As for ordinary regular files.

 In the no-MMU case: The filesystem providing the memory-backed file
 (such as ramfs or tmpfs) may choose to honour an open, truncate, mmap
 sequence by providing a contiguous sequence of pages to map. In that
 case, a shared-writable memory mapping will be possible. It will work
 as for the MMU case. If the filesystem does not provide any such
 support, then the mapping request will be denied.

 (*) Memory backed blockdev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

 In the MMU case: As for ordinary regular files.

 In the no-MMU case: As for memory backed regular files, but the
 blockdev must be able to provide a contiguous run of pages without
 truncate being called. The ramdisk driver could do this if it allocated
 all its memory as a contiguous array upfront.

 (*) Memory backed chardev, MAP_SHARED, PROT_READ / PROT_EXEC / PROT_WRITE

 In the MMU case: As for ordinary regular files.

 In the no-MMU case: The character device driver may choose to honour
 the mmap() by providing direct access to the underlying device if it
 provides memory or quasi-memory that can be accessed directly. Examples
 of such are frame buffers and flash devices. If the driver does not
 provide any such support, then the mapping request will be denied.

============================
FURTHER NOTES ON NO-MMU MMAP
============================

 (*) A request for a private mapping of less than a page in size may not return
     a page-aligned buffer. This is because the kernel calls kmalloc() to
     allocate the buffer, not get_free_page().

 (*) A list of all the mappings on the system is visible through /proc/maps in
     no-MMU mode.

 (*) A list of all the mappings in use by a process is visible through
     /proc/<pid>/maps in no-MMU mode.

 (*) Supplying MAP_FIXED or a requesting a particular mapping address will
     result in an error.

 (*) Files mapped privately usually have to have a read method provided by the
     driver or filesystem so that the contents can be read into the memory
     allocated if mmap() chooses not to map the backing device directly. An
     error will result if they don't. This is most likely to be encountered
     with character device files, pipes, fifos and sockets.

==========================
INTERPROCESS SHARED MEMORY
==========================

Both SYSV IPC SHM shared memory and POSIX shared memory is supported in NOMMU
mode.  The former through the usual mechanism, the latter through files created
on ramfs or tmpfs mounts.

=======
FUTEXES
=======

Futexes are supported in NOMMU mode if the arch supports them.  An error will
be given if an address passed to the futex system call lies outside the
mappings made by a process or if the mapping in which the address lies does not
support futexes (such as an I/O chardev mapping).

=============
NO-MMU MREMAP
=============

The mremap() function is partially supported.  It may change the size of a
mapping, and may move it[*] if MREMAP_MAYMOVE is specified and if the new size
of the mapping exceeds the size of the slab object currently occupied by the
memory to which the mapping refers, or if a smaller slab object could be used.

MREMAP_FIXED is not supported, though it is ignored if there's no change of
address and the object does not need to be moved.

Shared mappings may not be moved.  Shareable mappings may not be moved either,
even if they are not currently shared.

The mremap() function must be given an exact match for base address and size of
a previously mapped object.  It may not be used to create holes in existing
mappings, move parts of existing mappings or resize parts of mappings.  It must
act on a complete mapping.

[*] Not currently supported.

============================================
PROVIDING SHAREABLE CHARACTER DEVICE SUPPORT
============================================

To provide shareable character device support, a driver must provide a
file->f_op->get_unmapped_area() operation. The mmap() routines will call this
to get a proposed address for the mapping. This may return an error if it
doesn't wish to honour the mapping because it's too long, at a weird offset,
under some unsupported combination of flags or whatever.

The driver should also provide backing device information with capabilities set
to indicate the permitted types of mapping on such devices. The default is
assumed to be readable and writable, not executable, and only shareable
directly (can't be copied).

The file->f_op->mmap() operation will be called to actually inaugurate the
mapping. It can be rejected at that point. Returning the ENOSYS error will
cause the mapping to be copied instead if BDI_CAP_MAP_COPY is specified.

The vm_ops->close() routine will be invoked when the last mapping on a chardev
is removed. An existing mapping will be shared, partially or not, if possible
without notifying the driver.

It is permitted also for the file->f_op->get_unmapped_area() operation to
return -ENOSYS. This will be taken to mean that this operation just doesn't
want to handle it, despite the fact it's got an operation. For instance, it
might try directing the call to a secondary driver which turns out not to
implement it. Such is the case for the framebuffer driver which attempts to
direct the call to the device-specific driver. Under such circumstances, the
mapping request will be rejected if BDI_CAP_MAP_COPY is not specified, and a
copy mapped otherwise.

IMPORTANT NOTE:

 Some types of device may present a different appearance to anyone
 looking at them in certain modes. Flash chips can be like this; for
 instance if they're in programming or erase mode, you might see the
 status reflected in the mapping, instead of the data.

 In such a case, care must be taken lest userspace see a shared or a
 private mapping showing such information when the driver is busy
 controlling the device. Remember especially: private executable
 mappings may still be mapped directly off the device under some
 circumstances!

==============================================
PROVIDING SHAREABLE MEMORY-BACKED FILE SUPPORT
==============================================

Provision of shared mappings on memory backed files is similar to the provision
of support for shared mapped character devices. The main difference is that the
filesystem providing the service will probably allocate a contiguous collection
of pages and permit mappings to be made on that.

It is recommended that a truncate operation applied to such a file that
increases the file size, if that file is empty, be taken as a request to gather
enough pages to honour a mapping. This is required to support POSIX shared
memory.

Memory backed devices are indicated by the mapping's backing device info having
the memory_backed flag set.

========================================
PROVIDING SHAREABLE BLOCK DEVICE SUPPORT
========================================

Provision of shared mappings on block device files is exactly the same as for
character devices. If there isn't a real device underneath, then the driver
should allocate sufficient contiguous memory to honour any supported mapping.

Tools that manage md devices can be found at
   http://www.<country>.kernel.org/pub/linux/utils/raid/....

Boot time assembly of RAID arrays
---------------------------------

You can boot with your md device with the following kernel command
lines:

for old raid arrays without persistent superblocks:
  md=<md device no.>,<raid level>,<chunk size factor>,<fault level>,dev0,dev1,...,devn

for raid arrays with persistent superblocks
  md=<md device no.>,dev0,dev1,...,devn
or, to assemble a partitionable array:
  md=d<md device no.>,dev0,dev1,...,devn
 
md device no. = the number of the md device ...
              0 means md0,
       1 md1,
       2 md2,
       3 md3,
       4 md4

raid level = -1 linear mode
              0 striped mode
       other modes are only supported with persistent super blocks

chunk size factor = (raid-0 and raid-1 only)
              Set  the chunk size as 4k << n.
      
fault level = totally ignored
      
dev0-devn: e.g. /dev/hda1,/dev/hdc1,/dev/sda1,/dev/sdb1
      
A possible loadlin line (Harald Hoyer <HarryH@Royal.Net>)  looks like this:

e:\loadlin\loadlin e:\zimage root=/dev/md0 md=0,0,4,0,/dev/hdb2,/dev/hdc3 ro

Boot time autodetection of RAID arrays
--------------------------------------

When md is compiled into the kernel (not as module), partitions of
type 0xfd are scanned and automatically assembled into RAID arrays.
This autodetection may be suppressed with the kernel parameter
"raid=noautodetect".  As of kernel 2.6.9, only drives with a type 0
superblock can be autodetected and run at boot time.

The kernel parameter "raid=partitionable" (or "raid=part") means
that all auto-detected arrays are assembled as partitionable.

Boot time assembly of degraded/dirty arrays
-------------------------------------------

If a raid5 or raid6 array is both dirty and degraded, it could have
undetectable data corruption.  This is because the fact that it is
'dirty' means that the parity cannot be trusted, and the fact that it
is degraded means that some datablocks are missing and cannot reliably
be reconstructed (due to no parity).

For this reason, md will normally refuse to start such an array.  This
requires the sysadmin to take action to explicitly start the array
despite possible corruption.  This is normally done with
   mdadm --assemble --force ....

This option is not really available if the array has the root
filesystem on it.  In order to support this booting from such an
array, md supports a module parameter "start_dirty_degraded" which,
when set to 1, bypassed the checks and will allows dirty degraded
arrays to be started.

So, to boot with a root filesystem of a dirty degraded raid[56], use

   md-mod.start_dirty_degraded=1

Superblock formats
------------------

The md driver can support a variety of different superblock formats.
Currently, it supports superblock formats "0.90.0" and the "md-1" format
introduced in the 2.5 development series.

The kernel will autodetect which format superblock is being used.

Superblock format '0' is treated differently to others for legacy
reasons - it is the original superblock format.

General Rules - apply for all superblock formats
------------------------------------------------

An array is 'created' by writing appropriate superblocks to all
devices.

It is 'assembled' by associating each of these devices with an
particular md virtual device.  Once it is completely assembled, it can
be accessed.

An array should be created by a user-space tool.  This will write
superblocks to all devices.  It will usually mark the array as
'unclean', or with some devices missing so that the kernel md driver
can create appropriate redundancy (copying in raid1, parity
calculation in raid4/5).

When an array is assembled, it is first initialized with the
SET_ARRAY_INFO ioctl.  This contains, in particular, a major and minor
version number.  The major version number selects which superblock
format is to be used.  The minor number might be used to tune handling
of the format, such as suggesting where on each device to look for the
superblock.

Then each device is added using the ADD_NEW_DISK ioctl.  This
provides, in particular, a major and minor number identifying the
device to add.

The array is started with the RUN_ARRAY ioctl.

Once started, new devices can be added.  They should have an
appropriate superblock written to them, and then passed be in with
ADD_NEW_DISK.

Devices that have failed or are not yet active can be detached from an
array using HOT_REMOVE_DISK.

Specific Rules that apply to format-0 super block arrays, and
       arrays with no superblock (non-persistent).
-------------------------------------------------------------

An array can be 'created' by describing the array (level, chunksize
etc) in a SET_ARRAY_INFO ioctl.  This must has major_version==0 and
raid_disks != 0.

Then uninitialized devices can be added with ADD_NEW_DISK.  The
structure passed to ADD_NEW_DISK must specify the state of the device
and it's role in the array.

Once started with RUN_ARRAY, uninitialized spares can be added with
HOT_ADD_DISK.

 

MD devices in sysfs
-------------------
md devices appear in sysfs (/sys) as regular block devices,
e.g.
   /sys/block/md0

Each 'md' device will contain a subdirectory called 'md' which
contains further md-specific information about the device.

All md devices contain:
  level
     a text file indicating the 'raid level'. e.g. raid0, raid1,
     raid5, linear, multipath, faulty.
     If no raid level has been set yet (array is still being
     assembled), the value will reflect whatever has been written
     to it, which may be a name like the above, or may be a number
     such as '0', '5', etc.

  raid_disks
     a text file with a simple number indicating the number of devices
     in a fully functional array.  If this is not yet known, the file
     will be empty.  If an array is being resized (not currently
     possible) this will contain the larger of the old and new sizes.
     Some raid level (RAID1) allow this value to be set while the
     array is active.  This will reconfigure the array.   Otherwise
     it can only be set while assembling an array.

  chunk_size
     This is the size if bytes for 'chunks' and is only relevant to
     raid levels that involve striping (1,4,5,6,10). The address space
     of the array is conceptually divided into chunks and consecutive
     chunks are striped onto neighbouring devices.
     The size should be at least PAGE_SIZE (4k) and should be a power
     of 2.  This can only be set while assembling an array

  layout
     The "layout" for the array for the particular level.  This is
     simply a number that is interpretted differently by different
     levels.  It can be written while assembling an array.

  reshape_position
     This is either "none" or a sector number within the devices of
     the array where "reshape" is up to.  If this is set, the three
     attributes mentioned above (raid_disks, chunk_size, layout) can
     potentially have 2 values, an old and a new value.  If these
     values differ, reading the attribute returns
        new (old)
     and writing will effect the 'new' value, leaving the 'old'
     unchanged.

  component_size
     For arrays with data redundancy (i.e. not raid0, linear, faulty,
     multipath), all components must be the same size - or at least
     there must a size that they all provide space for.  This is a key
     part or the geometry of the array.  It is measured in sectors
     and can be read from here.  Writing to this value may resize
     the array if the personality supports it (raid1, raid5, raid6),
     and if the component drives are large enough.

  metadata_version
     This indicates the format that is being used to record metadata
     about the array.  It can be 0.90 (traditional format), 1.0, 1.1,
     1.2 (newer format in varying locations) or "none" indicating that
     the kernel isn't managing metadata at all.

  resync_start
     The point at which resync should start.  If no resync is needed,
     this will be a very large number.  At array creation it will
     default to 0, though starting the array as 'clean' will
     set it much larger.

   new_dev
     This file can be written but not read.  The value written should
     be a block device number as major:minor.  e.g. 8:0
     This will cause that device to be attached to the array, if it is
     available.  It will then appear at md/dev-XXX (depending on the
     name of the device) and further configuration is then possible.

   safe_mode_delay
     When an md array has seen no write requests for a certain period
     of time, it will be marked as 'clean'.  When another write
     request arrives, the array is marked as 'dirty' before the write
     commences.  This is known as 'safe_mode'.
     The 'certain period' is controlled by this file which stores the
     period as a number of seconds.  The default is 200msec (0.200).
     Writing a value of 0 disables safemode.

   array_state
     This file contains a single word which describes the current
     state of the array.  In many cases, the state can be set by
     writing the word for the desired state, however some states
     cannot be explicitly set, and some transitions are not allowed.

     clear
         No devices, no size, no level
         Writing is equivalent to STOP_ARRAY ioctl
     inactive
         May have some settings, but array is not active
            all IO results in error
         When written, doesn't tear down array, but just stops it
     suspended (not supported yet)
         All IO requests will block. The array can be reconfigured.
         Writing this, if accepted, will block until array is quiessent
     readonly
         no resync can happen.  no superblocks get written.
         write requests fail
     read-auto
         like readonly, but behaves like 'clean' on a write request.

     clean - no pending writes, but otherwise active.
         When written to inactive array, starts without resync
         If a write request arrives then
           if metadata is known, mark 'dirty' and switch to 'active'.
           if not known, block and switch to write-pending
         If written to an active array that has pending writes, then fails.
     active
         fully active: IO and resync can be happening.
         When written to inactive array, starts with resync

     write-pending
         clean, but writes are blocked waiting for 'active' to be written.

     active-idle
         like active, but no writes have been seen for a while (safe_mode_delay).

As component devices are added to an md array, they appear in the 'md'
directory as new directories named
      dev-XXX
where XXX is a name that the kernel knows for the device, e.g. hdb1.
Each directory contains:

      block
        a symlink to the block device in /sys/block, e.g.
      /sys/block/md0/md/dev-hdb1/block -> ../../../../block/hdb/hdb1

      super
        A file containing an image of the superblock read from, or
        written to, that device.

      state
        A file recording the current state of the device in the array
 which can be a comma separated list of
       faulty   - device has been kicked from active use due to
                         a detected fault
       in_sync  - device is a fully in-sync member of the array
       writemostly - device will only be subject to read
           requests if there are no other options.
    This applies only to raid1 arrays.
       spare    - device is working, but not a full member.
    This includes spares that are in the process
    of being recovered to
 This list may grow in future.
 This can be written to.
 Writing "faulty"  simulates a failure on the device.
 Writing "remove" removes the device from the array.
 Writing "writemostly" sets the writemostly flag.
 Writing "-writemostly" clears the writemostly flag.

      errors
 An approximate count of read errors that have been detected on
 this device but have not caused the device to be evicted from
 the array (either because they were corrected or because they
 happened while the array was read-only).  When using version-1
 metadata, this value persists across restarts of the array.

 This value can be written while assembling an array thus
 providing an ongoing count for arrays with metadata managed by
 userspace.

      slot
        This gives the role that the device has in the array.  It will
 either be 'none' if the device is not active in the array
        (i.e. is a spare or has failed) or an integer less than the
 'raid_disks' number for the array indicating which position
 it currently fills.  This can only be set while assembling an
 array.  A device for which this is set is assumed to be working.

      offset
        This gives the location in the device (in sectors from the
        start) where data from the array will be stored.  Any part of
        the device before this offset us not touched, unless it is
        used for storing metadata (Formats 1.1 and 1.2).

      size
        The amount of the device, after the offset, that can be used
        for storage of data.  This will normally be the same as the
 component_size.  This can be written while assembling an
        array.  If a value less than the current component_size is
        written, component_size will be reduced to this value.

An active md device will also contain and entry for each active device
in the array.  These are named

    rdNN

where 'NN' is the position in the array, starting from 0.
So for a 3 drive array there will be rd0, rd1, rd2.
These are symbolic links to the appropriate 'dev-XXX' entry.
Thus, for example,
       cat /sys/block/md*/md/rd*/state
will show 'in_sync' on every line.

 

Active md devices for levels that support data redundancy (1,4,5,6)
also have

   sync_action
     a text file that can be used to monitor and control the rebuild
     process.  It contains one word which can be one of:
       resync        - redundancy is being recalculated after unclean
                       shutdown or creation
       recover       - a hot spare is being built to replace a
                       failed/missing device
       idle          - nothing is happening
       check         - A full check of redundancy was requested and is
                       happening.  This reads all block and checks
                       them. A repair may also happen for some raid
                       levels.
       repair        - A full check and repair is happening.  This is
                       similar to 'resync', but was requested by the
                       user, and the write-intent bitmap is NOT used to
         optimise the process.

      This file is writable, and each of the strings that could be
      read are meaningful for writing.

       'idle' will stop an active resync/recovery etc.  There is no
           guarantee that another resync/recovery may not be automatically
    started again, though some event will be needed to trigger
           this.
 'resync' or 'recovery' can be used to restart the
           corresponding operation if it was stopped with 'idle'.
 'check' and 'repair' will start the appropriate process
           providing the current state is 'idle'.

   mismatch_count
      When performing 'check' and 'repair', and possibly when
      performing 'resync', md will count the number of errors that are
      found.  The count in 'mismatch_cnt' is the number of sectors
      that were re-written, or (for 'check') would have been
      re-written.  As most raid levels work in units of pages rather
      than sectors, this my be larger than the number of actual errors
      by a factor of the number of sectors in a page.

   bitmap_set_bits
      If the array has a write-intent bitmap, then writing to this
      attribute can set bits in the bitmap, indicating that a resync
      would need to check the corresponding blocks. Either individual
      numbers or start-end pairs can be written.  Multiple numbers
      can be separated by a space.
      Note that the numbers are 'bit' numbers, not 'block' numbers.
      They should be scaled by the bitmap_chunksize.

   sync_speed_min
   sync_speed_max
     This are similar to /proc/sys/dev/raid/speed_limit_{min,max}
     however they only apply to the particular array.
     If no value has been written to these, of if the word 'system'
     is written, then the system-wide value is used.  If a value,
     in kibibytes-per-second is written, then it is used.
     When the files are read, they show the currently active value
     followed by "(local)" or "(system)" depending on whether it is
     a locally set or system-wide value.

   sync_completed
     This shows the number of sectors that have been completed of
     whatever the current sync_action is, followed by the number of
     sectors in total that could need to be processed.  The two
     numbers are separated by a '/'  thus effectively showing one
     value, a fraction of the process that is complete.

   sync_speed
     This shows the current actual speed, in K/sec, of the current
     sync_action.  It is averaged over the last 30 seconds.

   suspend_lo
   suspend_hi
     The two values, given as numbers of sectors, indicate a range
     within the array where IO will be blocked.  This is currently
     only supported for raid4/5/6.

Each active md device may also have attributes specific to the
personality module that manages it.
These are specific to the implementation of the module and could
change substantially if the implementation changes.

These currently include

  stripe_cache_size  (currently raid5 only)
      number of entries in the stripe cache.  This is writable, but
      there are upper and lower limits (32768, 16).  Default is 128.
  strip_cache_active (currently raid5 only)
      number of active entries in the stripe cache