Function: comedi_close -- close a Comedi device
Retval: int
Param: comedi * device
Description:
 Close a device previously opened by comedi_open().
Returns:
 If sucessful, comedi_close returns 0.  On failure, -1 is returned.

Function: comedi_open -- open a Comedi device
Retval: comedi_t *
Param: const char * filename
Description:
 Open a Comedi device specified by the file filename.
Returns:
 If sucessful, comedi_open returns a pointer to a valid comedi_t
 structure.  This structure is transparent; the pointer should not
 be dereferenced by the application.  NULL is returned on failure.

Function: comedi_loglevel -- change Comedilib logging properties
Retval: int
Param: int loglevel
Description:
 This function affects the output of debugging and error messages
 from Comedilib.  By increasing the loglevel, additional debugging
 information will be printed.  Error and debugging messages are
 printed to the stream stderr.

 The default loglevel can be set by using the environment variable
 COMEDI_LOGLEVEL.  The default loglevel is 1.

 In order to conserve resources, some debugging information is
 disabled by default when Comedilib is compiled.

 The meaning of the loglevels is as follows:

 COMEDI_LOGLEVEL=0  Comedilib prints nothing.

 COMEDI_LOGLEVEL=1  (default) Comedilib prints error messages when
 there is a self-consistency error (i.e., an internal bug.)

 COMEDI_LOGLEVEL=2  Comedilib prints an error message when an invalid
 parameter is passed.

 COMEDI_LOGLEVEL=3  Comedilib prints an error message whenever an
 error is generated in the Comedilib library or in the C library,
 when called by Comedilib.

 COMEDI_LOGLEVEL=4  Comedilib prints a lot of junk.
Returns:
 This function returns the previous loglevel.

Function: comedi_perror -- print a Comedilib error message
Retval: void
Param: const char * s
Description:
 When a Comedilib function fails, it usually returns -1 or
 NULL, depending on the return type.  An internal library
 variable stores an error number, which can be retrieved with
 comedi_errno().  This error number can be converted to a
 human-readable form by the functions comedi_perror()
 and comedi_strerror().

 These functions are intended to mimic the behavior of the
 standard C library functions perror(), strerror(), and errno.
 In particular, Comedilib functions sometimes return an error
 that is generated inside the C library; the comedi error
 message in this case is the same as the C library.

 The function comedi_perror() prints an error message to stderr.
 The error message consists of the argument string, a colon, a
 space, a description of the error condition, and a new line.

Function: comedi_strerror -- return string describing Comedilib error code
Retval: char *
Param: int errnum
Description:
 When a Comedilib function fails, it usually returns -1 or
 NULL, depending on the return type.  An internal library
 variable stores an error number, which can be retrieved with
 comedi_errno().  This error number can be converted to a
 human-readable form by the functions comedi_perror()
 and comedi_strerror().

 These functions are intended to mimic the behavior of the
 standard C library functions perror(), strerror(), and errno.
 In particular, Comedilib functions sometimes return an error
 that is generated inside the C library; the comedi error
 message in this case is the same as the C library.

 The function comedi_strerror() returns a pointer to a
 character string
 describing the Comedilib error errnum.  The persistence
 of the returned pointer is undefined, and should not be trusted
 after the next Comedilib call.  An unrecognized error number will
 return a pointer to the string "undefined error", or similar.

Function: comedi_errno -- number of last Comedilib error
Retval: int
Param: void
Description:
 When a Comedilib function fails, it usually returns -1 or
 NULL, depending on the return type.  An internal library
 variable stores an error number, which can be retrieved with
 comedi_errno().  This error number can be converted to a
 human-readable form by the functions comedi_perror()
 and comedi_strerror().

 These functions are intended to mimic the behavior of the
 standard C library functions perror(), strerror(), and errno.
 In particular, Comedilib functions sometimes return an error
 that is generated inside the C library; the comedi error
 message in this case is the same as the C library.

 The function comedi_errno() returns an integer describing
 the most recent comedilib error.  This integer may be used
 as the errnum parameter for comedi_strerror().

 Note that comedi_errno() is deliberately different than the
 variable errno.  This is to overcome difficulties in making
 errno thread-safe.

Function: comedi_fileno -- integer descriptor of Comedilib device
Retval: int
Param: comedi_t * device
Description:
 The function comedi_fileno returns the integer descriptor for
 the device dev.  This descriptor can then be used as the
 file descriptor parameter of read(), write(), etc.
 This function is intended to mimic the standard C library
 function fileno().  If dev is an invalid comedi_t
 pointer, the function returns -1 and sets the appropriate
 Comedilib error value.

Function: comedi_get_n_subdevices -- number of subdevices 
Retval: int
Param: comedi_t * device
Description:
 Returns the number of subdevices belonging to the Comedi
 device referenced by the parameter device.

Function: comedi_get_version_code -- Comedi version code
Retval: int
Param: comedi_t * device
Description:
 Returns the Comedi kernel module version code.  A valid Comedi
 device referenced by the parameter device is necessary to
 communicate with the kernel module.  On error, -1 is returned.

 The version code is encoded as a bitfield of three 8-bit
 numbers.  For example, 0x00073d is the version code for
 version 0.7.61.

 This function is of limited usefulness.  A typical
 mis-application of this function is to use it to determine
 if a certain feature is supported.  If the application needs
 to know of the existence of a particular feature, an existence
 test function should be written and put in the Comedilib source.

Function: comedi_get_driver_name -- Comedi driver name
Retval: char *
Param: comedi_t * device
Description:
 The function comedi_get_driver_name returns a pointer
 to a string containing the name of the driver being used by comedi
 for the comedi device represented by device.  This pointer is
 valid until the device is closed.  This function returns NULL
 if there is an error.

Function: comedi_get_board_name -- Comedi device name
Retval: char *
Param: comedi_t * device
Description:
 The function comedi_get_board_name returns a pointer
 to a string containing the name of the device.  This pointer is
 valid until the comedi descriptor it is closed.  This
 function returns NULL if there is an error.

Function: comedi_get_subdevice_type -- type of subdevice
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_subdevice_type() returns an
 integer describing the type of subdevice that belongs to the comedi
 device device and has the index subdevice.  The
 function returns -1 if there is an error.
 
 XXX Subdevice type table

Function: comedi_find_subdevice_by_type -- search for subdevice type
Retval: int
Param: comedi_t * device
Param: int type
Param: unsigned int start_subdevice
Description:
 The function comedi_find_subdevice_by_type() tries to
 locate a subdevice belonging to comedi device device,
 having type type, starting with the subdevice
 start_subdevice.  If it finds a subdevice with the requested
 type, it returns its index.  If it does not locate the requested
 subdevice, it returns -1 and sets the Comedilib error number to
 XXX "subdevice not found".  If there is an error, the function
 returns -1 and sets the appropriate error.

Function: comedi_get_read_subdevice -- find streaming input subdevice
Retval: int
Param: comedi_t * device
Description:
 The function comedi_get_read_subdevice() returns the subdevice
 that allows streaming input for device dev.  If no subdevice
 supports streaming input, -1 is returned and the Comedilib error
 number is set to XXX "subdevice not found".

Function: comedi_get_write_subdevice -- find streaming output subdevice
Retval: int
Param: comedi_t * device
Description:
 The function comedi_get_write_subdevice() returns the subdevice
 that allows streaming output for device dev.  If no subdevice
 supports streaming output, -1 is returned and the Comedilib error
 number is set to XXX "subdevice not found".

Function: comedi_get_subdevice_flags -- properties of subdevice
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 This function returns a bitfield describing the capabilities of
 the specified subdevice.  If there is an error, -1 is returned,
 and the Comedilib error value is set.

 XXX table.

Function: comedi_get_n_channels -- number of subdevice channels
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_n_channels() returns the number
 of channels of the subdevice belonging to the comedi device device
 and having index subdevice.  This function returns -1 on error and
 the Comedilib error value is set.

Function: comedi_range_is_chan_specific -- range information depends on channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 If each channel of the specified subdevice has different range
 information, this function returns 1.  Otherwise, this function
 returns 0.  On error, this function returns -1.

Function: comedi_maxdata_is_chan_specific -- maximum sample depends on channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 If each channel of the specified subdevice has different maximum
 sample values, this function returns 1.  Otherwise, this function
 returns 0.  On error, this function returns -1.

Function: comedi_get_maxdata -- maximum sample of channel
Retval: lsampl_t
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Description:
 The function comedi_get_maxdata() returns the maximum
 valid data value for channel chan of subdevice
 subdevice belonging to the comedi device device
 This function returns 0 on error.

Function: comedi_get_n_ranges -- number of ranges of channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Description:
 The function comedi_get_n_ranges() returns the number
 of ranges of the channel chan belonging to the subdevice
 of the comedi device device.  This function returns -1 on error.

Function: comedi_get_range -- range information of channel
Retval: comedi_range *
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Description:
 The function comedi_get_range() returns a pointer to a
 comedi_range structure that contains information that can be used to
 convert sample values to or from physical units.  The pointer is valid
 until the Comedi device device is closed.  If there is an
 error, NULL is returned.

Function: comedi_find_range -- search for range
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int unit
Param: double min
Param: double max
Description:
 The function comedi_find_range() tries to
 locate the optimal (smallest) range for the channel chan
 belonging to a subdevice of the comedi device device,
 that includes both min and max in units.
 If a matching range is found, the index of the matching range is
 returned.  If no matching range is available, the function returns
 -1.

Function: comedi_get_buffer_size -- streaming buffer size of subdevice
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_buffer_size() returns the size (in bytes)
 of the streaming buffer for the subdevice specified by device and
 subdevice.  On error, -1 is returned.

Function: comedi_get_max_buffer_size -- maximum streaming buffer size
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_max_buffer_size() returns the maximum
 allowable size (in bytes) of the streaming buffer for the subdevice
 specified by device and subdevice.  Changing the maximum buffer
 size requires appropriate privileges.  On error, -1 is returned.

Function: comedi_set_buffer_size -- streaming buffer size of subdevice
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int size
Description:
 The function comedi_set_buffer_size() changes the size of the
 streaming buffer for the subdevice specified by device and subdevice.
 The parameter size must be a multiple of the virtual memory page
 size.

 The virtual memory page size can be determined using
 sysconf(_SC_PAGE_SIZE).

Function: comedi_trigger -- perform streaming input/output (deprecated)
Retval: int
Param: comedi_t * device
Param: comedi_trig * trig
Status: deprecated
Description:
 The function comedi_trigger() instructs Comedi to
 perform the command specified by the trigger structure trig.
 The return value depends on
 the particular trig being issued.  If there is an
 error, -1 is returned.

Function: comedi_do_insnlist -- perform multiple instructions
Retval: int
Param: comedi_t * device
Param: comedi_insnlist * list
Description:
 The function comedi_do_insnlist() performs multiple Comedi
 instructions as part of one system call.  In addition, Comedi
 attempts to perform the instructions atomically, that is, on
 standard Linux kernels, no process preemption should occur
 during the instructions.  However, the process may be preempted
 before or after the group of instructions.

 This function can be used to avoid the overhead of multiple
 system calls, or to ensure that multiple instructions occur
 without significant delay between them.
 
 Preemption may occur if any of the instructions or the data
 arrays of any of the instructions exist in non-resident or
 copy-on-write pages.
Returns:
 The function comedi_do_insnlist() returns the number of
 sucessfully completed instructions.  Error information for
 the unsucessful instruction is not available.  If there is
 an error before the first instruction can be executed, -1
 is returned.

Function: comedi_do_insn -- perform instruction
Retval: int
Param: comedi_t * device
Param: comedi_insn * instruction
Description:
 The function comedi_do_insn() performs a single instruction.
 If sucessful, comedi_do_insn() returns the number of samples
 measured, which may be less than the number of requested
 samples.  Comedi limits the number of requested samples in
 order to enforce fairness among processes.  If there is an
 error, -1 is returned.

Function: comedi_lock -- subdevice reservation
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_lock() reserves a subdevice for use by the
 current process.  While the lock is held, no other process is
 allowed to read, write, or configure that subdevice, although
 other processes can read information about the subdevice.  The
 lock is released when comedi_unlock() is called, or the device
 is closed.  If sucessful, 0 is returned.  If there is an error,
 -1 is returned.

Function: comedi_unlock -- subdevice reservation
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_unlock() released a subdevice lock acquired
 by comedi_lock().  If sucessful, 0 is returned, otherwise -1.

Function: comedi_to_phys -- convert sample to physical units
Retval: double
Param: lsampl_t data
Param: comedi_range * range
Param: lsampl_t maxdata
Description:
 Converts data given in sample values (lsampl_t, between 0 and
 maxdata) into physical units (double).  The parameter range
 represents the conversion information to use, and the parameter
 maxdata represents the maximum possible data value for the
 channel that the data was read.

 Conversion of endpoint sample values, that is, sample values
 equal to 0 or maxdata, is affected by the Comedilib out-of-range
 behavior.  If the out-of-range behavior is set to COMEDI_OOR_NAN, 
 endpoint values are converted to NAN.  If the out-of-range
 behavior is set to COMEDI_OOR_NUMBER, the endpoint values are
 converted similarly to other values.

 If there is an error, NAN is returned.

Function: comedi_from_phys -- convert physical units to sample
Retval: lsampl_t
Param: double data
Param: comedi_range * range
Param: lsampl_t maxdata
Description:
 Converts data given in physical units (data) into sample values
 (lsampl_t, between 0 and maxdata).  The parameter rng
 represents the conversion information to use, and the parameter
 maxdata represents the maximum possible data value for the
 channel that the data will be written to.

 Conversion is not affected by out-of-range behavior.  Out-of-range
 data parameters are silently truncated to the range 0 to maxdata.

Function: comedi_data_read -- read single sample from channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: lsampl_t * data
Description:
 Reads a single sample on the channel specified by the Comedi
 device device, the subdevice subdevice, and the channel channel.
 For the A/D conversion (if appropriate),
 the device is configured to use range specification
 range and (if appropriate) analog reference type
 aref. Analog reference types that are not supported
 by the device are silently ignored.

 The function comedi_data_read() reads one data value from
 the specified channel and places the data value in the
 location pointed to by data.

 WARNING: comedi_data_read() does not do any pausing to
 allow multiplexed analog inputs to settle before
 performing an analog to digital conversion.  If you are
 switching between different channels and need to allow
 your analog input to settle for an accurate reading,
 use comedi_data_read_delayed(), or set the
 input channel at an earlier time with
 comedi_data_read_hint().

 On sucess, comedi_data_read() returns 1 (the number of samples
 read).  If there is an error, -1 is returned.

 Data values returned by this function are unsigned integers
 less than or equal to the maximum sample value of the channel,
 which can be determined using the function comedi_get_maxdata().
 Conversion of data values to physical units can be performed
 by the function comedi_to_phys().

Function: comedi_data_read_delayed -- read single sample from channel after delaying for specified settling time
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: lsampl_t * data
Param: unsigned int nanosec
Description:
 Similar to comedi_data_read() except it will wait for the
 specified number of nanoseconds between setting the input
 channel and taking a sample.  For analog inputs, most
 boards have a single
 analog to digital converter which is multiplexed to be
 able to read multiple channels.  If the input is not allowed
 to settle after the multiplexer switches channels, the
 reading will be inaccurate.  This function is useful
 for allowing a multiplexed analog input to settle
 when switching channels.

 Although the settling time is specified in nanoseconds, the
 actual settling time will be rounded up to the nearest
 microsecond.

Function: comedi_data_read_hint -- tell driver which channel/range/aref you are going to read from next
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Description:
 Used to prepare an analog input for a subsequent call to
 comedi_data_read().  It is not necessary to use this
 function, but it can be useful for eliminating inaccuaracies
 caused by insufficient settling times when switching the
 channel
 or gain on an analog input.  This function sets an analog input
 to the channel, range, and aref specified but does not
 perform an actual analog to digital conversion.

 Alternatively, one can simply use comedi_data_read_delayed(),
 which sets up the
 input, pauses to allow settling, then performs a conversion.

Function: comedi_data_write -- write single sample to channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: lsampl_t data
Description:
 Writes a single sample on the channel that is specified by the
 Comedi device device, the subdevice subdevice, and the channel
 channel.  If appropriate, the device is configured to use range
 specification range and analog reference type aref.  Analog
 reference types that are not supported by the device are
 silently ignored.

 The function comedi_data_write() writes the data value specified
 by the parameter data to the specified channel.

 On sucess, comedi_data_write() returns 1 (the number of samples
 written).  If there is an error, -1 is returned.

Function: comedi_dio_config -- change input/output properties of channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int direction
Description:
 The function comedi_dio_config() configures individual channels
 in a digital I/O subdevice to be either input or output, depending
 on the value of parameter direction.  Valid directions are
 COMEDI_INPUT or COMEDI_OUTPUT.

 Depending on the capabilities of the hardware device, multiple
 channels may be grouped together to determine direction.  In this
 case, a single call to comedi_dio_config() for any channel in the
 group will affect the entire group.

 If sucessful, 1 is returned, otherwise -1.

Function: comedi_dio_read -- read single bit from digital channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int * bit
Description:
 The function reads the channel channel belonging to the
 subdevice subdevice of device device.  The data value that is
 read is stored in the location pointed to by bit.  This function
 is equivalent to comedi_data_read(device,subdevice,channel,0,0,bit).
 This function does not require a digital subdevice or a subdevice
 with a maximum data value of 1 to work properly.

 Return values and errors are the same as comedi_data_read().

Function: comedi_dio_write -- write single bit to digital channel
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int bit
Description:
 The function writes the value bit to the channel channel belonging
 to the subdevice subdevice of device device.  This function
 is equivalent to comedi_data_write(device,subdevice,channel,0,0,bit).
 This function does not require a digital subdevice or a subdevice
 with a maximum data value of 1 to work properly.

 Return values and errors are the same as comedi_data_write().

Function: comedi_dio_bitfield -- read/write multiple digital channels
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int write_mask
Param: unsigned int * bits
Description:
 The function comedi_dio_bitfield() allows multiple channels to
 be read simultaneously from a digital input or digital I/O device.
 The parameter write_mask and the value pointed to by bits
 are interpreted as bit fields, with the least significant bit
 representing channel 0.  For each bit in write_mask that is
 set to 1, the cooresponding bit in *bits is written to the digital
 output channel.  After writing all the output channels, each
 channel is read, and the result placed in the approprate bits in
 *bits.  The result of reading an output channel is undefined.
 It is not possible to access channels greater than 31 using this
 function.

Function: comedi_sv_init -- slowly-varying inputs
Retval: int
Param: comedi_sv_t * sv
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Status: deprecated
Description:
 The function comedi_sv_init() initializes the slow varying Comedi
 structure sv to use the device device, the analog input subdevice
 subdevice, and the channel channel.  The slow varying Comedi
 structure is used by comedi_sv_measure() to accurately measure
 an analog input by averaging over many samples.  The default
 number of samples is 100.  This function returns 0 on success,
 -1 on error.

Function: comedi_sv_update -- slowly-varying inputs
Retval: int
Param: comedi_sv_t * sv
Status: deprecated
Description:
 The function comedi_sv_update() updates internal parameters of
 the slowly varying Comedi structure sv.

Function: comedi_sv_measure -- slowly-varying inputs
Retval: int
Param: comedi_sv_t * sv
Param: double * data
Status: deprecated
Description:
 The function comedi_sv_measure() uses the slowly varying Comedi
 structure sv to measure a slowly varying signal.  If sucessful,
 the result (in physical units) is stored in the location pointed
 to by data, and the number of samples is returned.  On error, -1
 is returned.

Function: comedi_get_cmd_src_mask -- streaming input/output capabilities
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: comedi_cmd * command
Description:
 The command capabilities of the subdevice indicated by the parameters
 device and subdevice are probed, and the results placed in the
 command structure pointed to by the parameter command.  The trigger
 source elements of the command structure are set to the logical OR
 value of possible trigger sources.  Other elements in the structure
 are undefined.  If sucessful, 0 is returned, otherwise -1.

Function: comedi_get_cmd_generic_timed -- streaming input/output capabilities
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: comedi_cmd * command
Param: unsigned int period_ns
Description:
 The command capabilities of the subdevice indicated by the parameters
 device and subdevice are probed, and the results placed in the
 command structure pointed to by the parameter command.  The command
 structure pointed to by the parameter command is modified to be a
 valid command that can be used as a parameter to comedi_command().
 The command measures samples at a rate that corresponds to the
 period period_ns.  The rate is adjusted to a rate that the device
 can handle.  If sucessful, 0 is returned, otherwise -1.

Function: comedi_cancel -- stop streaming input/output in progress
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_cancel() can be used to stop a Comedi command
 previously started by comedi_command() that is still in progress
 on the subdevice indicated by the parameters device and subdevice.
 This may not return the subdevice to a ready state, since there may
 be samples in the buffer that need to be read.

 If sucessful, 0 is returned, otherwise -1.

Function: comedi_command -- start streaming input/output
Retval: int
Param: comedi_t * device
Param: comedi_cmd * command
Description:
 The function comedi_command() starts streaming input or output.  The
 command structure pointed to by the parameter command specifies the
 acquisition.  The command must be able to pass comedi_command_test()
 with a return value of 0, or comedi_command() will fail.
 For input subdevices, sample values are read using the
 function read().  For output subdevices, sample values are written
 using the function write().

 If sucessful, 0 is returned, otherwise -1.

Function: comedi_command_test -- test streaming input/output configuration
Retval: int
Param: comedi_t * device
Param: comedi_cmd * command
Description:
 The function comedi_command_test() tests the command structure pointed
 to by the parameter command and returns an integer describing the
 testing stages that were sucessfully passed.  In addition, if elements
 of the command structure are invalid, they may be modified.  Source
 elements are modified to remove invalid source triggers.  Argument
 elements are adjusted or rounded to the nearest valid value.

 The meanings of the return value are as follows.

 0 indicates a valid command.

 1 indicates that one of the *_src
 members of the command contained an
 unsupported trigger.  The bits corresponding to the unsupported
 triggers are zeroed.

 2 indicates that the particular combination
 of *_src settings is not supported by the driver, or that
 one of the *_src members has the bit corresponding to
 multiple trigger sources set at the same time.

 3 indicates that one of the *_arg members
 of the command is set outside the range of allowable values.
 For instance, an argument for a TRIG_TIMER source which
 exceeds the board's maximum speed.  The invalid *_arg
 members will be adjusted to valid values.

 4 indicates that one of the *_arg members
 required adjustment.  For instance, the argument of a
 TRIG_TIMER source may have been rounded to the nearest
 timing period supported by the board.

 5 indicates that some aspect of the
 command's chanlist is unsupported by the board.  For example,
 some board's require that all channels in the chanlist
 use the same range.

Function: comedi_poll -- force updating of streaming buffer
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_poll() is used on a subdevice that has a
 Comedi command in progress in order to update the streaming buffer.
 If supported by the driver, all available samples are copied to
 the streaming buffer.  These samples may be pending in DMA buffers
 or device FIFOs.  If sucessful, the number of additional bytes
 available is returned.  If there is an error, -1 is returned.

Function: comedi_set_max_buffer_size -- streaming buffer size of subdevice
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int max_size
Description:
 The function comedi_set_max_buffer_size() changes the maximum
 allowable size (in bytes) of the streaming buffer for the subdevice
 specified by device and subdevice.  Changing the maximum buffer
 size requires appropriate privileges.  If sucessful, the old buffer
 size is returned.  On error, -1 is returned.

Function: comedi_get_buffer_contents -- streaming buffer status
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_buffer_contents() is used on a subdevice
 that has a Comedi command in progress.  The number of bytes that
 are available in the streaming buffer is returned.  If there is
 an error, -1 is returned.

Function: comedi_mark_buffer_read -- streaming buffer status
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int num_bytes
Description:
 The function comedi_mark_buffer_read() is used on a subdevice
 that has a Comedi command in progress.  This function can be
 used to indicate that the next num_bytes bytes in the buffer
 are no longer needed and may be discarded.
 If there is an error, -1 is returned.

Function: comedi_get_buffer_offset -- streaming buffer status
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Description:
 The function comedi_get_buffer_offset() is used on a subdevice
 that has a Comedi command in progress.  This function returns
 the offset in bytes of the read pointer in the streaming buffer.
 This offset is only useful for memory mapped buffers.
 If there is an error, -1 is returned.

Function: comedi_get_timer -- timer information (deprecated)
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: double frequency
Param: unsigned int * trigvar
Param: double * actual_frequency
Status: deprecated
Description:
 The function comedi_get_timer converts the frequency frequency
 to a number suitable to send to the driver in a comedi_trig
 structure.  This function remains for compatibility with very
 old versions of Comedi, that converted sampling rates to timer
 values in the libary.  This conversion is now done in the kernel,
 and every device has the timer type nanosec_timer, indicating
 that timer values are simply a time specified in nanoseconds.

Function: comedi_timed_1chan -- streaming input (deprecated)
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: double frequency
Param: unsigned int num_samples
Param: double * data
Status: deprecated
Description:
 Not documented.

Function: comedi_set_global_oor_behavior -- out-of-range behavior
Retval: int
Param: enum comedi_oor_behavior behavior
Status: alpha
Description:
 This function changes the Comedilib out-of-range behavior.
 This currently affects the behavior of comedi_to_phys() when
 converting endpoint sample values, that is, sample values
 equal to 0 or maxdata.  If the out-of-range behavior is set to
 COMEDI_OOR_NAN, endpoint values are converted to NAN.  If the
 out-of-range behavior is set to COMEDI_OOR_NUMBER, the endpoint
 values are converted similarly to other values.

 The previous out-of-range behavior is returned.

Function: comedi_apply_calibration -- set calibration from file
Retval: int
Param: comedi_t *device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: const char *file_path
Status: alpha
Description:
 This function sets the calibration of the specified subdevice
 so that it is in proper calibration when using the specified
 channel, range and aref.  It does so by performing writes
 to the appropriate channels of the board's calibration
 subdevice(s).  Depending on the hardware, the
 calibration settings used may or may not depend on the channel,
 range, or aref.  Furthermore, the calibrations for different
 channels, ranges, or arefs may not be independent.
 For example, some boards cannot have their analog inputs calibrated
 for multiple input
 ranges simultaneously.  Applying a calibration for range 1 may
 blow away a previously applied calibration for range 0.  Applying
 a calibration for analog input channel 0 may cause the same
 calibration to be applied to all the
 other analog input channels as well.
 Your only guarantee is that calls to comedi_apply_calibration()
 on different subdevices will not interfere with each other.

 In practice, their are some rules of thumb on how
 calibrations behave.  No calibrations depend on the aref.
 A multiplexed analog input will have calibration settings that
 do not depend on the channel, and applying a setting for one
 channel will affect
 all channels equally.  Analog outputs, and analog inputs
 with independent a/d converters for each input channel, will have
 calibrations settings which do depend on the channel, and the
 settings for each channel will be independent of the other
 channels.

 If you wish to investigate exactly what comedi_apply_calibration()
 is doing, you can perform reads on your board's calibration
 subdevice to see which calibration channels it is changing.
 You can also try to decipher the calibration file directly (it's a
 text file).

 The file_path parameter can be used
 to specify the file which contains the calibration information.
 If <parameter>file_path</parameter> is NULL, then comedilib
 will use a default
 file location.  The calibration information used by this function
 is generated by the comedi_calibrate program (see its man page).

 The functions comedi_parse_calibration_file(),
 comedi_apply_parsed_calibration(), and comedi_cleanup_calibration()
 provide the same functionality at a slightly lower level.
Returns:
 Zero on success, a negative number on failure.

Function: comedi_apply_parsed_calibration -- set calibration from memory
Retval: int
Param: comedi_t * device
Param: unsigned int subdevice
Param: unsigned int channel
Param: unsigned int range
Param: unsigned int aref
Param: const comedi_calibration_t *calibration
Status: alpha
Description:
 This function is similar to comedi_apply_calibration()
 except the calibration information is read from memory
 instead of a file.  This function can be more
 efficient than comedi_apply_calibration() since the
 calibration file does not need to be reparsed with
 every call.  The <parameter>calibration</parameter> is
 obtained by a call to comedi_parse_calibration_file().

Returns:
 Zero on success, a negative number on failure.

Function: comedi_cleanup_calibration_file -- free calibration resources
Retval: void
Param: comedi_calibration_t *calibration
Status: alpha
Description:
 This function frees the resources associated with a
 <parameter>calibration</parameter> obtained from
 comedi_parse_calibration_file().  <parameter>calibration</parameter>
 can not be used again after calling this function.

Function: comedi_get_default_calibration_path -- get default calibration file path
Retval: char*
Param: comedi_t *dev
Status: alpha
Description:
 This function returns a string containing a default calibration file
 path appropriate for <parameter>dev</parameter>.  Memory for the
 string is allocated by the function, and should be freed when
 the string is no longer needed.
Returns:
 A string which contains a file path useable by
 comedi_parse_calibration_file().  On error, NULL is returned.

Function: comedi_parse_calibration_file -- set calibration
Retval: comedi_calibration_t*
Param: const char *file_path
Status: alpha
Description:
 This function parses a calibration file (produced by the
 comedi_calibrate program) and returns a pointer to a
 comedi_calibration_t which can be passed to the
 comedi_apply_parsed_calibration() function.  When you are
 finished using the comedi_calibration_t, you should
 call comedi_cleanup_calibration() to free the resources
 associated with the comedi_calibration_t.

 The comedi_get_default_calibration_path() function may
 be useful in conjunction with this function.
Returns:
 A pointer to parsed calibration information on success,  or NULL on failure.
