Change the assignment of *is_signed in iioutils_get_type() to a one-liner,
as already done with *be.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Return directly, if no common cleanup is required.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Fully document public functions and elements.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Add error handling to calls which can indicate a major problem by
returning an error code.
This also sets ret to -ENOENT in iioutils_get_type() and
iioutils_get_param_float() to indicate if no matching directory entry was
found.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

This patch catches errors in string allocation in generic_buffer.c,
iio_event_monitor.c, iio_utils.c and lsiio.c.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Functions _write_sysfs_int() and _write_sysfs_string() are supposed to
be called only by public wrappers, so make them static.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Previously, the return value of sscanf() was treated as an indication of
the digits it would have read. Yet, sscanf() only returns the amount of
valid matches.
Therefore, introduce a function to calculate the decimal digits of the
read number and use this one to commence a colon search, as originally
intended.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

fscanf() usually returns the number of input items successfully matched
and assigned, which can be fewer than provided (or even zero).
Add a check in iioutils_get_type() to make sure all items are matched.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

The man-page of errno states, that errno should be saved before doing any
library call, as that call may have changed the value of errno. So, when
encountering any error, save errno first.
This patch affects generic_buffer.c, iio_event_monitor.c and iio_utils.c.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

In build_channel_array(), count needs to be decreased in more places since
current->name and current->generic_name would be freed on the error path,
although they have not been allocated, yet.
This also requires to free current->name, when it is allocated, but
current->generic_name is not yet allocated.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

In iio_utils.c build_channel_array() dynamically allocates the string
generic_name in the current iio_channel_info, which doesn't got freed in
case of an error.
This dynamically allocated channel-array is used by generic_buffer, and
needs to be freed on the error/exit path.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

In the error path, the string scan_el_dir got freed, while it was missing when
build_channel_array() finished without errors.
Signed-off-by: NHartmut Knaack <knaack.h@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

When using generic_buffer to read data, the scale is not properly
detected for scale shared by type. This is caused by a problem
with the generation of generic name out of the full name.
E.g.: for current->name in_accel_z, the extracted generic name
is "in" (when it should be "in_accel"). This is used in generic_buffer
to generate scale and offset paths (in_accel_scale).

Consider the in_ or out_ prefix when extracting the generic name
from the full name.
Signed-off-by: NIrina Tirdea <irina.tirdea@intel.com>
Reviewed-by: NDaniel Baluta <daniel.baluta@intel.com>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

This patch moves iio userspace applications out of staging, to tools/iio/
and adds a Makefile in order to compile them easily. It also adds tools/iio/
to MAINTAINERS file.
Signed-off-by: NRoberta Dobrescu <roberta.dobrescu@gmail.com>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

This patch removes inline functions from iio_utils.h in order to clean the
code. iio_utils.c contains the implementation of the functions used by
iio_event_monitor.c, lsiio.c or generic_buffer.c and iio_utils.h contains
the declarations of these functions.

Since iio_utils.h is modified, generic_buffer.c and iio_event_monitor.c
must include stdlib.h.
Signed-off-by: NRoberta Dobrescu <roberta.dobrescu@gmail.com>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

The following files were added define guards to prevent multiple
inclusion.
Signed-off-by: NCristina Opriceana <cristina.opriceana@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Fixes "Missing a blank line after declarations" checkpatch.pl warning in
Documentation/iio_utils.h
Signed-off-by: NMelike Yurtoglu <aysemelikeyurtoglu@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

spotted by cppcheck
Signed-off-by: NToralf Förster <toralf.foerster@gmx.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NManuel Stahl <manuel.stahl@iis.fraunhofer.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

This was originally reported by Craig Markwardt on Zubair Lutfullah's
blog and Zubair forwarded it to linux-iio@vger.kernel.org. No email
address known.

The code first counted the number of enabled channels, then created an
array to hold information about them.  The code that filled this array then
stored whether a given element was enabled inside the array.  Curriously
this element was never used.  Craig's patch added a local temporary variable
to avoid the buffer overrun.  Jonathan then removed the original enabled
element of the structure as it was not needed at all.
Signed-off-by: NZubair Lutfullah <zubair.lutfullah@gmail.com>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

The `ret´ variable is only initialized in the error case. For some reason
it was always != 0 while I played with generic_buffer so here is a patch.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Signed-off-by: NJonathan Cameron <jic23@kernel.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Acked-by: NJonathan Cameron <jic23@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

iio_utils.h:367:10: warning: ‘iioutils_break_up_name’ is static but used in inline function ‘build_channel_array’ which is not static [enabled by default]
Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Acked-by: NJonathan Cameron <jic23@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Acked-by: NJonathan Cameron <jic23@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NPeter Meerwald <pmeerw@pmeerw.net>
Acked-by: NJonathan Cameron <jic23@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Eating the endian description for now.
Signed-off-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

iio_utils.h uses opendir and friends which need dirent.h
Signed-off-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Sorry all, this one is very invasive, though the driver changes are
just trivial interface fixes. Not all done yet.

V2 - bring the sca3000 with us.
V3 - fix ade7758 bugs in conversion.
V4 - add ad5933
Signed-off-by: NJonathan Cameron <jic23@cam.ac.uk>
Acked-by: NMichael Hennerich <michael.hennerich@analog.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Trivial space before newline fix incorporated.

Additional fixes related to handling of sign extension and shifted
data.
Signed-off-by: NJonathan Cameron <jic23@cam.ac.uk>

squash into buffer handling update.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Fixes generated by 'codespell' and manually reviewed.
Signed-off-by: NLucas De Marchi <lucas.demarchi@profusion.mobi>

The previous implementation flawed, in case some channels are not enabled.
The sorted array would then include channel information of disabled channels,
And misses the enabled ones, when the count is reached.
More troublesome, the loop would not even terminate.

The fix is twofold:

First we skip channels that are not enabled.
Then we use a tested bubble sort algorithm to sort the array.
Since we already allocated exactly the number of bytes we need.
We can exercise bubble sort on the original memory.
In all cases I've seen, the array is already sorted, so this sort
terminates immediately.

Changes since V1:
Fix coding style issues.
Signed-off-by: NMichael Hennerich <michael.hennerich@analog.com>
Acked-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Variable sizeint is used uninitialized.
Remove sizeint completely and use bits_used instead.
Signed-off-by: NMichael Hennerich <michael.hennerich@analog.com>
Acked-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The first part of build_channel_array()identifies the number of enabled channels.
Further down this count is used to allocate the ci_array. The next section parses the
scan_elements directory again, and fills ci_array regardless if the channel is enabled or not.
So if less than available channels are enabled ci_array memory is overflowed.

This fix makes sure that we allocate enough memory. But the whole approach looks a bit
cumbersome to me. Why not allocate memory for MAX_CHANNLES, less say 64
(I never seen a part with more than that channels). And skip the first part entirely.
Signed-off-by: NMichael Hennerich <michael.hennerich@analog.com>
Acked-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

No functional changes
Signed-off-by: NMichael Hennerich <michael.hennerich@analog.com>
Acked-by: NJonathan Cameron <jic23@cam.ac.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>