The elan driver converts frames into a different format. These frames
are only needed to assemable the image and should be free'ed afterwards.
Fixes: #213
While timeout was already cleared for parent, we didn't properly delete the
cancellable.
Although we'd warn anyways when starting the SSM, is still better to clear
any delayed action also for the sub-SSM
In order to be able to test the private device code (used by drivers) we
need to have that split a part in a different .c file so that we can compile
it alone and link with it both the shared library and the test executables.
Redefine fp_image_device_get_instance_private for private usage, not to move
the private struct as part of FpDevice.
In order to be able to test the private device code (used by drivers) we
need to have that split a part in a different .c file so that we can compile
it alone and link with it both the shared library and the test executables.
Redefine fp_device_get_instance_private for private usage, not to move
the private struct as part of FpDevice.
A Fp-device use an union to track the handle to the lower-level device, and
the value depends on the object type.
So in case of using a virtual device, the priv->usb_device location matches
the priv->virtual_env string location, and thus we'd end up unreffing a
string location as it was a GObject, while we'd leak the string.
To avoid such errors, instead of just checking the device type when we
finalize the device, let's just use different pointers to avoid other
possible clashes.
This fixes the the problem that the sensor becomes unresponsive after
pressing the wrong fingerprint. We fix the problem by making sure that
the non-match report is delayed until the finger is removed. With this
we cannot run into the situation that the next match request fails
immediately as the finger is still present.
Fixes: #208
The FPI_MATCH_ERROR constant was set to 0, however it is propagated to
the task completion using g_task_propagate_int. As g_task_propagate_int
will always return -1 on error, we either need to add an explicit -1
check or we just need to match the semantics.
Change the constant to -1, also rearange FP_MATCH_SUCCESS so that it
does not end up being 0.
Add commands to disable automatic finger reporting for images and to
send a specific finger report. This is useful to test more code paths
inside the image-device code.
Add fpi_ssm_new_full() that allows to pass a name to the state-machine
constructor, not to change all the drivers, provide a fpi_ssm_new() macro
that stringifies the nr_parameters value (usually an enum value) as the name
so that we can use it for better debugging.
The IRQ handler will re-register itself automatically. However, if this
happens after the callback is called, then the check whether the IRQ
handler is running fails.
Re-start the IRQ handler before calling the callback. This way the state
changes happening from the callback will see the correct IRQ handler
registration state.
See: #205
At the end of enroll, the image device would put the driver into the
AWAIT_FINGER_ON state and then deactivate the device afterwards. Doing
this adds additional complexity to drivers which would need to handle
both cancellation and normal deactivation from that state.
Only put the device into the AWAIT_FINGER_ON state when we know that
another enroll stage is needed. This avoids the critical state
transition simplifying the driver state machine.
Fixes: #203
Fixes: 689aff0232
During calibration, the internal state was stored as INACTIVE. This is
not true though, the device is actively running state machines.
Move the state update to the start of the operation, therefore ensuring
we don't deactivate without completing the SSM.
Note that this will prevent a crash, but the driver still does not
behave quite correctly when such a state change does happen. However,
this is just a safety measure as the state change should not happen in
the first place.
See: #203
These were probably added in previous iterations, but they are not uneeded
anymore as the GSource embeds already a callback function.
So make just this clearer in the dispatch function.
Now that nbis is a static library it should be possible to compile it
without any fprint-built dependency, although since it included fp_internal
there was a compile-time dependency on the fp-enums that can be generated at
later times.
So:
- Move nbis-helpers to nbis includes (and remove inclusion in fp_internal)
- Move the Minutiae definitions inside a standalone fpi-minutiae header
- Include fpi-minutiae.h in fp_internal.h
- Include nbis-hepers.h and fpi-minutiae.h in nbis' lfs.h
- Adapt missing definitions in libfprint
The nbis headers are full of redundant declarations, we can't fix them all
now, so in the mean time let's use an header using pragma to ignore such
errors.
Add the error to nbis private include folder that should be used only by
headers of libfprint-nbis.
As nbis is an external source bundle, it does not necessarily make sense
to enable/fix all warnings to the extend we do for our own library code.
As such, separate the build process into multiple stages.
We were passing around the common cflags and setting them for each library
or executable, but this is just a repetition given we can just use
add_project_arguments for this.
Properly follow function signature using a temporary gsize variable address
to make the function use the same pointer type and avoid troubles at
deferencing it, while use automatic-casting to switch to signed one if
transfer succeeded.
Add a GCancellable parameter to fpi_ssm_nex_state_delayed and
fpi_ssm_jump_to_state_delayed() so that it's possible to cancel an action
from the caller and in case the driver wants to cancel a delayed operation
when a device action has been cancelled.
This allows to have an automatic cleanup of the timeout source when the
the callback is reached and to avoid to do further state changes in the
middle.
Since GSource data can be automatically cleaned up on source destruction, we
can mimic this for the devices timeout easily as well.
Add an extra parameter, and let's use this cocci file to adapt all the
drivers like magic:
@@
expression e1, e2, e3, e4;
@@
fpi_device_add_timeout (e1, e2, e3, e4
+ , NULL
)
When using fpi_usb_transfer_submit_sync we still need to unref the transfer
once done with it, so let's use an auto pointer so we free it also on
errors and early returns without having to handle this manually.
In rare occasions it could happen that the driver was reading
insufficient data. Fix this by using g_input_stream_read_all_async
which will ensure that incomplete data will not be misinterpreted.
This fixes rare test failures seen in fprintd.
Seems like the older uncrustify versions did not find these indentation
issues. Fix them.
Old versions of uncrustify will leave things as is, so this is not a
problem if developers are using an old version of uncrustify.
Using floating point causes architecture dependent results due to
accuracy/rounding differences. It is not hard to switch to fixed point,
and while this does cause quite different rounding errors, the
difference is small.
Fixes: #200
g_variant_new_from_data() allows to destroy some other user_data passed as
parameter that might be different from the aligned_data itself.
But since in this case they match, pass it to be set as g_free parameter
or it won't be free'd.
When we notify the enroll progress with a print, this needs to be unreffed
once we're done, but this only was happening in case of error.
Since it's not up to the callback function to free it, let's do it at the
end of the function.
As per this, clarify the docs for FpEnrollProgress marking it as transfer
none.
Allow to pass a double-pointer to be nullified as the transfer data in order
to mark it as NULL when the transfer is done.
This is useful if we're keeping the transfer around in order to check that
no one is currently running.
When a transfer is completed, we automatically unref it since we can't
consider it valid anymore since this point.
Update the drivers not to free the transfer after submitting anymore.
When a machine is completed, we automatically free it since we can't
consider it valid anymore since this point.
Update the drivers not to free the SSM on completion callback anymore.
As per commit 201b5a961 we use g_date_copy() to copy the date, however the
GLib implementation is done assuming that the GDate getters are always used
as the copy function doesn't preserve the original format of the date
(whether is using julian days or dmy), and the synaptics driver access to
the dmy values directly, without using the getter that would recompute the
proper values.
Causing a read error of unset values.
So, to avoid this, just use the g_date_get_* getters to retrieve the day
month and year for for defining the print enroll id.
This is the same logic we apply to fp-device by default: any completed
action should trigger the subsequent one when it is finished.
So in case we want reactivate after a deactivation, let's do it in an idle,
after removing the current pending timeout.
Meson files are normally using 4-spaces to indent and functions use first
parameter on the same line while others at next indentation level, not
following the parenthesis indentation.
So adapt libfprint to follow the meson standard.
Use the same approach of GTask, making possible to set the data from a
function. Givent the fact that a SSM has now a device parameter, it's
generally not needed to pass an extra data value.
In such case make it possible to set it and to define a destroy-notify
function to handle its destruction when freeing the SSM.
With the spatch to use GLib memory functions a lot of error paths were
removed. This causes an unused variable warning, so drop in a further
patch to remove the unused variable.
When drivers signal that an action is completed, they might be still in the
middle of a SSM and so its callback might not be called properly as part of
the completion.
This could make impossible to chain operations like open->enroll/list with
some drivers (hey, synaptics, I'm looking at you!) when the next operation
is called from the GAsyncReadyCallback of the previous call.
To avoid this to happen, ensure that when a driver completes an operation,
we handle the notification to the caller in a next idle iteration, not to
end up in a possible broken state.
Abstract this by using a function that handles the task return for each
used task type to avoid duplicating similar functions doing all the same
thing
We are actually using the cancel idle source in case the device supports
cancellation, so only connect to the cancellable in such case, and use an
utility function to do it and disconnect and reset the state everywhere.
fpi_do_movement_estimation is always called with num_stripes set to the
length of the list. Rather than using the passed value, assume we should
consume all stripes from the list.
Closes: #132
As the driver is not a normal image device, we need to add a custom
script to test it. Note that the ioctl dump must also be manually
modified unfortunately as the state is tracked incorrectly for the
device by umockdev-record.
This changes the cancellation logic a bit to ensure we always deactivate
the device (equivalent to the AWAIT_OFF state in the driver). All
commands except for the deactivation command should be cancelled when an
operation is stopped, this is to ensure that the LED is turned off at
the end of an operation.
This allows us to properly extract metadata for prints that are stored
on the device. We could for example delete the oldest prints first with
this information.
Heavily modified by Benjamin Berg <bberg@redhat.com> to port it to the
new libfprint API and adjust the coding style to follow more closely
other drivers.
When the device is deactivated while it is still active then the exit SM
needs to be executed from the SM that was active at the time. This is
signalled by is_active being set to FALSE while the active SM completes.
Call m_exit_start in those cases to ensure proper device deactivation.
This driver has a rather odd state machine and also used to mess iwth
the internal state of the image device. This code has been removed, but
is untested unfortunately due to a lack of hardware.
Most likely, this driver is not quite functional currently.
This is a rewrite of the core based on GObject and Gio. This commit
breaks the build in a lot of ways, but basic functionality will start
working again with the next commits.
Apply the previously added spatch/coccinell file to replace all
free/malloc/realloc calls with g_free/g_malloc/g_realloc. It also removes
all the error code paths that we do not need to check anymore.
This means that the returned data must be free'ed using g_free rather
than free, making memory management more consistent.
Add an spatch/coccinelle file to replace all free/malloc/realloc calls
with g_free/g_malloc/g_realloc. It also removes all the error code paths
that we do not need to check anymore.
We will use GUsb rather than libusb directly in the future. This should
simplify a lot of the integration work and changes such as supporting
hotplugging. It will also require quite a lot of internal changes.
If one of the callbacks called from fpi_imgdev_deactivate_complete()
was reactivating the device, then we would be overwriting whichever
"action" got set in the callback, leading to
fpi_imgdev_activate_complete() failing as it doesn't handle the "none"
action.
Reset the action before calling the callbacks.
If a USB transfer is started but not completed in one go, the wdata we
pass to continue_write_regv() will already be freed by the time we try
to use it again.
Only free() the wdata on error, or when the USB transfer is completed.
Closes: #180
The state was always AWAIT_FINGER and it was never used by any driver
(except for error checking). So remove it, in particular as a correct
state change will be done after activation anyway.
The only driver with code that actually did anything based on this was
the URU4000 driver. However, all it did was an explicit state change
execution. This is not necessary, as the state_change handler is called
anyway (i.e. we now only write the AWAIT_FINGER register once rather
than twice).
Manual changes plus:
@ init @
identifier driver_name;
identifier activate_func;
@@
struct fp_img_driver driver_name = {
...,
.activate = activate_func,
...,
};
@ remove_arg @
identifier dev;
identifier state;
identifier init.activate_func;
@@
activate_func (
struct fp_img_dev *dev
- , enum fp_imgdev_state state
)
{
<...
- if (state != IMGDEV_STATE_AWAIT_FINGER_ON) { ... }
...>
}
The driver was never ported to the new asynchronous model, meaning it
has been defunct since some time in 2008. Remove the driver, as
seemingly no one has complained about this and we have no proper way to
even verify a port is correct.
The upekts driver needs upek_proto.c while the upektc driver does not.
Move the corresponding source file entries so that both drivers compile
standalone.
The only API user currently seems to be the examples. fprintd has its
own storage and that will be a good idea in general.
So deprecate the API, we'll need to find a different solution for the
examples eventually.
The function was committed by accident as part of commit d18e1053
(lib: Add a way to name timeouts). It is not used anywhere and
fpi_timeout_cancel_all_for_dev exists, is exported and used and serves
the same purpose.
If users put their finger on the sensor between the bulge and
"un-bulge" area first and then swipe, the captured image would
be bad.
Skipping more frames can reduce the impact, so bump
ELAN_SKIP_LAST_FRAMES to 2.
Check for the mean calibration being outside of range to know whether we
require a recalibration. Continue with the usual checks if the
calibration value is within range.
The dimensions some sensors return is the maximum zero-based index
rather than the number of pixels. Assuming every sensor has an
even number of pixels is safe.
This commit fixes a hang in gnome-settings when trying to enroll a finger.
The same issue could be seen in the enroll example. Previously the enroll
example would hang on "deactivating" because at some point dev->is_active
is set to false and m_exit_start is never called.
Work-around SELinux AVC warnings caused by p11-kit (which is an NSS
dependency) trying to load the root user's p11-kit configs. We disable
this feature using the P11_KIT_NO_USER_CONFIG envvar.
See https://bugzilla.redhat.com/show_bug.cgi?id=1688583
‘img->key_number’ variable is originally from the device through bulk
endpoint of USB. The variable is immediately assigned to ‘buf[0]’ for
sending to control endpoint of the device. Here, integer overflow may
occur when the ‘img->key_number’ attempts to assign a value that is
outside of type range of ‘char’ to the ‘buf[0]’
Remove the sanity check added in commit b1afa9d, having a poll setup
isn't necessary to use timeouts, as long as only sync functions are
used.
Closes: #139
This will allow us to segragate helper functions that we want to use
within NBIS. The first helper will be a macro checking for overflow in
multiplications.
Free the maps if we're not returning them.
libfprint/nbis/mindtct/maps.c:176:14: warning: Potential leak of memory pointed to by 'direction_map'
return(ret);
^~~
libfprint/nbis/mindtct/maps.c:195:14: warning: Potential leak of memory pointed to by 'low_contrast_map'
return(ret);
^~~
libfprint/nbis/mindtct/maps.c:195:14: warning: Potential leak of memory pointed to by 'low_flow_map'
return(ret);
^~~
libfprint/nbis/bozorth3/bozorth3.c:704:1: warning: Value stored to 'kx' is never read
kx = 0;
^ ~
libfprint/nbis/bozorth3/bozorth3.c:820:4: warning: Value stored to 'l' is never read
l = 1;
^ ~
The first contour wasn't freed if we failed on the second contour.
libfprint/nbis/mindtct/contour.c:521:14: warning: Potential leak of memory pointed to by 'half1_ex'
return(ret);
^~~
libfprint/nbis/mindtct/contour.c:521:14: warning: Potential leak of memory pointed to by 'half1_ey'
return(ret);
^~~
libfprint/nbis/mindtct/contour.c:521:14: warning: Potential leak of memory pointed to by 'half1_x'
return(ret);
^~~
libfprint/nbis/mindtct/contour.c:521:14: warning: Potential leak of memory pointed to by 'half1_y'
return(ret);
^~~
libfprint/drivers/elan.c:351:4: warning: 2nd function call argument is an uninitialized value
dbg_buf(elandev->last_read, transfer->actual_length);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:46:5: note: expanded from macro 'dbg_buf'
fp_dbg("%02x", buf[0]); \
^~~~~~~~~~~~~~~~~~~~~~
../../../../../../Projects/jhbuild/libfprint/libfprint/fpi-log.h:52:16: note: expanded from macro 'fp_dbg'
#define fp_dbg g_debug
^
libfprint/drivers/elan.c:351:4: warning: The left operand of '<<' is a garbage value
dbg_buf(elandev->last_read, transfer->actual_length);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:48:27: note: expanded from macro 'dbg_buf'
fp_dbg("%04x", buf[0] << 8 | buf[1]); \
~~~~~~ ^
libfprint/drivers/elan.c:351:4: warning: The left operand of '<<' is a garbage value
dbg_buf(elandev->last_read, transfer->actual_length);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:50:41: note: expanded from macro 'dbg_buf'
fp_dbg("%04x... (%d bytes)", buf[0] << 8 | buf[1], len)
~~~~~~ ^
By asserting if we have a zero line width.
libfprint/drivers/upeksonly.c:118:7: warning: Division by zero
mean /= (ctx->line_width / 2);
~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
We don't need to assign urudev if we only want the size of one of its
members.
libfprint/drivers/uru4000.c:554:20: warning: Value stored to 'urudev' during its initialization is never read
struct uru4k_dev *urudev = FP_INSTANCE_DATA(FP_DEV(dev));
^~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Use GPtrArray to simplify the implementation.
libfprint/fpi-data.c:777:13: warning: Access to field 'data' results in a dereference of a null pointer (loaded from variable 'elem')
list[i] = elem->data;
^~~~~~~~~~
libfprint/fpi-data.c:497:7: warning: 2nd function call argument is an uninitialized value
if (!fp_dev_supports_print_data(dev, fdata)) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/fpi-poll.c:398:11: warning: The left operand of '==' is a garbage value
else if (timercmp(&fprint_timeout, &libusb_timeout, <))
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/include/sys/time.h:162:17: note: expanded from macro 'timercmp'
(((a)->tv_sec == (b)->tv_sec) ? \
~~~~~~~~~~~ ^
libfprint/fpi-ssm.c:244:9: warning: Access to field 'completed' results in a dereference of a null pointer (loaded from variable 'machine')
BUG_ON(machine->completed);
^~~~~~~~~~~~~~~~~~
libfprint/fpi-log.h:84:6: note: expanded from macro 'BUG_ON'
if (condition) { \
^~~~~~~~~
See dda6857fee
and https://bugzilla.redhat.com/show_bug.cgi?id=1656518
When the number of lines to assemble is 1, the median_filter() function
would be passed -1 as its size as it was calculated with:
(num_lines / 2) - 1
so (1 / 2) - 1 = 0 - 1 = -1
Add a guard to stop drivers trying to assemble single lines. This
doesn't however fix the vfs5011 driver that tried to do that.
...and add "Returns" docs to fpi-dev.c
We might need to do this more, documenting what the function does in
prose, as well as a short explanation of the returned data.
If get_next_timeout_expiry() fails, and libusb_get_next_timeout()
has no timeouts or failed, fprint_timeout will not be set, and
we cannot compare it to libusb_timeout.
Exit early if both failed or have empty queues.
libfprint/drivers/elan.c:351:12: warning: format specifies type 'unsigned short' but the argument has type 'unsigned char' [-Wformat]
dbg_buf(elandev->last_read, transfer->actual_length);
~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:46:21: note: expanded from macro 'dbg_buf'
fp_dbg("%02hx", buf[0]); \
~~~~~~~~~~~~~~~~^~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/elan.c:351:12: warning: format specifies type 'unsigned short' but the argument has type 'int' [-Wformat]
dbg_buf(elandev->last_read, transfer->actual_length);
~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:48:21: note: expanded from macro 'dbg_buf'
fp_dbg("%04hx", buf[0] << 8 | buf[1]); \
~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/elan.c:351:12: warning: format specifies type 'unsigned short' but the argument has type 'int' [-Wformat]
dbg_buf(elandev->last_read, transfer->actual_length);
~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/drivers/elan.c:50:35: note: expanded from macro 'dbg_buf'
fp_dbg("%04hx... (%d bytes)", buf[0] << 8 | buf[1], len)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/elan.c:413:10: warning: format specifies type 'unsigned short' but the argument has type 'unsigned char' [-Wformat]
dbg_buf(cmd->cmd, 2);
~~~~~~~~^~~~~~~~~~~~
libfprint/drivers/elan.c:46:21: note: expanded from macro 'dbg_buf'
fp_dbg("%02hx", buf[0]); \
~~~~~~~~~~~~~~~~^~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/elan.c:413:10: warning: format specifies type 'unsigned short' but the argument has type 'int' [-Wformat]
dbg_buf(cmd->cmd, 2);
~~~~~~~~^~~~~~~~~~~~
libfprint/drivers/elan.c:48:21: note: expanded from macro 'dbg_buf'
fp_dbg("%04hx", buf[0] << 8 | buf[1]); \
~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/elan.c:413:10: warning: format specifies type 'unsigned short' but the argument has type 'int' [-Wformat]
dbg_buf(cmd->cmd, 2);
~~~~~~~~^~~~~~~~~~~~
libfprint/drivers/elan.c:50:35: note: expanded from macro 'dbg_buf'
fp_dbg("%04hx... (%d bytes)", buf[0] << 8 | buf[1], len)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
include/glib-2.0/glib/gmessages.h:345:32: note: expanded from macro 'g_debug'
__VA_ARGS__)
^~~~~~~~~~~
libfprint/drivers/vfs101.c:854:6: warning: absolute value function 'abs' given an argument of type 'long' but has parameter of type 'int' which may cause truncation of value [-Wabsolute-value]
if (abs(count - VFS_IMG_BEST_CONTRAST) < abs(vdev->best_clevel - VFS_IMG_BEST_CONTRAST))
^
libfprint/drivers/vfs101.c:854:6: note: use function 'labs' instead
if (abs(count - VFS_IMG_BEST_CONTRAST) < abs(vdev->best_clevel - VFS_IMG_BEST_CONTRAST))
^~~
labs
libfprint/fpi-dev-img.c:255:6: warning: variable 'print' is used uninitialized whenever 'if' condition is false [-Wsometimes-uninitialized]
if (imgdev->action != IMG_ACTION_CAPTURE) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/fpi-dev-img.c:271:25: note: uninitialized use occurs here
imgdev->acquire_data = print;
^~~~~
libfprint/fpi-dev-img.c:255:2: note: remove the 'if' if its condition is always true
if (imgdev->action != IMG_ACTION_CAPTURE) {
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
libfprint/fpi-dev-img.c:232:29: note: initialize the variable 'print' to silence this warning
struct fp_print_data *print;
^
= NULL
This reverts commit 700c5791f8.
We don't need a poll setup to be able to call sync functions, which then
use async functions for implementation internally.
Closes: #124
Conflicts:
libfprint/fpi-async.c
struct xyt_struct uses a fixed-sized array to fit MAX_BOZORTH_MINUTIAE (200)
minutiae. MAX_FILE_MINUTIAE is 1000. So if we detected more than
MAX_BOZORTH_MINUTIAE, we would crash copying the data from the capture
to the structure.
We might want to use dynamically allocated arrays in the future (or
bigger ones) so that we don't lose minutiae.
Closes: #116
The checks weren't:
- checking whether the width or height were negative
- whether img->width * img->height would overflow, or
was bigger than G_MAXINT
- whether img->width * img->height was bigger than the total
length of the buffer
The last one looks like a thinko, it checked for:
(img->length * img->height) < img->length
which is equivalent to:
img->height < 1
which we already check for earlier.
Closes: #85
Add fp_minutia_get_coords() so that debugging applications can show the
detected minutiae along with the captured image.
This will also help fix the positively ancient fprint_demo.
See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=907470
Group global variables, structure definitions, and functions, and
group functions by which C source they're defined in.
This also removes the FP_IMGDRV_SUPPORTS_UNCONDITIONAL_CAPTURE
definition that's already available in the drivers API-accessible
fpi-core.h
Pass the struct fp_dev and user_data to fpi_ssm callbacks, so that we
might be able to get rid of the fpi_ssm_get_user_data(), and
fpi_ssm_get_dev() as most drivers just get those from the ssm anyway
in their callbacks.
We can access the img_dev struct member directly, and we don't have
to double check that the img_dev member is actually an img_dev type,
as it always is.
Not precise enough, and it was only used in upekts to get/set the
instance private struct and keep it around, so add a new, more precisely
defined field to fp_dev for that.
The enum values need to be documented before the enum itself, otherwise
this warning is thrown:
warning: Value descriptions for fp_capture_result are missing in source code comment block.
mindtct/loop.c: In function ‘process_loop_V2’:
mindtct/loop.c:713:8: warning: variable ‘halfway’ set but not used [-Wunused-but-set-variable]
int halfway;
^~~~~~~
mindtct/morph.c: In function ‘get_south8_2’:
mindtct/morph.c:173:4: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
if (row >= ih-1) /* catch case where image is undefined southwards */
^~
mindtct/morph.c:176:7: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’
return *(ptr+iw);
^~~~~~
mindtct/morph.c: In function ‘get_north8_2’:
mindtct/morph.c:197:4: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
if (row < 1) /* catch case where image is undefined northwards */
^~
mindtct/morph.c:200:7: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’
return *(ptr-iw);
^~~~~~
mindtct/morph.c: In function ‘get_east8_2’:
mindtct/morph.c:221:4: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
if (col >= iw-1) /* catch case where image is undefined eastwards */
^~
mindtct/morph.c:224:7: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’
return *(ptr+ 1);
^~~~~~
mindtct/morph.c: In function ‘get_west8_2’:
mindtct/morph.c:243:4: warning: this ‘if’ clause does not guard... [-Wmisleading-indentation]
if (col < 1) /* catch case where image is undefined westwards */
^~
mindtct/morph.c:246:7: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the ‘if’
return *(ptr- 1);
^~~~~~
mindtct/block.c: In function ‘block_offsets’:
mindtct/block.c:108:18: warning: variable ‘ph’ set but not used [-Wunused-but-set-variable]
int pad2, pw, ph;
^~
NBIS is pretty complicated to update, seeing as we've made quite a few
changes to get it to compile as a library. With those scripts, we can
easily trim headers to remove functions we don't use, rename global
variables, and do any sort of fixups that are necessary right now.
In the future, removing unused NBIS functions might be as easy as
updating that script, re-running it, and pushing the changes.
Note that remove-function.lua is a very crude parser that only supports
NBIS' style of declaration, with the return type on the same line as the
function name. I wouldn't recommend trying to use it in another project.
Callcatcher (https://github.com/caolanm/callcatcher) was also used to
remove additional unused functions.
There has been a report that a 0x0c03 was installed straight (other
readers so far have produced images that are 90 degrees rotated). There
is no way for the dirver to know how a device is installed, so for now just
make an exception.
Sensor reset code has been removed because it is not needed during normal
operation.
Calibration and frame processing logic has been improved according to
recommendations from Elantech.
Make each command a separate struct to get finer control over which
commands are called on which devices. Update ssm's accordingly. Add
sensor_reset and fuse_load commands.
Support 0x0903, 0x0c03, 0x0c16, 0x0c1a, 0x0c26
Similarly to b1ac865abd, downgrade
fp_err() to be non-fatal. A number of drivers would spit out an error
when encountering this call, but not crash, carry on and most of the
time recover.
Make sure we don't assert in those cases.
BUG() and BUG_ON() didn't use to assert, but only print an error if
debugging was enabled. This was hiding a lot of state bugs in drivers,
and transforming those into assertions causes crashes.
Downgrade the assertion to only print a warning, and hope that those
eventually get fixed in the drivers so we can re-enable them.
Closes: #77
The driver will at least need to close its hardware resources, and
free memory, so it must have had one.
This case was never actually used as can be seen from the fact that
we would assert in fpi_drvcb_close_complete() if the state was wrong
but never set it to the expected value.
Which avoids passing zero lines to fpi_assemble_lines()
"gmem.c:130: failed to allocate 18446744073709551612 bytes"
#3 0x00007fe4f6ef428f in g_log (log_domain=log_domain@entry=0x7fe4f6f3506e "GLib", log_level=log_level@entry=G_LOG_LEVEL_ERROR, format=format@entry=0x7fe4f6f3e610 "%s: failed to allocate %lu bytes") at gmessages.c:1398
#4 0x00007fe4f6ef2ac4 in g_malloc0 (n_bytes=n_bytes@entry=18446744073709551612) at gmem.c:129
#5 0x00007fe4f8052020 in median_filter (filtersize=25, size=-1, data=0x0) at assembling.c:309
#6 fpi_assemble_lines (ctx=ctx@entry=0x7fe4f82ac3c0 <assembling_ctx>, lines=0x0, lines_len=0) at assembling.c:389
#7 0x00007fe4f805f3db in submit_image (ssm=ssm@entry=0x16c3cba360, data=data@entry=0x16c3cb9cc0) at drivers/vfs5011.c:412
See https://bugzilla.redhat.com/show_bug.cgi?id=1484812Closes: #42