Commit graph

1025 commits

Author SHA1 Message Date
Benjamin Berg 8cc0fd321f assembling: Use fixed point for image assembly
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
2019-11-28 20:41:45 +00:00
Benjamin Berg a7541b1f76 tests: Fix endianness issue in test suite
The test suite needs to compare greyscale images and was picking an
undefined byte in the pixel data on big-endian. Select a byte that works
on any endian instead.

See: #200
2019-11-28 20:41:45 +00:00
Marco Trevisan (Treviño) b9ff75c4e9 fp-print: Set the aligned_data as the data used by the cleanup function
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.
2019-11-28 21:30:17 +01:00
Marco Trevisan (Treviño) 4447a0d183 ci: Add a test case where we run tests with valgrind 2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 545af23536 tests: Use a loop for generating drivers tests and use suites
So we can just run drivers tests with --suite=drivers
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) db905a2048 fp-device: Use g_clear_error instead of check + free 2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 712853d1e3 fp-device: Mark user data in FpEnrollProgress as transfer none
The data has its own DestroyNotify set, so while no generic DestroyNotify
exists for generic data, let's make it clear.
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) c6298ede72 fp-device: Unref the print once we've notified the progress
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.
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) b1d99e7608 fp-device: Use an autopointer and steal the print when passed
Make it clearer that we're stealing the print when passing it away, instead
of just doing this silently.
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 5927a205e3 virtual-image: Also unref the object when closing a the stream
While a stream is closed when completely unreffed, the other way around
isn't true, so always unref the object.
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 8c05f3b78c fp-print: Unref print data and get static strings when deserializing 2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 92db82e3d4 fp-print: Assert the prints aren't set when initialized 2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) f6f689f9cd fp-print: Unref the prints on finalize 2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 5dc3edf07c fp-context: Run dispose on the usb context to deal with circular refs
Ensure that we dispose the USB context before unreffing it, so that it will
release any reference it has and destroy the internal libusb context.
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) 71625ec1cf fp-device: Unref the usb device on finalize
Each device adds a ref to the underlying usb device, but it doesn't remove
the reference on finalization.

So clear the object to fix the leak
2019-11-27 21:40:44 +01:00
Marco Trevisan (Treviño) c9216cf96c tests: Add setup mode to run tests using valgrind
In such case we need to ignore the python errors, so including a suppression
file when using --setup=valgrind.
2019-11-27 21:40:43 +01:00
Marco Trevisan (Treviño) 53713c0098 tests: Add 'gdb' setup to run tests using gdb
When using --setup=gdb the tests will be running using gdb, however this
as per meson limitation allows only running a test when using verbose mode.
2019-11-27 21:40:43 +01:00
Marco Trevisan (Treviño) 222c33ec32 virtual-image: Re-run the test using the defined wrapper if any
In case a LIBFPRINT_TEST_WRAPPER is defined, execute again the script using
the same python processor but using the passed wrapper command.
2019-11-27 21:40:43 +01:00
Marco Trevisan (Treviño) 19a50cfdc3 umockdev-test: Make possible to use a wrapper to run tests
Support LIBFPRINT_TEST_WRAPPER env variable to run tests with a wrapper.
2019-11-27 21:40:43 +01:00
Marco Trevisan (Treviño) 9892eb1c03 fpi-ssm: Make clearer that data is unused in fpi_ssm_usb_transfer_cb 2019-11-27 21:02:20 +01:00
Marco Trevisan (Treviño) 587131a6bd drivers: Use more fpi_ssm_usb_transfer_cb when possible
Replace all the transfer callbacks where we just switch to the next state or
fail with fpi_ssm_usb_transfer_cb.
2019-11-27 21:02:20 +01:00
Marco Trevisan (Treviño) 65d0d5e3e0 fpi-ssm: Add a usb transfer callback with data as weak pointer
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.
2019-11-27 21:02:19 +01:00
Marco Trevisan (Treviño) 2642fc6560 fpi-usb-transfer: Take ownership of the transfer when submitting it
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.
2019-11-27 21:02:19 +01:00
Marco Trevisan (Treviño) a855c0cc79 fpi-ssm: Take ownership of the SSM when completing it
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.
2019-11-27 21:02:19 +01:00
Marco Trevisan (Treviño) 876924df6a examples: Handle the cases where the print date is not set 2019-11-27 19:14:35 +01:00
Marco Trevisan (Treviño) 519b5acc91 synaptics: Initialize user_id autoptr to NULL 2019-11-27 18:54:21 +01:00
Marco Trevisan (Treviño) e812653acd synaptics: Use GDate getters to retrieve the DMY values
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.
2019-11-27 18:54:21 +01:00
Marco Trevisan (Treviño) 1b23f0efe1 drivers: Use clearer messages using parameters 2019-11-26 12:18:57 +00:00
Marco Trevisan (Treviño) f6559ba8b1 fp-device: Support variadic arguments to error functions
Make possible to generate a formatted message when creating an error from
a device, without having save it first.
2019-11-26 12:18:57 +00:00
Marco Trevisan (Treviño) 3b72b925b0 gitlab-ci: Check indentation in an initial check-source stage 2019-11-25 19:08:31 +01:00
Marco Trevisan (Treviño) 15d218a112 fpi-log: Set fp_error as equal to g_critical 2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 0a08a24896 fpi-ssm: Remove any reference to fpi_timeout_add()
This doesn't exist anymore, while fpi_device_add_timeout does exists.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) cca6d3b04b fp-image-device: Use a GObject signal to notify image state changed
This is more GObject-friendly and we have the automatic call of the vfunc if
one is set.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) be367988ae fp-image-device: Add private "fp-image-device-state" property
In this way drivers may get this without having to keep a copy of it
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) ea4da08af0 fp-image-device: Reactivate in idle on deactivation completed
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.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 60ad1ab9e3 fp-image-device: Clear the pending activation timeout on finalize 2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 201b5a9614 fp-print: Use g_date_copy
As per previous commit we depend on glib 2.56, we can use this utility
function as well.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 8b270141f3 image-device: Use g_clear_handle_id for timeouts
As per this depend on glib 2.56: it has been released almost 2 years ago,
I suppose we're fine with that.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) ceb62d7617 meson: Avoid repeating the needed glib version multiple times
Just define once and modify its syntax when needed.
Use a more verbose definition for the min/max version (instead of just
join the split version) so that in case we may depend on a specifc glib
micro release during development.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 099fa9f005 meson: Use preferred syntax everywhere
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.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) dd7d1baece meson: Use multiline-array for default dirvers listing
It will make reviews and diffs nicer to handle when adding new drivers.
2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) d8efa336e5 fpi-ssm, fp-device: Add missing copyright 2019-11-25 18:59:46 +01:00
Marco Trevisan (Treviño) 76dd4066f3 verify: Ensure we set set the autoptr value to NULL at definition 2019-11-25 18:59:46 +01:00
Benjamin Berg 9b48864c5b print: Ensure xyt struct is not leaked during deserialization
In the unlikely event of an error, the variable may have been leaked.
Fix this by using g_autoptr combined with a g_steal_pointer.
2019-11-25 18:46:14 +01:00
Benjamin Berg 14a41bdd48 print: Free temporary col variable
The variable was leaked during serialization. Free it.
2019-11-25 18:46:14 +01:00
Benjamin Berg 25bc89a4f5 image-device: Remove unused fpi_device_get_current_action call
There is a later call in the function which is sufficient. Simply remove
the first call.
2019-11-25 18:46:14 +01:00
Benjamin Berg 2f0824ab88 upeksonly: Add default clauses to switch statements
This effectively only annotates the code to make it clear that variables
set in the switch are always initialized.
2019-11-25 18:46:14 +01:00
Benjamin Berg 8ba6f4dad2 synaptics: Add an explicit assert on the response
The response must be non-NULL in the function. Add an explicit assert to
appease to static code analysis tools.
2019-11-25 18:46:14 +01:00
Benjamin Berg ada5d488fa synaptics: Correctly unref pointer array
The pointer arrays were unref'ed using g_ptr_array_free rather than
g_ptr_array_unref from g_clear_pointer. Switch to the correct function.
2019-11-25 18:46:14 +01:00
Benjamin Berg b16245ad58 elan: Fix switch in change_state
The switch in change_state had a useless break and a useless if clause.
Remove both.
2019-11-25 18:46:14 +01:00