This fixes the following defects reported by covscan tool:
- client/Wayland/wlf_input.c:251: overrun-local: Overrunning array "contacts" of 10 32-byte elements at element index 10 (byte offset 351) using index "i" (which evaluates to 10).
- client/Wayland/wlf_input.c:308: overrun-local: Overrunning array "contacts" of 10 32-byte elements at element index 10 (byte offset 351) using index "i" (which evaluates to 10).
- client/Wayland/wlf_input.c:360: overrun-local: Overrunning array "contacts" of 10 32-byte elements at element index 10 (byte offset 351) using index "i" (which evaluates to 10).
- client/Wayland/wlf_input.c:251: error[arrayIndexOutOfBounds]: Array 'contacts[10]' accessed at index 10, which is out of bounds.
- client/Wayland/wlf_input.c:308: error[arrayIndexOutOfBounds]: Array 'contacts[10]' accessed at index 10, which is out of bounds.
- client/Wayland/wlf_input.c:360: error[arrayIndexOutOfBounds]: Array 'contacts[10]' accessed at index 10, which is out of bounds.
- client/Wayland/wlf_input.c:246: uninit_use_in_call: Using uninitialized value "y" when calling "wlf_scale_coordinates".
- client/Wayland/wlf_input.c:246: uninit_use_in_call: Using uninitialized value "x" when calling "wlf_scale_coordinates".
The maximal number of touches can be higher then 10, see:
https://wayland.freedesktop.org/libinput/doc/latest/touchpads.html
Let's increse the MAX_CONTACTS count and add checks to prevent usage of
uninitialized values.
Thanks to @yol and @SaschaWessel a bug in scroll step conversion
was uncovered. The RDP value ranges are inverted when scrolling
in negative direction.
In `wlf_cliprdr_server_format_data_request()` `ConvertToUnicode()` may return 0
while not allocating memory for `cdata` and not setting `cdata` to a valid
address. In this case, `data` points to a random address.
Settting `cdata` to NULL before calling `ConvertToUnicode()` fixes this issues.
According to the channel docs, this field is only used in format data
request. Therefore, there's no need to hold it in the response. cliprdr
server code was copy-pasted from client code, therefore this must be
some leftover.
The compositor only samples rects that we damage (and have therefore
been overwritten). As we are doing damage tracking and reporting,
memcpy of our old drawing buffer serves no purpose.
For future GFX channel functions an image scaling function is required.
This moves the implementation from wayland client to core library
and adds support for the much faster SWScale library.
The file clipboard delegate needs a base URI to operate on for
systems that are not WIN32. Added that to the context and abort
conversion, if that is not set. (currently not fully implemented)
found by cppcheck
[client/Wayland/wlf_cliprdr.c:143] -> [client/Wayland/wlf_cliprdr.c:158]: (warning) Either the condition 'clipboard' is redundant or there is possible null pointer dereference: clipboard.
[client/Wayland/wlf_cliprdr.c:163] -> [client/Wayland/wlf_cliprdr.c:178]: (warning) Either the condition 'clipboard' is redundant or there is possible null pointer dereference: clipboard.
The wayland buffer size always matches the current window size.
That might be different from the actual remote framebuffer size,
to when copying always use the correct strides to avoid distorted
screen content.
