I'd prefer to move to C99 and use dynamic array sizes than use alloca, just cause it seems more conventional.
The s_grail_class_map array has a bunch of changes to unaffected gesture types. It appears there's a comma added near the end of each line. I think the comma is superfluous and should be removed so there's no change.
grail_mask_clear could just use memset, which would make the code more readable.
Why does grail_mask_setbit only concern itself with the lower 5 bits of each uint32_t?
Although grail exposes velocity attributes for touch gestures, I don't want to expose the attributes through GEIS. I propose masking them out.
I'd prefer to move to C99 and use dynamic array sizes than use alloca, just cause it seems more conventional.
The s_grail_class_map array has a bunch of changes to unaffected gesture types. It appears there's a comma added near the end of each line. I think the comma is superfluous and should be removed so there's no change.
grail_mask_clear could just use memset, which would make the code more readable.
Why does grail_mask_setbit only concern itself with the lower 5 bits of each uint32_t?
Although grail exposes velocity attributes for touch gestures, I don't want to expose the attributes through GEIS. I propose masking them out.
Looks good otherwise!