Add API variants to support filling-colour-with-mask operations for mirroring the mask first

[GorgonMeducer]

The existing filling-colour-with-mask operations do not support mirroring the mask.

by

• Adding API varients for filling-colour-with-mask-and-mirroring
• Adding API varients for filling-colour-with-mask-mirroring-and-opacity

We

• no longer need an intermedia A8 buffer for mirroring the mask first(hence save SRAM)
• finish the operation in just one step rather than two (hence improve the overall performance)

[SamSamYou]

Thanks to Arm-2D, my project achieves smooth animations under tight resource conditions(168Mhz Cortex-M4 64k-RAM 64k-Flash for 160x160 rgb565 lcd display). I have a symmetrical icon that needs to be displayed, but it is limited by MCU resources, and I need a more efficient rendering method to achieve smooth (~30fps) animation display.If I can use filling_colour_with_a2_mask_and_mirroring(X/Y/XY mirror),it can save SRAM and improve the overall performance of the system.Thank you for all you have done to bring the GUI to more devices.

[GorgonMeducer]

This feature is supported.