Unpack the pixel
A 32-bit colour packs four 8-bit channels into one word, laid out 0xAARRGGBB: alpha (how solid the pixel is) in the top byte, then red, then green, and blue in the lowest byte.
The starter reads a colour typed in hex into pixel_r. Pull each channel into its own register with ubfx, short for unsigned bitfield extract. For example, ubfx w9, w10, 4, 12 copies the 12 bits of w10 that start at bit 4 down to the bottom of w9 and fills the rest of w9 with zeros.
Then make swapped_r a copy of the colour with its red and blue bytes traded, using bfi, short for bitfield insert. For example, bfi w9, w10, 20, 4 copies the low 4 bits of w10 into bits 20 to 23 of w9 and leaves every other bit of w9 as it was.
With the sample input 4a90e2ff the output is:
alpha 74, red 144, green 226, blue 255
swapped 0x4affe290
What is checked
- both lines for the sample input
- the program exits cleanly
- the channels come out with
ubfxand go back in withbfi - the right answer for other colours, including all zeros, all ones, and a colour whose top bit is set
specification
ubfxbfiWe run your program on the input above and on the hidden ones, and compare what it does. Nothing is matched against a stored solution.