Some buttons not mapped correctly

[jeffythedragonslayer]

I think some of the KEY_ constants are not mapped correctly. I added a .proc to wait until all controller buttons are unpressed in this repo:

https://github.com/jeffythedragonslayer/lorom-template-with-waitpadup

Here is the control flow when holding down, and then releasing each of the buttons.

START		loops at wait_button_start, then fallsthrough the rest
SELECT(back)	loops at wait_button_select, then fallsthrough the rest
A		loops at wait_button_b, then fallsthrough
B		fallsthrough everything
X		loops at wait_button_y, then fallsthrough the rest
Y		fallsthrough everything
LEFT		loops at wait_button_up, then fallsthrough the rest
RIGHT		loops at wait_button_right, then fallsthrough the rest
UP		loops at up, then fallsthrough the rest
DOWN		loops at wait_button_down, then fallsthrough the rest
RBUMP		fallsthrough everything
LBUMP		fallsthrough everything

Tested on Mesen-SX with an XOTICPC GX7 gamepad. Also, I suggest renaming these constants to BUTTON_

[jeffythedragonslayer]

The XOTICPC GX7 gamepad has A swapped with B and X swapped with Y compared to a snes controller.

[pinobatch]

Your underlying bug at wait_pad_allup in revision 7898970 is that you are using the high byte address (cur_keys, defined as JOY1CUR + 1, which is $4219) with constants intended for the low byte (JOY1CUR, which is $4218). To give other users of this template a place I can point should they make the same mistake, I will add a new set of high byte constant definitions to snes.inc.

[jeffythedragonslayer]

Test results on an 8bitdo: (from before you committed d8770ad)

UP		loops at wait_button_up, then fallsthrough the rest
LEFT		loops at wait_button_left, then fallsthrough the rest
RIGHT		loops at wait_button_right, then fallsthrough the rest
SELECT		loops at wait_button_select, then fallsthrough the rest
START		loops at wait_button_start, then fallsthrough the rest
DOWN		loops at wait_button_down, then fallsthrough the rest
LBUMP		fallsthrough everything
RBUMP		fallsthrough everything
A		fallsthrough everything
B		loops at wait_button_b, then fallsthrough everything
X		fallsthrough everything
Y		loops at wait_button_y, then fallshtrough the rest

[pinobatch]

That's more like it. All 8 bits in the high byte (which correspond to the buttons of the NES controller) are causing the expected. Is there a remaining problem?