[diegoparrilla]

What

A compile-time choice for the app between two ways of sharing the ST's time between the frame copy and the sound (APP_PROFILE, rp/src/include/profile.h):

	PROFILE_50FPS (default)	PROFILE_25FPS
Frames	every VBL, 50 fps	every second VBL, 25 fps on every machine
YM (plain ST, Mega ST)	5,585 Hz	21,943 Hz (Timer-B count 28)
DMA chip (STE, Mega STE)	12,517 Hz	25,033 Hz
Suits	arcade games, full-screen effects	story games with rich sound (laserdisc-style)

APP_PROFILE=PROFILE_25FPS ./build.sh pico_w release <uuid>

• One ST image for both profiles. fb_init() writes the profile into shared slot 8 ($FA4030), and userfw reads it once at boot. From it, the ST sets Timer-B's count, the DMA chip's mode and lead, and the VBLs per frame. A new frame is taken no sooner than that many VBLs after the last one. The constants of both profiles are named on both sides, and test_layout.py checks that they agree.
• Bigger audio buffers.
  • The audio buffer grows to 4 KB, with YM slices of 1 KB (878 bytes per VBL at 21,943 Hz).
  • The DMA ring grows to 4 KB at $6F000, with its position reported by 16 bytes.
  • APP_FREE now starts at $FA5180 (about 12.4 KB).
• Sound assets are unchanged. YM sources (.YMS files, the built-in jingle) stay at 5,585 Hz and are resampled to the output rate.
• RAM. The two outputs' FIFOs share their memory, which keeps the 25 fps debug build within the RP's RAM.
• Build IDs. A 25 fps build's ID ends in +25fps, and tools/dev/flash.sh builds it in its own folder. Both examples' CMake files pass APP_PROFILE on.
• The ST's stopwatch replaces the slack report (both used Timer-A).
  • It's TIME_STUDY = 1 in userfw.s, off by default: Timer-A runs free at 4.07 µs per tick.
  • The VBL and four points of the loop are reported on one timeline ($FB8B / $FB8C / $FB8F).
  • tools/dev/swd.py stopwatch prints them.

Tests

• Host tests: make -C tests/host test passes. test_audio_dma.c is built and run under both profiles, and test_layout.py checks the constants of both.
• Builds: debug and release, each in both profiles, compile without warnings. So does the game example at 25 fps.
• Mega ST (YM):
  • At 25 fps: 24.9 fps. The full copy with 22 kHz sound takes 29.9 ms and leaves 9.75 ms before the next frame. With the mouse moved flat out: 8.9 ms, 1,429 packets, no IKBD error, no late audio slice or underrun. The sound was clean.
  • At 50 fps (release): 50.0 fps with the same trade-offs as before.
  • The 25 fps release build was verified separately.
• Mega STE (DMA and blitter): at 25 fps, the DMA chip plays at 25 kHz and the sound is clean. At 50 fps: 50.1 fps with the mouse moving and no errors.