summaryrefslogtreecommitdiffstatshomepage
diff options
context:
space:
mode:
author Angelo Salese <angelosa@users.noreply.github.com>2015-10-15 18:36:51 +0200
committer Angelo Salese <angelosa@users.noreply.github.com>2015-10-15 18:36:51 +0200
commit3ccb79bbbb80753fa320ea831f7e97098dc7f8fc (patch)
treeb4e2e368c27476ebb983772374a093d9a2aba7de
parentd9ad061e2a626c68379253ed61286c7de8f677ec (diff)
Notes
-rw-r--r--src/mame/drivers/cntsteer.c12
1 files changed, 7 insertions, 5 deletions
diff --git a/src/mame/drivers/cntsteer.c b/src/mame/drivers/cntsteer.c
index 229ec937897..3766eb129a2 100644
--- a/src/mame/drivers/cntsteer.c
+++ b/src/mame/drivers/cntsteer.c
@@ -16,11 +16,13 @@
- according to a side-by-side test, sound should be "darker" by some octaves,
likely that a sound filter is needed;
cntsteer specific:
- - proper bus request support, starting from understanding the role of DP register in M6809 core
- (having that returning non-zero in NMI for subcpu forces a reset?);
- - Understand how irq communication works between CPUs. Buffer $415-6 seems involved in the protocol.
- - understand who's master and who's slave between "back" and "mix" CPUs (order might be actually inverted);
- - understand why background mirroring causes wrong gfxs on title screen (wrong tilemap paging);
+ - In Back Rotate Test, rotation is tested with the following arrangement (upper bits of rotation parameter):
+ 04 -> 05 -> 02 -> 03 -> 00 -> 01 -> 06 -> 07 -> 04 and backwards
+ Anything with bit 0 set is tested from 0xff to 0, with bit 0 clear that's 0 -> 0xff, fun.
+ - Understand how irq communication works between CPUs. Buffer $415-6 seems involved in the protocol.
+ We currently have slave CPU irq hooked up to vblank, might or might not be correct.
+ - invert order between maincpu and subcpu, subcpu is clearly the master CPU here.
+ - understand why background mirroring causes wrong gfxs on title screen (wrong tilemap paging, missing video bit or it's actually a RMW thing);
cleanup
- split into driver/video;