summaryrefslogtreecommitdiffstatshomepage
diff options
context:
space:
mode:
author angelosa <salese_corp_ltd@email.it>2017-11-27 00:48:21 +0100
committer angelosa <salese_corp_ltd@email.it>2017-11-27 00:48:21 +0100
commit80b4a1272a13b188c891cf218f2070bb038a02c9 (patch)
tree8483481fdfe264dfeb9187809f3a7625fffa185d
parent14c2e61b406d7b2e08fff12ee0b29377cbc19666 (diff)
metro.cpp: resolve conflict (nw)
-rw-r--r--src/devices/video/imagetek_i4100.cpp9
1 files changed, 9 insertions, 0 deletions
diff --git a/src/devices/video/imagetek_i4100.cpp b/src/devices/video/imagetek_i4100.cpp
index b3813d88c25..1897f5951fc 100644
--- a/src/devices/video/imagetek_i4100.cpp
+++ b/src/devices/video/imagetek_i4100.cpp
@@ -13,6 +13,15 @@
- inputs for i4300;
- hyprduel.cpp uses scanline attribute which crawls to unusable state
with current video routines here;
+ - CRT Controller, also understand why it needs so many writes before actual parameters;
+ - Wrong color bars in service mode (e.g. balcube, toride2g). They use solid color tiles (80xx),
+ but the right palette is not at 00-ff.
+ Related to the unknown table in the RAM mapped just before the palette?
+ Update: the colors should have a common bank of 0xb (so 0x8bxx), it's unknown why the values
+ diverges, the blitter is responsible of that upload fwiw;
+ - Some gfx problems in ladykill, 3kokushi, puzzli, gakusai, seem related to how we handle
+ windows, wrapping, read-modify-write areas;
+ - puzzli: emulate hblank irq and fix video routines here (water effect not emulated?);
============================================================================