Jaklyy
|
7590c48471
|
Revert "revert arm9 main ram dma again (again (again))"
This reverts commit c90f10d056.
|
2024-12-16 11:45:14 -05:00 |
|
Jaklyy
|
67198a72bd
|
why did i remove that
This reverts commit 45be951a0f.
|
2024-12-16 10:32:33 -05:00 |
|
Jaklyy
|
5c5f4364db
|
run arm 7 first?
it's faster...?
|
2024-12-16 10:20:10 -05:00 |
|
Jaklyy
|
c90f10d056
|
revert arm9 main ram dma again (again (again))
|
2024-12-16 10:19:51 -05:00 |
|
Jaklyy
|
29421f1d27
|
re-enable mainram dma
it's not bugged, it's also not inaccurate.
something else is the issue...
|
2024-12-16 09:23:54 -05:00 |
|
Jaklyy
|
e77c2011bc
|
slightly optimize main loop
|
2024-12-15 22:02:41 -05:00 |
|
Jaklyy
|
c40efab62f
|
revert main ram dma
broke stuff
|
2024-12-15 22:01:44 -05:00 |
|
Jaklyy
|
c96b49e9cd
|
Revert "improve(?) irq handling"
This reverts commit 443ecb313d.
|
2024-12-15 21:48:27 -05:00 |
|
Jaklyy
|
93242e1189
|
revert most scheduler changes
|
2024-12-15 21:48:22 -05:00 |
|
Jaklyy
|
443ecb313d
|
improve(?) irq handling
...i dont think this fixes anything
|
2024-12-15 20:51:13 -05:00 |
|
Jaklyy
|
45be951a0f
|
this should be smarter
|
2024-12-15 18:48:35 -05:00 |
|
Jaklyy
|
db1991276a
|
tweak scheduling to be a little less gross
|
2024-12-15 18:40:21 -05:00 |
|
Jaklyy
|
5b07765eb5
|
misc tweaks to dma
|
2024-12-15 10:42:52 -05:00 |
|
Jaklyy
|
2247f17f4f
|
implement a main ram burst restart behavior
|
2024-12-15 06:34:36 -05:00 |
|
Jaklyy
|
d912429d8c
|
comment out some replaced stuff
|
2024-12-14 23:36:28 -05:00 |
|
Jaklyy
|
c902dcfc98
|
improve main ram dma timings
|
2024-12-14 23:23:00 -05:00 |
|
Jaklyy
|
2051d412d1
|
implement MR cont. for arm7 dma; also a hack?
the hack is to make arm9 dma contention work with prior improvements to synchronization
|
2024-12-14 17:17:44 -05:00 |
|
Jaklyy
|
4ea0e60e18
|
minor fix(?)
this should fix something?
|
2024-12-14 13:32:54 -05:00 |
|
Jaklyy
|
5e945669f5
|
hopefully reduce desync potential a little?
|
2024-12-14 10:45:08 -05:00 |
|
Jaklyy
|
610ac2491a
|
disable main ram contention for arm9 dma
caused innumerable issues
will need a more comprehensive rewrite later
|
2024-12-14 09:38:02 -05:00 |
|
Jaklyy
|
ac1d790d7e
|
fix the system timestamp being run wayyyy too fast
oh no that was covering up SO many bugs hhhhsdfghhg
|
2024-12-14 00:15:17 -05:00 |
|
Jaklyy
|
a445c0d32c
|
this makes a bit more sense
|
2024-12-13 22:08:34 -05:00 |
|
Jaklyy
|
cce5070077
|
probably not any faster
|
2024-12-13 21:32:15 -05:00 |
|
Jaklyy
|
456d07da03
|
unbork gxfifo stalls
|
2024-12-13 15:35:44 -05:00 |
|
Jaklyy
|
642f085975
|
probably unborks gxfifo stalls
|
2024-12-13 13:15:38 -05:00 |
|
Jaklyy
|
73be2f3e01
|
tweak dmas to be more accurate (actually less?)
|
2024-12-13 13:09:42 -05:00 |
|
Jaklyy
|
d341260e5a
|
dma rewrite 1
|
2024-12-10 21:23:02 -05:00 |
|
Jaklyy
|
feb1cd562d
|
clarify some more write buffer details
|
2024-12-10 12:04:43 -05:00 |
|
Jaklyy
|
96c8f67d5f
|
implement bit 10 of exmemcnt
|
2024-12-10 08:04:00 -05:00 |
|
Jaklyy
|
b048e0cbfd
|
improve ExMemCnt handling and defaults
|
2024-12-09 19:30:47 -05:00 |
|
Jaklyy
|
8382769073
|
fix a main loop freeze; exmemcnt bit 15 starts set
fixes twilight menu
|
2024-12-09 15:52:50 -05:00 |
|
Jaklyy
|
52e14612b1
|
probably faster to directly access main ram?
|
2024-12-09 12:25:23 -05:00 |
|
Jaklyy
|
0111ee7fac
|
micro-optimization
|
2024-12-09 11:11:24 -05:00 |
|
Jaklyy
|
918df047b8
|
cache line boundary align condition lut table
|
2024-12-09 10:03:47 -05:00 |
|
Jaklyy
|
cbdd6a0faf
|
cacheline align register array
IM SORRY GENERIC
|
2024-12-09 09:10:24 -05:00 |
|
Jaklyy
|
fe9a9ee27d
|
actually those do literally nothing
|
2024-12-09 00:39:24 -05:00 |
|
Jaklyy
|
33f6218972
|
avoid checking T bit every instruction
|
2024-12-09 00:31:21 -05:00 |
|
Jaklyy
|
aa2cdc37a1
|
optimize one of the main loops
|
2024-12-08 23:10:53 -05:00 |
|
Jaklyy
|
f823a92020
|
fix branches being able to break the queue system
fixes bw2
|
2024-12-08 22:41:32 -05:00 |
|
Jaklyy
|
7a4234dcd8
|
fix writeback when rn is also rd in ldr
something *has* to rely on this, as stupid as it seems
|
2024-12-08 19:49:08 -05:00 |
|
Jaklyy
|
1a1934df00
|
...removing the (s32) fixes sign extension? ig???
|
2024-12-08 19:24:19 -05:00 |
|
Jaklyy
|
0df4369305
|
tweak scheduler for better performance
might be less accurate
|
2024-12-08 15:25:18 -05:00 |
|
Jaklyy
|
91752c1925
|
fix emulator hanging under certain circumstances
|
2024-12-08 14:24:33 -05:00 |
|
Jaklyy
|
8e6755ce2c
|
jakly pls
|
2024-12-08 11:20:36 -05:00 |
|
Jaklyy
|
bda05a756e
|
only recalc mpu lut if it changed
|
2024-12-08 11:19:49 -05:00 |
|
Jaklyy
|
8209fdebb4
|
fix main ram timestamp
i hate order of operations
|
2024-12-08 10:02:25 -05:00 |
|
Jaklyy
|
e69a2aa1b5
|
write buffer shouldn't continue resolving main ram accesses if it passes the a7 ts
|
2024-12-08 09:05:33 -05:00 |
|
Jaklyy
|
68b4d96f0d
|
Queue ICache Prefetch
|
2024-12-08 00:25:44 -05:00 |
|
Jaklyy
|
b40c6bc41d
|
implement write buffer
|
2024-12-08 00:19:43 -05:00 |
|
Jaklyy
|
d14c5ea246
|
re-add itcm delay for ldm/stm
|
2024-12-07 10:07:57 -05:00 |