Commit Graph

1081 Commits

Author SHA1 Message Date
Hans-Kristian Arntzen
824f5f3a5f Experiment with GPU readbacks.
Avoids having to rely on software raster to run in parallel with HW rendering.
2018-12-14 17:54:13 +01:00
iCatButler
ea51d22616 Enhance detection of flipped sprites
- Use PGXP to better detect and reject 3D elements
- Greatly reduces false positives which caused visible popping on 3D elements as their UVs align on screen.
2018-12-08 20:43:00 +00:00
iCatButler
b9dbeb7b54 Allow partial culling of quads
- Flag which triangles in a quad are culled due to being too large and only draw the remaining ones (if any)
2018-12-08 20:23:35 +00:00
iCatButler
1e0e774ad2 Refactor UV offset and limit code to be shared by all renderers.
- Calculate and apply offset for quads with flipped UVs earlier.
- Reduce offsets applied in software renderer as resolution scaling increases (fixes line artefacts).
- Calculate UV limits used for filtering earlier and pass these to the individual renderers.
2018-11-30 12:11:39 +00:00
iCatButler
8d15a7823c Fix PGXP "+CPU" integration (#448)
Add PGXP calls for instructions that were previously missed. Should improve support for games like Croc, MGS, Alundra 2 and others.
2018-11-25 08:47:26 -06:00
FrozenFish24
b6cfa5ec51 Prevent crash when cue sheet references a non-existant file 2018-10-18 18:06:21 +08:00
360 CodeSafe
16f1f8d5aa
add checking against NULL
in line 240, the condition expression, pointer 'vi' may be a null pointer.
in line 245 (246 repaired), directly deference it as the condition of 'for' statement, which may cause null pointer deference.
2018-09-26 14:49:37 +08:00
hunterk
cc3b2db364 fix up lineRender core option 2018-08-27 22:19:57 -05:00
hunterk
7ed2a2d7cd initial line-to-quad option 2018-08-26 21:36:01 -05:00
twinaphex
da85ca2406 Use experimental PGXP for software - should at least provide
subpixel correction
2018-08-06 18:14:14 +02:00
iCatButler
2f3316ddf7 - New more aggressive variant of the MakeQuad/FindLine Hack
- Stores modified vertices for second triangle before rendering the first (software renderer modifies vertex order)
- Added code to toggle between disabled/default/aggressive (no UI support though)

Required for games which draw lines but use UVs that form a triangle in texture space (Dark Forces and Duke Nukem).
These rely on PS1 sampling rules to follow the upper or left side UVs when drawing the line of pixels.
New version of the hack forces these UVs into a line, both for the original triangle and the secondary one needed to make a quad.

Retains the older more conservative method as it works fine for most games and produces fewer artifacts.
2018-08-06 16:58:03 +02:00
iCatButler
474e5709fe Complete rewrite of line Hack_MakeQuad
Uses assumption that each end of the line has identical UVs to find and orient candidates much more easily.
Even with PGXP disabled this produces minimal artifacts.
2018-08-03 01:22:09 +02:00
iCatButler
22a36febb6 Add hack to fix line rendering trick (Doom, Soul Blade etc.)
These games render single pixel width axis aligned triangles which at native resolution become pixel wide lines. At higher resolutions they no longer resolve as lines and so gaps appear in the image.

This hack attempts to detect these triangles and then construct and render a second complimentary triangle that makes them a quad. Thus filling the whole space that would be taken if the intended line were scaled up.
2018-08-03 01:22:09 +02:00
twinaphex
4a75947e0f Update tremor 2018-04-24 06:47:48 +02:00
Dwedit
32fc5d370b Fixes for Savestates
* Savestate Performance Fix - no longer does giant memory allocations and copies
* Loadstate Performance Fix - no longer does N^2 string comparisons
* Fast Savestates (excludes text labels from the savestates, 20% speedup)
* Minor Savestate Performance Fix, no longer uses snprintf when strncpy will do.  Only applies to the text labels, no effect on fast savestates.
2018-04-23 20:16:53 -05:00
Dwedit
e10019acb8 Add missing __fastcall 2018-04-18 19:55:26 -05:00
Dwedit
9086bf32c0 MSVC code fixes for the GL/VULKAN build 2018-04-18 17:54:32 -05:00
Dwedit
58c57d35e7 Fixes to code to compile for MSVC 2018-04-18 14:34:16 -05:00
Lionel Flandrin
99a808fc3d Fix broken "Dithering pattern" option for the SW renderer 2018-04-08 23:13:57 +02:00
Lionel Flandrin
ce9d7d8aeb Fix PGXP regression in GTE NCLIP
Fixes #323
2018-04-07 21:45:48 +02:00
iCatButler
6475c19172 Fix PGXP bugs
- Pass clamped precise Z value to prevent divide by 0 on GPU (causes errors on Nvidia GPUs)
- Offset texture coordinates by 1/2 texel to avoid error on 1D look ups using Nvidia GPUs (issues #85)
- Prevent Vulkan renderer offsetting texcoord lookups by 1 during 1D lookups (caused error with above fix)
2018-03-02 21:46:44 +00:00
Lionel Flandrin
7a7ae882ff GTE: fix regression in RTPT 2018-02-23 23:53:09 +01:00
Lionel Flandrin
0b67670b71 GTE: fix regression in RTPS 2018-02-23 23:50:01 +01:00
Lionel Flandrin
390b07cc6b GTE: fix regression in SQR 2018-02-23 23:34:13 +01:00
twinaphex
80a21f2ba8 Put some debug conditionals around some printfs 2018-02-22 22:23:10 +01:00
twinaphex
15386cb9b0 Only trigger ILH mode and Debug Mode with DEBUG builds 2018-02-22 20:58:02 +01:00
twinaphex
9d2cc61b2d Only use decltype if HAS_CXX11 is defined 2018-02-22 20:20:04 +01:00
Twinaphex
4d482d8e66 Some ios buildfixes 2018-02-22 00:08:45 +01:00
Lionel Flandrin
76341b5672 Attempt to fix OSX build 2018-02-21 23:53:30 +01:00
Lionel Flandrin
b537b04866 Remove C++11 dependency 2018-02-21 23:45:25 +01:00
twinaphex
6800cd5c96 Buildfix 2018-02-21 23:24:38 +01:00
Lionel Flandrin
b505f54aec Update CPU code to match upstream 2018-02-21 23:17:22 +01:00
Lionel Flandrin
895767268b Re-enable PGXP option without accuracy regression 2018-02-21 22:02:25 +01:00
twinaphex
83a9841715 Use true/false 2018-02-21 17:34:14 +01:00
twinaphex
0d38a5de1e All GTE basic tests pass now - simias - try backporting the PGXP
patches to the non-GTE_SPEEDHACKS paths - you can decide for yourself
whether or not to kill the GTE_SPEEDHACKS code after this
2018-02-21 17:16:29 +01:00
Lionel Flandrin
bddafd31ea Disable threaded CD code when HAVE_THREADS is unset 2018-02-17 17:58:02 +01:00
Lionel Flandrin
4826c06e4d
Added GPU rasterizer overclock option 2018-02-09 11:09:28 +01:00
Lionel Flandrin
8e7ae9c61f Implement true CPU frequency scaling/overclocking 2018-02-08 22:06:35 +01:00
Lionel Flandrin
ac3fec5c36 Rename "CPU Overclock" option into "GTE Overclock"
It's still not very accurate (it's not really an overclock and it also
removes instruction cache/memory access latencies) but it's probably
easier to understand that way and it won't clash with true CPU
overclock.
2018-02-08 19:22:53 +01:00
twinaphex
8a0e087851 Update 2018-01-07 15:49:28 +01:00
twinaphex
3b4bbdde80 Use _WIN32 define 2018-01-07 15:38:42 +01:00
Tommy Nguyen
af262e12d1
Use correct conversion specifier 2018-01-07 06:09:56 -05:00
twinaphex
48fd8c568d Update 2017-12-20 03:51:00 +01:00
twinaphex
30f9e1bcc8 Cleanups 2017-12-20 03:36:17 +01:00
twinaphex
1332311781 Simplify state.cpp 2017-12-20 03:30:26 +01:00
twinaphex
f0781a41d1 Cleanups 2017-12-20 03:22:54 +01:00
twinaphex
f18c48e7b2 Get rid of file.c - use plain filestream 2017-12-19 03:19:58 +01:00
twinaphex
78836b5d1c Start using filestream directly where possible 2017-12-19 01:42:59 +01:00
twinaphex
456862089a Get rid of OpenedMode 2017-12-18 22:49:42 +01:00
twinaphex
6aae474026 Get rid of original_path 2017-12-18 22:47:08 +01:00
twinaphex
650b454e7b Buildfix 2017-12-18 16:07:52 +01:00
Alberto Fustinoni
8a226e3d51 Removed dependency on file_path 2017-12-18 22:35:54 +09:00
twinaphex
b144f717e2 Add file header include 2017-12-17 16:15:28 +01:00
Alberto Fustinoni
c6ad4b9c92 Bugfix 2017-12-17 17:40:44 +09:00
twinaphex
c11a85405c Cleanups 2017-12-17 02:29:51 +01:00
twinaphex
36d33b2334 Update 2017-12-16 14:45:56 +01:00
twinaphex
f9fd587119 Update API 2017-12-16 14:21:47 +01:00
twinaphex
7a66d0c6bd Update core - use VFS interface 2017-12-15 23:20:53 +01:00
twinaphex
018e009006 Simplify Filestream 2017-12-15 21:43:29 +01:00
twinaphex
42d9158102 Simplify Stream 2017-12-15 21:41:23 +01:00
twinaphex
022c5efdfd This can go now 2017-12-15 21:33:18 +01:00
David Walters
2ebd4e5023 Gun crosshair shapes 2017-11-23 13:08:26 +00:00
David Walters
a64d4ddfe2 Mouse sensitivity setting 2017-11-22 23:53:59 +00:00
Garrett Brown
5274d723ac Fix compilation on Android
Error was:

mednafen/cdrom/CDAccess_PBP.cpp:675:79: error:
      cannot pass object of non-trivial type 'std::string' (aka 'basic_string<char>') through variadic function; call will abort at
      runtime [-Wnon-pod-varargs]
      log_cb(RETRO_LOG_WARN, "[PBP] Invalid path/filename for SBI file %s\n", sbi_path);
2017-11-09 10:00:18 -08:00
Zapeth
cc344b22ba Make logging for SBI file loading more helpful 2017-10-18 20:38:47 +02:00
Brian Koropoff
6e14459d0a Detect and report slow CD reads
When not using async, first try read with a timeout.  If it fails,
display a message suggesting a different CD access method.
2017-10-15 14:42:07 -07:00
Brian Koropoff
0e6c67a736 Control CD image precache/async behavior through unified setting 2017-10-15 14:42:07 -07:00
Brian Koropoff
c06673bd7f Support subchannel-only reads from CCD 2017-10-15 14:42:07 -07:00
Brian Koropoff
5d227ea4b8 Avoid reading entire sector to get subchannel data
Since the subchannel data is either computed from scratch or
read from a separate sbi file, it's entirely unnecessary.

This avoids a synchronous file read in PreSeekHack in the CD
controller code which otherwise completely defeats async read
behavior.  It also makes the sector read hint provided to the
background thread more useful.
2017-10-15 00:31:41 -07:00
Brian Koropoff
2ee4fdbd2f Read CD sectors asynchronously to avoid stuttering
Rather than blocking the main thread waiting for data, try again at a
later emulation update.
2017-10-14 23:21:12 -07:00
r5
38370798ad (core/timer) Port timer emulation changes from 0.9.38.7
Attempt to fix #52
2017-10-11 02:58:13 +01:00
Lionel Flandrin
e84285d447 Don't enable CD fastload when streaming compressed audio to the SPU 2017-10-09 23:09:09 +02:00
Lionel Flandrin
b6868d47da Fix compilation with C++98 2017-10-09 11:45:52 +02:00
Lionel Flandrin
82f449bcd6 Add a core option to increase the CD loading speed 2017-10-08 23:05:00 +02:00
Lionel Flandrin
ed0a329572 Enable GPU cache
Fixes #12
2017-10-08 12:38:05 +02:00
Lionel Flandrin
11e3ee9bc6 Backport new triangle drawing code from upstream mednafen
The increased precision of the new code fixes some issues in a few
games. In particular it fixes a flickering glitch in FFVIII.

The original code used a few C++11 features, they've been rewritten to
make the code C++98 compliant.

This should fix #39
2017-10-08 01:09:48 +02:00
Brian Koropoff
2acf9af74a cdrom:chd: cache compressed data with disc cache option
This seems sufficient to avoid stutter without taking
as absurdly long on startup as decompressing the entire
file.
2017-10-01 12:44:14 -07:00
Brian Koropoff
e5cd40bdd1 Revert "cdrom: respect disc cache setting for chd files"
This reverts commit 0bb8b28179.
2017-10-01 12:37:01 -07:00
Brian Koropoff
0bb8b28179 cdrom: respect disc cache setting for chd files 2017-09-30 10:24:42 -07:00
rz5
ea93f24371 Update gpu.cpp 2017-09-06 17:25:17 +01:00
Zapeth
44e6e19e92 Merge with upstream to resolve Makefile.common merge conflict 2017-09-03 15:08:31 +02:00
Zapeth
790a3d3478 CHD EOL normalization and cleanup 2017-09-02 16:06:08 +02:00
Build
a44baaefa8 Add CHD dependencies to Makefiles and namespace mednafen's md5 implementation to avoid conflicts 2017-09-01 11:25:39 +02:00
Zapeth
052a22cd8c Add support for CHD files 2017-08-31 19:01:39 +02:00
rz5
fe418892cc (gpu.cpp) Remove a loose comment 2017-08-24 18:14:09 +01:00
r5
447dcdf4cf (gpu.cpp) Added some comments 2017-08-23 19:47:21 +01:00
r5
fad9372427 Fix issue #211
GPU is now a struct and a global variable in the stack, like
in upstream.
The GPU's vram is no longer 'flexible', it's still at the end
of the struct but it's a ptr type rather than a zero size array.

Changed some libretro calls so SET_GEOMETRY is used in favor of
SET_AV_INFO, as per the API's recommendation.

Style nits here and there as well.
2017-08-23 19:38:52 +01:00
twinaphex
5dab432b0d Build fix #1 2017-08-14 07:01:06 +02:00
orbea
b06629c280 Silence -Wunused-but-set-variable warnings 2017-08-10 15:08:01 -07:00
twinaphex
34321f59fc Revert "Turn texel_put into static function"
This reverts commit 1e6440db45.
2017-07-31 10:26:16 +02:00
twinaphex
1e6440db45 Turn texel_put into static function 2017-07-31 10:21:06 +02:00
r5
ae09b521b1 (SW Renderer) Properly get and set internal/dither mask resolutions.
Should fix issue #211
2017-07-30 18:27:25 +01:00
twinaphex
37fc17bf82 Simplifications 2017-07-29 18:11:32 +02:00
twinaphex
c0aec41da3 Remove this 2017-07-29 17:56:05 +02:00
rz5
46f8e2577d (gpu_line.cpp) Style nits 2017-07-29 15:31:27 +01:00
rz5
aefcac9a93 (rsx_intf) More corrections 2017-07-29 15:25:15 +01:00
rz5
bcbffe67ab (GL Renderer) Missed some rsx_intf calls here
Trying to pass MaskSetOR as a uint32_t for the GL Renderer. Also noticed a MaskEval_TA which is a bool and the vk renderer consumes it just like it does to MaskEvalAND != 0
2017-07-29 15:08:57 +01:00
r5
929cdf6937 Buildfix and more style nits. 2017-07-29 01:36:19 +01:00
twinaphex
ccb65c3395 (gpu.cpp) Refactor this code 2017-07-27 21:39:21 +02:00
twinaphex
93c35be4d9 Update libretro API header 2017-07-26 02:19:30 +02:00
twinaphex
f339a6e50b Update portions of libretro-common 2017-07-26 02:13:37 +02:00
Zapeth
a869219c33 Use workaround for sloppily converted pbp files 2017-05-07 17:07:42 +02:00
twinaphex
8a4c12d9d3 Move BlitterFIFO outside of PS_GPU class 2017-04-22 23:08:03 +02:00
twinaphex
8eacbae4af Move CTEntry Commands outside of class/struct 2017-04-22 21:33:20 +02:00
twinaphex
8230afa066 Further rewrites 2017-04-22 21:19:16 +02:00
twinaphex
afb7744507 Create GPU_Build static function 2017-04-22 20:26:49 +02:00
twinaphex
52b67179a9 Turn GPU_Alloc into static function 2017-04-22 20:23:32 +02:00
twinaphex
49b4d2a1e6 Turn more functions into static functions 2017-04-22 20:16:39 +02:00
twinaphex
662b9f1f33 Create UPSCALE macro 2017-04-22 20:06:50 +02:00
twinaphex
7d79b3bdf7 Get rid of vram_npixels 2017-04-22 20:02:34 +02:00
twinaphex
3bd2230a29 Turn DitherEnabled into macro 2017-04-22 19:59:22 +02:00
twinaphex
037adf7152 Create another static function 2017-04-22 19:49:11 +02:00
twinaphex
1883dbd509 Turn ReadData into static inline function 2017-04-22 18:53:35 +02:00
twinaphex
07a6fbf6c0 Simplify GPU some more 2017-04-22 18:47:01 +02:00
twinaphex
31b2b08c9e Get rid of ModTexel member 2017-04-22 18:03:25 +02:00
twinaphex
be1f8be48f Turn ModTexel into macro 2017-04-22 18:01:44 +02:00
twinaphex
027d7d4911 Get rid of PlotPixelBlend member function 2017-04-22 17:53:40 +02:00
twinaphex
459fa07c9a Turn CalcFIFOReadyBit into static inline function 2017-04-22 17:50:06 +02:00
twinaphex
bd0ebd8623 PS_GPU Refactor pt. 2 2017-04-22 17:43:39 +02:00
twinaphex
7bf0144c2a Start refactoring GPU code inspired by latest Mednafen code 2017-04-22 17:05:39 +02:00
iLag
d57283ae5e Fix support for true Multiline cheats. Backports relevant code from newer versions of standalone mednafen. 2017-04-06 14:27:12 -07:00
twinaphex
3ff904a09c Comment this out for non-software rendering mode 2017-03-29 23:11:20 +02:00
twinaphex
9a6363bc5d This is not necessary to be done in non-software mode 2017-03-29 23:00:56 +02:00
twinaphex
a73d3953af Remove CanRead() entirely 2017-03-26 18:24:11 +02:00
twinaphex
d4318fd564 Don't use CanRead 2017-03-26 18:21:01 +02:00
twinaphex
5ad744b3d5 Simplifications 2017-03-26 17:35:31 +02:00
twinaphex
6f75732f50 Revert "Cleanups"
This reverts commit 4bf8710c19.
2017-02-18 16:07:19 +01:00
twinaphex
4bf8710c19 Cleanups 2017-02-18 16:03:40 +01:00
twinaphex
c73df8090a Take out mednafen-endian.c 2017-02-18 03:14:24 +01:00
twinaphex
567bf5ae1c Update 2017-01-22 12:30:23 +01:00
twinaphex
5f8e68ba77 Get rid of 'register' keyword 2017-01-22 12:18:14 +01:00
twinaphex
d01e41dbcb Rewrite 'Sweep clock' function 2017-01-21 14:22:10 +01:00
twinaphex
c4cf61416c ReadS/ReadN - get rid of one early return 2017-01-21 13:19:05 +01:00
twinaphex
8e6c1e5e3f CDC - Cleanups 2017-01-21 13:14:57 +01:00
twinaphex
4d7aa422da Get rid of unneeded variables, implicit memsets 2017-01-21 13:08:36 +01:00
twinaphex
b3f9223a26 Get rid of some implicit memsets; get rid of early return 2017-01-21 12:57:16 +01:00
iLag
272f5c2e0c Enable Cheat Support
I had a lot of difficulties on this one due to my test case being non-functional.  Hopefully what there's no unnecessary changes left in this.
2016-12-23 09:27:19 -08:00
Tiny Tiger
697f9dd3fa Fix MDEC resolution in Soul Blade. 2016-12-04 17:11:39 +01:00
Tiny Tiger
fa039d53c1 Add parallel-psx and dumping. 2016-12-03 17:09:45 +01:00
twinaphex
2f3057cafc Add libretro-common float_minmax.h 2016-11-01 11:30:35 +01:00
twinaphex
fa181e1cd6 Update gte.cpp 2016-11-01 10:48:25 +01:00
twinaphex
9640288992 Use some SSE2 intrinsics 2016-11-01 09:45:23 +01:00
Twinaphex
dcb42463a2 Merge pull request #101 from iCatButler/master
Clamp depth values to near/2.f to prevent the projection factor becom…
2016-10-02 00:09:45 +02:00
iCatButler
7801d7845c Clamp depth values to near/2.f to prevent the projection factor becoming too large (or div_by_0)
Fixes RRR.
2016-10-01 22:57:32 +01:00
Twinaphex
b02e5afa8f Merge pull request #98 from iCatButler/master
Stop quads with partially bad W components from getting rendered.
2016-10-01 20:49:46 +02:00
iCatButler
5c73e07129 Stop quads with partially bad W components from getting through.
Should fix stretched textures in games like Spyro.
2016-10-01 19:34:05 +01:00
rz5
e5b11bf647 Fix compilation on MinGW/GCC
Fix the following compilation errors:
´´´
In file included from mednafen/psx/gte.cpp:24:0:
mednafen/psx/gte.cpp: In function 'void TransformXY(int64_t, float, uint16)':
mednafen/psx/gte.cpp:1076:48: error: 'min' was not declared in this scope
    precise_x = max(-0x400, min(precise_x, 0x3ff));
                                                ^
./mednafen/../pgxp/pgxp_gte.h:67:30: note: in definition of macro 'max'
 #   define max(a, b) ((a) > (b) ? (a) : (b))
                              ^
mednafen/psx/gte.cpp:1076:48: note: suggested alternative:
    precise_x = max(-0x400, min(precise_x, 0x3ff));
                                                ^
./mednafen/../pgxp/pgxp_gte.h:67:30: note: in definition of macro 'max'
 #   define max(a, b) ((a) > (b) ? (a) : (b))
                              ^
In file included from C:/msys64/mingw64/include/c++/6.1.0/bits/char_traits.h:39:0,
                 from C:/msys64/mingw64/include/c++/6.1.0/string:40,
                 from mednafen/psx/../git.h:4,
                 from mednafen/psx/../mednafen.h:12,
                 from mednafen/psx/psx.h:4,
                 from mednafen/psx/gte.cpp:21:
C:/msys64/mingw64/include/c++/6.1.0/bits/stl_algobase.h:243:5: note:   'std::min'
     min(const _Tp& __a, const _Tp& __b, _Compare __comp)
     ^~~
make: *** [Makefile:378: mednafen/psx/gte.o] Error 1

´´´
2016-10-01 12:23:21 +01:00
rz5
15931882ac Merge pull request #96 from iCatButler/master
GTE integration fixes
2016-10-01 11:56:15 +01:00
iCatButler
61876ee5bd GTE integration fixes
- Clamp precise X and Y values to psx limits (fixes crash in RR Revolution)
- Pass Z value as an unsigned 16-bit int (fixes far values wrapping to negative)
2016-10-01 10:40:01 +01:00
Lionel Flandrin
7bad2bd522 Don't reload software framebuffer when saving state (fixes #92) 2016-09-29 16:23:41 +02:00
iCatButler
66977786bb Merge pull request #4 from twinaphex/master
Fix widescreen hack with PGXP (fixes #91)
2016-09-27 15:49:35 +01:00
iCatButler
f074140e9f Tidy up GPU side code
- Remove unused functions
- Replace debug flags with a single valid_w
- Allow vertex caching to count as PGXP_enabled on its own
2016-09-27 15:42:26 +01:00
Lionel Flandrin
19b7481df6 Fix widescreen hack with PGXP (fixes #91) 2016-09-27 13:32:57 +02:00
Lionel Flandrin
ea10a60a06 Fix HW mode when PGXP is disabled 2016-09-26 01:45:39 +02:00
Lionel Flandrin
edf2c2b253 Fixed SW renderer 2016-09-26 00:46:25 +02:00
Lionel Flandrin
7d06b55d11 Always use native-precision values for software renderer 2016-09-25 22:43:16 +02:00
Lionel Flandrin
24d1559d77 Remove old subpixel-precision implementation 2016-09-25 22:24:53 +02:00
Lionel Flandrin
48ae05061b Fix compilation warnings 2016-09-25 22:17:45 +02:00
iCatButler
fba684f8c4 Fix a few CPU mode bugs
- Get register value after LUI
- Get Rs (not Rd) in MTHI and MTLO
2016-09-25 19:45:43 +01:00
iCatButler
e9d2688c74 Initial PGXP integration
- Create VS 2015 files
- Fix various compiler and linker errors (not sure if I'm missing some existing work arounds)

- Intergate all current operations (CPU currently broken)
- Integrate vertex caching
- Implement perspective correct texturing
- Update command vertices to use floating point positions
- Add menu options to toggle modes
2016-09-25 11:52:57 +01:00
twinaphex
582b867954 Update gte.cpp 2016-09-24 05:50:42 +02:00
twinaphex
754b5ec4cd cpu - update 2016-09-24 05:27:13 +02:00
twinaphex
12f99e491a Update cpu.cpp 2016-09-24 05:21:19 +02:00
twinaphex
09f1f88a67 Update cpu.cpp 2016-09-24 05:06:28 +02:00
twinaphex
3048a71b6d Update to 0.9.39.2 2016-09-23 21:51:50 +02:00
Alcaro
7a9fdfc1ff Merge pull request #90 from orbea/permissions
Code does not need to be executable
2016-09-21 18:03:47 +02:00
orbea
d0fcd58593 Code does not need to be executable 2016-09-21 09:03:02 -07:00
Twinaphex
cab6ef52a1 Merge pull request #88 from GeneralFailer/master
Correct option name mention for consistency
2016-09-19 19:48:22 +02:00
GeneralFailer
d5e9aa9d79 Update dualshock.cpp 2016-09-19 21:40:29 +04:00
Lionel Flandrin
b911183d22 Use delete [] to free the error message in MDFN_Error 2016-09-17 17:45:00 +02:00
Lionel Flandrin
5dc311ede7 Don't duplicate vertices when drawing a quad or sprite
Since we now have an index array we might as well use it...
2016-09-16 20:43:38 +02:00
twinaphex
f0ac0ddd09 Get rid of override 2016-08-15 15:37:18 +02:00
twinaphex
18f1c42802 Backport Fantastic Pinball Kyutenkai fix 2016-08-15 15:31:42 +02:00
twinaphex
9b3f092698 (CDROM) update 2016-08-15 06:35:34 +02:00
twinaphex
a2847fea49 Update 2016-08-15 04:43:51 +02:00
twinaphex
92a58b907c (CDROM) Update 2016-08-15 04:19:50 +02:00
twinaphex
6e61f46396 (CDROM) Update 2016-08-15 04:04:32 +02:00
twinaphex
4581c50784 Update CD-ROM code 2016-08-14 17:17:24 +02:00
twinaphex
05be10f124 Updates 2016-08-14 17:13:35 +02:00
twinaphex
0072138d23 Start using log_cb instead of going through MDFN_Error 2016-08-14 14:55:16 +02:00
twinaphex
be2f7e95cb Update HAVE_CDROM_NEW code 2016-08-14 14:28:02 +02:00
twinaphex
9410afc4b9 Update 2016-08-14 14:17:05 +02:00
twinaphex
6b929637e9 Updates 2016-08-14 13:54:20 +02:00
twinaphex
6f369f817e Updates 2016-08-14 13:43:27 +02:00
twinaphex
542fd6abc6 Add new CDROM code but don't yet use it 2016-08-14 13:22:51 +02:00
twinaphex
c508e48741 Update cdrom code 2016-08-14 13:15:13 +02:00
twinaphex
aa0dbd3dcd Update l-ec.c 2016-08-14 04:02:39 +02:00
twinaphex
f841c7f4f8 Remove SimpleFIFO 2016-08-14 03:51:42 +02:00
twinaphex
777efe34e3 Get rid of MDFNGameInfo->name 2016-08-09 07:27:55 +02:00
twinaphex
453d384cc1 Start getting rid of dependence on MDFNGameInfo 2016-08-09 07:24:22 +02:00
twinaphex
e55cbf9319 Cleanups 2016-08-09 05:38:51 +02:00
twinaphex
6bfeade60b Cleanup 2016-08-08 05:56:14 +02:00
twinaphex
34e7684111 Backport CDUtility changes 2016-08-07 08:00:23 +02:00
twinaphex
5c41cff4b5 Cleanups 2016-08-07 04:23:07 +02:00
twinaphex
8d7cd0123c Update 2016-08-06 20:53:46 +02:00
twinaphex
89704498f3 Update 2016-08-06 20:50:42 +02:00
twinaphex
7433c6c39a Add retro_inline.h include 2016-08-06 17:10:30 +02:00
twinaphex
43022bdc1b Backport b37ec0775d 2016-08-06 16:53:43 +02:00
twinaphex
44417fca27 Backport changes to SimpleFIFO 2016-08-06 16:45:44 +02:00
twinaphex
0c3429d67b Revert "(CDC) Backports"
This reverts commit b628c71c58.
2016-08-06 07:32:04 +02:00