Commit Graph

16462 Commits

Author SHA1 Message Date
magumagu
074397c12d Explicitly set up AllocateTexture configuration for palette conversion.
No functional change.
2015-02-19 15:57:05 -08:00
magumagu
ddc815dd7a Remove TextureAddress struct. 2015-02-19 15:36:32 -08:00
magumagu
c0a4760f0e Decode EFB copies used as paletted textures.
A number of games make an EFB copy in I4/I8 format, then use it as a
texture in C4/C8 format.  Detect when this happens, and decode the copy on
the GPU using the specified palette.

This has a few advantages: it allows using EFB2Tex for a few more games,
it, it preserves the resolution of scaled EFB copies, and it's probably a
bit faster.

D3D only at the moment, but porting to OpenGL should be straightforward..
2015-02-19 15:09:27 -08:00
Pierre Bourdon
fbbbad98e9 Merge pull request #2079 from lioncash/tables
PowerPC: Minor cleanups.
2015-02-19 22:51:34 +01:00
magumagu
d0c675af83 Merge pull request #2080 from lioncash/interp
Interpreter: Remove unnecessary prototypes and includes
2015-02-19 11:19:20 -08:00
Lioncash
8401b5a1f2 Interpreter: Remove unnecessary prototypes and includes 2015-02-19 13:20:24 -05:00
Lioncash
a60d3306b1 PowerPC: Get rid of magic numbers related to interp/JIT initialization. 2015-02-19 12:16:53 -05:00
Lioncash
d32359c915 PPCTables: Simplify UsesFPU
We can just retrieve the op info and check if the FPU flag is present
2015-02-19 12:16:44 -05:00
Pierre Bourdon
6a14300c9c Merge pull request #2078 from lioncash/fwd
PPCTables: Remove an unnecessary forward declaration
2015-02-19 15:51:15 +01:00
Pierre Bourdon
5393082117 Merge pull request #2077 from lioncash/docs
PPCTables: Correct some flag descriptions
2015-02-19 15:51:00 +01:00
Lioncash
e4df6026c4 PPCTables: Remove an unnecessary forward declaration
This class name isn't used anymore.
2015-02-19 09:08:00 -05:00
Lioncash
0fc3166b2f PPCTables: Correct some flag descriptions 2015-02-19 08:56:18 -05:00
Lioncash
c8c61041ba Merge pull request #2007 from lioncash/si
Core: Move several files into the correct VS directory filter.
2015-02-18 21:33:39 -05:00
Lioncash
bdca0dac29 Merge pull request #2071 from Sonicadvance1/Android_fix_unittests
[Android] Fix building unit tests.
2015-02-18 21:19:18 -05:00
Lioncash
2906f64e13 Merge pull request #2073 from Sonicadvance1/android_fix_formatchanges
[Android] Fix ignore format changes option.
2015-02-18 21:17:46 -05:00
Markus Wick
1a3bc8f286 Merge pull request #2001 from mimimi085181/multiple-texture-cache-entries
Allow multiple texture cache entries for textures at the same address
2015-02-19 01:08:07 +01:00
mimimi085181
2f8e0c9bb9 Allow multiple texture cache entries for textures at the same address
This is the same trick which is used for Metroid's fonts/texts, but for all textures. If 2 different textures at the same address are loaded during the same frame, create a 2nd entry instead of overwriting the existing one. If the entry was overwritten in this case, there wouldn't be any caching, which results in a big performance drop.

The restriction to textures, which are loaded during the same frame, prevents creating lots of textures when textures are used in the regular way. This restriction is new. Overwriting textures, instead of creating new ones is faster, if the old ones are unlikely to be used again.

Since this would break efb copies, don't do it for efb copies.

Castlevania 3 goes from 80 fps to 115 fps for me.

There might be games that need a higher texture cache accuracy with this, but those games should also see a performance boost from this PR.

Some games, which use paletted textures, which are not efb copies, might be faster now. And also not require a higher texture cache accuracy anymore. (similar sitation as PR https://github.com/dolphin-emu/dolphin/pull/1916)
2015-02-18 23:54:40 +01:00
magumagu
dfac2ca055 Merge pull request #2072 from Sonicadvance1/fix_inst_table_mulli
Fix mulli's flags in the instruction tables.
2015-02-18 14:44:09 -08:00
Ryan Houdek
09e8adb199 [Android] Fix ignore format changes option.
This option was inversed, so it was always set to the opposite of what was wanted.
2015-02-18 16:27:37 -06:00
Ryan Houdek
19f8524304 Fix mulli's flags in the instruction tables.
It doesn't set RC, so this flag is incorrect.
2015-02-18 16:26:12 -06:00
Ryan Houdek
0567b28ba1 [Android] Fix building unit tests. 2015-02-18 16:24:30 -06:00
shuffle2
861d341f4c Merge pull request #2063 from shuffle2/libusb-vcxproj
make libusb vcxproj use standard dolphin template and don't compile strerror.c.
2015-02-18 13:53:09 -08:00
Shawn Hoffman
f91e8e9423 make libusb vcxproj use standard dolphin template and don't compile strerror.c. 2015-02-18 13:29:39 -08:00
Shawn Hoffman
8645f33034 Revert "Externals: add Dolphin-specific patches to libusb"
This reverts commit 7a305bef86.
2015-02-18 13:23:37 -08:00
Markus Wick
86226cb2b9 Merge pull request #1979 from Mullin/head
Use clock_gettime instead of gettimeofday, it's more efficiency on low power system
2015-02-18 21:54:16 +01:00
Mullin
797fc14414 Use clock_gettime instead of gettimeofday, it's more effiency on low power system
Add def for mac(They don't support clock_gettime)

Fix my mistake

Fix my mistake 2
2015-02-18 21:44:58 +01:00
Pierre Bourdon
8cc6e5cff9 Merge pull request #1854 from JosJuice/read-disc-after-delay
DVDInterface: Read disc after delay, not before
2015-02-18 20:35:07 +00:00
Pierre Bourdon
042fcbcbca Merge pull request #2067 from magumagu/cputhread-assert
Disable an IsCPUThread() assert in Release builds.
2015-02-18 16:22:23 +00:00
Pierre Bourdon
96041a5aca Merge pull request #2068 from mimimi085181/hash-fix
GetCRC32: Fix the hash for the last byte(s)
2015-02-18 13:43:06 +00:00
mimimi085181
f6c9b8243e GetCRC32: Fix the hash for the last byte(s)
This fixes issue 8227
https://code.google.com/p/dolphin-emu/issues/detail?id=8227
2015-02-18 14:26:28 +01:00
Ryan Houdek
ef2fba2e85 Merge pull request #2070 from lioncash/unused
PowerPC: Remove unused enum values
2015-02-17 23:42:46 -06:00
Lioncash
78d917b0b3 PowerPC: Remove unused enum values 2015-02-18 00:37:32 -05:00
Lioncash
373b1d9915 Merge pull request #2069 from lioncash/docs
PowerPC: Document the instruction table flags
2015-02-18 00:26:42 -05:00
Lioncash
0b3a2ae4ca PowerPC: Document the instruction table flags
Should make it easier for newcomers to get up to speed.
2015-02-18 00:17:51 -05:00
Ryan Houdek
127e742992 Merge pull request #2046 from Mullin/Fix-unsigned/signed-stores
[ARMv7] Fix unsigned/signed stores
2015-02-17 14:38:36 -06:00
Mullin
39aa2432a6 Fix signed/unsigned stores 2015-02-17 21:29:01 +01:00
Fiora
53bf21e15a Merge pull request #2066 from lpereira/regcache-with-variadic-tpl
Jit64: Use variadic templates for RegCache locking
2015-02-17 08:53:07 -08:00
Lioncash
39e121e551 Merge pull request #2064 from magcius/fix-tilt
WiimoteEmu: Add back the calibrated accel data to the EEPROM
2015-02-17 11:14:09 -05:00
JosJuice
3ac0977bc6 DVDInterface: Read disc after delay, not before 2015-02-17 14:57:36 +01:00
Leandro Pereira
1162c5344b Jit64: Use variadic templates for RegCache locking 2015-02-17 10:01:42 -02:00
Ryan Houdek
3aa605236d Merge pull request #2041 from Sonicadvance1/AArch64_vertex_loader
[AArch64] Vertex loader and things
2015-02-17 00:51:51 -06:00
Ryan Houdek
ed008c3a69 [AArch64] Change the vertex loader over to using unscaled loadstores.
In nearly all direct loadstore cases we can use unscaled loadstores.
Still have a fallback in case we hit a situation that we /can't/ do a unscaled loadstore.
2015-02-16 22:03:09 -06:00
Ryan Houdek
120df4c688 [AArch64] Implement loadstore unscaled. 2015-02-16 22:00:43 -06:00
magumagu
77a0bfdd8f Merge pull request #2055 from magumagu/fakevmem-code
Allow executing code out of FakeVMEM.
2015-02-16 19:13:14 -08:00
magumagu
cce503c879 Disable an IsCPUThread() assert in Release builds.
IsCPUThread is extremely slow at the moment, and this code runs
frequently.
2015-02-16 19:07:42 -08:00
Jasper St. Pierre
e53268be1b WiimoteEmu: Add back the calibrated accel data to the EEPROM
It seems that games actually read this to determine how to act, which
isn't particularly strange. I was imagining it was more for the
Wiimote's internal normalization, but I was wrong.
2015-02-16 15:16:23 -08:00
Ryan Houdek
2ebe57ed3f Convert our vertex loader unit test to the new RunVertices arguments arrangement. 2015-02-16 16:51:39 -06:00
Ryan Houdek
f54a0d3ff4 [AArch64] Minor floating point loadstore improvements.
These use the Vector x Element FMUL instruction to improve speeds slightly.
Should give ~3cycle improvement per quantized loadstore done.
2015-02-16 16:51:39 -06:00
Ryan Houdek
b4b03641b3 [AArch64] Implement vertex loader recompiler.
Shows a noticeable reduction in time spent in the vertex loader.
2015-02-16 16:51:32 -06:00
magumagu
5b6a947e8f Merge pull request #2035 from magumagu/remove-bat-option
Remove the BAT checkbox/setting/implementation.
2015-02-15 17:27:18 -08:00