47d1b07abb
Add a thread safe variant of invalidating the icache
...
This is used by the next commit.
2016-08-23 07:37:41 -04:00
f1964f90d6
Merge pull request #4129 from RisingFog/hahahahahahahahahaha
...
Fix a really stupid GLSL version parsing bug
2016-08-20 08:48:28 +12:00
da0204a85c
Fix a really stupid GLSL version parsing bug
2016-08-19 08:53:27 -04:00
31c530c7b3
Merge pull request #3386 from lioncash/memory
...
Common: Namespace MemoryUtil
2016-08-19 11:04:45 +02:00
40f4308dc2
Merge pull request #4085 from Armada651/vertex-depth
...
VideoCommon: Implement depth range equation in vertex shader.
2016-08-19 02:11:37 +02:00
e86d7cbc99
OGL: Workaround gl_ClipDistance bug on Mesa i965.
2016-08-18 01:08:39 +02:00
7078216b61
Improve documentation.
2016-08-16 21:09:58 +02:00
8c1c7fc2da
Cosmetics.
2016-08-15 13:11:30 +02:00
959d1879e5
VertexShaderGen: Fix far clipping.
...
We should only check whether z > 0, we don't care about w.
2016-08-15 13:11:29 +02:00
afa251af42
DriverDetails: Add bug for broken gl_ClipDistance on i965.
2016-08-15 13:11:28 +02:00
94927f360f
VideoCommon: Add a user-defined far clipping plane.
2016-08-15 13:11:28 +02:00
6e2052fae6
OGL: Disable clip distance on when not in a game-like state.
2016-08-15 13:11:27 +02:00
9596ec8971
GeometryShaderGen: Pass-through clipping distance.
2016-08-15 13:11:27 +02:00
a141e91dd2
OGL: Check for GL_DEPTH_CLAMP support.
...
It's not available in OpenGL ES and officially it's not supported on OpenGL 3.0/3.1.
Fallback to old depth range code if there is no method to disable depth clipping.
It's more important to have correct clipping than to have accurate depth values.
Inaccurate depth values can be fixed by slow depth.
2016-08-15 13:11:26 +02:00
4582853af4
VertexShaderGen: Use reversed depth range.
2016-08-15 13:11:26 +02:00
e9e81ece65
VideoBackends: Enable depth clamping.
2016-08-15 13:11:25 +02:00
b1ed7e80fb
VertexShaderGen: Clip z using user-defined clipping planes.
2016-08-15 13:11:25 +02:00
159247f4ad
VertexShaderGen: Clamp to the console depth range.
...
This fixes the Mii Channel among others.
2016-08-15 13:11:24 +02:00
92aa7669b5
VertexShaderManager: Use a more accurate depth range.
...
This fixes the gxtest_depth hwtest.
2016-08-15 13:11:24 +02:00
c223bd47b9
VideoCommon: Implement depth range equation in vertex shader.
2016-08-15 13:11:23 +02:00
0015d2e86b
Merge pull request #4121 from leoetlino/osreport-sjis
...
Convert OSREPORT text from SJIS to UTF-8
2016-08-15 16:04:07 +12:00
c0be228c71
Convert OSREPORT text from SJIS to UTF-8
...
I'm not sure this is the correct fix, but it looks like OSREPORT output
is Shift-JIS, so we need to convert it to UTF-8. Most characters work
fine without and with this conversion, but Japanese text completely
fails and results in outputting invalid UTF-8 (which gets shown as �).
2016-08-13 17:19:21 +02:00
0bd5db3e05
Merge pull request #3894 from hthh/optimize-fifo-bug
...
Jit: Check the FIFO on EIEIO instructions
2016-08-12 13:25:25 +02:00
d841d9c7b3
JitArm64: Check the FIFO on EIEIO instructions
...
Copied from the Jit64 version
2016-08-12 21:09:15 +10:00
bbc0f0c744
Jit: Check the FIFO on EIEIO instructions
...
The gather pipe optimization postpones checking the FIFO until the end
of the current block (or 32 bytes have been written). This is usually
safe, but is not correct across EIEIO instructions.
This is inferred from a block in NBA2K11 which synchronizes the FIFO
by writing a byte to it, executing eieio, and checking if PI_FIFO_WPTR
has changed. This is not currently an issue, but will become an issue
if the gather pipe optimization is applied to more stores.
2016-08-12 21:03:41 +10:00
4c9c456846
Merge pull request #4116 from leoetlino/lint-real-fix
...
Tools: Fix the lint script (for real, this time -- and the last time)
2016-08-11 21:27:16 +02:00
c1184957a5
Run clang-format on missed files
...
`clang-format`s files that lint missed because of the bug. Fortunately,
not much.
2016-08-11 21:14:39 +02:00
1fe31f6f19
Tools: Fix the lint script
...
It turns out that last fix was only partial, and it didn't fix the
issue completely as paths were fixed, but staged changes were not
in the list. This should fix it for real, this time. Sorry about that…
(git diff needs --cached to include staged changes, unlike git status.)
2016-08-11 21:11:17 +02:00
cef71afc27
Merge pull request #3987 from JosJuice/scheduleevent-cleanup
...
CoreTiming: ScheduleEvent cleanup
2016-08-11 16:18:14 +02:00
c6fed0565d
Merge pull request #4109 from JosJuice/remove-am-baseboard
...
Remove AM Baseboard
2016-08-11 16:16:23 +02:00
5eab50d52c
Merge pull request #4113 from stenzek/externals-vulkan
...
Externals: Add Vulkan headers and glslang
2016-08-11 16:15:32 +02:00
951fc44d86
Externals: Add glslang from Vulkan SDK v1.0.21.1
2016-08-11 22:40:32 +10:00
49e57df2b6
Externals: Add Vulkan headers from SDK v1.0.21.1
2016-08-11 22:40:32 +10:00
3bb0e7e3f5
Remove AM Baseboard
...
Let's stop pretending that we support Triforce emulation.
Keeping this code around just in case someone will make
major improvements in the future isn't really worth it.
I'm keeping the Triforce game INIs so users will know that
the compatibility rating for Triforce games is 1 star (broken).
2016-08-11 12:53:44 +02:00
3443a10030
CoreTiming: Merge ScheduleEvent variants into one function
...
Now Core::IsCPUThread() only gets called once when using the AnyThread
variant. Also, I think the enum approach makes calling code clearer.
2016-08-11 12:45:57 +02:00
5697032ec7
Merge pull request #4103 from degasus/dynamic-bat
...
JitArm64: Fix two issues.
2016-08-11 12:42:02 +02:00
088f7eaa3d
Merge pull request #4110 from JosJuice/fix-exi-translations
...
Fix choosing certain localized EXI devices
2016-08-11 12:39:36 +12:00
e577ffaee5
Merge pull request #4107 from JosJuice/no-immediate-scheduleevent
...
Remove Immediate variants of ScheduleEvent
2016-08-11 12:13:49 +12:00
c9fa6318a2
Fix choosing certain localized EXI devices
2016-08-10 20:04:33 +02:00
8552ff5789
Merge pull request #4106 from RisingFog/dtm_gc_language
...
Add Gamecube Language to DTM Header
2016-08-10 12:16:20 -04:00
7914e4d19d
Merge pull request #3458 from JosJuice/seconds-not-ticks
...
Replace some unnamed tick constants with GetTicksPerSecond()
2016-08-10 17:43:51 +02:00
b8e04c9513
Replace some unnamed tick constants with GetTicksPerSecond()
...
These operations should always take the same amount of time,
not the same amount of ticks. The number of ticks per second
is different for GameCube and Wii.
2016-08-10 17:34:23 +02:00
47c8bb26f8
Remove Immediate variants of ScheduleEvent
...
Usage of these are replaced with regular equivalents in order
to avoid executing event handlers in the middle of JIT blocks.
2016-08-10 16:34:06 +02:00
caa7ff7c25
Merge pull request #4093 from leoetlino/flags-and-events
...
Use Common::Flag and Common::Event when possible
2016-08-10 16:15:52 +02:00
dca22e08eb
Use Common::Flag and Common::Event when possible
...
Replaces old and simple usages of std::atomic<bool> with Common::Flag
(which was introduced after the initial usage), so it's clear that
the variable is a flag and because Common::Flag is well tested.
This also replaces the ready logic in WiimoteReal with Common::Event
since it was basically just unnecessarily reimplementing Common::Event.
2016-08-10 16:08:15 +02:00
c6a0e543a5
Merge pull request #3861 from stenzek/d3d11-mip-dumping
...
D3D11: Support texture dumping of non-zero mipmap levels
2016-08-10 15:50:48 +02:00
bce8097712
D3D11: Support texture dumping of non-zero mipmap levels
2016-08-10 23:45:25 +10:00
cb2b110d42
Merge pull request #4097 from lioncash/sysconf
...
SysConf: Minor simplifications to initial generating
2016-08-10 14:41:35 +02:00
b075d18d2e
Merge pull request #4096 from lioncash/codeblock-init
...
CodeBlock: In-class initialize variables
2016-08-10 14:39:51 +02:00
0fecd2f6b0
Merge pull request #4104 from degasus/master
...
PixelEngine: Cleanup PE Token.
2016-08-10 13:36:34 +02:00
aldelaro5