Previously we had decided to busy loop on systems due to Windows' scheduler being terrible and moving us around CPU cores when we yielded.
Along with context switching being a hot spot.
We had decided to busy loop in these situations instead, which allows us greater CPU performance on the video thread.
This can be attributed to multiple things, CPU not downclocking while busy looping, context switches happening less often, yielding taking more time
than a busy loop, etc.
One thing we had considered when moving over to a busy loop is the issues that dual core systems would now face due to Dolphin eating all of their CPU
resources. Effectively we are starving a dual core system of any time to do anything else due to the CPU thread always being pinned at 100% and then
the GPU thread also always at 100% just spinning around. We noted the potential for a performance regression, but dismissed it as most computers are
now becoming quad core or higher.
This change in particular has performance advantages on the dual core Nvidia Denver due to its architecture being nonstandard. If both CPU cores are
maxed out, the CPU can't effectively take any idle time to recompile host code blocks to its native VLIW architecture.
It can still do so, but it does less frequently which results in performance issues in Dolphin due to most code just running through the in-order
instruction decoder instead of the native VLIW architecture.
In one particular example, yielding moves the performance from 35-40FPS to 50-55FPS. So it is far more noticeable on Denver than any other system.
Of course once a triple or quad core Denver system comes out this will no longer be an issue on this architecture since it'll have a free core to do
all of this work.
Seems to be pretty high in the profile in some geometry-heavy games like The
Last Story, and the compiler-generated assembly is terrifyingly bad, so
SSE-ize it.
Just use regular boolean negation in our pixel shader's depth test everywhere except on Qualcomm.
This works around a bug in the Intel Windows driver where comparing a boolean value against true or false fails but boolean negation works fine.
Quite silly.
Should fix issues #7830 and #7899.
This shader constant was previously used for depth remapping in D3D and for pixel center correction. Now it only serves one purpose and the new name makes it clear.
This particular issue was fixed in the v66 (07-08-2014) development drivers from Qualcomm.
To make sure we cover all drivers that may or may not have the issue fixed, make sure to mandate v95 minimum to work around the issue.
The next commit is the actual work around for post processing for this.
Due to changes in how we render to the final framebuffer we no longer encounter this bug.
With the change to post processing being enabled at all times and no longer using glBlitFramebuffer, Qualcomm no longer has the chance to rotate our
framebuffer underneath of us.
This particular bug from our friends over at Qualcomm manifests itself due to our alpha testing code having a conditional if statement in it.
This is a fairly recent breakage this time around, it was introduced in the v95 driver which comes with Android 5.0 on the Nexus 5.
So to break this issue down; In our alpha testing code we have two comparisons that happen and if they are true we will continue rendering, but if
they aren't true we do an early discard and return. This is summed up with a fairly simple if statement.
if (!(condition_1 <logic op> condition_2)) { /* discard and return */ }
This particular issue isn't actually due to the conditions within the if statement, but the negation of the result. This is the particular issue that
causes Qualcomm to fall flat on its face while doing so.
I've got two simple test cases that demonstrate this.
Non-working: http://hastebin.com/evugohixov.avrasm
Working: http://hastebin.com/afimesuwen.avrasm
As one can see, the disassembled output between the two shaders is different even though in reality it should have the same visual result.
I'm currently writing up a simple test program for Qualcomm to enjoy, since they will be asking for one when I tell them about the bug.
It will be tracked in our video driver failure spreadsheet along with the others.
We will now rely on Memory::CopyFromEmu to do bounds checking.
Some games actually load palettes from 0x00000000, despite the
fact no valid palette data should ever be there.
Fixes Issue 7792.
The timing information is set on s_scaled_frame->pts, giving precise
timing information to the encoder. Frames arriving too early (less than
one tick after the previous frame) are droped. The setting of packet's
timestamps and flags is done after the call to avcodec_encode_video2()
as this function resets these fields according to its documentation.
This is good hygiene, and also happens to be required to build Dolphin
using Clang modules.
(Under this setup, each header file becomes a module, and each #include
is automatically translated to a module import. Recursive includes
still leak through (by default), but modules are compiled independently,
and can't depend on defines or types having previously been set up. The
main reason to retrofit it onto Dolphin is compilation performance - no
more textual includes whatsoever, rather than putting a few blessed
common headers into a PCH. Unfortunately, I found multiple Clang bugs
while trying to build Dolphin this way, so it's not ready yet, but I can
start with this prerequisite.)
ShaderConstantProfile and ShaderUid now have an empty implementation
of Write() that uses variadic templates instead of varargs. MSVC is now
able to inline and optimize away this when necessary.
It only ever did anything on 32-bit OS X.
Anyway, it wasn't even on the right functions, and these days
ABI_PushRegistersAndAdjustStack should handle maintaining the ABI
correctly.
It now affects the GPU determinism mode as well as some miscellaneous
things that were calling IsNetPlayRunning. Probably incomplete.
Notably, this can change while paused, if the user starts recording a
movie. The movie code appears to have been missing locking between
setting g_playMode and doing other things, which probably had a small
chance of causing crashes or even desynced movies; fix that with
PauseAndLock.
The next commit will add a hidden config variable to override GPU
determinism mode.
It's a relatively big commit (less big with -w), but it's hard to test
any of this separately...
The basic problem is that in netplay or movies, the state of the CPU
must be deterministic, including when the game receives notification
that the GPU has processed FIFO data. Dual core mode notifies the game
whenever the GPU thread actually gets around to doing the work, so it
isn't deterministic. Single core mode is because it notifies the game
'instantly' (after processing the data synchronously), but it's too slow
for many systems and games.
My old dc-netplay branch worked as follows: everything worked as normal
except the state of the CP registers was a lie, and the CPU thread only
delivered results when idle detection triggered (waiting for the GPU if
they weren't ready at that point). Usually, a game is idle iff all the
work for the frame has been done, except for a small amount of work
depending on the GPU result, so neither the CPU or the GPU waiting on
the other affected performance much. However, it's possible that the
game could be waiting for some earlier interrupt, and any of several
games which, for whatever reason, never went into a detectable idle
(even when I tried to improve the detection) would never receive results
at all. (The current method should have better compatibility, but it
also has slightly higher overhead and breaks some other things, so I
want to reimplement this, hopefully with less impact on the code, in the
future.)
With this commit, the basic idea is that the CPU thread acts as if the
work has been done instantly, like single core mode, but actually hands
it off asynchronously to the GPU thread (after backing up some data that
the game might change in memory before it's actually done). Since the
work isn't done, any feedback from the GPU to the CPU, such as real
XFB/EFB copies (virtual are OK), EFB pokes, performance queries, etc. is
broken; but most games work with these options disabled, and there is no
need to try to detect what the CPU thread is doing.
Technically: when the flag g_use_deterministic_gpu_thread (currently
stuck on) is on, the CPU thread calls RunGpu like in single core mode.
This function synchronously copies the data from the FIFO to the
internal video buffer and updates the CP registers, interrupts, etc.
However, instead of the regular ReadDataFromFifo followed by running the
opcode decoder, it runs ReadDataFromFifoOnCPU ->
OpcodeDecoder_Preprocess, which relatively quickly scans through the
FIFO data, detects SetFinish calls etc., which are immediately fired,
and saves certain associated data from memory (e.g. display lists) in
AuxBuffers (a parallel stream to the main FIFO, which is a bit slow at
the moment), before handing the data off to the GPU thread to actually
render. That makes up the bulk of this commit.
In various circumstances, including the aforementioned EFB pokes and
performance queries as well as swap requests (i.e. the end of a frame -
we don't want the CPU potentially pumping out frames too quickly and the
GPU falling behind*), SyncGPU is called to wait for actual completion.
The overhead mainly comes from OpcodeDecoder_Preprocess (which is,
again, synchronous), as well as the actual copying.
Currently, display lists and such are escrowed from main memory even
though they usually won't change over the course of a frame, and
textures are not even though they might, resulting in a small chance of
graphical glitches. When the texture locking (i.e. fault on write) code
lands, I can make this all correct and maybe a little faster.
* This suggests an alternate determinism method of just delaying results
until a short time before the end of each frame. For all I know this
might mostly work - I haven't tried it - but if any significant work
hinges on the competion of render to texture etc., the frame will be
missed.
videoBuffer -> s_video_buffer
size -> s_video_buffer_write_ptr
g_pVideoData -> g_video_buffer_read_ptr (impl moved to Fifo.cpp)
This eradicates the wonderful use of 'size' as a global name, and makes
it clear that s_video_buffer_write_ptr and g_video_buffer_read_ptr are
the two ends of the FIFO buffer s_video_buffer.
Oh, and remove a useless namespace {}.
This state will be used to calculate sizes for skipping over commands on
a separate thread. An alternative to having these state variables would
be to have the preprocessor stash "state as we go" somewhere, but I
think that would be much uglier.
GetVertexSize now takes an extra argument to determine which state to
use, as does FifoCommandRunnable, which calls it. While I'm modifying
FifoCommandRunnable, I also change it to take a buffer and size as
parameters rather than using g_pVideoData, which will also be necessary
later. I also get rid of an unused overload.
VertexLoader::VertexLoader was setting loop_counter, a *static*
variable, to 0. This was nonsensical, but harmless until I started to
run it on a separate thread, where it had a chance of interfering with a
running vertex translator.
Switch to just using a register for the loop counter.
- Lazily create the native vertex format (which involves GL calls) from
RunVertices rather than RefreshLoader itself, freeing the latter to be
run from the CPU thread (hopefully).
- In order to avoid useless allocations while doing so, store the native
format inside the VertexLoader rather than using a cache entry.
- Wrap the s_vertex_loader_map in a lock, for similar reasons.
To avoid FPRs being pushed unnecessarily, I checked the uses: DSPEmitter
doesn't use FPRs, and VertexLoader doesn't use anything but RAX, so I
specified the register list accordingly. The regular JIT, however, does
use FPRs, and as far as I can tell, it was incorrect not to save them in
the outer routine. Since the dispatcher loop is only exited when
pausing or stopping, this should have no noticeable performance impact.
- Factor common work into a helper function.
- Replace confusingly named "noProlog" with "rsp_alignment". Now that
x86 is not supported, we can just specify it explicitly as 8 for
clarity.
- Add the option to include more frame size, which I'll need later.
- Revert a change by magumagu in March which replaced MOVAPD with MOVUPD
on account of 32-bit Windows, since it's no longer supported. True,
apparently recent processors don't execute the former any faster if the
pointer is, in fact, aligned, but there's no point using MOVUPD for
something that's guaranteed to be aligned...
(I discovered that GenFrsqrte and GenFres were incorrectly passing false
to noProlog - they were, in fact, functions without prologs, the
original meaning of the parameter - which caused the previous change to
break. This is now fixed.)
For a long time, we've had ugly and inconsistent function names here as
helpers, names like "decodebytesRGB5A3rgba" which are absolutely
incomprehensible to understand. Fix this by introducing a new consistent
naming scheme, where the above function now becomes "DecodeBytes_RGB5A3".
Instead of having three separate functions and checking the tlutfmt in a
variety of places, just do it once in a helper method. This is already
for the slow path either in our Generic decoder or in our Software
renderer, so it doesn't matter that this is slower.
x64 will continue using the separate functions for speed.
The D3D / OGL backends only ever used RGBA textures, and the Software
backend uses its own custom code for sampling. The ARGB path seems to
just be dead code.
Since ARGB and RGBA formats are similar, I don't think this will make
the code more difficult to read or unable to be used as
reference. Somebody who wants to use this code to output ARGB can simply
modify the MakeRGBA function to put the shift at the other end.
This pulls all the duplicate code from TextureDecoder_Generic /
TextureDecoder_x64 out and puts it in a common file. Out custom font
used for debugging the texture cache is also pulled out and put in a
common "sfont.inc" file. At some point we should also combine this font
with the other six binary fonts we ship.
GetPC_TexFormat was never used. It was added in commit d02426a, with the
only user being commented out code. The commented out code was later
removed in 9893122, but the implementation stayed.
We were decoding to BGRA32 textures in our RGBA32 texture decoder. Since
this is the same for the BGRA32 decoder implementation, this is most
likely a copy/paste typo, rather than the texture actually being
bit-swapped. Fix this.
I'm not sure of any games that use the C14X2 texture format, so I'm not
sure this fixes any games, but it does make the code cleaner for when we
clean it up in the future, and merge some of these similar loops.
Separated out from my gpu-determinism branch by request. It's not a big
commit; I just like to write long commit messages.
The main reason to kill it is hopefully a slight performance improvement
from avoiding the double switch (especially in single core mode);
however, this also improves cycle calculation, as described below.
- FifoCommandRunnable is removed; in its stead, Decode returns the
number of cycles (which only matters for "sync" GPU mode), or 0 if there
was not enough data, and is also responsible for unknown opcode alerts.
Decode and DecodeSemiNop are almost identical, so the latter is replaced
with a skipped_frame parameter to Decode. Doesn't mean we can't improve
skipped_frame mode to do less work; if, at such a point, branching on it
has too much overhead (it certainly won't now), it can always be changed
to a template parameter.
- FifoCommandRunnable used a fixed, large cycle count for display lists,
regardless of the contents. Presumably the actual hardware's processing
time is mostly the processing time of whatever commands are in the list,
and with this change InterpretDisplayList can just return the list's
cycle count to be added to the total. (Since the calculation for this
is part of Decode, it didn't seem easy to split this change up.)
To facilitate this, Decode also gains an explicit 'end' parameter in
lieu of FifoCommandRunnable's call to GetVideoBufferEndPtr, which can
point to there or to the end of a display list (or elsewhere in
gpu-determinism, but that's another story). Also, as a small
optimization, InterpretDisplayList now calls OpcodeDecoder_Run rather
than having its own Decode loop, to allow Decode to be inlined (haven't
checked whether this actually happens though).
skipped_frame mode still does not traverse display lists and uses the
old fake value of 45 cycles. degasus has suggested that this hack is
not essential for performance and can be removed, but I want to separate
any potential performance impact of that from this commit.
This is required to make packing consistent between compilers: with u32, MSVC
would not allocate a bitfield that spans two u32s (it would leave a "hole").
The only possible functionality change is that s_efbAccessRequested and
s_swapRequested are no longer reset at init and shutdown of the OGL
backend (only; this is the only interaction any files other than
MainBase.cpp have with them). I am fairly certain this was entirely
vestigial.
Possible performance implications: efbAccessReady now uses an Event
rather than spinning, which might be slightly slower, but considering
the slow loop the flags are being checked in from the GPU thread, I
doubt it's noticeable.
Also, this uses sequentially consistent rather than release/acquire
memory order, which might be slightly slower, especially on ARM...
something to improve in Event/Flag, really.
This shouldn't affect functionality. I'm not sure if the breakpoint
distinction is actually necessary (my commit messages from the old
dc-netplay last year claim that breakpoints are broken anyway, but I
don't remember why), but I don't actually need to change this part of
the code (yet), so I'll stick with the trimmings change for now.
This is effectively unused, as the window handles that we pass to the
GLInterface are window handles for the frame which isn't ever a real
toplevel window. Host_UpdateTitle is what actually sets the proper title
on the render window.
Now that MainNoGUI is properly architected and GLX doesn't need to
sometimes craft its own windows sometimes which we have to thread back
into MainNoGUI, we don't need to thread the window handle that GLX
creates at all.
This removes the reference to pass back here, and the g_pWindowHandle
always be the same as the window returned by Host_GetRenderHandle().
A future cleanup could remove g_pWindowHandle entirely.
The framebuffer is no longer rotated the wrong way around in Qualcomm's latest development drivers.
They did something right, only took them over a year.
This catches most instances of configuration failures that can happen in a post processing shader.
Gives a user a helpful error message that lets them know what they have failed to set up correctly
This class loads all the common PP shader configuration options and passes those options through to a inherited class that OpenGL or D3D will have.
Makes it so all the common code for PP shaders is in VideoCommon instead of duplicating the code across each backend.
Fixes a bug where "Use Fullscreen" would initialize into exclusive fullscreen regardless of the borderless fullscreen setting.
Also relieves the need for the video renderer to check the borderless fullscreen setting each time.
The hack was needed because the Nvidia 3D Vision heuristics are documented to only support surfaces that are the same size as the backbuffer. This would be the case if you enabled the hack and selected the "Auto (Window Size)" internal resolution.
However, on recent drivers the same effect is achieved by selecting the "Auto (Multiple)" internal resolution. Therefore the hack is no longer required.
Also have the renderer remember its own fullscreen state. This is done to prevent a case where we exit exclusive fullscreen through the configuration and a focus shift at the same time. In this case the renderer would fail to detect that the fullscreen state was changed.
In the cases where we support the binding layout keyword, use it for more than binding UBO location.
This changes it so it is supported for samplers as well.
Instances when this is enabled is if a device supports GL_ARB_shading_language_420pack, or if it supports GLES 3.10.
ffmpeg 2.0 changed requirements for the FFV1 encoder and made them more strict,
requiring more fields of the input frame to be initialized. Explicitly setting
pixfmt, width and height solve the EINVAL issues with FFV1 encoding.
Original fix from http://ffmpeg.org/pipermail/libav-user/2013-October/005759.html
We are used to have a 1:1 mapping of GX vertex formats and the native (OGL + D3D) ones, but there are by far more GX ones.
This new cache maps them directly so that we don't flush on GX vertex format changes as long as the native one doesn't change.
The idea is stolen from galop1n.
- Isolate it into it's own namespace
- Shorten function names, the namespace self-documents.
- Just use the std I/O, we can just write directly to the stream for
logging.
Some headers where using #ifndef to guard being including multiple times. But most were using pragma once. So for consistency I changed them all to use pragma once.
When I dropped ARM from a generic target, this caused the vertexloader to try using the JIT path.
Instead of !_M_GENERIC, check for _M_X86 instead. Since it is only for the x86 target
These were only compiled in on Windows and x86_32.
They provided "optimized" copies and compares based on blocksizes for the AMD Athlon and Duron CPU families.
The code was taken from something that AMD provides with a as-is license.
Just get rid of this crap.
For whatever reason, the hardware doesn't do a full divide by 255, but
instead uses an approximation with shifting, similar to the way it is done
in TEV.
Noticed this while messing with EFB to RAM.
We were having an implicit conversion from integer to float, GLSL ES doesn't allow implicit conversion.
Changes it to a explicit conversion to float.
This matches how ARM handles their naming in their drivers for different models.
Really it's that way because both Mali-T6xx and Mali-T7xx fall under Midgard.
While everything else (except Mali-55) fall under Utgard.
This was a mistake of mine when translating floating point values to integer values.
Also, the max() part of that line was just completely redundant because the sign of an absolute value is always greater than or equal to zero.
Fixes issue 7178.
They are similar enough that they will share bugs with their drivers, so make them fall under the same Mali-Txxx umbrella of bug issues.
If there is ever a need in the future for having separate bugs depending on family, we can support that then.
Fixes issue 6990.
This uses a bit of templating to remove the duplicate code that is the CodeBlocks in each emitter headers.
No actual functionality change in this.
GLSL ES 3.10 adds implicit support for the binding layout qualifier that we use.
Changes our GLSL version enums to bit values so we can check for both ES versions easily.
This is undefined behavior in C++, and a clang warning suggests it is
actually producing bad code as a result:
../Source/Core/VideoCommon/BPFunctions.cpp:164:45: warning: comparison of constant 4294967295 with expression of type 'PEControl::PixelFormat' is always false [-Wtautological-constant-out-of-range-compare]
if (new_format == old_format || old_format == (unsigned int)-1)
Fixes all warnings on Android build except for what is in externals.
Removes a function from TextureDecoder_Generic since it is unused and generates a warning.
Metroid: Other M was the only game which required this field, but the
issue in that game can be fixed properly by enabling format change
emulation. Hence, there's no point in having this around anymore.
Fixes issue 6644.
Our defines were never clear between what meant 64bit or x86_64
This makes a clear cut between bitness and architecture.
This commit also has the side effect of bringing up aarch64 compiling support.
- remove unused variables
- reduce the scope where it makes sense
- correct limits (did you know that strcat()'s last parameter does not
include the \0 that is always added?)
- set some free()'d pointers to NULL
inline halfxb
So we know which is the first pixel by masking.
inline xl
inline xb a bit
inline yl
inline uv1.x shift
remove likely wrong guessed ternary operator
add pixel layout comment
inline xel
optimize the shifts a bit
inline xb
optimize shifts in a second step
extract xb
rename all variables
calculate cache line by position.x
Revert 5115b459f40d53044cd7a858f52e6e876e1211b4 "optimize the shifts a bit"
It seems I was wrong, the other way is the more natural.
use x_virtual_position instead of uv1.x for x_offset_in_block
This looks more natural and the offset should be masked anyway.
substitude factor with cache_lines
move 32bit logic in a conditional block
This option was known to break every second game and only boost a bit.
It also seems to be broken because of streaming into pinned memory and buffer storage buffers.
v2: also remove dlc_desc
Older Qualcomm drivers rotated the framebuffer 90 degrees and this fix didn't work.
Now for some obscene reason it rotates a full 180 degrees.
This can at least be worked around by flipping around the image on our end.
On Windows, nvidia don't give us their driver version, so we can't workaround any issues.
As buffer_storage is broken on some drivers, we wanted to disble it for them.
So we can't.
Luckyly only "some" released driver versions are affected as this extension is only available since some months. Let's hope that nobody have to use one of this driver version, else they will get a black screen ...
OSX has their own driver, so performance issues aren't shared with the nvidia driver (unlike the closed source linux and windows nvidia driver). So now they'll also use the MapAndSync backend like all other osx drivers.
fixes issue 6596
I've also cleaned up the if/else block selecting the best backend a bit.
The new chapter title in Paper Mario TTYD had a small graphical bug due to the new code because it read one extra pixel, this fixes it.
I hope this gets everything, I though I had checked most bugs and yet here I am, commit-spamming...
Fixes some remaining bbox related bugs in Mickey's Magical Mirror and a slight graphical glitch in Paper Mario: TTYD when flipping and Vivian as your companion (I've been scratching my head for days to find this one).
Instead of being vertex-based, it is now primitive (point, line or dissected triangle) based, with proper clipping.
Also, screen position is now calculated based on viewport values, instead of "guesstimating".
This fixes many graphical glitches in Paper Mario: TTYD and Super Paper Mario.
Also, the new code allows Mickey's Magical Mirror and Disney's Hide & Sneak to work (mostly) bug-free. I changed their inis to use bbox.
These changes have a slight cost in performance when bbox is being used (rare), mostly due to the new clipping algorithm.
Please check for any regressions or crashes.
This branch is the final step of fully supporting both OpenGL and OpenGL ES in the same binary.
This of course only applies to EGL and won't work for GLX/AGL/WGL since they don't really support GL ES.
The changes here actually aren't too terrible, basically change every #ifdef USE_GLES to a runtime check.
This adds a DetectMode() function to the EGL context backend.
EGL will iterate through each of the configs and check for GL, GLES3_KHR, and GLES2 bits
After that it'll change the mode from _DETECT to whichever one is the best supported.
After that point we'll just create a context with the mode that was detected
As we do lots of writes to *Iptr, the compiler isn't allowed to cache any shared variable (neither index nor Iptr itself).
This commit inlines Iptr + index into the index generator functions, so the compiler know that they are const.
We are used to render them out of order as long as everything else matches, but rendering order does matter, so we have to flush on primitive switch. This commit implements this flush.
Also as we flush on primitive switch, we don't have to create three different index buffers. All indices are now stored in one buffer.
This will slow down games which switch often primitive types (eg ztp), but it should be more accurate.
This "u32 components" is a list of flags which attributes of the vertex loader are present.
We are used to append this variable to lots of vertex generation functions, but some of them don't need it at all.
The usual way to handle this kind of request is to rise a flag which the gpu thread polls.
The gpu thread itself either generates the result or just write zeros if disabled.
After this, it rise another flag which says that this work is done.
So if disabled, we still have the cpu-gpu round trip time. This commit just returns 0 on the cpu thread
instead of playing ping pong...
Some information on this bug since this isn't quite true.
Seemingly with the v53 driver, Qualcomm has actually fixed this bug. So we can dynamically access UBO array members.
The issue that is cropping up is actually converting our attribute 'fposmtx' to an integer.
int posmtx = int(fpostmtx);
This line causes some seemingly garbage values to enter in to the posmtx variable.
Not sure exactly why it is failing, probably them just not actually converting the float to an integer and just handling the float directly as a integer.
So the bug is going to stay active with Qualcomm devices until we convert this vertex attribute from a float to a integer.
Let's talk a bit about this bug. 12nd oldest bug not fixed in Dolphin, it was a
lot of fun to debug and it kept me busy for a while :)
Shoutout to Nintendo for framework.map, without which this could have taken a
lot longer.
Basic debugging using apitrace shows that the heat effect is rendered in an
interesting way:
* An EFB copy texture is created, using the hardware scaler to divide the
texture resolution by two and that way create the blur effect.
* This texture is then warped using indirect texturing: a deformation map is
used to "move" the texture coordinates used to sample the framebuffer copy.
Pixel shader: http://pastie.org/private/25oe1pqn6s0h5yieks1jfw
Interestingly, when looking at apitrace, the deformation texture was only 4x4
pixels... weird. It also does not have any feature that you would expect from a
deformation map. Seeing how the heat effect glitches, this deformation texture
being wrong looks like a good candidate for the problem. Let's see how it's
loaded!
By NOPing random calls to GXSetTevIndirect, we find a call that when removed
breaks the effect completely. The parameters used for this call come from the
results of methods of JPAExTexShapeArc objects. 3 different objects go through
this code path, by breaking each one we can notice that the one "controlling"
the heat effect is the one at 0x81575b98.
Following the path of this object a bit more, we can see that it has a method
called "getIndTexId". When this is called, the returned texture ID is used to
index a map and get a JPATextureArc object stored at 0x81577bec.
Nice feature of JPATextureArc: they have a getName method. For this object, it
returns "AK_kagerouInd01". We can probably use that to see how this texture
should look like, by loading it "manually" from the Wind Waker DVD.
Unfortunately I don't know how to do that. Fortunately @Abahbob got me the
texture I wanted in less than 10min after I asked him on Twitter.
AK_kagerouInd01 is a 32x32 texture that really looks like a deformation map:
http://i.imgur.com/0TfZEVj.png . Fun fact: "kagerou" means "heat haze" in JP.
So apparently we're not using the right texture object when rendering! The
GXTexObj that maps to the JPATextureArc is at offset 0x81577bf0 and points to
data at 0x80ed0460, but we're loading texture data from 0x0039d860 instead.
I started to suspect the BP write that loads the texture parameters "did not
work" somehow. Logged that and yes: nothing gets loaded to texture stage 1! ...
but it turns out this is normal, the deformation map is loaded to texture stage
5 (hardcoded in the DOL). Wait, why is the TextureCache trying to load from
texture stage 1 then?!
Because someone sucked at hex.
Fixes issue 2338.
At the moment, custom textures with:
- invalid mipmap size
- invalid aspect ratio
- non-fractional scaling factors
are allowed. But they can't be loaded fine by the backend, so generate a warning if someone trys to load them.
fixes issue 6898
OpenGL defaults are GL_REPEAT, which is even more unlikely than GL_CLAMP_TO_EDGE.
As I can't test the behavoir of the real hardware, I changed it to how it works before,
but I guess just clip the texture makes more sense.
This flag wasn't cleared at all, so we set our constants dirty every time...
This could fix some performance regressions because of revision 6798a4763e
This removes the redundant code and also implements this feature for OSX and Wayland.
But so it's dropped for non-wx builds...
imo DolphinWX still isn't the best place for this, but now it's in the same file as all other hotkeys. Maybe they'll be moved to InputCommon sometimes at once ...
Now OGL doesn't rely on WX for PNG saving.
FlipImageData supports (pixel data len > 3) now.
TextureToPng is now in ImageWrite.cpp/h
Video Common depends on zlib and png.
D3D no longer depends on zlib and png.
Revert "Actually, filename really does need to be a parameter because of some random debug thing."
Revert "fix non-HAVE_WX case"
Revert "Handle screenshot saving in RenderBase. Removes dependency on D3DX11 for screenshots (texture dumping is still broken)."
This reverts commits 00fe5057f1, 74b5fb3ab4, cd46138d29 and 5f72542e06 because taking screenshots in D3D still crashed for me so there was no point in the code changes (which I found ugly anyway).
This reverts commit 6cece6b486.
In fact, there was a _huge_ speedup on lots of games (mostly on nvidia+ogl), but there are some crashes on D3D.
I have to fix this crash and then I'll commit something like this again :-)
Conflicts:
Source/Core/VideoCommon/Src/TextureCacheBase.cpp
We often need the same native texture objects for new textures. This commit
try to avoid destroying and creation of this textures by pooling them.
This should be a big performance gain for some efb2ram games as they may
overwrites partially a cached texture (which would be deleted) and afterwards
try to read it.
Creating/destroying sounds like an easy task, but it isn't. eg the nvidia ogl
driver synchonize their threads do avoid use-after-free issues.
D3D doesn't allow bigger viewports than rendertargets. But flipper does, so the viewport will be clipped and the transformation matrix will be changed.
This was done in the D3D backend itself. This is now moved into VideoCommon. This don't reduce code, but in this way, VideoCommon doesn't depend on the backends.
This isn't needed for both OGL+D3D11 as they support sample shading directly. So we
could use the common MSAA util shaders instead of writing custom ones.
* Currently there is no DEBUGFAST configuration. Defining DEBUGFAST as a preprocessor definition in Base.props (or a global header) enables it for now, pending a better method. This was done to make managing the build harder to screw up. However it may not even be an issue anymore with the new .props usage.
* D3DX11SaveTextureToFile usage is dropped and not replaced.
* If you have $(DXSDK_DIR) in your global property sheets (Microsoft.Cpp.$(PlatformName).user), you need to remove it. The build will error out with a message if it's configured incorrectly.
* If you are on Windows 8 or above, you no longer need the June 2010 DirectX SDK installed to build dolphin. If you are in this situation, it is still required if you want your built binaries to be able to use XAudio2 and XInput on previous Windows versions.
* GLew updated to 1.10.0
* compiler switches added: /volatile:iso, /d2Zi+
* LTCG available via msbuild property: DolphinRelease
* SDL updated to 2.0.0
* All Externals (excl. OpenAL and SDL) are built from source.
* Now uses STL version of std::{mutex,condition_variable,thread}
* Now uses Build as root directory for *all* intermediate files
* Binary directory is populated as post-build msbuild action
* .gitignore is simplified
* UnitTests project is no longer compiled
D3D9 only supports 8 texcoords. But we need a new one for ppl, so we just store it in the first 4 texcoords in the free 4th component.
This isn't needed for both d3d11 and ogl3, so just remove it.
This isn't needed in VertexShaderManager as it's still in the old dirty flag way.
But it's very importend for PixelShaderManager as some float4s aren't initialized as 0.0f
The old way was to use a dirty flag per setter. Now we just update the const buffer per setter directly.
The old optimization isn't needed any more as the setters don't call the backend any more.
The follow parts are rewritten:
Alpha
ZTextureType
zbias
FogParam
FogColor
Color
TexDim
IndMatrix
MaterialColor
FogRangeAdjust
Lights
The upload in the backend isn't done, it's just pushed by the mostly removed SetMulti*SConstant4fv.
Also no optimizations was done on VideoCommon side, but I can start now :-)
Sorry for the hacky way, but I think this is a nice (working) snapshot for a much bigger change.
It changes the string in the Android backend select to just OpenGL ES.
Adds a check in the Android code to check for Tegra 4 and to enable the option to select the OpenGL ES backend.
Adds a DriverDetails bug under BUG_ISTEGRA as a blanket case of Tegra 4 support.
The changes that effects most lines in this change. Removing all float suffixes in the pixel/vertex/util shaders since OpenGL ES 2 doesn't support float suffixes.
Disables the shaders for reinterpreting the EFB format since Tegra 4 doesn't support integers.
Changes GLFunctions.cpp to grab the correct Tegra extension functions.
Readds the GLSL 1.2 'hacks' as GLSLES2 'hacks' since they are required for GLSL ES 2
Adds a GLSLES2 to the GLSL_VERSION enum.
Disable the SamplerCache on Tegra since Tegra doesn't support samplers...
Enable glBufferSubData on Tegra since it is the only mobile GPU to correctly work with it.
Disable glDrawRangeElements on Tegra since it doesn't support it, This uses glDrawElements instead.
- Call ABI_AlignStack even on x86-64.
- Have ABI_AlignStack respect the difference in current alignment
between the root JIT function, which has a prolog, and
ProtectFunction thunks, which do not. This was causing many games
to crash on start on OS X. Since this might otherwise mean changing
the stack pointer before every call...
- Have one prolog/epilog function rather than two (one of which
definitely did not do what it was thought to do), and make it
actually work like a normal one, so that the stack frame shows up
properly in the debugger. There should be no performance impact.
attn is sometimes very big (eg 1e27), so attn*attn doesn't fit into a float.
So the funny part here is: 0.0 * (1e27*1e27) = 0.0 * Inf = NaN
As the shader compiler is allowed to change the order of multiplications,
this issue isn't fixed completely.
Use strings internally, use a multimap and std::function for callbacks (instead
of a flat vector + loop over the vector to find the right callback type), fix
coding style issues. Simplify MainAndroid code a bit.
It's possible to configure to use the vertex color as lightning source without enabling the vertex color at all.
The old implementation will use zero, but it seems to be wrong (prooven by THPS3), more likely is to disable
the lightning and just return the global color.
This fixes THPS3 on OpenGL, but it isn't verifed on hardware
- rewrite loops to not use divisions and multiplications
- remove warnings as the current implementations seems to be correct (ignore additional vertices)
use always ppd is a huge gpu performance drop: 20%-50%
and always disable it cause some rendering issues
so there is an option again
But this time it's called "Fast Depth Calculation"
fix issue 6282
The Last Story seems to render a fan with two vertices. It is non-sense as it
shouldn't do anything, but the code underflows at (u32)numVerts-3
Block braces on new lines.
Also killed off trailing whitespace and dangling elses.
Spaced some things out to make them more readable (only in places where it looked like a bit of a clusterfuck).
See Render.cpp, PixelShaderGen.cpp, and PixelShaderManager.cpp for most of the changes.
See VertexShaderManager.cpp for a logging typo fix.
See SWRenderer.cpp for a small typo fix for a message that gets swprintf'd in DrawDebugText.
See SWVertexLoader.cpp for a typo fix of an assert message.
Should slightly improve the readability of some of those files.
This commit mainly elaborates on some messages a little more. Also fixes some typos that slipped through the last commit.
A large change in text can be seen in EXI_DeviceMemoryCard.cpp. I added more info as to why a write to a memory card may fail. (This actually was a reason I was unable to write to a memcard recently).
Elaborations can be seen in WGL.cpp
I did change some comments in some files that I was correcting logging messages in, however this is only if I spot a typo or if an abbreviation is lower-cased. Even in that case, the amount of changes done to comments is very minimal.
Sorry for a direct commit to the main branch but i need fast feedback, and i don't want to leave problematic code in the main branch for a long time.
if this approach does not work for the drivers with problems will transform dual source blend to an option in the D3D9 backend.
I appreciate the help of the people that tested my last commit and thanks to neobrain for pointing this solution.
Convert all quads+triangles into trangle_strip and uses primitive restart to split them.
Speed up triangle_strip, but slows down all others primitive formats.
Only implemented in ogl.
this implementation does not work in windows xp (sorry no support for dual source blending there).
this should improve speed on older hardware or in newer hardware using super sampling.
disable partial fix for 4x supersampling as I'm interested in knowing the original issue with the implementation to fix it correctly.
remove the deprecation label from the plugin while I'm working on it.
* Fast-EE:
Forced the exception check only for ARAM DMA transfers. Removed the Eternal Darkness boot hack and replaced it with an exception check.
Reverted rd76ca5783743 as it was made obsolete by r1d550f4496e4.
Removed the tracking of the FIFO Writes as it was made obsolete by r1d550f4496e4.
Forced the external exception check to occur sooner by changing the downcount.
# By skidau (30) and Pierre Bourdon (1)
* FIFO-BP: (31 commits)
Set g_bSignalTokenInterrupt on the main thread. Fixes the random hang in Harry Potter: Prisoner of Azkaban.
Used a scheduled event to generate the ARAM DMA interrupt if the DMA is greater than a certain size. Fixes NFS:HP2 GC.
Bumped up the disc transfer speed enough to prevent audio stuttering in Gauntlet: Dark Legacy.
Enabled Synchronise GPU on "SPEED CHALLENGE - Jacques Villeneuve's Racing Vision". Required to go in-game.
Added direct GameCube controller commands to the Serial Interface emulation. Fixes the controls in MaxPlay Classic Games Volume 1 and the Action Replay disc.
Increased the FIFO buffer size to 2MB from 1MB. Fixes Killer 7's Angel boss.
Used an immediate GenerateDSPInterrupt when transferring data from ARAM to MRAM and a scheduled DSP interrupt when transferring data from MRAM to ARAM.
Fixes the audio cutting in and out in the Resident Evil GC games using DSP HLE. Triggered the ARAM interrupt by the scheduler instead of directly in function.
Implemented proper timing for the sample counter in the AudioInterface, removing the previous hack. Cleaned up some of the audio streaming code.
Skipped the EE check if there is a CP interrupt pending.
Disabled "Speed up disc transfer" from the ZTP GC game ini.
Removed the disc seek times for GC games and removed the disc speed option on Wii games. Checked for external exceptions only in mtmsr.
Delayed the interrupts in the EXI Channel.
Merge aram-dma-fixes (r76a13604ef49b522281af75675f044d59a74e871)
Added a patch that bypasses the FIFO reset code in Wallace and Gromit: Project Zoo, allowing it to go in-game.
Made vertex loading take constant time.
Increased the cycle time of the vertex command. Fixes "Speed Challenge: Jacques Villeneuve's Racing Vision".
Moved the setting of the Finish interrupt signal back to the main thread as it was causing Wii games like Resident Evil 4 (Wii) to hang.
Profile stores, fp stores and ps stores only to the fifo write addresses list. This should make the JIT a little faster as it will not be checking for external exceptions unnecessarily.
...
Conflicts:
Source/Core/VideoCommon/Src/PixelEngine.cpp
Adds support for PE performance metrics.
Used in Super Mario Sunshine's "Scrubbing Sirena Beach" level to determine when enough goop has been cleaned up to finish the level.
Also used in TimeSplitters: Future Perfect to determine the appearance of flares around light sources (e.g. sun).
OpenGL and D3D11 only. D3D9 support unlikely to be added unless anyone bothers to do the work.
Initial work and D3D11 support by me. Kudos go to Billiard for adding the OpenGL support and reviving development of this branch that way :D
Slightly (~7%) decreases performance when performance metrics are used (and only then).
Fixes issue 1498.
Fixes issue 5368.
# By Jordan Woyak (46) and others
# Via Jordan Woyak (2) and others
* master: (70 commits)
Fixes two memory leaks, one is pretty bad for OSX. Yell at pauldachz if this doesn't work. Or... say thanks.
Added a BluetoothEnumerateInstalledServices call so that the wiimote remembers the pairing.
Make ARMJit core default CPU core on ARM architecture
Fix a StringUtil regression from the arm-noglsl merge
Small improvement to cmpli/cmpi in ARMJit.
Merge latest ArmEmitter changes from ppsspp while we're at it.
Ah. I blame vim on this typo entirely.
Add disabled code for authenticating wiimotes on Windows.
Add the missing FPR cache
Buildfix.
Yell at the user if they change window size while dumping frames, and some other avi dumping stuff.
Not sure if this is the right way to handle this, but it makes the save states perfectly stable. That's all that really matters, right?
Abort loading states from incompatible graphics backends.
ARM Support without GLSL
Improve VideoSoftware save states. They are fairly stable, but not perfect. OpcodeDecoder::DoState() needs to be fixed.
Begin implementing save states to video software. Kind of works, sometimes.
Make error message for loading save state with wrong dsp engine shorter.
Abort load state if it uses a different dsp engine, instead of crashing.
Update the gameini of F-zero. Efb to Ram is no longer the default choice.
fix last commit by neobrain
...
Conflicts:
Source/Core/VideoCommon/Src/Fifo.cpp
This reverts commit 4653adecf1.
Also dx9 isn't able to hanlde more than 11 varying registers.
More frustrating is the lightning issue by this commit. I don't know why it happens...
# By Jordan Woyak (9) and others
* master:
Fixed a buffer overflow in the OpenAL buffer.
TextureCache: Fix D3D backends crashing when a game uses multiple 1x1-sized LODs.
WII_IPC_HLE_Device_FileIO: don't rebuild the filename on every operation.
Some cleanup of CWII_IPC_HLE_Device_FileIO: The real file was never kept open for longer than a single operation so there was no point in dealing with it in DoState. Saving the real path in the savestate was also probably a bad idea. Savestates should be a bit more portable now.
Removing destination on rename when source isn't present doesn't make sense. IOCTL_RENAME_FILE still might not be totally correct.
Change some CNANDContentLoader logic to what was probably intended. Kills some warn logs when opening Dolphin.
Let's not CreateDir an empty string every time CreateFullPath is used, logging an error every time.
Fix a memleak. Probably/maybe improve USBGecko performance.
Remove the core count from the cpu info OSD message. It was often wrong and not rather important.
Use omp_get_num_procs to set the number of OpenMP threads rather than our core count detection.
Bulk send TCP data to the client with the emulated USB Gecko.
Added the ability to reverse the direction of the force feedback by allowing negative range values.
Changes/cleanup to TextureCache::Load and other mipmap related code. The significant change is what is now line 520 of TextureCacheBase.cpp: ((std::max(mipWidth, bsw) * std::max(mipHeight, bsh) * bsdepth) >> 1) to TexDecoder_GetTextureSizeInBytes(expanded_mip_width, expanded_mip_height, texformat);
The significant change is what is now line 520 of TextureCacheBase.cpp:
((std::max(mipWidth, bsw) * std::max(mipHeight, bsh) * bsdepth) >> 1)
to
TexDecoder_GetTextureSizeInBytes(expanded_mip_width, expanded_mip_height, texformat);
Fixes issue 5328.
Fixes issue 5461.
# By Jordan Woyak (24) and others
# Via Jordan Woyak (3) and others
* master: (66 commits)
Reduce some DI command delays. Fix DKCR hanging with DSP HLE. My other games continue to work.
Video_Software: Fix ZComploc option breaking stuff.
Video_Software: Fix the ZFreeze option doing nothing.
Video_Software: Toggable zfreeze and early_z support for testing.
Fix header guard and definitions not being set to 1
Add the option to turn on only the EGL interface to use desktop OpenGL with it.
Change the ugly "no banner" banner to the sexy "X" from the website.
Fix a crash in the FifoPlayer dialog.
Use different reply delays for various DI commands. Fixes issue 5983.
Revert "[bugfix] DX9::TextureCache: Use max_lod instead of min_lod where necessary."
Fix some potential issues when blending on EFB formats without alpha. Clean up state transition tables.
Disable play and record buttons if an iso was selected, but is later deselected.
Disable start/play recording buttons when no iso is selected.
Only delay DI and fs IPC replies. Fixes issue 5982.
Fix compilation with SDL2. (based on a patch from matthewharveys) Fixes issue 5971.
"Fix" using SDL from externals.
Clean up SDL includes a bit. Maybe fix an SDL2 problem.
Number "unknown" axes in OSX rather than call them all "unk".
Revert "Only delay DI command replies." Fix "Wii Party" again.
Hopefully make wiimote speaker less crappy.
...
Conflicts:
Source/Core/VideoCommon/Src/PixelShaderGen.cpp
Source/Plugins/Plugin_VideoSoftware/Src/SWRenderer.cpp
immediate-removal is a new created branch seperated from master but reverted the revert of immediate-removal
so we get less conflicts by merging
Depth calculations are always done in the pixel shader now.
Due to the unpredictability of our zcomploc hacks this commit probably changes the behavior of some games which use zcomploc.
That's what would've been done in the next TCB::Load() call, anyway.
Fixes issue 5742.
Additionally, change efb copies to specify 1 as the number of mipmaps because that makes more sense than anything else.
change naming in all the backends vertex managers to make more easy to continue with the merge an some future improvements.
please test this as i'm interested in knowing the performance in linux and windows with the different hardware platforms.
So to compensate lets bring back some speed to the emulation.
change a little the way the vertex are send to the gpu,
This first implementation changes dx9 a lot and dx11 a little to increase the parallelism between the cpu and gpu.
ogl: is my next step in ogl is a little more trickier so i have to take a little more time.
the original concept is Marcos idea, with my little touch to make it even more faster.
what to look for: SPEEEEEDDD :).
please test it a lot and let me know if you see any problem.
in dx9 the code is prepared to fall back to the previous implementation if your card does not support the amount of buffers needed.
So if you did not experience any speed gains you know where is the problem :).
for the ones with more experience and compression of the code please test changing the amount and size of the buffers to tune this for your specific machine.
The current values are the sweet spot for my machine.
All must Thanks Marcos, I hate him for giving good ideas when I'm full of work.
PS and VS making. Untested and won't work for now.
Add in program shader cache files.
Readd NativeVertexFormat stuffs.
Add in PS and VS cache things.
SetShaders in places.
Fixed EFB cache index computations in OpenGL renderer.
The previous computation was very likely to go out of array bounds,
which could result in crashes on EFB access.
Also, the cache size was rounded down instead of up. This is a problem
since EFB_HEIGHT (528) is not a multiple of EFB_CACHE_RECT_SIZE (64).
Very useful to compare performance between two builds, check the impact of
a configuration option, etc. FPS log is stored in User/Logs/fps.txt and is
reset each time you launch a game. Only enabled if you check the "Log FPS
to file" option in your graphics settings.
Could be improved a bit: currently logs only every 1s (so you can't really
see small variations), maybe output more infos to the fps.txt like
average/stddev (but Excel/Libreoffice/Google Docs can compute that easily
too).
Super Mario Sunshine is using a cool trick: To determine how much goop has been cleaned in ep. 6 of Sirena Beach, it counts the number of pixels that are input to the blending stage. For that it's using the PE performance registers ;)
Fixes issue 1498.
Reason:
- It's wrong, zcomploc can't be emulated perfectly in HW backends without severely impacting performance.
- It provides virtually no advantages over the previous hack while introducing lots of code.
- There is a better alternative: If people insist on having some sort of valid zcomploc emulation, I suggest rendering each primitive separately while using a _clean_ dual-pass approach to emulate zcomploc.
This reverts commit 0efd4e5c29.
This reverts commit b4ec836aca.
This reverts commit bb4c9e2205.
This reverts commit 146b02615c.
the intent is to replace the haphazard scheduling and finger-crossing associated with saving/loading with the correct and minimal necessary wait for each thread to reach a known safe location before commencing the savestate operation, and for any already-paused components to not need to be resumed to do so.
This branch reduces the number of useless state flushes in the video
emulation layer by checking whether a BP/XF change will have an effect
or not. Greatly reduces the number of GL calls per frame.
Thanks to degasus for his help!
When a game writes the same value that was already configured to a BP
register, Dolphin previously flushed the GPU pipeline and reconfigured
the internal video state (calling SetScissor/SetLineWidth/SetDepthMode).
Some of these useless writes still need to perform actions, for example
writes to the EFB copy trigger or the texture preload registers (which
need to reload the texture from memory).
This adds an "Analyzer" tab to the fifoplayer dialog which allows to conveniently browse through all register pokes that are being sent by the game each frame.
There's also a search function, but it doesn't work all that well for anything but simple searches at the moment. However, I'm merging this anyway since I'm not sure if I'm going to finish this.
Note that due to recent fifo changes, it's not yet possible to run fifoplayer in dual-core mode.
please test for regressions, speed and for other issues fixed, as a example, the black color in water splash in super mario galaxy are fixed with this rev.
please as soon as yo find a bug let me know.
a little code cleaning to avoid duplicated execution of AlphaPreTest and a little correction to some comments from the previous commits.
this change must behave exactly like last revision, if something is broken please let me know
to marcosvitali.
Added an external exception check when the CPU writes to the FIFO. This allows
the CPU time to service FIFO overflows. Fixes random hangs caused by FIFO
overflows and desyncs like in "The Last Story" and "Battalion Wars 2". Thanks
to marcosvitali for the research.
Added some code to unlink invalidated blocks so that the recompiled block can be
linked (speed-up).
This release still fixed the hangs produced by fifo overflow without sacrifice
performance. For example you can test Tutorial moves at the beginning of The last history now
is fluid 30/60.
Fixed possibles random hangs in DC mode.
Fixed hangs in DC mode in (Simpsons, Monkey Island, Pokemon XD, etc)
Implemented accurate management of Pixel Engine Interrupts. Now the GPU loop
is stopped when a PE Interrupt needs to be managed and resumed when Pixel Engine
finish.
Fixed Metroid Prime 3 and 2 desync. And other games with desync because of
FIFO Reset. That happens because FIFO_RW_DISTANCE_HI must be written first, for checking
fifo.CPReadWriteDistance == 0, so some fifo resets was not managed in the right
way.
Fixed Super Monkey Ball in some cases when the game write the
WriteReadDistance need to be safe like the SafeCPRead.
Improved the CheckException for the GatherPipe writes in JIT, now only the
External Exceptions are processed.
Fixed definitely Pokemon XD in dual core mode. This game is doing something
not allowed. It attach to CPU the same fifo attached to the GPU in multibuffer
mode. I added a check to prevent overwrite the GPU FIFO with the CPU FIFO. If
the game do that on breakpoint the solution can fail.
Fixed ReadWriteDistance calc when CPRead > CPWrite.
Added Token and Finish cause to GP Jit checking.
Additional cleanup in CommandProcessor.
Fixes issue 5209
Fixes issue 5055
Fixes issue 4889
Fixes issue 4061
Fixes issue 4010
Fixes issue 3902
This fix is not related with the previous commits, but the previous commits help me to see that because in the new scenery SMB was hanging. May be in the past also doesn't boot some times because of that.
Please Test FZero boot also. Thanks.
The external exceptions in dolphin are checking frequently but is different to real HW, so sometime the game is in a loop checking GPU STATUS, the exceptions doesn't checked, and the game hang.\
For solve this I need a trick: still waiting for the exception handler be linked but if CommandProcecsor is reading the GPStatus, resume this.
This fixed "TimeSplitters: Future Perfect" broken in the Revision c2e6fdf09f and surely others games.
That happens because FIFO_RW_DISTANCE_HI must be written first, for checking fifo.CPReadWriteDistance == 0, so some fifo resets was not managed in the right way.
I didn't test Metroid 2 desync reported in Issue 4336 but I think is the same.
About the flickering in MP2, I don't know for my is not related or yes, but you can test anyway.
Fixed Issue 3902
Well now the FIFO is 99.99% finished :)
- I've fixed hangs in DC mode in (Simpsons, Monkey Island, Pokemon XD, etc)
- I've implemented accurate manage of Pixel Engine Interrupts, now the GPU loop is stopped when a PE Interrupt needs to be managed and resume when Pixel Engine finish,
I think now, the Fifo in DC mode is more accurate than SC mode. :)
Time to close the big fifo Issue 3694 (snif), please if you have a possible fifo issue report this like a game issue.
I was working with Skid_AU together, especially thanks for him.
Test a lot all games, and compare the performance with the master maybe this accuracy has a cost (not a lot).
I think now the fifo is very stable, overflow fixed, random hang fixed, if you have a game with a hang with this rev and not in master please report this.
For example you can test Tutorial moves at the beginning of The last history now is fluid 30/60.
Shuffle2: I've delete the hacky line, I think is not necessary anymore. Additional some clean in CommandProcessor.
Please test The Last Story and others games affected in the previous commits and give me a feedback.
PS and VS making. Untested and won't work for now.
Add in program shader cache files.
Readd NativeVertexFormat stuffs.
Add in PS and VS cache things.
SetShaders in places.
Fixed EFB cache index computations in OpenGL renderer.
The previous computation was very likely to go out of array bounds,
which could result in crashes on EFB access.
Also, the cache size was rounded down instead of up. This is a problem
since EFB_HEIGHT (528) is not a multiple of EFB_CACHE_RECT_SIZE (64).
- Fixes all (I hope) BBox-related unknown pointer crashes.
- Fixes wrong BBox values with Frame Skip on (and the resulting unknown pointer crashes).
- Fixes a small oversight on the change I made to the ISO Properties dialog.
This should also be a (very very little) bit faster than the previous version.
Should fix most graphical issues with Paper Mario: TTYD and Super Paper Mario. Fixes issue 360.
Since only those two games seem to require BBox support, and as per ector's suggestion, BBox is only enabled for those two games.
BBoxes and Display List Caches don't get along too well, causing Paper Mario: TTYD to hang during certain effects where BBoxes are used. For now, I disabled DList Cache for the Paper Mario games, hopefully both will be compatible in the future.
zcomploc is a feature of the BP which switch depth test from before the alpha
test to after the alpha test. This way, transparent fragments are written to
the depth buffer too.
The current implementation is quite hacky and does not cover all cases but is
enough to fix problems in a lot of game. A complete implementation would
require a multipass rendering method and is attempted in the
zcomploc-experimental branch.
According to testers feedback, fixes bugs in the following games:
- Baten Kaitos
- Baten Kaitos Origins
- 007: Everything or Nothing
- Ty the Tasmanian Tiger
- Tony Hawk's Pro Skater 3
And probably other games too.
Conflicts (because of new-shadercache-uids):
Source/Core/VideoCommon/Src/PixelShaderGen.cpp
In previous revisions of Dolphin, changing texture generation settings (via
GX_SetTexCoordGen for example) did not regenerate the vertex shader and flush
the vertices. Adding this flush should fix texture problems in a few game, for
example in Superman: Shadow of Apokolips:
Before: http://i.imgur.com/mHmfb.jpg
After: http://i.imgur.com/2ThES.png
- don't load shader cache from disk in d3d9/11 if shader debugging is enabled (we won't have any info about the source shader code otherwise, etc)
- dump shader source codes on safe UIDs mismatch
Thanks to LordMark and [SS] for reporting those to me ;)
some little corrections in dlist cache code, i thing they will not have effect in the remaining issues but now the code behaves more correctly in some situations.
i discover the solution for the remaining issues bur requires some changes in the vertex translator so will work on them after the upcoming Release.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7556 8ced0084-cf51-0410-be5f-012b33b47a6e
- Fast Mipmaps enabled
- Safe Texture Cache enabled at "fast"
- showing shader errors by default
- using Native as internal resolution
- using EFB to Texture by default
- ignoring EFB format changes by default (turns out r7436 made this in Super Mario Sunshine unnecessary anyway :P)
Any game which depends on other settings has that specified in the game ini, so I think this should be fine.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7497 8ced0084-cf51-0410-be5f-012b33b47a6e
Show FPS, Input Display, Statistics, Projection Statistics, EFB Copy Regions, Dump Textures, Dump EFB target, Dump Frames, Free Look, Use FFV1, Show Shader Errors, Texture Format Overlay, WireFrame, Hotkeys and Adapter
- Also removed various game ini config values which were still parsed from the Video section.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7489 8ced0084-cf51-0410-be5f-012b33b47a6e
Note: Newest nVidia drivers break Dolphins 3D (270.xx), not sure why yet, gonna ask around. It really confused me though when trying to fix this issue, spent at least an hour trying to figure out what the hell was going on. I eventually realised a build I knew worked in 3D was having issues...
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7487 8ced0084-cf51-0410-be5f-012b33b47a6e
- Added final support to string type (planned implementation reaches 100%)
- Added support to post processing shaders (OpenGL backend)
- Code Maintenance:
- Reorganized the class hierarchy
- General polishing... deleted some redundant code
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7437 8ced0084-cf51-0410-be5f-012b33b47a6e
Also, if the render window size changes while frame dumping, scale the resulting video to prevent clipping on linux. This is a complete fix for issue 4316 on linux. I don't know how to implement this on windows though.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7412 8ced0084-cf51-0410-be5f-012b33b47a6e
Hotkeys now work when running 3D Vision. You've just got to select the render window first (as in click on the fullscreen image), don't ask me why, it's something to do with RendererHasFocus in Frame.cpp. Annoyingly, escape crashes the emulator and the screenshot button doesn't do anything, though I guess any other hotkey should be working. I wouldn't usually commit this but I remember someone really wanting to be able to connect their Wiimote while in 3D, so I figured I might aswell.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7396 8ced0084-cf51-0410-be5f-012b33b47a6e
- Removed recent ghanges about separate per-game Video Settings
- Added "One Window 3-State Support" (very close to full implementation)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7391 8ced0084-cf51-0410-be5f-012b33b47a6e
It now works in the debug builds, but not in release builds for some reason. I'm not used to big projects like this so it's probably something obvious that I'm missing. Exclusive full screen mode seems to need the fullscreen resolution for the backbuffer when the DX device is created, so I used adapters[adapter].resolutions[f].xres and yres to get that, but in release builds, that gives me the windowed resolution (in debug it give me the fullscreen res, like I believe it's supposed to), which it reallly doesn't like.
It's not a difficult fix probably, but I've been staring at my monitor in confusion for too long now and need to get to bed. If anyone wants to test, you can still enable the 3D Vision option, it just wouldn't be 3D (obviously).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7385 8ced0084-cf51-0410-be5f-012b33b47a6e
With wxWidgets 2.9.2 now in Externals, SCons was the only
external dependency other than Xcode and building just SCons
from Macports caused a fair mount of extra stuff to be built
unnecessarily, so just include a copy of scons-local.
Always build wxWidgets from Externals which allows us to get
rid of the home-grown autoconf machinery in SconsTests.
The main use case for args.cache was the wxconfig option,
which is no longer relevant after the above changes and
caching options has a lot of potential for confusion,
especially since other build systems don't tend to do this,
so we no longer do it now.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7381 8ced0084-cf51-0410-be5f-012b33b47a6e
* Moved the config validity check to reflect the emulator enforced settings at the config dialog
* Removed some obsolete validity checks
* Corrected "EnablePixelLighting" typo in the ini files
Fixes issue 4277.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7374 8ced0084-cf51-0410-be5f-012b33b47a6e
Second:
A experiment. implemented parallelization in texture decoding using openmp. is most a experiment to test the performance in different os/plataforms. in my system (windows x64 amd 1055t) give a speedup in large textures, but i tested in in intel dual core and gives a slowdown. o i limited the use for large textures and cpus with more than 3 cores.
please test an let me know if it improves or degrades the speed.
please for linux and osx user. to enable this you will have to enable your compiler support for openmp to test this code.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7284 8ced0084-cf51-0410-be5f-012b33b47a6e
in windows sleep(0) behaves more like what is defined in the yield instruction so use it instead. In my amd 6 core system brings a nice 8% speedup so please test I'm interested in knowing the behavior in different systems.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7261 8ced0084-cf51-0410-be5f-012b33b47a6e
some code additions for future use ;).
please gcc user test this as i don't have opportunity to test it myself i only use reference code to.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7247 8ced0084-cf51-0410-be5f-012b33b47a6e
- ReImplementing Single Core Mode like Dual Core Mode Style.
- Stage 1: My goal is, we have the Fifo, CommandProccessor code the more clear, maintenible and documented possible. When I quit dolphin I want any developer can continue with the work only reading the code.
* Big Refactoring: A lot of functions was changed the names, and modularized.
Now the FifoLoop and CatchUpGPU does not exist, was replaced by RunGpu() and RunGpuLoop().
The general idea is modeling the code like the real HW. The fifo is only a buffer where the Write Gather Pipe write the commands and from the Graphic Processor read these.
* Big Clean UP a lot of obsolete code and comments was deleted, like DcFakeWachDog, "Fifo very soon hack", etc.
In the stage 2, I will refactoring more code doing emphasis in the division of CommandProcessor, Fifo, Gpu Emulation. Beside I will comment all functions and variables in the code (Don't worry I will ask for English help for this part ;) )
Please test a lot SC mode and DC mode :)
Thank you so much for testing always and the patience. I don't like broke your favorite game but... you must believe me this part is very sensible, I only try to contribute for have a better and stable dolphin emulator.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7185 8ced0084-cf51-0410-be5f-012b33b47a6e
Changes:
* Allow events to be scheduled when the emulator is not running. This allows the save state event to be added before the emulator starts.
* Removed the Audio back-end init flag from the save state. This value should not be saved as it is not data relevant to guest machine.
* Allow a recording to be started at any time (apart from when a recording is already being made).
* Updated the status bar and title bar when an on-screen message is shown
* Removed the saving of PEToken from the save state as the FIFO will save this information
* Added a couple Pixel Engine interrupt states to the save state
* Added the copyright notice to the GCPadStatus.h file.
This function is preliminary. Let us know of any bugs you find or any UI quirks.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7175 8ced0084-cf51-0410-be5f-012b33b47a6e
This allows ALL settings in the gfx configuration dialog to be adjusted _per game_ in a more user-friendly way than before.
Obsoletes the gfx related options in the ISO properties, but I kept those since the configuration profiles use different ini file keys (i.e. convert the game inis to the new naming, please!).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7167 8ced0084-cf51-0410-be5f-012b33b47a6e
- This game read the CPRead pointer when the GP is processing data and after detach the fifo and attach again an use this pointer for continue, so implement fifo.SafeCPReadPointer pointer.
- fifo.SafeCPReadPointer is updated when (FAKE_GetFifoEndPtr() - g_pVideoData) == 0) to have a pointer without GFX Commands Cut in the beginning of 32 block.
- Reset video buffer when the CPReadWriteDistance is written by the software.
- This commit is experimental please test a lot your games.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7165 8ced0084-cf51-0410-be5f-012b33b47a6e
Also verify the validity of gfx configuration directly after loading rather than in the GUI code (which was kinda stupid anyway).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7164 8ced0084-cf51-0410-be5f-012b33b47a6e
- In the Command Processor Control Register when GPReadEnable is OFF, wait until the fifo lopp finish with thi actual 32 bytes block. This is necessary to have Safe Disabble GP Read in this way AbortFrame is performed by the software, beside this can help to have accurate GPReadEnable OFF.
Please test Metroid Prime 1/2, Guilty Gear XX, X men gc, SMG, and shuffle you can test Harry Potter Prince for WII. :)
PD: The fifo code need a clean very soon.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7158 8ced0084-cf51-0410-be5f-012b33b47a6e
I think that isFifoBusy bring better sync with VI (video interface) because the CPU emulated threads are waiting for DrawDone in BP Register. So, I do some modifications.
1) Rename "IsFifoBusy" by "isPossibleWaitingSetDrawDone"
2) Only activate isPossibleWaitingSetDrawDone when bFF_GPLinkEnable is true in fifo loop "Inmediate mode" that is because in theory this drawsync function is using in this mode.
3) Deactivate isPossibleWaitingSetDrawDone also in SetFinish in PixelEngine, beside when 32 block is finish.
Please regression in yours games thats can bring some FPS more above all with VPS frame limiter ON (Auto, 60, 50, etc).
- Fix waiting in AbortFrame(), please test mp1/mp2 is fixed again.
Good look!
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7123 8ced0084-cf51-0410-be5f-012b33b47a6e
this is marcos idea so the credits go to him.
please test the broken games in my last commits to see if this solve all the issues.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7114 8ced0084-cf51-0410-be5f-012b33b47a6e
buildfix
move much of the build settings to .props files
-please use them as much as possible in the future, instead of changing individual projects
NOTE: to avoid left over blobs, clean your builds *before* applying these changes.
TODO: add DebugFast target for projects that are lacking it. Lack of DebugFast targets cause the linker to use LTCG when we don't want it.
please test for regressions which could be caused by being too happy with compiler flags :)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7109 8ced0084-cf51-0410-be5f-012b33b47a6e
whit this code hashing is completely disabled for the dlists content itself this must bring some fps more but i must know if this causes any glitches.
i suspect that no game modify the content of the dlist so hash is time wasted but the only way to test it is with this commit.
please test as many games s you can and give me feedback for any glitch.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7106 8ced0084-cf51-0410-be5f-012b33b47a6e
problems, whole functions in .h files need to also be static in
case they are included in several .cpp files.
Also a few other minor LTO fixes.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7082 8ced0084-cf51-0410-be5f-012b33b47a6e
Changed non-hi-res textures to use MurmurHash3, which has better performance that the previous hash.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7080 8ced0084-cf51-0410-be5f-012b33b47a6e
Re-fixed custom textures higher than 1024x1024. (It must have accidentally got reverted somewhere during the video plugin merge)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7064 8ced0084-cf51-0410-be5f-012b33b47a6e
and just operate on lists of object files instead.
This helps with LTO since LLVM/clang LTO is completely broken
by static libraries. It also helps identify symbol clashes
between components like the former plugins.
Many linkers also expect static libraries to form a strict DAG
which turns out be a difficult rule to uphold in practice,
especially since some of our platforms aren't picky about this.
LTO builds currently appears to crash at runtime because of
the static wx libs.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7056 8ced0084-cf51-0410-be5f-012b33b47a6e
fix all build targets (they've all built here - you may have to manually delete the intermediate directories if you have conflicts after this commit).
set the debug path to $(TargetDir)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7022 8ced0084-cf51-0410-be5f-012b33b47a6e
Properly pause the core when saving/loading savestates, previously we used PowerPC::Pause() and Start() which only update the state but doesn't properly set m_StepEvent and caused random hangs.
Fixed hang when creating a savestate while the game was paused or in Frame Advance mode, now the thing works (just remember to press play duh).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6992 8ced0084-cf51-0410-be5f-012b33b47a6e
- dislocating all sensible stuff related to "values per game/pattern" outside the sourcecode.
- giving more control to end-users across the user-friendly interface.
- deleting/cleaning some dead variables.
- updating all gameconfig.ini data to reflect new PHack concept (pending upload).
* Updated Italian translation
- Includes corrections, a better strings translating and suggestions directly by Google Code's people.
+ Minor old pending changes...
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6973 8ced0084-cf51-0410-be5f-012b33b47a6e
- Reimplemented AbortFrame. Now the Write Gather Pipe buffer is reseted and Read Write Distances is reseted before the game do it instead the process all GP CPRWDistance for prevent CP wrong pointers.
This fifo reset should be more accurate, efficient and smooth the FPS when happens. Please, test regresion of SMG1 and SMG2 or others games with FIFO RESETS.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6971 8ced0084-cf51-0410-be5f-012b33b47a6e
This makes the OS X build more robust and should help pave the
way for the integration of the video plugins as well as LTO.
There are now no more global class level namespace conflicts left,
as evidenced by the fact that Dolphin can be linked with -all_load,
not that you would want to.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6958 8ced0084-cf51-0410-be5f-012b33b47a6e
This WILL temporarily break the Linux and MacOSX builds but should be easy to fix.
Things left to do:
* The UI on the new Audio tab for the LLE/HLE choice is ugly
* At times the code still look "plugin-y" and needs cleanup
* The two plugins should be merged further. DSPHLE should use the emulated memory etc of DSPLLE as much as possible, so that simply saving the DSPLLE state is enough. This would also bring the possibility of savestate compatibility between the two plugins.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6947 8ced0084-cf51-0410-be5f-012b33b47a6e
Profiler.{cpp,h} also happened to be the only dupliated
filenames left in Source, the absence of which should make
link-time optimization easier to get working.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6935 8ced0084-cf51-0410-be5f-012b33b47a6e
some optimization for vertex color loading, please test for any change in game color, this is a optimization so nothing should change :)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6908 8ced0084-cf51-0410-be5f-012b33b47a6e
Move shader and input layout management from EmuGfxState to Vertex/PixelShaderCache and D3DVertexFormat.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6906 8ced0084-cf51-0410-be5f-012b33b47a6e
Move depth state management from EmuGfxState to Renderer.
Call stateman->Apply in Renderer::ApplyState instead of EmuGfxState::ApplyState.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6903 8ced0084-cf51-0410-be5f-012b33b47a6e
more accurate means more slow so no complains about speed lost, will optimize later.
please test for regression, but i hope you will get a nice surprise about this commit :).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6881 8ced0084-cf51-0410-be5f-012b33b47a6e
* Change the process ProcessAllDistace to ClearRegister
* Fix a bug in ProcessFifoEvents() in commandProcessro (thanks Skid)
-- The last 2 items should be improve the speed lost in r6572 and maybe improve a little bit because ProcessFifoEvents() is right. (Test car selection in GxZero)
* Alliviate Fifo desyncs, In the meant time I still Researching for real solution. (This fix RE0 again and maybe others
"FIFOs linked but out of sync" and "GFX FIFO: Unknown Opcode (0x%x).\n" games suffers these for desync.)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6837 8ced0084-cf51-0410-be5f-012b33b47a6e
please test for regressions and fixes.
some little changes to make pixel shader more dx9 sm2.0 friendly. the condition is not to use pixel lighting ( sorry no hardware support for the quantity of parameters needed).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6777 8ced0084-cf51-0410-be5f-012b33b47a6e
The code now uses direct pointer access from C code to write the colors to the destination texture instead of trying to force them back up into an __m128i and a single write call. This is what produces the major speed-up.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6761 8ced0084-cf51-0410-be5f-012b33b47a6e
* Fixed a bug on Texture Converter when Copy EFB to RAM's cache is disabled, so...
* ...accordingly, I revert my latest commit because now it becomes useless.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6760 8ced0084-cf51-0410-be5f-012b33b47a6e
- Assign width and height to the actual powers of two rather than to the exponents...
- Clean up FramebufferManager()
- Make use of more depth buffer formats to prevent some devices from failing to create a depth buffer
Should fix issue 3256.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6730 8ced0084-cf51-0410-be5f-012b33b47a6e
the private SVN_REV_STR which relies on finicky build mechanics.
With SCons, kill the old svnrev.h in Source if still present.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6715 8ced0084-cf51-0410-be5f-012b33b47a6e
GX_TF_RGBA8 texture decoder optimized with SSE2 producing a ~68% speed increase over reference C implementation.
TABified the entire document per NeoBrainX. :)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6706 8ced0084-cf51-0410-be5f-012b33b47a6e
~25% speed improvement in decoding GX_TF_RGB5A3 textures which aren't used very much. I thought it would help for movie playback but I misled myself. Video playback has nothing to do with this texture format.
Next I'll see if I can knock out some of these other texture decoders. Byte swizzling I'm sure can somehow be accomplished using _mm_unpacklo_epi8 trickery, so that'd be another big win I hope.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6698 8ced0084-cf51-0410-be5f-012b33b47a6e
TextureDecoder.cpp: merged two functionally identical decode5A3RGBA and decode5A3rgba methods.
OpcodeDecoding.cpp and DLCache.cpp: optimization for GX_LOAD_XF_REG. The PSUHFB solution sounds better for SSSE3, but this is a small win for the default case.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6692 8ced0084-cf51-0410-be5f-012b33b47a6e
Whatever that means, it fixes that stupid Super Mario Sunshine glitch and possibly lots of other stuff, so test as many glitchy games as possible with this ;)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6669 8ced0084-cf51-0410-be5f-012b33b47a6e
Should fix almost all regressions of the recent ClearScreen changes and keep the fixed stuff.
The Super Mario Sunshine glitch is caused by another issue and will be addressed in my next commit.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6668 8ced0084-cf51-0410-be5f-012b33b47a6e
corrected peek color and peek z to correctly emulate real hardware formats.
implements native gamma correction.(i don't own any game that uses this functionality so i will appreciate feedback)
i need a lot of feedback in this changes please
enjoy
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6664 8ced0084-cf51-0410-be5f-012b33b47a6e
make some changes to the Clear code. please test a lot , the point of this commit is to determine the correct behavior of the efb clearing so feedback is welcome
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6663 8ced0084-cf51-0410-be5f-012b33b47a6e
1) What is the FIFO? The fifo is a ring queue for write (CPU) and read (GPU) the graphics commands.
2) What is the Brakpoints? The breakpoint is the FIFO mark to allow parallel work (CPU-GPU) When the GPU reached the breakpoint must stop read immediately until this Breakpoint will be removed for the CPU.
3) What is an overflow? The CPU write all room FIFO possible, and like a ring overwrite commands not processed yet.
4) ¿Why you have an overflow? In theory should not have an overflow never because the fifo has another mark (High Watermark) When the CPU Write reach this mark raise a CP interruption and the FIFO CPU WRITES should stop write until distance between READ POINTER AND WRITE POINTER will be equal to another mark (LO Watemark to prevent and overflow.
5) ¡So if impossible why you have overflows? Simple, the CP interruption is processed later and the Overflow happens. (there is a lot of theories about this)
6) ¿Why is no so simple like when CPU WRITE POINTER is near to the end of the FIFO only process pending graphics command?
Because when this happens sometimes we are in BREAKPOINT and is IMPOSIBLE process the graphics commands.
- This HACK process the pending data when CPU WRITE POINTER is 32 bytes before the end of the fifo, and if there is a Breakpoint force the situation to process the commands and prevent an overflown.
In theory you have not see "FIFO is overflown by GatherPipe nCPU thread is too fast!" anymore. But if you have a hang in game where you had this please read the NOTICE LOG in user\logs, I've added this message "FIFO is almost in overflown, BreakPoint" when the hack is activated. (I will delete this message very soon)
Good Luck!! PD: Shuffle sorry for the large commit description :P
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6662 8ced0084-cf51-0410-be5f-012b33b47a6e
logging: turn off write to file
core: don't insert anything in memcard b slot
video: efb scale to integral, AF to 1x
audio: 100% volume, 48KHz sample rate
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6650 8ced0084-cf51-0410-be5f-012b33b47a6e
- Clear alpha channel to 0xFF instead of 0x00 whenever we (re-)create the EFB texture
- Disable color/alpha channels whenever the actual EFB format doesn't use them
Fixes issue 2826.
Fixes issue 1879.
Please check if issue 3014 is fixed now, as well as any other games which might have broken due to ClearScreen changes.
For what it's worth, Super Mario Sunshine still shows a small glitch even with these changes.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6635 8ced0084-cf51-0410-be5f-012b33b47a6e
Will fix some games and break some others.
Assist trophies in SSBB work fine now, Super Mario Sunshine is a little broken again.
Still needs some more work...
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6627 8ced0084-cf51-0410-be5f-012b33b47a6e
skid_au had implemented this in a previous revision and enabled it by default, but it caused glitches if STC wasn't set to "safe" (which kinda defeated the purpose since it slowed down stuff again).
Also renamed the "safe texture cache" to "accurate texture cache", since setting the "safe" texture cache to "safe" sounds kind of silly..
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6625 8ced0084-cf51-0410-be5f-012b33b47a6e
this is mainly so they fit within ps_2_0 limits, but I made this separate from my last checkin to make it easier to test separately in case it somehow causes problems
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6619 8ced0084-cf51-0410-be5f-012b33b47a6e
(probably nobody else cares, but I need at least one video plugin that actually works on this computer)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6618 8ced0084-cf51-0410-be5f-012b33b47a6e
- Merge the EmuWindow class to VideoCommon.
- Implement free look in the D3D plugins. Seems like it had broken before though; only rotating the view works..
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6616 8ced0084-cf51-0410-be5f-012b33b47a6e
Fixes issue 2805. Fixes issue 3235.
If this causes problems on Windows, just use this intead:
#if defined(HAVE_X11) && HAVE_X11 || defined __APPLE__
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6614 8ced0084-cf51-0410-be5f-012b33b47a6e
To do so I had to re-add the casting bloat removed in revision 6102. Also, for some odd reason the NVidia OpenCL drivers don't like 8 bit rotations, but are okay with 2, 4 bit rotations. These are apparently bugs in the NVidia drivers that are hopefully fixed in future versions.
Also, on linux make sure the TextureDecoder.cl file is copied from the shared data directory to the users directory.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6611 8ced0084-cf51-0410-be5f-012b33b47a6e
* More formats recognition for "Clear Screen" step. Improved the ability handling the alpha channel (% color translucency).
* Related to CPUID detecting:
- Added support displaying correct SysInfo for recent multicore Intel CPUs x86-64 with HT/SMT (Core i3/i5/i7 and Xeon) based on Nehalem architecture.
+ minor and passive changes.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6604 8ced0084-cf51-0410-be5f-012b33b47a6e
Also set up cmake/cpack to create prebuilt binary packages and source packages.
Also change the vertex shader dump file name in the gfx debugger to something that makes more sense.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6587 8ced0084-cf51-0410-be5f-012b33b47a6e
Make sure the gfx debugger unpauses when the emulator is stopped.
When a wad is installed make sure directories exist.
For the cmake build if a header is not provided in the check_lib macro don't check for it, and assume pkg-config was supposed to work.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6581 8ced0084-cf51-0410-be5f-012b33b47a6e
DX9: Revert to the old EFB coordinate scaling. Glitches caused by higher EFB scales probably can't even be fixed properly in DX9, so let's not even mess with it...
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6573 8ced0084-cf51-0410-be5f-012b33b47a6e
- I've implemented calling to ProcessFifoEvents when is there is a pending event in the main queue from CP, PE & GP
- I've implemented FifoIntReset(TRUE, TRUE) in write Clear Register.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6572 8ced0084-cf51-0410-be5f-012b33b47a6e
Also fix some compiler warnings.
On linux don't install the license.txt file. Most distributions handle license files via a package. For example on debian based systems this gives a lintian warning.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6568 8ced0084-cf51-0410-be5f-012b33b47a6e
- Improved the GP Register Status: now this is all the time from the fifo loop.
- Improved the Interrupts manage:
1) Removed All UpdateInturrupts from CommandProcessor Writes and Read.
2) now the CP interrupts are schedule from the video thread and the fifo loop waiting until this happens
3) considering Inmediate mode for the CP interrupts
3) Implemented Interrupt CP Cache State
4) Implemented only Overflow interrupt in GatherPipeCheck because this need to be detected quickly.
- Implemented Overflow handling like a real HW, when Hiwatermark interrupt happens this write ClearRegister with True en HI and False in LO (FifoIntReset) after that a Control Register is writed and the FIFO is processed to LO Watermark.
- Removed all ugly code from LO and HI watermark manage
- Removed all ugly code from BP manage
- Change >= by == in the BP clauses
- Removed speed hack (1024 chunk) for better GP Status Control.
- Commented GXSetGPFifo very soon hack
- Commented FackWatchDog hack
- Commented FIFO_RW_DISTANCE = WritePointer hack
This is the beginning and the base for the future., If this broke your favorite game my apologize, only report this and will try solve it. If you have a Overflown don't worry, I've implemented the real solution code using the Overflow Interruption only need continue working for a perfect protection. Why I did it? Because is preferable a accurate and clean fifo instead hack y fifo for improve that. Thanks to DONKO for you awesome Video Plug in and skid for the chatting. PD: I have 7-10 fps more in the star fox video. bye :P
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6554 8ced0084-cf51-0410-be5f-012b33b47a6e
Should fix problems caused by EFB scales other than Native (excluding fractional!). Test whether this fixes games which work fine with native EFB resolution but show glitches with higher internal resolutions.
Also fixed numerous warnings.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6549 8ced0084-cf51-0410-be5f-012b33b47a6e
Fix the default wiimote emulation mouse button settings on linux.
Fix a little c++ operator precedence issue.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6534 8ced0084-cf51-0410-be5f-012b33b47a6e
Most of the code dealing with the LogTypes namespace was C which lead to a
lot of nonsensical casting, so I have dumbed LOG_TYPE and LOG_LEVEL down to
plain C even though the move of wiiuse into Source means we don't currently
call GenericLog from C.
Set logging threshold to MAX_LOGLEVEL at startup so debug builds will also
p rint debugging messages before the GUI is running.
For some reason the way we use SetDefaultStyle doesn't play nice with wx 2.9
so we just get the default black text on a black background. Using a gray
background works around that problem, but I found it to also be much easier
on the eyes so I have switched the background color on all versions.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6528 8ced0084-cf51-0410-be5f-012b33b47a6e
so the recent addition of VideoCommon/Src/Debugger.h clashed with
DebuggerWX/Src/Debugger.h.
The latter had atrophied anyway, so just get rid of it to resolve the
conflict.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6526 8ced0084-cf51-0410-be5f-012b33b47a6e
Separated UI from debugger functionality. Generally cleaned up that stuff.
Most functionality needs to be reimplemented now, but will be available to D3D9, D3D11 as well as OpenGL then.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6523 8ced0084-cf51-0410-be5f-012b33b47a6e
first fix for issues introduced in sms in r6501, please test for a regressions in The Calling
Second Fix for Issue 3539, by making pixel Dept calculation an option. in games with this issue Enabling pixel depth will solve the issue, in other games disabling the option will not cause speed lost
some cleanup in dlist cache
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6517 8ced0084-cf51-0410-be5f-012b33b47a6e
* I've fixed concurrency problem with fifo.bFF_GPReadEnable when the breakpoint is achieved
* I modified when fifo.bFF_Breakpoint is turn off to prevent that the interruption happens
* I reverted r6483
* I improved the commit r6495 adding another protection when the GPRead is enabled in CTRL_REGISTER
You can test Pokemon XD in dual core mode is more estable now :P This commit could solve others hangs fifo related in DKC, Sport Resorts, etc. I dont have these games.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6503 8ced0084-cf51-0410-be5f-012b33b47a6e
Plus: I added a _assert_msg_ in FIFO loop when the WriteDistance is negative. Please if that happens report it, this can help to solve strange issues with wrong WriteDistances.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6495 8ced0084-cf51-0410-be5f-012b33b47a6e
In this way it prevents wrong fifo.CPReadWriteDistance in the future. That should fix RE0 Issue 2846 and others "Warning: GFX FIFO: Unknown Opcode (0x4d)" in dual core mode.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6485 8ced0084-cf51-0410-be5f-012b33b47a6e
- Speed up MSAA by using custom resolve shaders instead of ID3D11DeviceContext::ResolveSubresource().
- Fix EFB access when MSAA is enabled.
- Implement the "force linear filtering" option.
- Compile shaders using shader model 4.1 or 5.0 if available.
- Some minor cleanups.
Who cares about DX9 and OGL anyway :P
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6481 8ced0084-cf51-0410-be5f-012b33b47a6e
DX11: Fix a DX11 debug runtime error caused by the vertex shader output signature and the pixel shader input signature not matching if per-pixel lighting is disabled. Someone forgot disabling those pixel shader parameters ;P
DX11: Enable use of dynamic textures wherever possible.
DX9: Remove two unnecessary TODOs. Someone forgot removing those since they've been obsolete for a long time ;P
Anyway, at least working mipmaps and dynamic textures should increase DX11 performance a bit. And "a bit" is more than my usual "marginally" ;)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6473 8ced0084-cf51-0410-be5f-012b33b47a6e
* revert back to proper upscale mantissa of parameters A and C
* properly downscale magnitude to 0.24 bits instead than ≡0.23
* Z Eyespace conversion for projection by original patent concept
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6463 8ced0084-cf51-0410-be5f-012b33b47a6e
- Add some info about a backend's feature set (MSAA, Real XFB, EFB to RAM, ..) to VideoConfig
- Gray out options if they aren't supported by the backend or if changing them doesn't affect anything (e.g. changing STC mode if STC is disabled)
- Allow signed bytes for D3D11. Not sure if this causes glitches, but it should work
- Call wxEvent.Skip() in the event function handlers, not sure if this fixes any bugs but the old code definitely caused bugs during development of this patch
- Fix a memory leak in the configuration dialog if D3D11 is used
- Other minor stuff that doesn't need to be mentioned or which I forgot
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6450 8ced0084-cf51-0410-be5f-012b33b47a6e
Also added an option to explicitly disable building with MPG framedumps. To use it add -DENCODE_FRAMEDUMPS=OFF to cmake on the command line.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6429 8ced0084-cf51-0410-be5f-012b33b47a6e
Add a version check for avcodec and swscale libraries.
Fix a bug in the xlib wiimote emulation button configuration. If you are using a mouse button for wiimote emulation you may need to adjust your settings.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6426 8ced0084-cf51-0410-be5f-012b33b47a6e
a little bugfix in vertex loading and some fixes.
not much time to work these days but at least i can spend a little time fixing thing.
please test for regressions.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6409 8ced0084-cf51-0410-be5f-012b33b47a6e
DX9: When clearing the back buffer in Renderer::Swap(), use IDirect3DDevice9::Clear instead of a clear quad if possible. Should improve performance marginally.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6373 8ced0084-cf51-0410-be5f-012b33b47a6e
Also some general clean up of the main CMakeLists.txt file. I think it is nice to keep the checks and analysis of the results together.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6363 8ced0084-cf51-0410-be5f-012b33b47a6e
Not sure how to fix Dolphins hotkeys when 3D Vision is enabled (I do know a way, but it's messy and I don't want 3D Vision messing with Dolphin's cleanliness).
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6362 8ced0084-cf51-0410-be5f-012b33b47a6e
The finding was that 99% of the time, textures are static in memory. However, Dolphin would pessimistically continue to decode and copy the same texture every time. The optimisation is to check if the texture matches what is in the cache, and if it does, Dolphin should early exit.
The result is the speed in New Super Mario Bros Wii increased 35% with spinning coins. Still not as fast as EFB copy to texture though.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6352 8ced0084-cf51-0410-be5f-012b33b47a6e
Note that most users will want to use the release build to get a -O3 optimized build. This can be enabled by adding "-D CMAKE_BUILD_TYPE=Release".
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6344 8ced0084-cf51-0410-be5f-012b33b47a6e
Some notes about this:
- config.h doesn't get written yet, so you'll have to use one generated by SCons
- dependency checking isn't really implemented, yet. Just some basic checks for OpenGL or ALSA, we need something more sophisticated though.
- the OpenGL plugin fails to load for some reason which I can't debug right now due to the libc debuginfo package version in openSUSE not matching the runtime packages
- there's even some support for generating install packages (rpm/deb/.. packages, NSIS installer, etc). It doesn't work properly right now though, since some paths seem to be hardcoded into Dolphin's source
- probably lots of other stuff I forgot... Just take a look at all the TODOs in the CMakeLists.txt files for more information ;P
Additionally, I added various files to the svn:ignore list.
tl;dr: Unless you are a dev or you're building binary packages, this commit shouldn't bother you :P
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6326 8ced0084-cf51-0410-be5f-012b33b47a6e
Upon texture reloading, the cache entry hash wasn't updated and thus we effectively disabled any texture caching in that case.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6308 8ced0084-cf51-0410-be5f-012b33b47a6e
Games affected by this change: Mario Smash Football, Mario Strikers Charged Football, Monster Hunter Tri.
Other games possibly affected: Shaun White Snowboarding, Resident Evil Code: Veronica, Baten Kaitos.
This implementation will decrease performance if the game uses this feature, but the glitches will be gone. I'll add an option for this in a later commit. EFB pokes are somewhat slow in DX11 right now, speed should be okayish in DX9 though.
Other changes:
- SOMEWHAT cleaned up the EFB access code in DX9
- Fixed incompatible parameter list of AccessEFB and TVideo_AccessEFB.
- Fixed a theoretical bug in ReplaceRGBATexture2D, add support for STAGING textures
- Removed unused parameters in various DX9 functions
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6300 8ced0084-cf51-0410-be5f-012b33b47a6e
Now, there are two:
Real dimensions: Width and height of the original GameCube texture
Virtual dimensions: Width and height of the texture used by dolphin-emu's renderer
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6291 8ced0084-cf51-0410-be5f-012b33b47a6e
Work around different gcc versions giving conflicting warnings
about signed/unsigned comparisons involving bit fields.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6290 8ced0084-cf51-0410-be5f-012b33b47a6e
Renamed EFB Scale to Internal Resolution.
Removed Auto Scale option (it is now always on).
Added on-the-fly changing of the Internal Resolution in the OpenGL and Direct3D9 plug-ins.
Further consolidated the code in the video plug-ins.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6240 8ced0084-cf51-0410-be5f-012b33b47a6e
final rendering of (at least) Sonic Unleashed.
This should be (nearly) invisible in Z16 depth copies and in games, but there is
some chance that it collides with other such border cases. There probably is some
room to decrease the adjustment but 9999999/10000000 is not enough. A static offset
may be an option, too. I don't know if the game can/does set something like that.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6225 8ced0084-cf51-0410-be5f-012b33b47a6e
for all the plugins implemented per pixel lighting, this will make games that uses lighting a lot nice. (just look at mario sunshine and compare :))
for dx9: implemented temporal anaglyph stereo: just grab your red-cyan glasses and enjoy.
stereo calibration: use stereo separation ( distance of the point from you are looking) and Focal Angle: the angle necessary to focus in one particular object.
this settings are different in every games as they use different depth ranges.
please for any regression and bug introduced by this commit.
if you ask why i did not implement stereo in dx11 and opengl the reason is one: they don't work right when i have more time will try to find a way to make them work.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6224 8ced0084-cf51-0410-be5f-012b33b47a6e
Main Issues:
DX11 is functional with a ~2MB/s mem leak.
OpenGL/DirectX9 have a black display while game runs. (DirectX 9 flashes good display on emulation stop)
Too many virtual function calls. (once everything is working, I will work on removing them)
Won't build on non-Windows in its current state. (mainly EmuWindow will need changes for Linux/OS X)
Probably other stuff.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6219 8ced0084-cf51-0410-be5f-012b33b47a6e
include referenced array data in the hash to avoid problems in some games.
this version is a slower than the last version but is more stable, still much work to do but so little time to do it :)
Added Dlist Caching Option to the plugin configuration so anyone can compare the performance change.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6205 8ced0084-cf51-0410-be5f-012b33b47a6e
still experimental, not totally optimized but must bring a nice speed up
please test for regressions an error. an please Linux people fix scons :)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6149 8ced0084-cf51-0410-be5f-012b33b47a6e
use plain vertex arrays instead of VBOs to render in Opengl plugin as the nature of the data make VBOs slower. This must bring, depending on the implementation, a good speedup in opengl.
in my system now opengl and d3d9 have a difference of 1 to 5 fps depending of the game.
some cleanup and a little work pointing to future improvements in the way of rendering.
please test and check for any errors.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6139 8ced0084-cf51-0410-be5f-012b33b47a6e
fixed one of the last graphic problems in smg, now the glow in the plants and planets must be correct.
the error was caused by a error in lighting calculations. so it must affect other games.
please test and let me know the results.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6103 8ced0084-cf51-0410-be5f-012b33b47a6e
Really that happens because the fifo.CPReadWriteDistance is negative.
Example: CPReadWriteDistance: -864 CPEnd: 10092672 fifo.CPBase: 9568416
In SMG this is because PI_FIFO_RESET is writing and after fifo.CPReadWriteDistance will be setted to 0.
To Prevent that, I've Implemented AbortFrame function in the CommmandProcessor. It should fix overflown because of that.
Note: There is other issue (Issue 2846) where the fifo.CPReadWriteDistance is negative too but the effect is different.
I'm working to solve this.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6083 8ced0084-cf51-0410-be5f-012b33b47a6e
I am not sure i am understanding what the pipeline really does, and more so what
the GC/WII expects here. If my comments are incorrect, please let me know.
This was tested with MP2:E, ZWW on r6075.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6078 8ced0084-cf51-0410-be5f-012b33b47a6e
Remove a member variable which I introduced in r5907 although it never actually got used. Restores binary compatibility (at least in that regard) to pre-r5907 video plugins, but breaks any binaries after that :P
Update FIFO watermark tightness recommendations. 1000 is quite a high value I guess, but some people seem to need it.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6060 8ced0084-cf51-0410-be5f-012b33b47a6e
The code changes disable the usual pipeline flush for certain BP Writes that occur while the minimap is being drawn in Zelda: twilight princess. This significantly increases speed while in hyrule field. The way this is accomplished is described more in depth on page 42 of Xtreme's thread. Big thanks to Xtreme for doing a great job hosting that thread, and Kiesel-stein for initial work on the hack
Also, I used the resource editor in Visual studio to generate the GUI code for the DX11 plugin, and some code appeared to be removed, although the behavior of the GUI did not seem to change. Hopefully someone more experienced with resource files (forms?) can double check that no code was damaged
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6057 8ced0084-cf51-0410-be5f-012b33b47a6e
(those originating from a thread other than the gpu thread)
glxMakeCurrent is supposed to error out when the context is
still current in another thread. NVIDIA handles that, but mesa3d errors.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6054 8ced0084-cf51-0410-be5f-012b33b47a6e
The GCC model for extended instructions like these is that you compile
with -msse3 etc. These affect code generation for whole compilation units,
so the idea is that you have a separate .c file for each instruction set
class and then indirect to the desired one at runtime.
Without e.g. -msse4.1, the GCC built-ins used by <foointrin.h> are not
available. However, in our specific case of compiling with -msse2 and
wanting to use SSE3.1 code, enough built-ins are available that we only
need to provide a little hack for pshufb.
Upgrading this to also use SSE4.1 instructions doesn't appear feasible
without a lot of undesirable duplication of GCC built-in functions and
headers, so we'd probably have to move to the GCC model of separate
source files for that.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6014 8ced0084-cf51-0410-be5f-012b33b47a6e
is required to correctly link with C++ libraries like wxw.
Tell OS X that the user is doing something when using the wiimote
so that the screensaver doesn't come on when playing a Wii game.
Replace unnecessary floating point with integer math.
Remove unnecessary <sys/stat.h>'s.
Correct a few type nits.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5995 8ced0084-cf51-0410-be5f-012b33b47a6e
This is EXPERIMENTAL, in theory don't break any game but if so Revert this commit immediately please.
Note: Beside Eternal Darkness needs DirectX 11 or DirectX 9 with Zelda Hack to works fine.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5986 8ced0084-cf51-0410-be5f-012b33b47a6e
Use SCons' convenient # shorthand for the root of our tree.
Also various minor cleanup of SConscripts.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5915 8ced0084-cf51-0410-be5f-012b33b47a6e
To change it, right click the affected game in the iso list, select Properties, and enter some constant for "Watermark tightness". Reasonable values range from 20 to 200.
FIFO seems unoverflowable on my computer no matter what I set this value to, so test whether tuning the value helps you ;P
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5907 8ced0084-cf51-0410-be5f-012b33b47a6e
- fix numerous warnings
- make some variables unsigned
- remove redundant code and header inclusions
- make code more compact in lots of cases
Committing some additional changes by myself:
- additional header cleanups
- cleanup DX11 initialization/shutdown process (hinted at by avindra)
- Remove the cgD3D9 stuff from Externals since it's no longer needed
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5903 8ced0084-cf51-0410-be5f-012b33b47a6e
Because we only ever call Pos_ReadDirect (and through that, DataRead<T>)
from JIT generated code, the compiler may not get the heads-up to properly
prepare for run-time instantiation of those template functions.
Explicitly instantiating Pos_ReadDirect gets around that issue.
Also force DataRead* inline as gcc didn't always do that itself when the
DataRead functions in turn were called from (other) template functions.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5902 8ced0084-cf51-0410-be5f-012b33b47a6e
from JIT generated code, the compiler may not get the heads-up to properly
prepare for run-time instantiation of those template functions.
Explicitly instantiating Pos_ReadDirect gets around that issue.
Also force DataRead* inline as gcc didn't always do that itself when the
DataRead functions in turn were called from (other) template functions.
I am far from a C++ language lawyer, so I cannot speak learnedly about the
correctness of this solution, but it works.
Dolphin.app built on OS X 10.6 now actually works on 10.5 as well.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5901 8ced0084-cf51-0410-be5f-012b33b47a6e
- Eliminate the useless check for cpu modifications option from efb to ram as it must be enabled always
- use constant names in dx11 for buffer length calculation instead to simplify code reading
- implemented scaled efb copy in opengl, still bugy in some games, the option is not in the gui but will add it when it works perfect
- Change the depth calculation behavior:
if the game use z textures is exactly the same as before.
if the game do not use z texture calculate z values in the vertex shader. the advantage id this approach is that early z culling is applied,
improving fill rate. this mus speed up things, even with ssaa and msaa enabled.
please test for regression and enjoy.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5896 8ced0084-cf51-0410-be5f-012b33b47a6e
implemented new efb to ram in opengl
modified TextureConversionShader preparing the implementation of new efb to ram in dx11
please test for regressions
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5874 8ced0084-cf51-0410-be5f-012b33b47a6e
fix frame dumping and screenshots in d3d9 and opengl
some improvements to the new efb to ram
please test to see if the issues introduced by the new efb to ram are solved by this commit
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5862 8ced0084-cf51-0410-be5f-012b33b47a6e
