This is the bare minimum required to run a few games on AArch64.
Was able to run starfield and Animal Crossing to the Nintendo logo.
QEmu emulation is literally the slowest thing in the world, it maxes out at around 12mhz on my Core i7-4930MX.
I've tested a few instruction encodings and am expecting most to work as long as one stays away from VFP/SIMD.
This implements mostly instructions to bring up an initial JIT with integer support.
This can be improved to allow ease of use functions in the future, dealing with the raw imms/immr encodings is probably the worst thing ever.
Uses are split into three categories:
- Arbitrary (except for size savings) - constants like RSCRATCH are
used.
- ABI (i.e. RAX as return value) - ABI_RETURN is used.
- Fixed by architecture (RCX shifts, RDX/RAX for some instructions) -
explicit register is kept.
In theory this allows the assignments to be modified easily. I verified
that I was able to run Melee with all the registers changed, although
there may be issues if RSCRATCH[2] and ABI_PARAM{1,2} conflict.
And switch to a register order that consistently prefers callee-save to
caller-save. phire suggested putting rdi/rsi first, even though they're
caller-save, to save code space; this is more conservative and I can do
that later.
Rather than using a variety of registers including RSI, ABI_PARAM1
(either RCX or RDI), RCX, and RDX, the rule is:
- RDI and RSI are never used. This allows them to be allocated on Unix,
bringing parity with Windows.
- RDX is a permanent temporary register along with RAX (and is thus not
FlushLocked). It's used frequently enough that allocating it would
probably be a bad idea, as it would constantly get flushed.
- RCX is allocatable, but is flushed in two situations:
- Non-immediate shifts (rlwnm), because x86 requires RCX to be used.
- Paired single loads and stores, because they require three
temporary registers: the helper functions take two integer
arguments, and another register is used as an index to get the
function address.
These should be relatively rare.
While we're at it, in stores, use the registers directly where possible
rather than always using temporaries (by making SafeWriteRegToReg
clobber less). The address doesn't need to be clobbered in the usual
case, and on CPUs with MOVBE, neither does the value.
Oh, and get rid of a useless MEMCHECK.
This commit does not actually add new registers to the allocation order;
it is intended to test for any performance or correctness issues
separately.
The special case is where the registers are actually to be swapped (i.e.
func(ABI_PARAM2, ABI_PARAM1); this was previously impossible but would
be ugly not to handle anyway.
In two cases, my old code was using a temporary register but not saving
it properly; it basically worked by accident (an otherwise useless
FlushLock was causing CallerSavedRegistersInUse to think it was in use
by the GPR cache, even though it was actually a temporary).
I'm going to modify this in the next commit to use RDX, but I didn't
want to leave a broken revision in the middle.
The register is RBP, previously in the GPR allocation order. The next
commit will investigate whether there are too few GPRs (now or before),
but for now there is no replacement.
Previously, it was accessed RIP relatively; using RBP, anything in the
first 0x100 bytes of ppcState (including all the GPRs) can be accessed
with three fewer bytes. Code to access ppcState is generated constantly
(mostly by register save/load), so in principle, this should improve
instruction cache footprint significantly. It seems that this makes a
significant performance difference in practice.
The vast majority of this commit is mechanically replacing
M(&PowerPC::ppcState.x) with a new macro PPCSTATE(x).
Version 2: gets most of the cases which were using the register access
macros.
GetOpInfo was returning null pointers for invalid ops in subtables
instead of asserting an error. This was causing segfaults when the
jit tried to jit invalid code.
Prior to this change, it was possible to cause an infinite loop by making the string to be replaced and the replacing string the same thing.
e.g.
std::string some_str = "test";
ReplaceAll(some_str, "test", "test");
This also changes the replacing in a way that doesn't require starting from the beginning of the string on each replacement iteration.
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.
The person who wrote this seemed to misunderstand how XPending and
XNextEvent actually work. XNextEvent will wait in poll if there's
no event yet, meaning that we don't need to sleep after we process
all the events; the kernel will sleep for us.
This changes indentation, so view with -w or a similar feature to
understand what's actually changed here.
Added the option to handle whether the user wants to iterate through the
assignment of button mappings or assign them one at a time.
fixed formatting issues and code style.
I excluded this option from the config file. This stopped the check box value and the boolean from becoming offset. Since the option should always start as false.
This still causes an issue with the Wiimote input, since the class variable that keeps the state will be wiped, but the check box value will stay the same after closing/reopening without closing the entire Wiimote configuration. I am looking for a way to resolve this.
I also reduced wait time to 2.5 seconds vs. the 5 seconds previously. Seemed to be a little long.
These changes apparently did not go through.
This should fix the Wiimote issue.
Allows user to map all inputs seamlessly without having to
click on each button.
Also increased button timeout to 5 seconds from 1.5 due to pita.
Motion controls are not included since they will be special cases.
As far as I can tell, this has literally been here since the start of the git
history; maybe it was stubbed out because the author wasn't sure it was right?
It matches the PPC/Broadway manuals perfectly, though.
This value was "helpful" for debugging when the stack got corrupted.
Helpful that if gpr[1](Which is the stack pointer with PPC ABI) is zero then the interpreter would spam huge amounts of annoy text saying that we
managed to get in to a "corrupted" state.
This is incremented every instruction on the interpreter, or every block run on the JIT64....Only if debugging is enabled(JIT64 it is a const
variable)
The message is only outputted when interpreter is used and debugging is enabled.
The old method would always evict the first suitable register, i.e. the
same register every time once the cache got full. The cache doesn't get
terribly often, but the result is pathological...
Faster, of course, since we avoid the interpreter, but also means we can
get more a more accurate timer in long blocks by adding the offset from the
start of the block to the retrieved timer. I don't know if this will actually
fix any issues, but it's more correct and a nearly-free improvement.
This causes glDrawArrays to fail in core profile, and thus on OS X, see:
http://renderingpipeline.com/2012/03/attribute-less-rendering/
There must be something bound, even though it is not used.
Fixes#7599. I'm not sure this is actually the best way to fix it,
since AFAICT it makes a nonobvious assumption that *something* will be
bound before the first attributeless rendering in
TextureConverter::DecodeToTexture, but it's what degasus suggested and
seems to work.
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").
OSD messages can be disabled, while still leaving them in the status bar. This is incredibly useful for certain users, who may wish to see the messages, but do not wish to have them cover up half of the screen. In particular TASers will generally have OSD messages on the screen 100% of the time, and they cover up useful information, making it critical to turn them off. However the messages are still very useful to them, so it's important to have them somewhere.
This reverts 4a16211bae.
This time, check the address carefully beforehand, since apparently some games
do horrible things like running it on non-RAM addresses, or at the very least
virtual addresses.
This is no longer required since we don't support x86_32 anymore.
x86_64 implies SSE2 support.
Also this check was a bit messed up and was hitting on Generic builds.
Only show the JIT cores on x86_64(Will have its own issues once we reach that point)
Show AArch64 JIT if running on a AArch64 device(Good luck with that for now. Future proofing though)
This is available in a PR here: https://github.com/taka-no-me/android-cmake/pull/23
The maintainer of the android toolchain cmake file seems to be AWOL for now.
I have tested this file personally and it works, it just isn't merged in yet
A bug that seems to have been uncovered by allowing immediate-address loads.
Super Monkey Ball 2 crashes without this change -- it's possible, however, that
the game actually requires the MMU hack, since it crashed due to accessing an
address in the 0x20000000-0x3fffffff range.
Decreases total Wii state save time (not counting compression) from
~570ms to ~18ms.
The compiler can't remove this check because of potential aliasing; this
might be fixable (e.g. by making mode const), but there is no reason to
have the code work in such a braindead way in the first place.
- DoVoid now uses memcpy.
- DoArray now uses DoVoid on the whole rather than Doing each element
(would fail for an array of STL structures, but we don't have any of
those).
- Do also now uses DoVoid. (In the previous version, it replicated
DoVoid's code in order to ensure each type gets its own implementation,
which for small types then becomes a simple load/store in any modern
compiler. Now DoVoid is __forceinline, which addresses that issue and
shouldn't make a big difference otherwise - perhaps a few extra copies
of the code inlined into DoArray or whatever.)
We were generating a texture without ever setting the data to a known value.
This happened on the old code as well, just that PP shaders are receiving some love and people are using it and noticing some of its issues.
Doesn't support all the FPSCR flags, just the FPRF ones.
Add PPCAnalyzer support to remove unnecessary FPRF calculations.
POV-ray benchmark with enableFPRF forced on for an extreme comparison:
Before: 1500s
After, fmul/fmadd only: 728s
After, all float: 753s
In real games that use FPRF, like F-Zero GX, FPRF previously cost a few percent
of total runtime.
Since FPRF is so much faster now, if enableFPRF is set, just do it for every
float instruction, not just fmul/fmadd like before. I don't know if this will
fix any games, but there's little good reason not to.
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.
While we're at it, support a bunch of float load/store variants that weren't
implemented in the JIT. Might not have a big speed impact on typical games but
they're used at least a bit in povray and luabench.
694 -> 644 seconds on povray.
In particular, even in code that only runs on x86-64, you can't use
PRIx64 for size_t because, on OS X, one is unsigned long and the other
is unsigned long long and clang whines about the difference. I guess
you could make a size_t specifier macro, but those are horribly ugly, so
I just used casting.
Anyone want to make a nice (and slow) template-based printf?
Now without bare 'unsigned'.
Thanks to magumagu's softfp experiments, we know a lot more about the Wii's
strange floating point unit than we used to. In particular, when doing a
single-precision floating point multiply (fmulsx), it rounds the right hand
side's mantissa so as to lose the low 28 bits (of the 53-bit mantissa).
Emulating this behavior in Dolphin fixes a bunch of issues with games that
require extremely precise emulation of floating point hardware, especially
game replays. Fortunately, we can do this with rather little CPU cost; just ~5
extra instructions per multiply, instead of the vast load of a pure-software
float implementation.
This doesn't make floating-point behavior at all perfect. I still suspect
fmadd rounding might not be quite right, since the Wii uses fused instructions
and Dolphin doesn't, and NaN/infinity/exception handling is probably off in
various ways... but it's definitely way better than before.
This appears to fix replays in Mario Kart Wii, Mario Kart Double Dash, and
Super Smash Brothers Brawl. I wouldn't be surprised if it fixes a bunch of
other stuff too.
The changes to instructions other than fmulsx may not be strictly necessary,
but I included them for completeness, since it feels wrong to fix some
instructions but not others, since some games we didn't test might rely on
them.
(1) Rename ABI_ALL_CALLEE_SAVED to ABI_ALL_CALLER_SAVED, because that's
what it was actually defined as (and used as). Derp.
(2) RegistersInUse is always used for the purpose of saving registers
before calling a C++ function in the middle of a JIT block (without
flushing). There is no need to save callee-saved registers in this
case. Change the name to CallerSavedRegistersInUse and mask with
ABI_ALL_CALLER_SAVED.
Nothing obvious broke when starting up a Melee game. (I added a test
for anything actually being masked out; it happens, but in this
particular case seemed to occur at most a few dozen times per second, so
the actual performance benefit is probably negligible.)
The PPC_FP conversion code can be made a lot simpler with the observation
that the only values that need to be sent through the slow x87 path are
denormals.
A whole bunch faster: 708->678 seconds on POV-RAY.
Note warnings may still be generated by code in Externals if it is included into warning-enabled code (this is the case for libusb).
Also, quiet an intellisense warning in wxWidgets caused by using __w64 for an inapplicable type.
strictStrings is not supported by debug libraries, and indeed breaks the build.
Drop wbemidl.h (incompatible with strictStrings) dependency by using SDL-style search for XInput GUIDs.
Zhuowei Zhang