Current order of operations (rotate then normalize) seems to produce a
lot more distortion than normalizing and then rotating. This makes Citra
results match pretty closesly with hardware, and indicates that hardware
may also be using lerp instead of slerp to interpolate the quaternions.
It is now known that result code description vary depending on the
module, and so they're best defined on a per-module basis. To support
this, allow passing in an arbitrary integer instead of limiting to the
ones in the ErrorDescription enum. These will be gradually migrated to
their individual users, but a few will be kept as "common" codes shared
by all modules.
Previously we were letting vadd flush the value to positive 0, but there are cases where this behavior is wrong, for example,
vsub: -0 - +0 = -0
vadd: -0 + +0 = +0
Now we'll flush the value to +0 inside vsub, and then negate it.
One of the later commits will enable writing to GS regs.
It turns out that on startup, most games will write 4096 GS program words.
The current limit of 1024 would hence result in 3072 (4096 - 1024) error messages:
```
HW.GPU <Error> video_core/shader/shader.cpp:WriteProgramCode:229: Invalid GS program offset 1024
```
New constants have been introduced to represent these limits.
The swizzle data size has also been raised. This matches the given field sizes of [GPUREG_SH_OPDESCS_INDEX](https://3dbrew.org/wiki/GPU/Internal_Registers#GPUREG_SH_OPDESCS_INDEX) and [GPUREG_SH_CODETRANSFER_INDEX](https://www.3dbrew.org/wiki/GPU/Internal_Registers#GPUREG_SH_CODETRANSFER_INDEX) (12 bit = [0; 4095]).
Instead of tying the QFileSystemWatcher to the GameList and updating in
the UI thread, this change moves it to the worker thread. Since it gets
deleted and recreated as part of the worker thread, this prevents it from
ever getting used from multiple threads (which is why it was originally
done on the UI thread)
This replaces the hardcoded VRAM/DSP mappings with ones made based on
the ExHeader ARM11 Kernel caps list. While this has no visible effect
for most applications (since they use a standard set of mappings) it
does improve support for system modules and n3DS exclusives.
Appears to be currently broken, and given the complexity of doing this
for ARM code without debugging information, should probably be left to
an external tool or library. Use the GDB stub instead.
Closes#586
It has performance problems, a very misleading UI, and is broken in
general. It has essentially been superceded by the GDB stub, but if we
wanted a built-in disassembler in the future it'd essentially need to be
rewritten from scratch anyway.
Closes#427, #1480
* Create a random console_id when config save_file is created
Added button in system config to refresh the console unique id
* Moved the connect for the button from .ui file to constructor of ConfigureSystem
* Added warning and info dialog
Fixup: Make use of qt5 style connects,
renamed the refresh button,
removed some duplicate code,
changed random device and moved all to the generate function
* Changed the random generator to reflect what a real 3DS stores as console unique id
Fixup: Changed the warning message
* Fixup: Set and Create
* Fixup: Added console id label, therfore removed second message box
* Fixup: fixed the endianess
* Fixup: more endianness fixes
* Fixup: Endianness the 3rd
Yuri Kunde Schlesner