So, a FifoRecorder instance is instantiated as a file-local variable and
used as a singleton (ugh). Most users likely don't regularly use the
FIFO player/FIFO recorder, so this is kind of a substantial waste of
memory.
FifoRecorder's internal RAM and ExRAM vectors are 33554432 and 67108864
bytes respectively, which is around 100.66MB in total.
Just on the game list view on a clean build with nothing loaded, this
knocks debug build memory usage down from ~232.4MB to ~137.5MB, and
release build memory usage down from ~101MB to ~5.7MB.
This moves all the byte swapping utilities into a header named Swap.h.
A dedicated header is much more preferable here due to the size of the
code itself. In general usage throughout the codebase, CommonFuncs.h was
generally only included for these functions anyway. These being in their
own header avoids dumping the lesser used utilities into scope. As well
as providing a localized area for more utilities related to byte
swapping in the future (should they be needed). This also makes it nicer
to identify which files depend on the byte swapping utilities in
particular.
Since this is a completely new header, moving the code uncovered a few
indirect includes, as well as making some other inclusions unnecessary.
Since in this case we're setting it based on the state at record start
time, not when a register is loaded, UseMemory would not be called, so
this could potentially wipe out texture memory that was valid.
This should ensure that when playing with loop enabled, the first frame is
in the same state each time. There is potentially still issues when the
start frame is set to something other than zero, but I'm not sure how we
could work around this without capturing the entire state on each frame.
Making changes to ConfigManager.h has always been a pain, because
it means rebuilding half of Dolphin, since a lot of files depend on
and include this header.
However, it turns out some includes are unnecessary. This commit
removes ConfigManager includes from files which don't contain
SConfig or GPUDeterminismMode or GPU_DETERMINISM (which means the
ConfigManager include is not used).
(I've also had to get rid of some indirect includes.)
Fundamentally, all this does is enforce the invariant that we always
translate effective addresses based on the current BAT registers and
page table before we do anything else with them.
This change can be logically divided into three parts. The first part is
creating a table to represent the current BAT state, and keeping it up to
date (PowerPC::IBATUpdated, PowerPC::DBATUpdated, etc.). This does
nothing by itself, but it's necessary for the other parts.
The second part (mostly in MMU.cpp) is simply removing all the hardcoded
checks for specific untranslated addresses, and consistently translating
addresses using the current BAT configuration. Very straightforward, but a
lot of code changes because we hardcoded assumptions all over the place.
The third part (mostly in Memmap.cpp) is making the fastmem arena reflect
the current BAT configuration. We do this by redoing the mapping (calling
memmap()) based on the BAT table whenever it changes.
One additional minor change is that translation can fail in two ways:
either the segment is a direct store segment, or page table lookup failed.
The difference doesn't usually matter, but the difference affects cache
instructions, like dcbz.
Fix Frame Advance and FifoPlayer pause/unpause/stop.
CPU::EnableStepping is not atomic but is called from multiple threads
which races and leaves the system in a random state; also instruction
stepping was unstable, m_StepEvent had an almost random value because
of the dual purpose it served which could cause races where CPU::Run
would SingleStep when it was supposed to be sleeping.
FifoPlayer never FinishStateMove()d which was causing it to deadlock.
Rather than partially reimplementing CPU::Run, just use CPUCoreBase
and then call CPU::Run(). More DRY and less likely to have weird bugs
specific to the player (i.e the previous freezing on pause/stop).
Refactor PowerPC::state into CPU since it manages the state of the
CPU Thread which is controlled by CPU, not PowerPC. This simplifies
the architecture somewhat and eliminates races that can be caused by
calling PowerPC state functions directly instead of using CPU's
(because they bypassed the EnableStepping lock).
Reading uninitalized memory is non-deterministic. We used to only
clear the memory when using EmulatedBS2_GC or FifoPlayer, but we
now do it during Memory::Init instead so it always gets done.
Detected by the previous commit. We had forgotten to make sure the
Vertex Loader's table of normal sizes is initilzied, resulting in
the wrong offsets/sizes being used for vertices with normals.
It wasn't working, I'm not really sure why.
Since #2997 we rely on video common to mark efb copies 'written'
during recording, and for old dffs we just ignore the bad texture
while playing back in the texture cache.
We actually discovered a bug while combining the two functions with
FifoRecordAnalzyer's vertex array loading code. If per-vertex
postion or texture matrices were enabled and vertex arrays in use
then the wrong data would be used to calculate the minimum/maxmium
indices, which would result in either too much or too little vertex
data being included in the dff.
So this commit also increments the dff version number, so we can
identify old broken dffs later.
