Implementations for SSE3 CPUs (prescott)
Instead of fixing some of the bugs with HostCall when paired with vectors, i'll simply remove as many host calls as I can within the most used vector instructions - then just minimize their usage to memory read/writes.
Emitting the raw assembly code is faster than doing a HostCall, HostCalls are VERY expensive. So this is the desired output anyways.
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3301
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
crueter will love this one
mcl nuked off externals, i did however copy some important headers (notably intrusive list and bit.hpp); because grand part of dynarmic still uses them
but i made the appropriate adjustments anyways
solves mcl for #3373
depends on #3718
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3777
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Lets do the quick math
There was 1 LUT for every fsize() instancing
Now... the number of functions on each lut was (fsize + 1), multiplied by 5 (number of rounding modes)
8 = 9 * 5 = 45
16 = 17 * 5 = 85
32 = 33 * 5 = 165
64 = 65 * 5 = 325
this is just pure insanity - look at what fucking nm reported:
```
0000000003dc39b8 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<16ul, false>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
0000000003dc3a18 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<16ul, true>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
0000000003dc39d8 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<32ul, false>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
0000000003dc3a38 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<32ul, true>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
0000000003dc39f8 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<64ul, false>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
0000000003dc3a58 0000000000000008 V guard variable for void Dynarmic::Backend::X64::EmitFPVectorToFixed<64ul, true>(Dynarmic::Backend::X64::BlockOfCode&, Dynarmic::Backend::X64::EmitContext&, Dynarmic::IR::Inst*)::lut
```
"ah its not bad" - OH MATE ITS JUST THE GUARD VARIABLES - i attached a file with just the functions generated for each case...
now with this PR only 6 * 6 functions are made (still not ideal, but way better), 36 is way better than 1156 FUCKING FUNCTIONS
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3718
Reviewed-by: DraVee <chimera@dravee.dev>
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
- our logging code was bigger than spdlog itself, why???? just keep it simple
- fix issues when logging before logging system is even started
- removes the "initialized logging twice" issue
- removes uneeded indirection in file logging
- uses direct formatting instead of jumping hoopla-around the fmt::format() ressult
- code duplication and dead code removal as usual
I did explore dup2() but I think it's not worth the hassle
I did try `fwopen()` but it's better if things are just kept as-is.
there is a lot of noise because I removed a bunch of redundant files on logging and just put everything in one file
now normally this wouldn't be a good idea, however consider: the complexity of logging; it's less than 500 lines... does it really need a whole subsystem?!?!?! ITS JUST LOGGING
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3688
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: DraVee <chimera@dravee.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
HLE macro does this, may as well do it on dynarm as well :)
could improve perf a bit
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3507
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: DraVee <dravee@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
interpreter was never called in practice and doesn't do anything other than just crash
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3547
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: crueter <crueter@eden-emu.dev>
Reviewed-by: DraVee <dravee@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
It seems to be related to the fact multimap on xbyak doesn't properly work if it's unordered
It also seems to NOT play nice at all with unstable iterators (like those found on ankerl) - either we use boost or just let xbyak use normal std
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3589
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Reviewed-by: crueter <crueter@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
standard std::unordered_map and std::unordered_set sucks, hadn't found drop-in replacement for multimap, but this change should bring a nice perf boost for x86_64 :)
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3438
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: crueter <crueter@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Originally was a test for Qt 6.9.3. Still broken, I have absolutely zero idea how to fix it.
Signed-off-by: crueter <crueter@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3559
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
should make JIT translation a bit faster - especially for non-fastmem
test if thumb still werks
previously:
we read 32-bits one by one, and do translation **for each u32 we read**
now:
one big read of 4096 bytes (aligned of course), only 1 VAddr translation
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3459
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
results in way better codegen (less noise for movs and all dat)
i think it was removed by accident by me, oops
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3519
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Allows regalloc to more aggressively take on registers for immediate codepath; also uses lea for stc/adc (since carry can equate +1 in some cases)
Signed-off-by: lizzie lizzie@eden-emu.dev
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/2958
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
this results in removed reloads for any code that uses those heavily :)
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3518
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
dont have AVX512CD to verify the change; but should[tm] work
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3499
Reviewed-by: DraVee <dravee@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
PLEASE test, & provide gdb logs kthx!
mainly because there is a bug with pop/push for r8-r15:
94f6fe6941
(may have been linked as to why Engage had SCREECH when doing xmm spills!)
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Co-authored-by: Caio Oliveira <caiooliveirafarias0@gmail.com>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3505
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: Maufeat <sahyno1996@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
mainly doing this to reduce memory footprint; we all know how nice ankerl::unordered_dense is
in theory 4x faster - in practice these maps arent that "hot" anyways so not likely to have much perf gained
i just want to reduce mem fragmentation to ease my porting process, plus it helps other platforms as well (ahem weak Mediatek devices) :)
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Co-authored-by: Caio Oliveira <caiooliveirafarias0@gmail.com>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3442
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Forgot to disable this. Will be re-enabled when UUID changing can become
a thing
Signed-off-by: crueter <crueter@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3428
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Only merge once we find out wth is causing the crashes [we know it's fsr]
This reverts commit 82f9d489e7.
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3250
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Support for bundled Qt, not through aqtinstall but rather my CI. Multimedia is
implemented too, works on both Windows and Linux, though we don't
actually use it so it doesn't really matter. Contains Declarative and all that so the Quick frontend will work once it becomes a thing.
Some options have changed, notably w.r.t LTO and faster
linker, which are now handled directly in the modules.
CPMUtil also has support for custom dirs (`PackageName_CUSTOM_DIR`) now. Probably most useful for adding external fragment shaders and whatnot.
Signed-off-by: crueter <crueter@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3289
Alder Lake/Tremont or superior feature only, based on archived UMWAIT by @Wunkolo. With some fixes due to register corruption/clobber (regalloc was blind to this)
Signed-off-by: lizzie <lizzie@eden-emu.dev>
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/2614
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
supercedes #3295
Intel Atom N450 6 billion fps
also for the note, host calls are expensive as fuck
please test i didn't break other games kthx
Test with both a pre-AVX2 CPU (like i5-3th gen) AND a AVX2 cpu (like a i7-4th gen)
Signed-off-by: lizzie lizzie@eden-emu.dev
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3297
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
Raises the size of each page entry to 32 bytes, however, it merges them into a single structure
THEORETICALLY this is better since the access pattern observed corresponds to the program wanting backing_addr/pointers/blocks immediately after one another.
This may improve performance at the cost of some extra memory.
Another implementation would be to structure only backing_addr/blocks within the same virtual buffer.
Alas spamming virtual buffers is evil since each of them is a cache trasher (imagine jumping from wildly different block to wildly different block immediately).
Signed-off-by: lizzie lizzie@eden-emu.dev
Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/3215
Reviewed-by: DraVee <dravee@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Co-authored-by: lizzie <lizzie@eden-emu.dev>
Co-committed-by: lizzie <lizzie@eden-emu.dev>
