steam-frame/FEX-Emu--FEX
1
0 Fork 0
mirror of https://github.com/FEX-Emu/FEX.git synced 2026-10-10 10:00:20 +02:00
Code Issues Packages Projects Releases 67 Wiki Activity
11,870 Commits 11 Branches 71 Tags
f371bf2bc4cfdf043dfd0d596f5dbf9a0d1792cb
Commit Graph
2453 Commits
Author SHA1 Message Date
Ryan Houdek 7ca757bb6d FEXCore: Move CPUInfo to FEX 
This is only ever used in the frontend now.
 2025-04-08 22:54:43 -07:00
Ryan Houdek c2d59b02bd FEXCore: Reduce stack usage in CalculateNumberOfCPUs 
Nothing crazy, just recalculate the maximum string length rather than
use PATH_MAX.
 2025-04-08 22:54:43 -07:00
Ryan Houdek 5767c61a91 FEXCore/Emitter: Stop creating a vector on the heap 
In the Push/Pop CalleeSavedRegisters these vectors were getting created
on the heap, allocating memory and then just iterating them.

Just use a std::array which makes it stop allocating memory and saves
the number of instructions.
 2025-04-08 18:36:47 -07:00
Ryan Houdek 4377603d7e FEXCore/Allocator: Removes old workaround for kernel 4.17 
This was only used for working around our old CI machines and now that
our minimum kernel requirement is 5.15 this isn't required anymore.
 2025-04-08 18:22:39 -07:00
Ryan Houdek 2d7f37386e Softfloat: Remove warnings 
These precision warnings are no longer true!
 2025-04-07 15:22:24 -07:00
Ryan Houdek 0fbb6aa02f cephes: Rewrite to use softfloat-3e 128-bit 
And also use it at the same time, since the function signatures changed.

Instead of relying on the host libc math libraries for `long double` ALU
operations, rewrite the entire thing to use softfloat-3e fixed width
float128_t types.

This is a very invasive change in cephes but is a necessary requirement
for getting the precision we require in environments that map `long
  double` to be the same as `double`, like Win32 and MacOS.

This fixes the precision issue in transcendental operations when running
under WINE.
 2025-04-07 15:13:04 -07:00
Ryan Houdek d8cd807520 Softfloat-3e: Moves to Externals 2025-04-05 17:26:15 -07:00
Ryan Houdek ffca27cbde FEXCore/Softfloat: Wire up cephes math library for transcendental operations 
This is solving a different problem than what #4411 is specifically
trying to solve.

For our transcendental operations, we can't currently guarantee that
these functions will actually operate at the 128-bit softfloat
precision. While this is true with glibc, this is /not/ true for musl
and likely more libraries.

Instead of relying on our libc implementation to implement these,
instead include the cephes math library directly which is what most
people use for this. Including musl even, but not for all operations.

With this we are no longer beholden to the standard libraries for
providing a correct implementation.
 2025-04-04 16:03:09 -07:00
LC 74cb225ccb Merge pull request #4478 from Sonicadvance1/sha256rnds2_for_reals 
OpcodeDispatcher: Implement support for sha256rnds2 using ARM instructions
 2025-04-03 11:26:58 -04:00
LC d5db2ccf18 Merge pull request #4475 from Sonicadvance1/x87_stack_bug 
x87OptimizationPass: Fixes {Inc,Dec}StackPop
 2025-04-03 11:25:53 -04:00
Ryan Houdek cebcf50c65 x87OptimizationPass: Fixes {Inc,Dec}StackPop 
On the slow path these were pushing and popping in the wrong direction.
Switch them around to ensure the unittests work.
 2025-04-02 14:25:02 -07:00
Ryan Houdek ec1c7797f4 OpcodeDispatcher: Implement support for sha256rnds2 using ARM instructions 
The big one.
 2025-04-02 12:51:03 -07:00
Tony Wasserka 0bd924eb7e Merge pull request #4460 from Sonicadvance1/dead_code 
FEXCore/Frontend: Remove logically dead code
 2025-04-02 09:20:41 +02:00
Ryan Houdek 9b0bb29d78 IR: Implement support for sha256h{2,} 
I keep carrying this patch around. Not yet wired up to the instruction
implementation yet, but I don't want to forget about it.
 2025-04-01 17:53:48 -07:00
Ryan Houdek c3e71de1d7 FEXCore/Frontend: Remove logically dead code 
This code can't get hit.
 2025-04-01 16:09:07 -07:00
Ryan Houdek 1b18bfaff5 Merge pull request #4473 from bylaws/win32-f 
Windows: Small fixups
 2025-04-01 10:53:28 -07:00
Ryan Houdek 8aecdc536c Merge pull request #4471 from alyssarosenzweig/opt/cvtss2si 
Optimize float->integer conversions with Feat_FRINTTS
 2025-04-01 08:52:50 -07:00
Billy Laws 8f50106187 Context: Fix incorrect ifdef on ARM64EC 2025-03-31 23:57:58 +01:00
Billy Laws 02d3a319f9 OpTables: Disable thunk opcodes on win32 
They are of no use here, and are quite frequent in never-taken blocks in Denuvo games
so treating them as invalid avoids wasting some time.
 2025-03-31 23:57:58 +01:00
Ryan Houdek cdaf1c5262 Arm64Emitter: Removes warning 2025-03-31 13:51:50 -07:00
Alyssa Rosenzweig c4f7b27459 OpcodeDispatcher: accelerate F->I conversions with FRINTTS 
this should significantly help perf on supported platforms.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-31 16:09:54 -04:00
Alyssa Rosenzweig 1f08f8df0d IR: allow VUShrNI with bitshift=0 
encodes to Xtn, we need this to narrow.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-31 16:03:18 -04:00
Alyssa Rosenzweig 0038a0b19c IR: plumb Vector_FToISized op 
this exposes the frint* opcodes in a new ir op

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-31 16:03:18 -04:00
Alyssa Rosenzweig 166a7c7e53 FEXCore: plumb Feat_FRINTTS 
we want these instructions to accelerate conversions.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-31 16:01:32 -04:00
Ryan Houdek f9b369c550 Telemetry: Removes unnecessary indirection 
Telemetry value address generation was forcing an indirection at all
times which was unnecessary. These values live in the BSS, zero
initialized at process start and is unnecessary.

Instead change the wrapper defines to directly operate on the enum
passed in which saves an indirection on all of these telemetry
operations (except for the ones in the JIT which are required to be PIC
compliant).

This also fixes an annoying warning about
`FEXCORE_TELEMETRY_STATIC_INIT` causing initialization and destruction
order being unspecified, so two wins.
 2025-03-29 15:10:37 -07:00
LC 949b205f42 Merge pull request #4462 from Sonicadvance1/passes_initialize_data 
x87StackOptimizationPass: Initialize a couple of arrays
 2025-03-29 18:00:47 -04:00
Ryan Houdek feab0bce4b x87StackOptimizationPass: Initialize a couple of arrays 
Just to silence some warnings that think these aren't zero initialized
before using.
 2025-03-29 14:02:46 -07:00
Ryan Houdek 0599d80b13 FEXCore/Frontend: Changes how VEX operand encoding flags are encoded 
These three options are mutually exclusive with each other and could
potentially result in invalid encodings of the table on accident.

Change over to a 2-bit bitfield to encode if the operand that consumes
the VEX option is none, destination, 1st src, or 2nd src.

This ensures the table can't ever be incorrectly encoded.
 2025-03-29 13:51:02 -07:00
Lioncache 88e6c48db7 X86Tables: Replace deprecated std::is_trivial template 
This is deprecated in C++26, so we can just use a more specific type trait.
 2025-03-29 00:55:15 -04:00
Lioncache bfba74dab9 IR: Replace use of deprecated std::is_trivial_v template 
This is deprecated in C++26
 2025-03-29 00:33:39 -04:00
Ryan Houdek 0ca34d11ad Merge pull request #4454 from alyssarosenzweig/silly-nop 
Fix 66 90 decoding to a nop
 2025-03-28 14:50:02 -07:00
Ryan Houdek bf3275ba4a Merge pull request #4451 from pmatos/SingleStepCheck 
Enable maxinst to 1 only if singlestep exists and is enabled
 2025-03-28 14:45:40 -07:00
Paulo Matos 68b5a90518 Enable maxinst to 1 only if singlestep exists and is enabled 2025-03-28 18:09:52 +01:00
Alyssa Rosenzweig 3a2ca41724 OpcodeDispatcher: handle 66 90 as a NOP 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-28 13:03:34 -04:00
Ryan Houdek d34c287f69 Merge pull request #4450 from lioncash/ilog 
Addressing: Remove unnecessary assert in LoadEffectiveAddress()
 2025-03-28 06:43:31 -07:00
Ryan Houdek b744ca16e1 Merge pull request #4449 from pmatos/ConfigWarnSA 
Remove some warnings from Config.cpp
 2025-03-28 06:43:01 -07:00
Paulo Matos 573c262bb2 Remove some warnings from Config.cpp 
- unused includes, and
- unused functions.
 2025-03-28 14:18:40 +01:00
Lioncache ff2c2f1e1f Addressing: Remove unnecessary assert in LoadEffectiveAddress() 
ilog2 already has an assert for this.
 2025-03-28 06:08:08 -04:00
Lioncache 54dbb9f248 OpcodeDispatcher: Fix potential for uninitialized value use in RCRSmallerOp() 
If Src isn't a constant, then no value is actually assigned to SrcConst, so this
can result in uninitialized arithmetic being performed
 2025-03-28 04:45:40 -04:00
Ryan Houdek 21233430aa Merge pull request #4446 from lioncash/cond 
OpcodeDispatcher: Remove unnecessary 128-bit check in VPGATHER()
 2025-03-27 20:30:49 -07:00
Ryan Houdek bd1d6820b7 Merge pull request #4438 from Sonicadvance1/static_analysis_wars 
Static analysis warning fixes
 2025-03-27 20:30:37 -07:00
Lioncache 40b1c32008 OpcodeDispatcher: Remove unnecessary 128-bit check in VPGATHER() 
This is already guaranteed to be true, since it's checked in the outer if,
so this can just be a regular else statement.
 2025-03-27 23:08:59 -04:00
Ryan Houdek b5ed804578 Merge pull request #4444 from lioncash/jitcond 
JIT: Simplify SVE 256 operation asserts
 2025-03-27 19:46:23 -07:00
Ryan Houdek fbe3a86c4e Merge pull request #4443 from alyssarosenzweig/ra/cleanup 
RA: small cleanups
 2025-03-27 19:46:14 -07:00
Ryan Houdek ec976f3f75 Merge pull request #4435 from alyssarosenzweig/opt/pall-pf-af 
Pair PF/AF when spilling static regs
 2025-03-27 19:45:29 -07:00
Lioncache 35ee12e7e9 Addressing: Amend binary AND into logical AND in SelectAddressMode() 
Bitwise AND here is a little odd and was likely intended to be a logical AND.
 2025-03-27 17:29:08 -04:00
Lioncache 4497ab8844 JIT: Simplify SVE 256 operation asserts 
We can make these slightly less verbose.
 2025-03-27 16:55:20 -04:00
Alyssa Rosenzweig 9f399f3313 RegisterAllocationPass: rm unused arg to DecodeSRAReg 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-27 16:45:31 -04:00
Alyssa Rosenzweig 3939213336 RegisterAllocationPass: rm useless assertion 
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
 2025-03-27 16:45:31 -04:00
Ryan Houdek b8165813b4 Merge pull request #4439 from pmatos/Init-SA 
Initialize class fields to null/zero
 2025-03-27 09:32:18 -07:00