steam-frame/FEX-Emu--FEX
1
0 Fork 0
mirror of https://github.com/FEX-Emu/FEX.git synced 2026-10-14 10:00:42 +02:00
Code Issues Packages Projects Releases 67 Wiki Activity
5,981 Commits 11 Branches 71 Tags
ad37c19043cd3ecfd0316fed3274b6a79ea7d55c
Commit Graph
5981 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Lioncache ad37c19043 ARMEmitter: Handle SVE integer compare scalar count and limit group 2023-02-27 14:09:40 -05:00
Ryan Houdek f2aa0026b5 Merge pull request #2439 from lioncash/log 
Emitter/ALUOps: Fix typos in log messages
 2023-02-23 16:44:19 -08:00
Lioncache 553efbeb29 Emitter/ALUOps: Fix typos in log messages 
Fixes a few incorrect instruction names in the logs.
 2023-02-23 19:14:45 -05:00
Ryan Houdek b39a882a2d Merge pull request #2438 from lioncash/restrict 
OpcodeDispatcher: Restrict partial XMM stores to FPRs in StoreResult_WithOpSize
 2023-02-23 16:06:53 -08:00
Lioncache 85f7f8e6c0 OpcodeDispatcher: Restrict partial XMM stores to FPRs in StoreResult_WithOpSize 
As far as I know, nothing actually uses this path. Partially resolves
the TODO of dealing with partial writes.
 2023-02-23 18:48:24 -05:00
Ryan Houdek 4d25de31de Merge pull request #2437 from lioncash/dup 
OpcodeDispatcher: Remove now unused _VDupElement path in LoadSource_WithOpSize
 2023-02-23 13:58:09 -08:00
Ryan Houdek 9e01730c6c Merge pull request #2436 from lioncash/builtin 
Arm64Emitter: Use bit utils wrapper over __builtin_ffs
 2023-02-23 13:51:24 -08:00
Lioncache fea3ee1298 OpcodeDispatcher: Remove now unused _VDupElement path in LoadSource_WithOpSize 
XMM instances can't use high indices anymore, since we've gotten rid of
the only flag that allows this scenario to occur.
 2023-02-23 16:09:03 -05:00
Lioncache e1c42315ed Arm64Emitter: Use bit utils wrapper over __builtin_ffs 
Just keeps the use of builtins contained to one place.
 2023-02-23 15:23:07 -05:00
Ryan Houdek 165db37c8d Merge pull request #2434 from lioncash/predmisc 
ARMEmitter: Finish off SVE Predicate Misc group
 2023-02-23 12:20:03 -08:00
Ryan Houdek 9b23ae9133 Merge pull request #2433 from lioncash/subsw 
OpcodeDispatcher: Handle VPHSUBSW
 2023-02-23 12:17:58 -08:00
Ryan Houdek f951a406e6 Merge pull request #2435 from lioncash/mov 
OpcodeDispatcher: Share MOVHPD implementation with MOVHPS
 2023-02-23 12:16:27 -08:00
Lioncache 497b5c0561 X86Tables: Reclaim FLAGS_SF_HIGH_XMM_REG as an unused flag 
Now that we've moved MOVHPS over to sharing the implementation of
MOVHPD, the FLAGS_SF_HIGH_XMM_REG is now unused.

Since we're supporting AVX, this flag is kind of weird in terms of
behavior, since what determines the high part of a register is now
situationally different.

Also it's much more explicit to perform the insert directly in the
implementation of instructions, than relying on a flag to do it for us.

So, instead of keeping it around, we can reclaim it as unused for use
with any necessary behavior that we would require in the future.
 2023-02-23 14:22:25 -05:00
Lioncache 95393b07fb OpcodeDispatcher: Share MOVHPD implementation with MOVHPS 
These instructions essentially have the same behavior. This also allows
us to remove the only used instance of FLAGS_SF_HIGH_XMM_REG, which,
given that we now support AVX, has ambiguous use.

While we're at it, we can expand the tests to make use of the store to
memory variant.

Also removes an erroneous copy-pasted comment about ZEXTing. This is
from the MOVQ implementation function. MOVHPS/MOVHPD don't do any
ZEXTing, they either store to memory or insert into a register.
 2023-02-23 14:07:22 -05:00
Lioncache 372da1b820 ARMEmitter: Handle PNEXT 
Now, with the helper in place, we can implement PNEXT and finish off the
SVE Predicate Misc group.
 2023-02-23 11:59:39 -05:00
Lioncache 61a59d0314 ARMEmitter: Unify SVE Predicate Misc group under single helper 
Centralizes the implementations and also gets rid of some code in the
process.
 2023-02-23 11:51:03 -05:00
Lioncache 1045e05870 OpcodeDispatcher: Handle VPHSUBSW 2023-02-23 10:55:57 -05:00
Lioncache 052872725c OpcodeDispatcher: Factor out PHSUBS implementation into helper 
This will allow it to be shared in the AVX implementation.
 2023-02-23 10:32:48 -05:00
Ryan Houdek 4d655218ab Merge pull request #2431 from lioncash/brk 
ARMEmitter: Handle SVE partition break categories
 2023-02-22 21:13:33 -08:00
Lioncache 78ba195b66 ARMEmitter: Handle SVE partition break condition category 2023-02-22 22:49:37 -05:00
Lioncache ae2b28716d ARMEmitter: Handle SVE propagate break to next partition category 2023-02-22 22:41:55 -05:00
Lioncache 326e5e8d57 ARMEmitter: Handle propagate break from previous partition category 2023-02-22 22:35:31 -05:00
Ryan Houdek 0a8fc2cbef Merge pull request #2430 from lioncash/assert 
ARMEmitter: Handle SVE integer compare with wide elements category
 2023-02-22 18:57:33 -08:00
Lioncache 552293b226 ARMEmitter: Handle SVE integer compare with wide elements category 
We can piggy-back on top of the existing SVEIntegerCompareVector to make
these trivial to implement.
 2023-02-22 21:03:35 -05:00
Lioncache 751a4c8019 ARMEmitter: Move assertion into SVEIntegerCompareVector 
Same behavior, but centralizes the assertion. While we're at it, we can
also add another assert to ensure that only predicates p0-p7 are used.
 2023-02-22 20:14:49 -05:00
Ryan Houdek 68b2072eab Merge pull request #2429 from lioncash/align 
OpcodeDispatcher: Handle alignment for MOVAPS a little better
 2023-02-22 14:40:28 -08:00
Lioncache e3cac40b1b OpcodeDispatcher: Fix SSE MOVAPS variants being treated as MOVUPS 
0x10/0x11 in the two byte op table corresponds to MOVUPS
0x28/0x29 in the two byte op table corresponds to MOVAPS
 2023-02-22 15:56:15 -05:00
Ryan Houdek 9b123353b3 Merge pull request #2428 from lioncash/hsub 
OpcodeDispatcher: Handle VHSUBPD/VHSUBPS
 2023-02-22 11:43:10 -08:00
Lioncache f2c0c55b9c OpcodeDispatcher: Handle VHSUBPS 2023-02-22 14:27:51 -05:00
Ryan Houdek a4c694ffc7 Merge pull request #2427 from lioncash/pred 
ARMEmitter: Finish off SVE Permute Vector - Predicated group
 2023-02-22 11:26:49 -08:00
Lioncache 1eb722dea7 OpcodeDispatcher: Handle VHSUBPD 2023-02-22 14:12:11 -05:00
Lioncache a6746988d7 x86_64/VectorOps: Fix behavior of UnZip2 with 64-bit element 256-bit vectors 
The 256-bit variant of vshufpd uses extra immediate bits rather than the
same bits for the lower lane.
 2023-02-22 14:12:11 -05:00
Lioncache 0a1707f1bd OpcodeDispatcher: Factor HSUBP implementation into helper 
Will be used for implementing the AVX variants of the same instructions.
 2023-02-22 12:12:55 -05:00
Lioncache 23b9d8e108 ARMEmitter: Add check for registers being consecutive in constructive SPLICE 
Will catch cases where registers aren't consecutive in the constructive
variant. While we're at it, we can also amend EXT's similar but slightly wrong
consecutive check.

Also adds tests to ensure these corner-cases hold.
 2023-02-22 11:59:58 -05:00
Lioncache 3fa44604ba ARMEmitter: Make SPLICE use SVEPermuteVectorPredicated 
These are in the same instruction category, so we can use the helper to
simplify the implementation.
 2023-02-22 11:48:11 -05:00
Lioncache d3bc0c084d ARMEmitter: Make CPY (SIMD&FP) and CPY (scalar) use SVEPermuteVectorPredicated 
These fall under the same instruction category, so we can use the helper
to simplify the implementation.
 2023-02-22 11:29:07 -05:00
Lioncache e206414919 ARMEmitter: Make COMPACT use SVEPermuteVectorPredicated 
This falls under the same category of instructions, so we can use it to
simplify the implementation.
 2023-02-22 11:23:37 -05:00
Lioncache e635cc5404 ARMEmitter: Use predicated helper with revb/revh/revw/rbit 
Since these are under the same category, we can merge these and get rid
of a now unnecessary helper.
 2023-02-22 11:17:12 -05:00
Lioncache 822d67467b ARMEmitter: Handle SVE conditionally extract element to GPR/scalar categories 2023-02-22 11:10:13 -05:00
Lioncache e043d2c0f5 ARMEmitter: Handle SVE conditionally broadcast element to vector category 2023-02-22 10:55:34 -05:00
Lioncache d58c4405f7 ARMEmitter: Handle extract element to general register/scalar categories 2023-02-22 10:47:49 -05:00
Mai 66d879f387 Merge pull request #2400 from Sonicadvance1/rip_reconstruct 
Dispatcher: Support reconstructing RIP from block entry
 2023-02-22 09:48:02 -05:00
Mai 55d3edb8e6 Merge pull request #2426 from Sonicadvance1/optimize_getemulatedpath 
FileManagement: Optimize GetEmulatedFDPath with an FD!
 2023-02-22 09:46:44 -05:00
Ryan Houdek 98f0f22f41 FileManagement: Optimize GetEmulatedFDPath with an FD! 
Performance stats up front:
This improves pressure-vessel startup time on my test device by 10.1%
Improving the startup time from 9.71425 seconds to 8.7421 seconds.

Most filesystem based syscalls support a file descriptor version with an
*at suffix. This allows us to do these syscalls with pathnames that are
relative to the directory FD that is passed to the syscall.

This is pretty much exactly what we want when we are searching for files
inside of our rootfs. The only quirk ends up being that we are getting
passed absolute paths. This ends up being very simple to workaround by
stripping off the front '/' character. Doing this is just offsetting the
pointer passed to the syscall by one byte.

This does require having two temporary buffers of size PATH_MAX passed
to the handler since just like in the other implementation, we need to
keep the previous result around. The difference being now that we aren't
doing a bunch of std::string temporary manipulation and now we are
returning one of the passed in buffers back depending on the result.
 2023-02-22 01:25:40 -08:00
Ryan Houdek 5f574fb935 Merge pull request #2425 from lioncash/xop 
VEXTables: Remove VPERMIL2PD and VPERMIL2PS entries
 2023-02-20 18:27:32 -08:00
Ryan Houdek 618f5bb869 Merge pull request #2424 from lioncash/permil 
OpcodeDispatcher: Handle register variants of VPERMILPD/VPERMILPS
 2023-02-20 18:02:14 -08:00
Lioncache b5ca5f173e VEXTables: Remove VPERMIL2PD and VPERMIL2PS entries 
These are actually XOP instructions. That, despite being so, are encoded
using a VEX prefix.
 2023-02-20 20:58:55 -05:00
Lioncache 5cf6a680bb OpcodeDispatcher: Handle register variants of VPERMILPD/VPERMILPS 2023-02-20 20:32:31 -05:00
Ryan Houdek 645f40bb96 Merge pull request #2423 from lioncash/permd 
OpcodeDispatcher: Handle VPERMD/VPERMPS
 2023-02-20 15:20:59 -08:00
Ryan Houdek 268deddd09 Merge pull request #2422 from lioncash/phadds 
OpcodeDispatcher: Handle VPHADDSW
 2023-02-20 15:20:20 -08:00