Commit graph

1271 commits

Author SHA1 Message Date
Lioncash
a1bc8ddb53 A64: Implement UADDLV 2020-04-22 20:46:21 +01:00
Lioncash
1dc1e3dcd8 fp: Use forward declarations where applicable
Minimizes the amount of files that need to be rebuilt if the headers
ever change.
2020-04-22 20:46:21 +01:00
Lioncash
46cb0d813b emit_x64_vector: Append 'v' prefix onto movq in AVX path
This is something I missed when adding in the AVX broadcast code.
2020-04-22 20:46:21 +01:00
Subv
4606a081c9 A64: The A64SetTPIDR IR instruction writes to a system register and should not be eliminated by the dead code elimination pass.
Previously this instruction was alway eliminated, resulting in incorrect values for TPIDR_EL0.
2020-04-22 20:46:21 +01:00
MerryMage
b53127600b fp: A64::FPCR -> FP::FPCR 2020-04-22 20:46:21 +01:00
MerryMage
084bf63a10 bit_util: Implement ClearBits and ModifyBits 2020-04-22 20:46:21 +01:00
MerryMage
699c5f36d5 system: Simplify static_cast 2020-04-22 20:46:21 +01:00
MerryMage
3f602129f4 system: Ensure value of CNTPCT_EL0 is accurate
Since we currently only update the host's tick count at the end of a
block, we force an end-of-block before executing a MRS %, CNTPCT_ELO
instruction.
2020-04-22 20:46:21 +01:00
Lioncash
84affdb260 safe_ops: Avoid cases where shift bases are invalid with signed values
For example, say the converted signed type is s64, shifting left  by 63
bits would be undefined behavior.

However, given an ASL is essentially the same behavior as an LSL
we can just use an unsigned type instead of converting to a signed type.
2020-04-22 20:46:21 +01:00
Lioncash
d0274f412a safe_ops: Avoid signed overflow in Negate()
Negation of values such as -9223372036854775808 can't be represented in
signed equivalents (such as long long), leading to signed overflow.
Therefore, we can just invert bits and add 1 to perform this behavior
with unsigned arithmetic.
2020-04-22 20:46:21 +01:00
Lioncash
af3e23b224 simd_scalar_shift_by_immediate: Implement FCVT{ZS, ZU} (vector, fixed-point)'s scalar double/single-precision variant 2020-04-22 20:46:21 +01:00
Lioncash
91abf87169 simd_scalar_two_register_misc: Implement FCVT{AS, AU, MS, MU, NS, NU, PS, PU, ZS, ZU} (vector)'s scalar double/single-precision variants
We can simply implement this in terms of the fixed-point IR opcodes.
2020-04-22 20:46:21 +01:00
Lioncash
0ec8dac660 emit_x64: Remove FPSCR_RoundTowardsZero() virtual function from EmitContext struct
This code was bugged in that we were comparing if the rounding mode was
not equal to rounding towards zero. Fortunately, however, nothing uses
this function anymore, and there's already the more general
FPSCR_RMode() available, so this can be removed entirely.
2020-04-22 20:46:21 +01:00
Lioncash
fd92e2f186 emit_x64: Add missing <array> include
Commit 755adef62e504a8d616de9dda8937d2428a9471b introduced a helper
alias for std::array, eliminating the need to manually type out sizes
for them, however I forgot to add the include for <array>
2020-04-22 20:46:21 +01:00
Lioncash
f939bd0228 emit_x64_vector{_floating_point}: Add helper alias for sizing arrays relative to vector width
Avoids needing to remember to specify the proper size of the arrays, all
that's needed is to specify the type of the array and the size will
automatically be deduced from it. This helps prevent potential oversized
or undersized arrays from being specified.
2020-04-22 20:46:21 +01:00
MerryMage
58f3399032 A64/PopRSBHint: Prevent RETing to a guest PC of ~0ull from crashing the jit 2020-04-22 20:46:21 +01:00
MerryMage
e18fca17dc A64: Implement FABD in terms of existing IR instructions
Fixes NaN issue. Closes #306.
2020-04-22 20:46:21 +01:00
MerryMage
1dbe9d95e6 FPRoundInt: Final FPRound based on new sign
While this shouldn't change any of the results in theory, it's just logically more consistent
2020-04-22 20:46:21 +01:00
MerryMage
83be491875 emit_x64_floating_point: SSE4.1 implementation of EmitFPRound 2020-04-22 20:46:20 +01:00
MerryMage
a40127a054 A64: Implement FRINTX, FRINTI (scalar) 2020-04-22 20:46:20 +01:00
MerryMage
962fa3b65e A64: Implement FRINTP, FRINTM, FRINTZ (scalar) 2020-04-22 20:46:20 +01:00
MerryMage
5200bf41cf A64: Implement FRINTN (scalar) 2020-04-22 20:46:20 +01:00
MerryMage
8718dc1692 A64: Implement FRINTA (scalar) 2020-04-22 20:46:20 +01:00
MerryMage
b228694012 IR: Implement FPRoundInt 2020-04-22 20:46:20 +01:00
MerryMage
e24054f4d7 fp: Implement FPRoundInt 2020-04-22 20:46:20 +01:00
MerryMage
f876e4afa2 fp: Implement FPProcessNaN 2020-04-22 20:46:20 +01:00
MerryMage
591adee443 fp/info: Add DefaultNaN 2020-04-22 20:46:20 +01:00
MerryMage
797e18cd97 fp: Move FPToFixed to its own file 2020-04-22 20:46:20 +01:00
MerryMage
295deb4035 a64_jit_state: Add FPSR.QC flag 2020-04-22 20:46:20 +01:00
Lioncash
7797bc2fb2 emit_x64_vector: Use non-scratch Use* variants of registers within EmitVectorUnsignedAbsoluteDifference()
In some cases, a register isn't modified, depending on the branch taken,
so we can signify this by using the non-scratch variants in certain
cases.
2020-04-22 20:46:20 +01:00
Lioncash
f7f83b76b7 simd_scalar_two_register_misc: Implement scalar double/single-precision variants of FCM{EQ, GE, GT, LE, LT} (zero) 2020-04-22 20:46:20 +01:00
Lioncash
9db6d1e98b translate_arm: Remove unnecessary rotr() function
We already have RotateRight() in our common code, so we can remove this
function and replace it with it. We can also implement ArmExpandImm_C()
in terms of ArmExpandImm().
2020-04-22 20:46:20 +01:00
Lioncash
9f8a44c982 cast_util: Remove unnecessary typename
Given we use std::aligned_storage_t, we don't need to specify
typename here. If we used std::aligned_storage, then we would need to.
2020-04-22 20:46:19 +01:00
MerryMage
89e43867c1 A64: Implement FADDP (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
33fa65de23 A64: Implement FADDP (vector) 2020-04-22 20:46:19 +01:00
MerryMage
9dba273a8c A64: Implement SADDLP 2020-04-22 20:46:19 +01:00
MerryMage
70ff2d73b5 A64: Implement UADDLP 2020-04-22 20:46:19 +01:00
MerryMage
5563bbbd79 A64: Implement EXT 2020-04-22 20:46:19 +01:00
MerryMage
304cc7f61e emit_x64_floating_point: SSE4.1 implementation for FP{Double,Single}ToFixed{S,U}{32,64} 2020-04-22 20:46:19 +01:00
MerryMage
3d9677d094 A64: Implement FCVTMU (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
79c9018d60 A64: Implement FCVTMS (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
49c4499a87 A64: Implement FCVTPU (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
af661ef5a6 A64: Implement FCVTPS (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
27319822bb A64: Implement FCVTAU (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
c0c7a26314 A64: Implement FCVTAS (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
a1965a74a0 A64: Implement FCVTNU (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
7d36dbcdfd A64: Implement FCVTNS (scalar) 2020-04-22 20:46:19 +01:00
MerryMage
617ca0adf0 floating_point_conversion_integer: Refactor implementation of FCVTZS_float_int and FCVTZU_float_int 2020-04-22 20:46:19 +01:00
MerryMage
caaf36dfd6 IR: Initial implementation of FP{Double,Single}ToFixed{S,U}{32,64}
This implementation just falls-back to the software floating point implementation.
2020-04-22 20:46:19 +01:00
MerryMage
760cc3ca89 EmitContext: Expose FPCR 2020-04-22 20:46:19 +01:00
MerryMage
9571269552 fp/op: Implement FPToFixed 2020-04-22 20:46:19 +01:00
MerryMage
8087e8df05 mantissa_util: Implement ResidualErrorOnRightShift
Accurately calculate residual error that is shifted out
2020-04-22 20:46:19 +01:00
MerryMage
8668d61881 fp/unpacked: Implement FPRound 2020-04-22 20:46:19 +01:00
MerryMage
55d590c01f FPCR: Add AHP setter and FZ16 getter 2020-04-22 20:46:19 +01:00
MerryMage
7360a2579b mp: Implement metaprogramming library 2020-04-22 20:46:19 +01:00
MerryMage
4ab029c114 fp: Implement FPUnpack 2020-04-22 20:46:19 +01:00
MerryMage
4875658917 fp: Implement FPProcessException 2020-04-22 20:46:19 +01:00
MerryMage
3cb98e1560 fp: Move fp_util to fp/util 2020-04-22 20:46:19 +01:00
MerryMage
c41a38b13e fp: Add FPSR 2020-04-22 20:46:19 +01:00
MerryMage
66381352f3 fp: Add FPInfo
Provides information about floating-point format for various bit sizes
2020-04-22 20:46:19 +01:00
MerryMage
d21659152c safe_ops: Implement safe shifting operations
Implement shifiting operations that perform consistently across architectures
without running into undefined or implemented-defined behaviour.
2020-04-22 20:46:19 +01:00
MerryMage
b00fe23b91 bit_util: Implement MostSignificantBit 2020-04-22 20:46:19 +01:00
MerryMage
95ad0d0a66 bit_util: Use Ones to implement Bits 2020-04-22 20:46:19 +01:00
MerryMage
62b640b2fa bit_util: Add ClearBit and ModifyBit 2020-04-22 20:46:19 +01:00
MerryMage
8651c2d10e u128: Implement u128
For when we need a 128-bit integer
2020-04-22 20:46:19 +01:00
Lioncash
e7409fdfe4 A64: Implement UCVTF (vector, integer)'s double/single-precision variant 2020-04-22 20:46:19 +01:00
Lioncash
4aa4885ba7 ir: Add opcodes for vector conversion of u32/u64 to floating-point 2020-04-22 20:46:19 +01:00
Lioncash
fcae4e2418 simd_three_different: Deduplicate common implementations
Generally, the only difference between the signed variants and the
unsigned variants is whether or not we use a sign-extension or
zero-extension, so we can simply use common functions to implement both
cases without totally duplicating code twice here.
2020-04-22 20:46:19 +01:00
Lioncash
9c0d5cf15c floating_point_conversion_integer: Handle S64/U64 -> F32 conversions in SCVTF_float_int and UCVTF_float_int 2020-04-22 20:46:19 +01:00
Lioncash
7a84b6e8d8 ir: Add opcodes for converting S64 and U64 to single-precision floating-point values 2020-04-22 20:46:19 +01:00
Lioncash
066061fa50 constant_pool: Remove unnecessary std::memset from constructor
AllocateFromCodeSpace() already zeroes out the allocated memory.
2020-04-22 20:46:19 +01:00
Lioncash
a1d6a86e8c A64: Implement ADDV 2020-04-22 20:46:19 +01:00
Lioncash
35026a6ce3 emit_x64_vector: Vectorize fallback path for EmitVectorMaxU32() 2020-04-22 20:46:19 +01:00
Lioncash
245c903129 simd_three_same: Join FPAbsoluteComparison() into FPCompareRegister()
These are part of the same comparison family, so there's no real point
in keeping them separate.
2020-04-22 20:46:19 +01:00
Lioncash
9912836b59 A64: Implement scalar double/single-precision variants of FACGE, FACGT, FCMEQ, FCMGE, FCMGT 2020-04-22 20:46:18 +01:00
MerryMage
0b97e9bd8d emit_x64_floating_point: Fix EmitFPU64ToDouble for TowardsMinusInfinity rounding mode 2020-04-22 20:46:18 +01:00
MerryMage
a2eb9a02e0 backend_x86: Add FPSCR_RMode to EmitContext 2020-04-22 20:46:18 +01:00
MerryMage
d875c08ebf fp: Extract common RoundingMode enum 2020-04-22 20:46:18 +01:00
Lioncash
3714bc0ed4 floating_point_conversion_integer: Use FPS64ToDouble and FPU64ToDouble in SCVTF_float_int and UCVTF_float_int
The opcodes introduced in 979b6f39f1621b80bd463645ec5b08661cb6b1bf can
also be used here, avoiding more falling back to the interpreter.
2020-04-22 20:46:18 +01:00
Lioncash
b97358075e simd_scalar_two_register_misc: Handle 64-bit case in SCVTF and UCVTF's scalar double/single-precision variant
Avoids falling back to the interpreter in the 64-bit case.
2020-04-22 20:46:18 +01:00
Lioncash
7252293184 emit_x64_floating_point: Correct use of UseGpr() in EmitFPU32ToDouble() and EmitFPU32ToSingle()
In the non-AVX512 path, the following code is present:

code.mov(from.cvt32(), from.cvt32());

since this potentially modifies 'from', we should be using
UseScratchGpr() instead.
2020-04-22 20:46:18 +01:00
Lioncash
fbd7623fe5 emit_x64_floating_point: Add AVX512F conversion operations to EmitFPU32ToSingle() and EmitFPU32ToDouble()
AVX-512F provides convenient instructions for these kinds of conversions
directly
2020-04-22 20:46:18 +01:00
Lioncash
3a41465eaf ir: Add opcodes for converting S64 and U64 to double-precision values 2020-04-22 20:46:18 +01:00
MerryMage
436ca80bcd Merge branch 'global_monitor' 2020-04-22 20:46:18 +01:00
Lioncash
0f4bf26e05 simd_two_register_misc: Utilize FPVectorAbs in FABS implementations
Since we already have opcodes introduced to implement FACGE and FACGT,
we can reutilize it for the FABS implementations.
2020-04-22 20:46:18 +01:00
MerryMage
821cff1227 A64: Add ClearExclusiveState method 2020-04-22 20:46:18 +01:00
Lioncash
81e572c78c ir: Extend FPVectorAbs opcode to also handle 16-bit elements for FP16 2020-04-22 20:46:18 +01:00
MerryMage
2a8de5f733 a64_emit_x64: Clear exclusive state in EmitA64CallSupervisor
The kernel would have to execute an ERET instruction to return to
userland; this clears exclusive state.
2020-04-22 20:46:18 +01:00
Lioncash
53dbb6a92a A64: Implement FACGE's vector single/double precision variants 2020-04-22 20:46:18 +01:00
MerryMage
57f7c7e1b0 Implement global exclusive monitor 2020-04-22 20:46:18 +01:00
Lioncash
6912a02d9b A64: Implement FACGT's vector single/double precision variants 2020-04-22 20:46:18 +01:00
MerryMage
85234338d3 a64_emit_x64: Simplify EmitExclusiveWrite 2020-04-22 20:46:18 +01:00
Lioncash
fc731dddae ir: Add opcodes for performing vector absolute floating-point values
This will be usable for implementing FACGE and FACGT
2020-04-22 20:46:18 +01:00
MerryMage
2fc6b33829 CMakeLists: Add missing files 2020-04-22 20:46:18 +01:00
Lioncash
0bee648b4f emit_x64_vector: Deduplicate a bit of code in EmitVectorSetElement{8, 32, 64} functions
Given both branches are the same, we can hoist out the common code.
2020-04-22 20:46:18 +01:00
Lioncash
d86fea0d28 A64: Implement FCMEQ (zero)'s vector single and double precision variant 2020-04-22 20:46:18 +01:00
Lioncash
593eca7fb1 A64: Implement load/store single structure instructions
Implements LD{1, 2, 3, 4}, LD{1, 2, 3, 4}R, and ST{1, 2, 3, 4} single
structure variants.
2020-04-22 20:46:18 +01:00
Lioncash
9bec354791 A64: Implement FCMEQ (register)'s vector single and double precision variant 2020-04-22 20:46:18 +01:00
Lioncash
b6e223fc58 emit_x64_vector: Deduplicate a bit of code within EmitVectorGetElement8()
Given both branches use the same destination register size, we can hoist
the common code out.
2020-04-22 20:46:18 +01:00
Lioncash
5ce187a54e ir: Add opcodes for floating-point vector equalities 2020-04-22 20:46:18 +01:00
MerryMage
be354dbfd0 ir/basic_block: Add missing U16 immediate type to DumpBlock 2020-04-22 20:46:18 +01:00
Lioncash
cf188448d4 emit_x64_vector: Vectorize fallback case in EmitVectorMultiply64()
Gets rid of the need to perform a fallback.
2020-04-22 20:46:18 +01:00
MerryMage
5503ff28c3 llvm_disassemble: Allow disassembly of invalid AArch64 instructions 2020-04-22 20:46:18 +01:00
Lioncash
954deff2d4 emit_x64_vector: Add break to final case in EmitVectorRoundingHalvingAddUnsigned()
This doesn't alter behavior but does make the code better if anything
else is ever added to this function in the future.
2020-04-22 20:46:18 +01:00
Lioncash
11a92eaaef A64: Implement SRHADD and URHADD 2020-04-22 20:46:18 +01:00
Lioncash
9e75d08860 A64: Implement FABD's scalar single/double precision variant 2020-04-22 20:46:18 +01:00
Lioncash
bc718c5b28 ir: Add opcodes for performing rounding halving adds 2020-04-22 20:46:18 +01:00
Lioncash
d898d1779d A64: Implement FABD's vector single/double precision variant 2020-04-22 20:46:18 +01:00
Lioncash
054549da35 emit_x64_vector: Simplify AVX-512 codepath in EmitVectorMultiply64
I realized I introduced a helper for simple AVX operation emitting, so
use that instead of writing it all out long-form.
2020-04-22 20:46:18 +01:00
Lioncash
8a4f8aed06 ir: Add opcode for performing FP vector absolute differences 2020-04-22 20:46:18 +01:00
Lioncash
cb456f914b A64: Implement UMLAL{2}, UMLSL{2}, and UMULL{2}
Now that we have the helper function set up for the signed variants, we
can also modify it to be used with the unigned ones by performing a zero
extension instead of a sign extension.
2020-04-22 20:46:18 +01:00
MerryMage
ba84e7a8de A64: Implement FNMSUB 2020-04-22 20:46:18 +01:00
Lioncash
3576c02d91 A64: Implement SMLSL{2} 2020-04-22 20:46:18 +01:00
MerryMage
a1042cfcd8 A64: Implement FNMADD 2020-04-22 20:46:18 +01:00
Lioncash
ada5c0b2fa A64: Implement SMLAL{2} 2020-04-22 20:46:18 +01:00
MerryMage
0d83032a6f A64: Implement FMSUB 2020-04-22 20:46:18 +01:00
Lioncash
2d1aca25e6 A64: Implement SMULL{2} 2020-04-22 20:46:18 +01:00
MerryMage
69e00d225c A64: Implement FMADD 2020-04-22 20:46:18 +01:00
MerryMage
8c90fcf58e IR: Implement FPMulAdd 2020-04-22 20:46:18 +01:00
Lioncash
c5ae9107a9 A64: Implement SABAL/SABAL2 and SABDL/SABDL2
Now that we have a helper function for the unsigned variants, we can
modify it to also be usable with the signed variants.
2020-04-22 20:46:18 +01:00
Lioncash
24e3299276 A64: Implement FCMGT, FCMGE (register) vector double and single precision variants 2020-04-22 20:46:18 +01:00
Lioncash
26d4473851 A64: Implement UABAL/UABAL2 2020-04-22 20:46:18 +01:00
Lioncash
350bc70be8 A64: Implement FCMGT, FCMGE, FCMLE, FCMLT (zero) vector double and single precision variants. 2020-04-22 20:46:18 +01:00
Lioncash
3397742c74 A64: Implement UABDL/UABDL2 2020-04-22 20:46:18 +01:00
Lioncash
c695da1cf3 ir: Add opcode for floating-point GE and GT comparisons
The rest of the comparisons can be implemented in terms of these two
2020-04-22 20:46:18 +01:00
Lioncash
6de5ed96e5 emit_x64_vector: Emit VPMULLQ in EmitVectorMultiply64 on AVX-512{DQ, VL} capable CPUs
Shortens code-gen down to a single instruction in the 64-bit path.
2020-04-22 20:46:18 +01:00
Lioncash
9054d1c20b A64: Implement LDR (literal, SIMD&FP) 2020-04-22 20:46:18 +01:00
Lioncash
0da5e949a8 Correct typo in DataCacheOperation enum
Fixes a typo for the InvalidateByVAToPoC enum entry. Given yuzu is the
only known user of 64-bit mode and it doesn't use this value, we can get
away with changing this.
2020-04-22 20:46:18 +01:00
Lioncash
9736e2cce2 A64: Implement FABS' half-precision variant 2020-04-22 20:46:18 +01:00
Lioncash
6e5750e4ec A64: Implement FABS' single and double precision variant 2020-04-22 20:46:18 +01:00
Lioncash
7bce8d8757 A64: Implement URSHR (scalar) and URSRA (scalar)
Now that the utility function is all set up from implementing SRSRA, the
unsigned variants can now be trivially implemented by modifying the
utility function to perform a logical shift right instead of an
arithmetical shift right for the unsigned case.
2020-04-22 20:46:18 +01:00
Lioncash
1e70a589b0 A64: Implement SRSRA (scalar) 2020-04-22 20:46:18 +01:00
Lioncash
998aef07f6 A64: Implement SRSHR (scalar) 2020-04-22 20:46:17 +01:00
Lioncash
7c0250e9f8 A64: Implement SABA 2020-04-22 20:46:17 +01:00
Lioncash
f00789e6f7 A64: Implement SABD 2020-04-22 20:46:17 +01:00
Lioncash
1e10017f4b ir: Add opcodes for signed absolute differences 2020-04-22 20:46:17 +01:00
Tillmann Karras
d3b44c1b5a decoder_detail: use structured bindings 2020-04-22 20:46:17 +01:00
Lioncash
f745eb28bf simd_two_register_misc: Handle 64-bit case for SCVTF_int_4 2020-04-22 20:46:17 +01:00
Lioncash
3f6c529da2 ir: Add opcode to perform the vector conversion S64->F64
Unfortunately x86 prior to AVX-512 doesn't really give us any convenient instruction to do the work for us
2020-04-22 20:46:17 +01:00
Lioncash
0e61ee6bf6 A64: Implement SHLL/SHLL2 2020-04-22 20:46:17 +01:00
Lioncash
43e6e98c3b A64: Add missing decoding for PRFM (unscaled offset) 2020-04-22 20:46:17 +01:00
Lioncash
f2a85d5601 A64: Implement UHSUB 2020-04-22 20:46:17 +01:00
Lioncash
b33360a324 A64: Implement SHSUB 2020-04-22 20:46:17 +01:00
Lioncash
44a5f8095a ir: Add opcodes for performing vector halving subtracts 2020-04-22 20:46:17 +01:00
Lioncash
4f37c0ec5a A64: Implement SM4EKEY 2020-04-22 20:46:17 +01:00
Lioncash
3bde3347a5 A64: Implement SM4E 2020-04-22 20:46:17 +01:00
Lioncash
b312d28295 ir: Add an opcode for doing an SM4 lookup table query 2020-04-22 20:46:17 +01:00
Lioncash
27a6d5f6ce emit_x64_vector: Use VPOPCNTB in EmitVectorPopulationCount() if AVX-512 BITALG is available 2020-04-22 20:46:17 +01:00
Lioncash
4dcc7724e0 A64: Implement UHADD 2020-04-22 20:46:17 +01:00
Lioncash
f8714f7250 A64: Implement SHADD 2020-04-22 20:46:17 +01:00
Lioncash
089096948a ir: Add opcodes for performing halving adds 2020-04-22 20:46:17 +01:00
Lioncash
3d00dd63b4 emit_x64_vector: Emit VPMINSQ and VPMINUQ for 64-bit vector min operations if AVX-512VL is available 2020-04-22 20:46:17 +01:00
Lioncash
b97b71b8aa emit_x64_vector: Emit VPMAXSQ and VPMAXUQ for 64-bit vector max operations if AVX-512VL is available 2020-04-22 20:46:17 +01:00
Lioncash
033e400df0 emit_x64_vector_floating_point: Deduplicate accurate NaN handling code
Allows the code to both be used from the 32 bit and 64 bit operations without duplicating code.
2020-04-22 20:46:17 +01:00
Lioncash
0f067b7330 emit_x64_vector: Emit VPABSQ in EmitVectorAbs() for the 64-bit case if AVX-512VL is available 2020-04-22 20:46:17 +01:00
Lioncash
d4ee878cbd emit_x64_vector: Use VPSRAQ in EmitVectorArithmeticShiftRight64() if AVX-512VL is available 2020-04-22 20:46:17 +01:00
Lioncash
b38dd191bd disassembler_arm: Remove rotation helper function in favor of Common::RotateRight
Mildly reduces the amount of duplicated behavior
2020-04-22 20:46:17 +01:00
Lioncash
51e4f1d9db emit_x64_vector: Vectorize fallback path of EmitVectorMaxS32() 2020-04-22 20:46:17 +01:00
Lioncash
c692ccdd6d emit_x64_vector: Vectorize fallback path of EmitVectorMaxS8() 2020-04-22 20:46:17 +01:00
Lioncash
b194313d8c emit_x64_vector: Vectorize fallback path in EmitVectorMinU32() 2020-04-22 20:46:17 +01:00
Lioncash
7ceda6d919 emit_x64_vector: Vectorize fallback path in EmitVectorMinU16() 2020-04-22 20:46:17 +01:00
Lioncash
cda85a1da0 emit_x64_vector: Vectorize fallback path in EmitVectorMinS32() 2020-04-22 20:46:17 +01:00
Lioncash
6e08eed210 emit_x64_vector: Vectorize fallback path in EmitVectorMinS8() 2020-04-22 20:46:17 +01:00
Lioncash
0fb6dce689 emit_x64_vector: Remove unnecessary if constexpr expression in LogicalVShift
This can simply be merged with the previous one.
2020-04-22 20:46:17 +01:00
Lioncash
5b71b1337b emit_x64_vector: Avoid left shift of negative value in LogicalVShift
Now that we handle the signed variants, we also have to be careful about left shifts with negative values,
as this is considered undefined behavior.
2020-04-22 20:46:17 +01:00
Lioncash
9954d28868 a64_jitstate: Zero SP and PC on construction of A64JitState
Given we zero out/reset everything else in the struct, do the same for these members to keep initialization consistent
2020-04-22 20:46:17 +01:00
Lioncash
4efbd40ea4 backend_x64/callback: Default virtual destructor in the cpp file
Prevents the vtable being generated in each translation unit that includes the header (and silences -Wweak-vtables warnings)
2020-04-22 20:46:17 +01:00
Lioncash
edd0b5c8c7 a32_interface/a64_interface: Change reinterpret_casts to static_casts in GetCurrentBlock thunks
It's well-defined to static_cast a void* to its proper type.
2020-04-22 20:46:17 +01:00
Lioncash
e71612d394 A64: Implement SSHL (scalar) 2020-04-22 20:46:17 +01:00
Lioncash
ef1e69a1e3 A64: Implement SSHL (vector) 2020-04-22 20:46:17 +01:00
Lioncash
21974ee57e backend_x64/ir: Amend generic LogicalVShift() template to also handle signed variants
Also adds IR opcodes to dispatch said variants
2020-04-22 20:46:17 +01:00
Lioncash
9fc89f0a0e emit_x64_vector_floating_point: Use arrays for retrieving size instead of hardcoding the size
Similar changes were done in emit_x64_vector, but these were missed.
2020-04-22 20:46:17 +01:00
Lioncash
af28e89a13 emit_x64_vector: Vectorize fallback path in EmitVectorMaxU16() 2020-04-22 20:46:17 +01:00
Lioncash
cda75e2079 A64: Implement CMTST's scalar variant 2020-04-22 20:46:17 +01:00
Lioncash
0d20423ad5 emit_x64_vector: Vectorize non-SSE4.1 fallback path for VectorMultiply32() 2020-04-22 20:46:17 +01:00
Lioncash
d70ee7c0d1 emit_x64_vector: Use VBPROADCAST where applicable and available
Uses the instruction that does what it says in its name if available. Allows avoiding the use
of a scratch register in EmitVectorBroadcast8() and EmitVectorBroadcastLower8()'s SSSE3 path.
2020-04-22 20:46:17 +01:00
Lioncash
bebe7235ae A64: Implement UZP1 and UZP2 2020-04-22 20:46:17 +01:00
Lioncash
26d77c6f09 ir: Add opcodes for performing vector deinterleaving 2020-04-22 20:46:17 +01:00
Lioncash
d6f9ed47d9 A64: Implement FNEG (half-precision) 2020-04-22 20:46:17 +01:00
Lioncash
7efbd73bac A64: Implement USHL (scalar) 2020-04-22 20:46:17 +01:00
Lioncash
41f4717f2b A64: Implement FNEG (vector) 2020-04-22 20:46:17 +01:00
Lioncash
ba1cc6366d A64: Implement RSUBHN/RSUBHN2 2020-04-22 20:46:17 +01:00
Lioncash
e41640fe33 A64: Implement RADDHN/RADDHN2 2020-04-22 20:46:17 +01:00
Lioncash
b719a6b3f7 A64: Implement XAR 2020-04-22 20:46:17 +01:00
Lioncash
0b1b131ec2 simd_two_register_misc: Factor out common comparison code
Gets rid of a tiny bit of duplicated code.
2020-04-22 20:46:17 +01:00
Lioncash
ed0b84da70 A64: Implement CMLE (zero)'s vector variant 2020-04-22 20:46:17 +01:00
Lioncash
b595a68ffa A64: Implement CMTST (vector) 2020-04-22 20:46:17 +01:00
Lioncash
48c7f8630c A64: Implement ADDHN{2} and SUBHN{2} 2020-04-22 20:46:17 +01:00
Lioncash
3acd9c9200 translate: zero extend result in Vpart when storing to lower part of vector 2020-04-22 20:46:17 +01:00
Lioncash
87ca63699f emit_x64_vector: Emit PMAXUD in EmitVectorMaxU32 on SSE4.1-capable CPUs 2020-04-22 20:46:17 +01:00
Lioncash
f17702f608 emit_x64_vector: Emit PMINUD in EmitVectorMinU32 on SSE4.1-capable CPUs 2020-04-22 20:46:17 +01:00
Lioncash
596a8dd1dd emit_x64_vector: Emit PMINSD in EmitVectorMinS32 on SSE4.1-capable CPUs
Provides a better alternative to a fallback operation.
2020-04-22 20:46:17 +01:00
Lioncash
75fd4eaaaa emit_x64_vector: Get rid of some magic numbers in loop bounds 2020-04-22 20:46:17 +01:00
Lioncash
7b80ac25eb emit_x64_vector: Generify variable shift functions 2020-04-22 20:46:17 +01:00
Lioncash
4ec735f707 A64: Implement CMLE (zero)'s scalar variant 2020-04-22 20:46:17 +01:00
Lioncash
6534184df2 A64: Implement CMLT (zero)'s scalar single/double-precision variant 2020-04-22 20:46:17 +01:00
Lioncash
8863c9bb4b A64: Implement SHA512H2 2020-04-22 20:46:17 +01:00
Lioncash
033b890e25 A64: Implement SHA512H 2020-04-22 20:46:17 +01:00
Lioncash
d1f5b084b4 A64: Handle S32->F32 case for SCVTF (vector) 2020-04-22 20:46:17 +01:00
Lioncash
38fa984b53 IR: Add opcode for packed word->f32 conversions 2020-04-22 20:46:16 +01:00