ryujinx-mirror/Ryujinx.Tests/Cpu/CpuTestSimdReg32.cs

424 lines
16 KiB
C#
Raw Normal View History

Add most of the A32 instruction set to ARMeilleure (#897) * Implement TEQ and MOV (Imm16) * Initial work on A32 instructions + SVC. No tests yet, hangs in rtld. * Implement CLZ, fix BFI and BFC Now stops on SIMD initialization. * Exclusive access instructions, fix to mul, system instructions. Now gets to a break after SignalProcessWideKey64. * Better impl of UBFX, add UDIV and SDIV Now boots way further - now stuck on VMOV instruction. * Many more instructions, start on SIMD and testing framework. * Fix build issues * svc: Rework 32 bit codepath Fixing once and for all argument ordering issues. * Fix 32 bits stacktrace * hle debug: Add 32 bits dynamic section parsing * Fix highCq mode, add many tests, fix some instruction bugs Still suffers from critical malloc failure :weary: * Fix incorrect opcode decoders and a few more instructions. * Add a few instructions and fix others. re-disable highCq for now. Disabled the svc memory clear since i'm not sure about it. * Fix build * Fix typo in ordered/exclusive stores. * Implement some more instructions, fix others. Uxtab16/Sxtab16 are untested. * Begin impl of pairwise, some other instructions. * Add a few more instructions, a quick hack to fix svcs for now. * Add tests and fix issues with VTRN, VZIP, VUZP * Add a few more instructions, fix Vmul_1 encoding. * Fix way too many instruction bugs, add tests for some of the more important ones. * Fix HighCq, enable FastFP paths for some floating point instructions (not entirely sure why these were disabled, so important to note this commit exists) Branching has been removed in A32 shifts until I figure out if it's worth it * Cleanup Part 1 There should be no functional change between these next few commits. Should is the key word. (except for removing break handler) * Implement 32 bits syscalls Co-authored-by: riperiperi <rhy3756547@hotmail.com> Implement all 32 bits counterparts of the 64 bits syscalls we currently have. * Refactor part 2: Move index/subindex logic to Operand May have inadvertently fixed one (1) bug * Add FlushProcessDataCache32 * Address jd's comments * Remove 16 bit encodings from OpCodeTable Still need to catch some edge cases (operands that use the "F" flag) and make Q encodings with non-even indexes undefined. * Correct Fpscr handling for FP vector slow paths WIP * Add StandardFPSCRValue behaviour for all Arithmetic instructions * Add StandardFPSCRValue behaviour to compare instructions. * Force passing of fpcr to FPProcessException and FPUnpack. Reduces potential for code error significantly * OpCode cleanup * Remove urgency from DMB comment in MRRC DMB is currently a no-op via the instruction, so it should likely still be a no-op here. * Test Cleanup * Fix FPDefaultNaN on Ryzen CPUs * Improve some tests, fix some shift instructions, add slow path for Vadd * Fix Typo * More test cleanup * Flip order of Fx and index, to indicate that the operand's is the "base" * Remove Simd32 register type, use Int32 and Int64 for scalars like A64 does. * Reintroduce alignment to DecoderHelper (removed by accident) * One more realign as reading diffs is hard * Use I32 registers in A32 (part 2) Swap default integer register type based on current execution mode. * FPSCR flags as Registers (part 1) Still need to change NativeContext and ExecutionContext to allow getting/setting with the flag values. * Use I32 registers in A32 (part 1) * FPSCR flags as registers (part 2) Only CMP flags are on the registers right now. It could be useful to use more of the space in non-fast-float when implementing A32 flags accurately in the fast path. * Address Feedback * Correct FP->Int behaviour (should saturate) * Make branches made by writing to PC eligible for Rejit Greatly improves performance in most games. * Remove unused branching for Vtbl * RejitRequest as a class rather than a tuple Makes a lot more sense than storing tuples on a dictionary. * Add VMOVN, VSHR (imm), VSHRN (imm) and related tests * Re-order InstEmitSystem32 Alphabetical sorting. * Address Feedback Feedback from Ac_K, remove and sort usings. * Address Feedback 2 * Address Feedback from LDj3SNuD Opcode table reordered to have alphabetical sorting within groups, Vmaxnm and Vminnm have split names to be less ambiguous, SoftFloat nits, Test nits and Test simplification with ValueSource. * Add Debug Asserts to A32 helpers Mainly to prevent the shift ones from being used on I64 operands, as they expect I32 input for most operations (eg. carry flag setting), and expect I32 input for shift and boolean amounts. Most other helper functions don't take Operands, throw on out of range values, and take specific types of OpCode, so didn't need any asserts. * Use ConstF rather than creating an operand. (useful for pooling in future) * Move exclusive load to helper, reference call flag rather than literal 1. * Address LDj feedback (minus table flatten) one final look before it's all gone. the world is so beautiful. * Flatten OpCodeTable oh no * Address more table ordering * Call Flag as int on A32 Co-authored-by: Natalie C. <cyuubiapps@gmail.com> Co-authored-by: Thog <thog@protonmail.com>
2020-02-23 21:20:40 +00:00
#define SimdReg32
using ARMeilleure.State;
using NUnit.Framework;
using System;
using System.Collections.Generic;
namespace Ryujinx.Tests.Cpu
{
[Category("SimdReg32")]
public sealed class CpuTestSimdReg32 : CpuTest32
{
#if SimdReg32
#region "ValueSource (Types)"
private static ulong[] _1B1H1S1D_()
{
return new ulong[] { 0x0000000000000000ul, 0x000000000000007Ful,
0x0000000000000080ul, 0x00000000000000FFul,
0x0000000000007FFFul, 0x0000000000008000ul,
0x000000000000FFFFul, 0x000000007FFFFFFFul,
0x0000000080000000ul, 0x00000000FFFFFFFFul,
0x7FFFFFFFFFFFFFFFul, 0x8000000000000000ul,
0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _1D_()
{
return new ulong[] { 0x0000000000000000ul, 0x7FFFFFFFFFFFFFFFul,
0x8000000000000000ul, 0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _1H1S_()
{
return new ulong[] { 0x0000000000000000ul, 0x0000000000007FFFul,
0x0000000000008000ul, 0x000000000000FFFFul,
0x000000007FFFFFFFul, 0x0000000080000000ul,
0x00000000FFFFFFFFul };
}
private static ulong[] _4H2S_()
{
return new ulong[] { 0x0000000000000000ul, 0x7FFF7FFF7FFF7FFFul,
0x8000800080008000ul, 0x7FFFFFFF7FFFFFFFul,
0x8000000080000000ul, 0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _4H2S1D_()
{
return new ulong[] { 0x0000000000000000ul, 0x7FFF7FFF7FFF7FFFul,
0x8000800080008000ul, 0x7FFFFFFF7FFFFFFFul,
0x8000000080000000ul, 0x7FFFFFFFFFFFFFFFul,
0x8000000000000000ul, 0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _8B_()
{
return new ulong[] { 0x0000000000000000ul, 0x7F7F7F7F7F7F7F7Ful,
0x8080808080808080ul, 0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _8B4H2S_()
{
return new ulong[] { 0x0000000000000000ul, 0x7F7F7F7F7F7F7F7Ful,
0x8080808080808080ul, 0x7FFF7FFF7FFF7FFFul,
0x8000800080008000ul, 0x7FFFFFFF7FFFFFFFul,
0x8000000080000000ul, 0xFFFFFFFFFFFFFFFFul };
}
private static ulong[] _8B4H2S1D_()
{
return new ulong[] { 0x0000000000000000ul, 0x7F7F7F7F7F7F7F7Ful,
0x8080808080808080ul, 0x7FFF7FFF7FFF7FFFul,
0x8000800080008000ul, 0x7FFFFFFF7FFFFFFFul,
0x8000000080000000ul, 0x7FFFFFFFFFFFFFFFul,
0x8000000000000000ul, 0xFFFFFFFFFFFFFFFFul };
}
private static IEnumerable<ulong> _1S_F_()
{
yield return 0x00000000FF7FFFFFul; // -Max Normal (float.MinValue)
yield return 0x0000000080800000ul; // -Min Normal
yield return 0x00000000807FFFFFul; // -Max Subnormal
yield return 0x0000000080000001ul; // -Min Subnormal (-float.Epsilon)
yield return 0x000000007F7FFFFFul; // +Max Normal (float.MaxValue)
yield return 0x0000000000800000ul; // +Min Normal
yield return 0x00000000007FFFFFul; // +Max Subnormal
yield return 0x0000000000000001ul; // +Min Subnormal (float.Epsilon)
if (!NoZeros)
{
yield return 0x0000000080000000ul; // -Zero
yield return 0x0000000000000000ul; // +Zero
}
if (!NoInfs)
{
yield return 0x00000000FF800000ul; // -Infinity
yield return 0x000000007F800000ul; // +Infinity
}
if (!NoNaNs)
{
yield return 0x00000000FFC00000ul; // -QNaN (all zeros payload) (float.NaN)
yield return 0x00000000FFBFFFFFul; // -SNaN (all ones payload)
yield return 0x000000007FC00000ul; // +QNaN (all zeros payload) (-float.NaN) (DefaultNaN)
yield return 0x000000007FBFFFFFul; // +SNaN (all ones payload)
}
for (int cnt = 1; cnt <= RndCnt; cnt++)
{
ulong grbg = TestContext.CurrentContext.Random.NextUInt();
ulong rnd1 = GenNormalS();
ulong rnd2 = GenSubnormalS();
yield return (grbg << 32) | rnd1;
yield return (grbg << 32) | rnd2;
}
}
private static IEnumerable<ulong> _2S_F_()
{
yield return 0xFF7FFFFFFF7FFFFFul; // -Max Normal (float.MinValue)
yield return 0x8080000080800000ul; // -Min Normal
yield return 0x807FFFFF807FFFFFul; // -Max Subnormal
yield return 0x8000000180000001ul; // -Min Subnormal (-float.Epsilon)
yield return 0x7F7FFFFF7F7FFFFFul; // +Max Normal (float.MaxValue)
yield return 0x0080000000800000ul; // +Min Normal
yield return 0x007FFFFF007FFFFFul; // +Max Subnormal
yield return 0x0000000100000001ul; // +Min Subnormal (float.Epsilon)
if (!NoZeros)
{
yield return 0x8000000080000000ul; // -Zero
yield return 0x0000000000000000ul; // +Zero
}
if (!NoInfs)
{
yield return 0xFF800000FF800000ul; // -Infinity
yield return 0x7F8000007F800000ul; // +Infinity
}
if (!NoNaNs)
{
yield return 0xFFC00000FFC00000ul; // -QNaN (all zeros payload) (float.NaN)
yield return 0xFFBFFFFFFFBFFFFFul; // -SNaN (all ones payload)
yield return 0x7FC000007FC00000ul; // +QNaN (all zeros payload) (-float.NaN) (DefaultNaN)
yield return 0x7FBFFFFF7FBFFFFFul; // +SNaN (all ones payload)
}
for (int cnt = 1; cnt <= RndCnt; cnt++)
{
ulong rnd1 = GenNormalS();
ulong rnd2 = GenSubnormalS();
yield return (rnd1 << 32) | rnd1;
yield return (rnd2 << 32) | rnd2;
}
}
private static IEnumerable<ulong> _1D_F_()
{
yield return 0xFFEFFFFFFFFFFFFFul; // -Max Normal (double.MinValue)
yield return 0x8010000000000000ul; // -Min Normal
yield return 0x800FFFFFFFFFFFFFul; // -Max Subnormal
yield return 0x8000000000000001ul; // -Min Subnormal (-double.Epsilon)
yield return 0x7FEFFFFFFFFFFFFFul; // +Max Normal (double.MaxValue)
yield return 0x0010000000000000ul; // +Min Normal
yield return 0x000FFFFFFFFFFFFFul; // +Max Subnormal
yield return 0x0000000000000001ul; // +Min Subnormal (double.Epsilon)
if (!NoZeros)
{
yield return 0x8000000000000000ul; // -Zero
yield return 0x0000000000000000ul; // +Zero
}
if (!NoInfs)
{
yield return 0xFFF0000000000000ul; // -Infinity
yield return 0x7FF0000000000000ul; // +Infinity
}
if (!NoNaNs)
{
yield return 0xFFF8000000000000ul; // -QNaN (all zeros payload) (double.NaN)
yield return 0xFFF7FFFFFFFFFFFFul; // -SNaN (all ones payload)
yield return 0x7FF8000000000000ul; // +QNaN (all zeros payload) (-double.NaN) (DefaultNaN)
yield return 0x7FF7FFFFFFFFFFFFul; // +SNaN (all ones payload)
}
for (int cnt = 1; cnt <= RndCnt; cnt++)
{
ulong rnd1 = GenNormalD();
ulong rnd2 = GenSubnormalD();
yield return rnd1;
yield return rnd2;
}
}
#endregion
private const int RndCnt = 2;
private static readonly bool NoZeros = false;
private static readonly bool NoInfs = false;
private static readonly bool NoNaNs = false;
[Explicit]
[Test, Pairwise, Description("VADD.f32 V0, V0, V0")]
public void Vadd_f32([Values(0u)] uint rd,
[Values(0u, 1u)] uint rn,
[Values(0u, 2u)] uint rm,
[ValueSource("_2S_F_")] ulong z0,
[ValueSource("_2S_F_")] ulong z1,
[ValueSource("_2S_F_")] ulong a0,
[ValueSource("_2S_F_")] ulong a1,
[ValueSource("_2S_F_")] ulong b0,
[ValueSource("_2S_F_")] ulong b1,
[Values] bool q)
{
uint opcode = 0xf2000d00u; // VADD.F32 D0, D0, D0
if (q)
{
opcode |= 1 << 6;
rm <<= 1;
rn <<= 1;
rd <<= 1;
}
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
opcode |= ((rn & 0xf) << 16) | ((rn & 0x10) << 3);
V128 v0 = MakeVectorE0E1(z0, z1);
V128 v1 = MakeVectorE0E1(a0, a1);
V128 v2 = MakeVectorE0E1(b0, b1);
SingleOpcode(opcode, v0: v0, v1: v1, v2: v2);
CompareAgainstUnicorn();
}
[Test, Pairwise, Description("VCMP.f<size> Vd, Vm")]
public void Vcmp([Values(2u, 3u)] uint size,
[ValueSource("_1S_F_")] ulong a,
[ValueSource("_1S_F_")] ulong b,
[Values] bool e)
{
uint opcode = 0xeeb40840u;
uint rm = 1;
uint rd = 2;
if (size == 3)
{
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
}
Add most of the A32 instruction set to ARMeilleure (#897) * Implement TEQ and MOV (Imm16) * Initial work on A32 instructions + SVC. No tests yet, hangs in rtld. * Implement CLZ, fix BFI and BFC Now stops on SIMD initialization. * Exclusive access instructions, fix to mul, system instructions. Now gets to a break after SignalProcessWideKey64. * Better impl of UBFX, add UDIV and SDIV Now boots way further - now stuck on VMOV instruction. * Many more instructions, start on SIMD and testing framework. * Fix build issues * svc: Rework 32 bit codepath Fixing once and for all argument ordering issues. * Fix 32 bits stacktrace * hle debug: Add 32 bits dynamic section parsing * Fix highCq mode, add many tests, fix some instruction bugs Still suffers from critical malloc failure :weary: * Fix incorrect opcode decoders and a few more instructions. * Add a few instructions and fix others. re-disable highCq for now. Disabled the svc memory clear since i'm not sure about it. * Fix build * Fix typo in ordered/exclusive stores. * Implement some more instructions, fix others. Uxtab16/Sxtab16 are untested. * Begin impl of pairwise, some other instructions. * Add a few more instructions, a quick hack to fix svcs for now. * Add tests and fix issues with VTRN, VZIP, VUZP * Add a few more instructions, fix Vmul_1 encoding. * Fix way too many instruction bugs, add tests for some of the more important ones. * Fix HighCq, enable FastFP paths for some floating point instructions (not entirely sure why these were disabled, so important to note this commit exists) Branching has been removed in A32 shifts until I figure out if it's worth it * Cleanup Part 1 There should be no functional change between these next few commits. Should is the key word. (except for removing break handler) * Implement 32 bits syscalls Co-authored-by: riperiperi <rhy3756547@hotmail.com> Implement all 32 bits counterparts of the 64 bits syscalls we currently have. * Refactor part 2: Move index/subindex logic to Operand May have inadvertently fixed one (1) bug * Add FlushProcessDataCache32 * Address jd's comments * Remove 16 bit encodings from OpCodeTable Still need to catch some edge cases (operands that use the "F" flag) and make Q encodings with non-even indexes undefined. * Correct Fpscr handling for FP vector slow paths WIP * Add StandardFPSCRValue behaviour for all Arithmetic instructions * Add StandardFPSCRValue behaviour to compare instructions. * Force passing of fpcr to FPProcessException and FPUnpack. Reduces potential for code error significantly * OpCode cleanup * Remove urgency from DMB comment in MRRC DMB is currently a no-op via the instruction, so it should likely still be a no-op here. * Test Cleanup * Fix FPDefaultNaN on Ryzen CPUs * Improve some tests, fix some shift instructions, add slow path for Vadd * Fix Typo * More test cleanup * Flip order of Fx and index, to indicate that the operand's is the "base" * Remove Simd32 register type, use Int32 and Int64 for scalars like A64 does. * Reintroduce alignment to DecoderHelper (removed by accident) * One more realign as reading diffs is hard * Use I32 registers in A32 (part 2) Swap default integer register type based on current execution mode. * FPSCR flags as Registers (part 1) Still need to change NativeContext and ExecutionContext to allow getting/setting with the flag values. * Use I32 registers in A32 (part 1) * FPSCR flags as registers (part 2) Only CMP flags are on the registers right now. It could be useful to use more of the space in non-fast-float when implementing A32 flags accurately in the fast path. * Address Feedback * Correct FP->Int behaviour (should saturate) * Make branches made by writing to PC eligible for Rejit Greatly improves performance in most games. * Remove unused branching for Vtbl * RejitRequest as a class rather than a tuple Makes a lot more sense than storing tuples on a dictionary. * Add VMOVN, VSHR (imm), VSHRN (imm) and related tests * Re-order InstEmitSystem32 Alphabetical sorting. * Address Feedback Feedback from Ac_K, remove and sort usings. * Address Feedback 2 * Address Feedback from LDj3SNuD Opcode table reordered to have alphabetical sorting within groups, Vmaxnm and Vminnm have split names to be less ambiguous, SoftFloat nits, Test nits and Test simplification with ValueSource. * Add Debug Asserts to A32 helpers Mainly to prevent the shift ones from being used on I64 operands, as they expect I32 input for most operations (eg. carry flag setting), and expect I32 input for shift and boolean amounts. Most other helper functions don't take Operands, throw on out of range values, and take specific types of OpCode, so didn't need any asserts. * Use ConstF rather than creating an operand. (useful for pooling in future) * Move exclusive load to helper, reference call flag rather than literal 1. * Address LDj feedback (minus table flatten) one final look before it's all gone. the world is so beautiful. * Flatten OpCodeTable oh no * Address more table ordering * Call Flag as int on A32 Co-authored-by: Natalie C. <cyuubiapps@gmail.com> Co-authored-by: Thog <thog@protonmail.com>
2020-02-23 21:20:40 +00:00
else
{
opcode |= ((rm & 0x1e) >> 1) | ((rm & 0x1) << 5);
opcode |= ((rd & 0x1e) << 11) | ((rd & 0x1) << 22);
}
opcode |= ((size & 3) << 8);
if (e)
{
opcode |= 1 << 7;
}
V128 v1 = MakeVectorE0(a);
V128 v2 = MakeVectorE0(b);
bool v = TestContext.CurrentContext.Random.NextBool();
bool c = TestContext.CurrentContext.Random.NextBool();
bool z = TestContext.CurrentContext.Random.NextBool();
bool n = TestContext.CurrentContext.Random.NextBool();
int fpscr = (int)(TestContext.CurrentContext.Random.NextUInt(0xf) << 28);
SingleOpcode(opcode, v1: v1, v2: v2, overflow: v, carry: c, zero: z, negative: n, fpscr: fpscr, copyFpFlags: true);
CompareAgainstUnicorn();
}
[Test, Pairwise, Description("VMLSL.<type><size> <Vd>, <Vn>, <Vm>")]
public void Vmlsl_I([Values(0u)] uint rd,
[Values(1u, 0u)] uint rn,
[Values(2u, 0u)] uint rm,
[Values(0u, 1u, 2u)] uint size,
[Random(RndCnt)] ulong z,
[Random(RndCnt)] ulong a,
[Random(RndCnt)] ulong b,
[Values] bool u)
{
uint opcode = 0xf2800a00u; // VMLSL.S8 Q0, D0, D0
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
opcode |= ((rn & 0xf) << 16) | ((rn & 0x10) << 3);
opcode |= size << 20;
if (u)
{
opcode |= 1 << 24;
}
V128 v0 = MakeVectorE0E1(z, z);
V128 v1 = MakeVectorE0E1(a, z);
V128 v2 = MakeVectorE0E1(b, z);
SingleOpcode(opcode, v0: v0, v1: v1, v2: v2);
CompareAgainstUnicorn();
}
[Test, Pairwise, Description("VMULL.<size> <Vd>, <Vn>, <Vm>")]
public void Vmull_I([Values(0u)] uint rd,
[Values(1u, 0u)] uint rn,
[Values(2u, 0u)] uint rm,
[Values(0u, 1u, 2u)] uint size,
[Random(RndCnt)] ulong z,
[Random(RndCnt)] ulong a,
[Random(RndCnt)] ulong b,
[Values] bool op,
[Values] bool u)
{
uint opcode = 0xf2800c00u; // VMULL.S8 Q0, D0, D0
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
opcode |= ((rn & 0xf) << 16) | ((rn & 0x10) << 3);
if (op)
{
opcode |= 1 << 9;
size = 0;
u = false;
}
opcode |= size << 20;
if (u)
{
opcode |= 1 << 24;
}
V128 v0 = MakeVectorE0E1(z, z);
V128 v1 = MakeVectorE0E1(a, z);
V128 v2 = MakeVectorE0E1(b, z);
SingleOpcode(opcode, v0: v0, v1: v1, v2: v2);
CompareAgainstUnicorn();
}
Add most of the A32 instruction set to ARMeilleure (#897) * Implement TEQ and MOV (Imm16) * Initial work on A32 instructions + SVC. No tests yet, hangs in rtld. * Implement CLZ, fix BFI and BFC Now stops on SIMD initialization. * Exclusive access instructions, fix to mul, system instructions. Now gets to a break after SignalProcessWideKey64. * Better impl of UBFX, add UDIV and SDIV Now boots way further - now stuck on VMOV instruction. * Many more instructions, start on SIMD and testing framework. * Fix build issues * svc: Rework 32 bit codepath Fixing once and for all argument ordering issues. * Fix 32 bits stacktrace * hle debug: Add 32 bits dynamic section parsing * Fix highCq mode, add many tests, fix some instruction bugs Still suffers from critical malloc failure :weary: * Fix incorrect opcode decoders and a few more instructions. * Add a few instructions and fix others. re-disable highCq for now. Disabled the svc memory clear since i'm not sure about it. * Fix build * Fix typo in ordered/exclusive stores. * Implement some more instructions, fix others. Uxtab16/Sxtab16 are untested. * Begin impl of pairwise, some other instructions. * Add a few more instructions, a quick hack to fix svcs for now. * Add tests and fix issues with VTRN, VZIP, VUZP * Add a few more instructions, fix Vmul_1 encoding. * Fix way too many instruction bugs, add tests for some of the more important ones. * Fix HighCq, enable FastFP paths for some floating point instructions (not entirely sure why these were disabled, so important to note this commit exists) Branching has been removed in A32 shifts until I figure out if it's worth it * Cleanup Part 1 There should be no functional change between these next few commits. Should is the key word. (except for removing break handler) * Implement 32 bits syscalls Co-authored-by: riperiperi <rhy3756547@hotmail.com> Implement all 32 bits counterparts of the 64 bits syscalls we currently have. * Refactor part 2: Move index/subindex logic to Operand May have inadvertently fixed one (1) bug * Add FlushProcessDataCache32 * Address jd's comments * Remove 16 bit encodings from OpCodeTable Still need to catch some edge cases (operands that use the "F" flag) and make Q encodings with non-even indexes undefined. * Correct Fpscr handling for FP vector slow paths WIP * Add StandardFPSCRValue behaviour for all Arithmetic instructions * Add StandardFPSCRValue behaviour to compare instructions. * Force passing of fpcr to FPProcessException and FPUnpack. Reduces potential for code error significantly * OpCode cleanup * Remove urgency from DMB comment in MRRC DMB is currently a no-op via the instruction, so it should likely still be a no-op here. * Test Cleanup * Fix FPDefaultNaN on Ryzen CPUs * Improve some tests, fix some shift instructions, add slow path for Vadd * Fix Typo * More test cleanup * Flip order of Fx and index, to indicate that the operand's is the "base" * Remove Simd32 register type, use Int32 and Int64 for scalars like A64 does. * Reintroduce alignment to DecoderHelper (removed by accident) * One more realign as reading diffs is hard * Use I32 registers in A32 (part 2) Swap default integer register type based on current execution mode. * FPSCR flags as Registers (part 1) Still need to change NativeContext and ExecutionContext to allow getting/setting with the flag values. * Use I32 registers in A32 (part 1) * FPSCR flags as registers (part 2) Only CMP flags are on the registers right now. It could be useful to use more of the space in non-fast-float when implementing A32 flags accurately in the fast path. * Address Feedback * Correct FP->Int behaviour (should saturate) * Make branches made by writing to PC eligible for Rejit Greatly improves performance in most games. * Remove unused branching for Vtbl * RejitRequest as a class rather than a tuple Makes a lot more sense than storing tuples on a dictionary. * Add VMOVN, VSHR (imm), VSHRN (imm) and related tests * Re-order InstEmitSystem32 Alphabetical sorting. * Address Feedback Feedback from Ac_K, remove and sort usings. * Address Feedback 2 * Address Feedback from LDj3SNuD Opcode table reordered to have alphabetical sorting within groups, Vmaxnm and Vminnm have split names to be less ambiguous, SoftFloat nits, Test nits and Test simplification with ValueSource. * Add Debug Asserts to A32 helpers Mainly to prevent the shift ones from being used on I64 operands, as they expect I32 input for most operations (eg. carry flag setting), and expect I32 input for shift and boolean amounts. Most other helper functions don't take Operands, throw on out of range values, and take specific types of OpCode, so didn't need any asserts. * Use ConstF rather than creating an operand. (useful for pooling in future) * Move exclusive load to helper, reference call flag rather than literal 1. * Address LDj feedback (minus table flatten) one final look before it's all gone. the world is so beautiful. * Flatten OpCodeTable oh no * Address more table ordering * Call Flag as int on A32 Co-authored-by: Natalie C. <cyuubiapps@gmail.com> Co-authored-by: Thog <thog@protonmail.com>
2020-02-23 21:20:40 +00:00
[Test, Pairwise, Description("VSHL.<size> {<Vd>}, <Vm>, <Vn>")]
public void Vshl([Values(0u)] uint rd,
[Values(1u, 0u)] uint rn,
[Values(2u, 0u)] uint rm,
[Values(0u, 1u, 2u, 3u)] uint size,
[Random(RndCnt)] ulong z,
[Random(RndCnt)] ulong a,
[Random(RndCnt)] ulong b,
[Values] bool q,
[Values] bool u)
{
uint opcode = 0xf2000400u; // VSHL.S8 D0, D0, D0
if (q)
{
opcode |= 1 << 6;
rm <<= 1;
rn <<= 1;
rd <<= 1;
}
if (u)
{
opcode |= 1 << 24;
}
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
opcode |= ((rn & 0xf) << 16) | ((rn & 0x10) << 3);
opcode |= size << 20;
V128 v0 = MakeVectorE0E1(z, z);
V128 v1 = MakeVectorE0E1(a, z);
V128 v2 = MakeVectorE0E1(b, z);
SingleOpcode(opcode, v0: v0, v1: v1, v2: v2);
CompareAgainstUnicorn();
}
[Explicit]
[Test, Pairwise, Description("VPADD.f32 V0, V0, V0")]
public void Vpadd_f32([Values(0u)] uint rd,
[Range(0u, 7u)] uint rn,
[Range(0u, 7u)] uint rm)
{
// not currently a slow path test - just a sanity check for pairwise
uint opcode = 0xf3000d00u; // VPADD.F32 D0, D0, D0
opcode |= ((rm & 0xf) << 0) | ((rm & 0x10) << 1);
opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
opcode |= ((rn & 0xf) << 16) | ((rn & 0x10) << 3);
var rnd = TestContext.CurrentContext.Random;
V128 v0 = new V128(rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue));
V128 v1 = new V128(rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue));
V128 v2 = new V128(rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue), rnd.NextFloat(int.MinValue, int.MaxValue));
SingleOpcode(opcode, v0: v0, v1: v1, v2: v2);
CompareAgainstUnicorn();
}
#endif
}
}