Commit graph

78 commits

Author SHA1 Message Date
riperiperi
9db73f74cf
ARMeilleure: Respect FZ/RM flags for all floating point operations (#4618)
* ARMeilleure: Respect Fz flag for all floating point operations.

This is a change in strategy for emulating the Fz FPCR flag. Before, it was set before instructions that "needed it" and reset after. However, this missed a few hot instructions like the multiplication instruction, and the entirety of A32.

The new strategy is to set the Fz flag only in the following circumstances:

- Set to match FPCR before translated functions/loop are executed.
- Reset when calling SoftFloat methods, set when returning.
- Reset when exiting execution.

This allows us to remove the code around the existing Fz aware instructions, and get the accuracy benefits on all floating point instructions executed while in translated code.

Single step executions now need to be called with a context wrapper - right now it just contains the Fz flag initialization, and won't actually do anything on ARM.

This fixes a bug in Breath of the Wild where some physics interactions could randomly crash the game due to subnormal values not flushing to zero.

This is draft right now because I need to answer the questions:
- Does dotnet avoid changing the value of Mxcsr?
- Is it a good idea to assume that? Or should the flag set/restore be done on every managed method call, not just softfloat?
- If we assume that, do we want a unit test to verify the behaviour?

I recommend testing a bunch of games, especially games affected when this was originally added, such as #1611.

* Remove unused method

* Use FMA for Fmadd, Fmsub, Fnmadd, Fnmsub, Fmla, Fmls

...when available.

Similar implementation to A32

* Use FMA for Frecps, Frsqrts

* Don't set DAZ.

* Add round mode to ARM FP mode

* Fix mistakes

* Add test for FP state when calling managed methods

* Add explanatory comment to test.

* Cleanup

* Add A64 FPCR flags

* Vrintx_S A32 fast path on A64 backend

* Address feedback 1, re-enable DAZ

* Fix FMA instructions By Elem

* Address feedback
2023-04-10 12:22:58 +02:00
Wunk
0992310b76
ARMeilleure: Check for XSAVE cpuid flag for AVX{2,512} (#4584)
Protection for the `xgetbv` instruction for systems that do not support
`xcr0` such as nehalem processors.

The `XSAVE` cpuid indicates support for `XSAVE`, `XRESTOR`, `XSETBV`,
`XGETBV` while `OSXSAVE` indicates if the operating system itself has
`XSAVE` turned on. Both must be checked at the same time.
2023-03-22 14:51:21 -03:00
Wunk
17620d18db
ARMeilleure: Add initial support for AVX512 (EVEX encoding) (cont) (#4147)
* ARMeilleure: Add AVX512{F,VL,DQ,BW} detection

Add `UseAvx512Ortho` and `UseAvx512OrthoFloat` optimization flags as
short-hands for `F+VL` and `F+VL+DQ`.

* ARMeilleure: Add initial support for EVEX instruction encoding

Does not implement rounding, or exception controls.

* ARMeilleure: Add `X86Vpternlogd`

Accelerates the vector-`Not` instruction.

* ARMeilleure: Add check for `OSXSAVE` for AVX{2,512}

* ARMeilleure: Add check for `XCR0` flags

Add XCR0 register checks for AVX and AVX512F, following the guidelines
from section 14.3 and 15.2 from the Intel Architecture Software
Developer's Manual.

* ARMeilleure: Remove redundant `ReProtect` and `Dispose`, formatting

* ARMeilleure: Move XCR0 procedure to GetXcr0Eax

* ARMeilleure: Add `XCR0` to `FeatureInfo` structure

* ARMeilleure: Utilize `ReadOnlySpan` for Xcr0 assembly

Avoids an additional allocation

* ARMeilleure: Formatting fixes

* ARMeilleure: Fix EVEX encoding src2 register index

> Just like in VEX prefix, vvvv is provided in inverted form.

* ARMeilleure: Add `X86Vpternlogd` acceleration to `Vmvn_I`

Passes unit tests, verified instruction utilization

* ARMeilleure: Fix EVEX register operand designations

Operand 2 was being sourced improperly.

EVEX encoded instructions source their operands like so:
Operand 1: ModRM:reg
Operand 2: EVEX.vvvvv
Operand 3: ModRM:r/m
Operand 4: Imm

This fixes the improper register designations when emitting vpternlog.
Now "dest", "src1", "src2" arguments emit in the proper order in EVEX instructions.

* ARMeilleure: Add `X86Vpternlogd` acceleration to `Orn_V`

* ARMeilleure: PTC version bump

* ARMeilleure: Update EVEX encoding Debug.Assert to Debug.Fail

* ARMeilleure: Update EVEX encoding comment capitalization
2023-03-20 16:09:24 -03:00
jhorv
5131b71437
Reducing memory allocations (#4537)
* add RecyclableMemoryStream dependency and MemoryStreamManager

* organize BinaryReader/BinaryWriter extensions

* add StreamExtensions to reduce need for BinaryWriter

* simple replacments of MemoryStream with RecyclableMemoryStream

* add write ReadOnlySequence<byte> support to IVirtualMemoryManager

* avoid 0-length array creation

* rework IpcMessage and related types to greatly reduce memory allocation by using RecylableMemoryStream, keeping streams around longer, avoiding their creation when possible, and avoiding creation of BinaryReader and BinaryWriter when possible

* reduce LINQ-induced memory allocations with custom methods to query KPriorityQueue

* use RecyclableMemoryStream in StreamUtils, and use StreamUtils in EmbeddedResources

* add constants for nanosecond/millisecond conversions

* code formatting

* XML doc adjustments

* fix: StreamExtension.WriteByte not writing non-zero values for lengths <= 16

* XML Doc improvements. Implement StreamExtensions.WriteByte() block writes for large-enough count values.

* add copyless path for StreamExtension.Write(ReadOnlySpan<int>)

* add default implementation of IVirtualMemoryManager.Write(ulong, ReadOnlySequence<byte>); remove previous explicit implementations

* code style fixes

* remove LINQ completely from KScheduler/KPriorityQueue by implementing a custom struct-based enumerator
2023-03-17 13:14:50 +01:00
jhorv
23c844b2aa
Misc performance tweaks (#4509)
* use Array.Empty() where instead of allocating new zero-length arrays

* structure for loops in a way that the JIT will elide array/Span bounds checking

* avoiding function calls in for loop condition tests

* avoid LINQ in a hot path

* conform with code style

* fix mistake in GetNextWaitingObject()

* fix GetNextWaitingObject() possibility of returning null if all list items have TimePoint == long.MaxValue

* make GetNextWaitingObject() behave FIFO behavior for multiple items with the same TimePoint
2023-03-11 17:05:48 -03:00
gdkchan
f0562b9c75
CPU: Avoid argument value copies on the JIT (#4484)
* Minor refactoring of the pre-allocator

* Avoid LoadArgument copies

* PPTC version bump
2023-03-08 23:25:35 +01:00
gdkchan
a1a4771ac1
Remove use of GetFunctionPointerForDelegate to get JIT cache function pointer (#4337)
* Remove use of GetFunctionPointerForDelegate to get JIT cache function pointer

* Rename FuncPtr to FuncPointer
2023-01-23 22:37:53 +00:00
merry
4f293f8cbe
Arm64: Simplify TryEncodeBitMask and use for constants (#4328)
* Arm64: Simplify TryEncodeBitMask

* CodeGenerator: Use TryEncodeBitMask in GenerateConstantCopy

* Ptc: Bump version
2023-01-22 14:15:49 +00:00
merry
a11784fcbf
Arm64: Cpu feature detection (#4264)
* Arm64: Cpu feature detection

* Ptc: Add Arm64 feature info

* nits

* simplify CheckSysctlName

* restore some macos flags

* feedback
2023-01-12 08:05:18 +01:00
gdkchan
5e0f8e8738
Implement JIT Arm64 backend (#4114)
* Implement JIT Arm64 backend

* PPTC version bump

* Address some feedback from Arm64 JIT PR

* Address even more PR feedback

* Remove unused IsPageAligned function

* Sync Qc flag before calls

* Fix comment and remove unused enum

* Address riperiperi PR feedback

* Delete Breakpoint IR instruction that was only implemented for Arm64
2023-01-10 19:16:59 -03:00
gdkchan
ee0f9b03a4
Eliminate zero-extension moves in more cases on 32-bit games (#4140)
* Eliminate zero-extension moves in more cases on 32-bit games

* PPTC version bump

* Revert X86Optimizer changes
2022-12-19 14:45:58 -03:00
gdkchan
f93c5f006a
Revert "ARMeilleure: Add initial support for AVX512(EVEX encoding) (#3663)" (#4145)
This reverts commit 295fbd0542.
2022-12-18 20:21:10 -03:00
Wunk
295fbd0542
ARMeilleure: Add initial support for AVX512(EVEX encoding) (#3663)
* ARMeilleure: Add AVX512{F,VL,DQ,BW} detection

Add `UseAvx512Ortho` and `UseAvx512OrthoFloat` optimization flags as
short-hands for `F+VL` and `F+VL+DQ`.

* ARMeilleure: Add initial support for EVEX instruction encoding

Does not implement rounding, or exception controls.

* ARMeilleure: Add `X86Vpternlogd`

Accelerates the vector-`Not` instruction.

* ARMeilleure: Add check for `OSXSAVE` for AVX{2,512}

* ARMeilleure: Add check for `XCR0` flags

Add XCR0 register checks for AVX and AVX512F, following the guidelines
from section 14.3 and 15.2 from the Intel Architecture Software
Developer's Manual.

* ARMeilleure: Increment InternalVersion

* ARMeilleure: Remove redundant `ReProtect` and `Dispose`, formatting

* ARMeilleure: Move XCR0 procedure to GetXcr0Eax

* ARMeilleure: Add `XCR0` to `FeatureInfo` structure

* ARMeilleure: Utilize `ReadOnlySpan` for Xcr0 assembly

Avoids an additional allocation

* ARMeilleure: Formatting fixes
2022-12-18 16:46:13 -03:00
Andrey Sukharev
4da44e09cb
Make structs readonly when applicable (#4002)
* Make all structs readonly when applicable. It should reduce amount of needless defensive copies

* Make structs with trivial boilerplate equality code record structs

* Remove unnecessary readonly modifiers from TextureCreateInfo

* Make BitMap structs readonly too
2022-12-05 14:47:39 +01:00
Wunk
45ce540b9b
ARMeilleure: Add gfni acceleration (#3669)
* ARMeilleure: Add `GFNI` detection

This is intended for utilizing the `gf2p8affineqb` instruction

* ARMeilleure: Add `gf2p8affineqb`

Not using the VEX or EVEX-form of this instruction is intentional. There
are `GFNI`-chips that do not support AVX(so no VEX encoding) such as
Tremont(Lakefield) chips as well as Jasper Lake.

13df339fe7/GenuineIntel/GenuineIntel00806A1_Lakefield_LC_InstLatX64.txt (L1297-L1299)

13df339fe7/GenuineIntel/GenuineIntel00906C0_JasperLake_InstLatX64.txt (L1252-L1254)

* ARMeilleure: Add `gfni` acceleration of `Rbit_V`

Passes all `Rbit_V*` unit tests on my `i9-11900k`

* ARMeilleure: Add `gfni` acceleration of `S{l,r}i_V`

Also added a fast-path for when the shift amount is greater than the
size of the element.

* ARMeilleure: Add `gfni` acceleration of `Shl_V` and `Sshr_V`

* ARMeilleure: Increment InternalVersion

* ARMeilleure: Fix Intrinsic and Assembler Table alignment

`gf2p8affineqb` is the longest instruction name I know of. It shouldn't
get any wider than this.

* ARMeilleure: Remove SSE2+SHA requirement for GFNI

* ARMeilleure Add `X86GetGf2p8LogicalShiftLeft`

Used to generate GF(2^8) 8x8 bit-matrices for bit-shifting for the `gf2p8affineqb` instruction.

* ARMeilleure: Append `FeatureInfo7Ecx` to `FeatureInfo`
2022-10-02 11:17:19 +02:00
merry
f5235fff29
ARMeilleure: Hardware accelerate SHA256 (#3585)
* ARMeilleure/HardwareCapabilities: Add Sha

* ARMeilleure/Intrinsic: Add X86Sha256Rnds2

* ARmeilleure: Hardware accelerate SHA256H/SHA256H2

* ARMeilleure/Intrinsic: Add X86Sha256Msg1, X86Sha256Msg2

* ARMeilleure/Intrinsic: Add X86Palignr

* ARMeilleure: Hardware accelerate SHA256SU0, SHA256SU1

* PTC: Bump InternalVersion
2022-08-25 10:12:13 +00:00
Nicholas Rodine
7defc59b9d
A few minor documentation fixes. (#3599)
* A few minor documentation fixes.

* Removed more invalid inheritdoc instances.
2022-08-19 18:21:06 -03:00
Nicholas Rodine
951700fdd8
Removed unused usings. (#3593)
* Removed unused usings.

* Added back using, now that it's used.

* Removed extra whitespace.
2022-08-18 18:04:54 +02:00
merry
6dfb6ccf8c
PreAllocator: Check if instruction supports a Vex prefix in IsVexSameOperandDestSrc1 (#3587) 2022-08-14 17:35:08 -03:00
gdkchan
26a881176e
Fix tail merge from block with conditional jump to multiple returns (#3267)
* Fix tail merge from block with conditional jump to multiple returns

* PPTC version bump
2022-04-09 16:56:50 +02:00
gdkchan
c3c3914ed3
Add a limit on the number of uses a constant may have (#3097) 2022-02-09 17:42:47 -03:00
gdkchan
f3bfd799e1
Fix calls passing V128 values on Linux (#3034)
* Fix calls passing V128 values on Linux

* PPTC version bump
2022-01-24 11:23:24 +01:00
gdkchan
f0824fde9f
Add host CPU memory barriers for DMB/DSB and ordered load/store (#3015)
* Add host CPU memory barriers for DMB/DSB and ordered load/store

* PPTC version bump

* Revert to old barrier order
2022-01-21 12:47:34 -03:00
Mary
f39fce8f54
misc: Migrate usage of RuntimeInformation to OperatingSystem (#2901)
Very basic migration across the codebase.
2021-12-04 20:02:30 -03:00
FICTURE7
fbf40424f4
Add an early TailMerge pass (#2721)
* Add an early `TailMerge` pass

Some translations can have a lot of guest calls and since for each guest
call there is a call guard which may return. This can produce a lot of
epilogue code for returns. This pass merges the epilogue into a single
block.

```
Using filter 'hcq'.
Using metric 'code size'.

Total diff: -1648111 (-7.19 %) (bytes):
  Base: 22913847
  Diff: 21265736

Improved: 4567, regressed: 14, unchanged: 144
```

* Set PTC version

* Address feedback

* Handle `void` returning functions

* Actually handle `void` returning functions

* Fix `RegisterToLocal` logging
2021-10-18 19:51:22 -03:00
FICTURE7
69093cf2d6
Optimize LSRA (#2563)
* Optimize `TryAllocateRegWithtoutSpill` a bit

* Add a fast path for when all registers are live.
* Do not query `GetOverlapPosition` if the register is already in use
  (i.e: free position is 0).

* Do not allocate child split list if not parent

* Turn `LiveRange` into a reference struct

`LiveRange` is now a reference wrapping struct like `Operand` and
`Operation`.

It has also been changed into a singly linked-list. In micro-benchmarks
traversing the linked-list was faster than binary search on `List<T>`.
Even for quite large input sizes (e.g: 1,000,000), surprisingly.

Could be because the code gen for traversing the linked-list is much
much cleaner and there is no virtual dispatch happening when checking if
intervals overlaps.

* Turn `LiveInterval` into an iterator

The LSRA allocates in forward order and never inspect previous
`LiveInterval` once they are expired. Something similar can be done for
the `LiveRange`s within the `LiveInterval`s themselves.

The `LiveInterval` is turned into a iterator which expires `LiveRange`
within it. The iterator is moved forward along with interval walking
code, i.e: AllocateInterval(context, interval, cIndex).

* Remove `LinearScanAllocator.Sources`

Local methods are less susceptible to do allocations than lambdas.

* Optimize `GetOverlapPosition(interval)` a bit

Time complexity should be in O(n+m) instead of O(nm) now.

* Optimize `NumberLocals` a bit

Use the same idea as in `HybridAllocator` to store the visited state
in the MSB of the Operand's value instead of using a `HashSet<T>`.

* Optimize `InsertSplitCopies` a bit

Avoid allocating a redundant `CopyResolver`.

* Optimize `InsertSplitCopiesAtEdges` a bit

Avoid redundant allocations of `CopyResolver`.

* Use stack allocation for `freePositions`

Avoid redundant computations.

* Add `UseList`

Replace `SortedIntegerList` with an even more specialized data
structure. It allocates memory on the arena allocators and does not
require copying use positions when splitting it.

* Turn `LiveInterval` into a reference struct

`LiveInterval` is now a reference wrapping struct like `Operand` and
`Operation`.

The rationale behind turning this in a reference wrapping struct is
because a `LiveInterval` is associated with each local variable, and
these intervals may themselves be split further. I've seen translations
having up to 8000 local variables.

To make the `LiveInterval` unmanaged, a new data structure called
`LiveIntervalList` was added to store child splits. This differs from
`SortedList<,>` because it can contain intervals with the same start
position.

Really wished we got some more of C++ template in C#. :^(

* Optimize `GetChildSplit` a bit

No need to inspect the remaining ranges if we've reached a range which
starts after position, since the split list is ordered.

* Optimize `CopyResolver` a bit

Lazily allocate the fill, spill and parallel copy structures since most
of the time only one of them is needed.

* Optimize `BitMap.Enumerator` a bit

Marking `MoveNext` as `AggressiveInlining` allows RyuJIT to promote the
`Enumerator` struct into registers completely, reducing load/store code
a lot since it does not have to store the struct on the stack for ABI
purposes.

* Use stack allocation for `use/blockedPositions`

* Optimize `AllocateWithSpill` a bit

* Address feedback

* Make `LiveInterval.AddRange(,)` more conservative

Produces no diff against master, but just for good measure.
2021-10-08 18:15:44 -03:00
FICTURE7
ecc64c934d
Add Operand.Label support to Assembler (#2680)
* Add `Operand.Label` support to `Assembler`

This adds label support to `Assembler` and enables branch tightening
when compiling with relocatables. Jump management and patching has been
moved to the `Assembler`.

* Move instruction table to `Assembler.Table`

* Set PTC internal version

* Rename `Assembler.Table` to `AssemblerTable`
2021-10-05 14:04:55 -03:00
FICTURE7
312be74861
Optimize HybridAllocator (#2637)
* Store constant `Operand`s in the `LocalInfo`

Since the spill slot and register assigned is fixed, we can just store
the `Operand` reference in the `LocalInfo` struct. This allows skipping
hitting the intern-table for a look up.

* Skip `Uses`/`Assignments` management

Since the `HybridAllocator` is the last pass and we do not care about
uses/assignments we can skip managing that when setting destinations or
sources.

* Make `GetLocalInfo` inlineable

Also fix a possible issue where with numbered locals. See or-assignment
operator in `SetVisited(local)` before patch.

* Do not run `BlockPlacement` in LCQ

With the host mapped memory manager, there is a lot less cold code to
split from hot code. So disabling this in LCQ gives some extra
throughput - where we need it.

* Address Mou-Ikkai's feedback

* Apply suggestions from code review

Co-authored-by: VocalFan <45863583+Mou-Ikkai@users.noreply.github.com>

* Move check to an assert

Co-authored-by: VocalFan <45863583+Mou-Ikkai@users.noreply.github.com>
2021-09-29 01:38:37 +02:00
FICTURE7
a9343c9364
Refactor PtcInfo (#2625)
* Refactor `PtcInfo`

This change reduces the coupling of `PtcInfo` by moving relocation
tracking to the backend. `RelocEntry`s remains as `RelocEntry`s through
out the pipeline until it actually needs to be written to the PTC
streams. Keeping this representation makes inspecting and manipulating
relocations after compilations less painful. This is something I needed
to do to patch relocations to 0 to diff dumps.

Contributes to #1125.

* Turn `Symbol` & `RelocInfo` into readonly structs

* Add documentation to `CompiledFunction`

* Remove `Compiler.Compile<T>`

Remove `Compiler.Compile<T>` and replace it by `Map<T>` of the
`CompiledFunction` returned.
2021-09-14 01:23:37 +02:00
FICTURE7
f2a7b300c4
Fix type mismatch in BitwiseAnd simplification (#2571)
* Fix type mismatch in `BitwiseAnd` simplification

`TryEliminateBitwiseAnd` would turn the `BitwiseAnd` operation into a
copy of the wrong type. E.g:

Before `Simplification`:
```llvm
i64 %0 = BitwiseAnd i64 0x0, %1
```

After `Simplication`:
```llvm
i64 %0 = Copy i32 0x0
```

Since the with the changes in #2515, we iterate in reverse order and
`Simplication`, `ConstantFolding` does not indicate if it modified
the CFG, the second pass to "retype" the copy into the proper
destination type does not happen.

This also blocked copy propagation since its destination type did not
match with its source type. But in the cases I've seen, the
`PreAllocator` would insert a copy for the propagated constant, which
results in no diffs.

Since the copy remained as is, asserts are fired when generating it.

* Set PPTC version
2021-08-20 14:42:00 -03:00
FICTURE7
22b2cb39af
Reduce JIT GC allocations (#2515)
* Turn `MemoryOperand` into a struct

* Remove `IntrinsicOperation`

* Remove `PhiNode`

* Remove `Node`

* Turn `Operand` into a struct

* Turn `Operation` into a struct

* Clean up pool management methods

* Add `Arena` allocator

* Move `OperationHelper` to `Operation.Factory`

* Move `OperandHelper` to `Operand.Factory`

* Optimize `Operation` a bit

* Fix `Arena` initialization

* Rename `NativeList<T>` to `ArenaList<T>`

* Reduce `Operand` size from 88 to 56 bytes

* Reduce `Operation` size from 56 to 40 bytes

* Add optimistic interning of Register & Constant operands

* Optimize `RegisterUsage` pass a bit

* Optimize `RemoveUnusedNodes` pass a bit

Iterating in reverse-order allows killing dependency chains in a single
pass.

* Fix PPTC symbols

* Optimize `BasicBlock` a bit

Reduce allocations from `_successor` & `DominanceFrontiers`

* Fix `Operation` resize

* Make `Arena` expandable

Change the arena allocator to be expandable by allocating in pages, with
some of them being pooled. Currently 32 pages are pooled. An LRU removal
mechanism should probably be added to it.

Apparently MHR can allocate bitmaps large enough to exceed the 16MB
limit for the type.

* Move `Arena` & `ArenaList` to `Common`

* Remove `ThreadStaticPool` & co

* Add `PhiOperation`

* Reduce `Operand` size from 56 from 48 bytes

* Add linear-probing to `Operand` intern table

* Optimize `HybridAllocator` a bit

* Add `Allocators` class

* Tune `ArenaAllocator` sizes

* Add page removal mechanism to `ArenaAllocator`

Remove pages which have not been used for more than 5s after each reset.

I am on fence if this would be better using a Gen2 callback object like
the one in System.Buffers.ArrayPool<T>, to trim the pool. Because right
now if a large translation happens, the pages will be freed only after a
reset. This reset may not happen for a while because no new translation
is hit, but the arena base sizes are rather small.

* Fix `OOM` when allocating larger than page size in `ArenaAllocator`

Tweak resizing mechanism for Operand.Uses and Assignemnts.

* Optimize `Optimizer` a bit

* Optimize `Operand.Add<T>/Remove<T>` a bit

* Clean up `PreAllocator`

* Fix phi insertion order

Reduce codegen diffs.

* Fix code alignment

* Use new heuristics for degree of parallelism

* Suppress warnings

* Address gdkchan's feedback

Renamed `GetValue()` to `GetValueUnsafe()` to make it more clear that
`Operand.Value` should usually not be modified directly.

* Add fast path to `ArenaAllocator`

* Assembly for `ArenaAllocator.Allocate(ulong)`:

  .L0:
    mov rax, [rcx+0x18]
    lea r8, [rax+rdx]
    cmp r8, [rcx+0x10]
    ja short .L2
  .L1:
    mov rdx, [rcx+8]
    add rax, [rdx+8]
    mov [rcx+0x18], r8
    ret
  .L2:
    jmp ArenaAllocator.AllocateSlow(UInt64)

  A few variable/field had to be changed to ulong so that RyuJIT avoids
  emitting zero-extends.

* Implement a new heuristic to free pooled pages.

  If an arena is used often, it is more likely that its pages will be
  needed, so the pages are kept for longer (e.g: during PPTC rebuild or
  burst sof compilations). If is not used often, then it is more likely
  that its pages will not be needed (e.g: after PPTC rebuild or bursts
  of compilations).

* Address riperiperi's feedback

* Use `EqualityComparer<T>` in `IntrusiveList<T>`

Avoids a potential GC hole in `Equals(T, T)`.
2021-08-17 15:08:34 -03:00
FICTURE7
9d7627af64
Add multi-level function table (#2228)
* Add AddressTable<T>

* Use AddressTable<T> for dispatch

* Remove JumpTable & co.

* Add fallback for out of range addresses

* Add PPTC support

* Add documentation to `AddressTable<T>`

* Make AddressTable<T> configurable

* Fix table walk

* Fix IsMapped check

* Remove CountTableCapacity

* Add PPTC support for fast path

* Rename IsMapped to IsValid

* Remove stale comment

* Change format of address in exception message

* Add TranslatorStubs

* Split DispatchStub

Avoids recompilation of stubs during tests.

* Add hint for 64bit or 32bit

* Add documentation to `Symbol`

* Add documentation to `TranslatorStubs`

Make `TranslatorStubs` disposable as well.

* Add documentation to `SymbolType`

* Add `AddressTableEventSource` to monitor function table size

Add an EventSource which measures the amount of unmanaged bytes
allocated by AddressTable<T> instances.

 dotnet-counters monitor -n Ryujinx --counters ARMeilleure

* Add `AllowLcqInFunctionTable` optimization toggle

This is to reduce the impact this change has on the test duration.
Before everytime a test was ran, the FunctionTable would be initialized
and populated so that the newly compiled test would get registered to
it.

* Implement unmanaged dispatcher

Uses the DispatchStub to dispatch into the next translation, which
allows execution to stay in unmanaged for longer and skips a
ConcurrentDictionary look up when the target translation has been
registered to the FunctionTable.

* Remove redundant null check

* Tune levels of FunctionTable

Uses 5 levels instead of 4 and change unit of AddressTableEventSource
from KB to MB.

* Use 64-bit function table

Improves codegen for direct branches:

    mov qword [rax+0x408],0x10603560
 -  mov rcx,sub_10603560_OFFSET
 -  mov ecx,[rcx]
 -  mov ecx,ecx
 -  mov rdx,JIT_CACHE_BASE
 -  add rdx,rcx
 +  mov rcx,sub_10603560
 +  mov rdx,[rcx]
    mov rcx,rax

Improves codegen for dispatch stub:

    and rax,byte +0x1f
 -  mov eax,[rcx+rax*4]
 -  mov eax,eax
 -  mov rcx,JIT_CACHE_BASE
 -  lea rax,[rcx+rax]
 +  mov rax,[rcx+rax*8]
    mov rcx,rbx

* Remove `JitCacheSymbol` & `JitCache.Offset`

* Turn `Translator.Translate` into an instance method

We do not have to add more parameter to this method and related ones as
new structures are added & needed for translation.

* Add symbol only when PTC is enabled

Address LDj3SNuD's feedback

* Change `NativeContext.Running` to a 32-bit integer

* Fix PageTable symbol for host mapped
2021-05-29 18:06:28 -03:00
FICTURE7
c805542b29
Allow LocalVariable to be assigned more than once (#2288)
* Allow `LocalVariable` to be assigned more than once

This allows us to write flow controls like loops and if-elses with
LocalVariables participating in phi nodes.

* Add `GetLocalNumber` to operand
2021-05-17 01:54:53 +02:00
gdkchan
e318022b89
Fold constant offsets and group constant addresses (#2285)
* Fold constant offsets and group constant addresses

* PPTC version bump
2021-05-13 21:26:57 +02:00
gdkchan
d43a56726c
(CPU) Fix CRC32 instruction when constant values are used as input (#2183) 2021-04-07 23:43:08 +02:00
LDj3SNuD
bcbf240d2e
PPTC: Fix unwanted propagation of a relocatable constant in a specific case. (#1990)
* Fix unwanted propagation of a relocatable constant in a specific case.

* Ptc.InternalVersion = 1990

* Nit to retrigger the Checks.
2021-02-23 13:15:45 +01:00
LDj3SNuD
dc0adb533d
PPTC & Pool Enhancements. (#1968)
* PPTC & Pool Enhancements.

* Avoid buffer allocations in CodeGenContext.GetCode(). Avoid stream allocations in PTC.PtcInfo.

Refactoring/nits.

* Use XXHash128, for Ptc.Load & Ptc.Save, x10 faster than Md5.

* Why not a nice Span.

* Added a simple PtcFormatter library for deserialization/serialization, which does not require reflection, in use at PtcJumpTable and PtcProfiler; improves maintainability and simplicity/readability of affected code.

* Nits.

* Revert #1987.

* Revert "Revert #1987."

This reverts commit 998be765cf.
2021-02-22 03:23:48 +01:00
FICTURE7
1586880114
Turn Copy into Fill in HybridAllocator (#2010)
* Turn Copy into Fill in HybridAllocator

* Set PTC internal verison
2021-02-21 18:33:59 +01:00
sharmander
40797a1283
Optimization | Modify Add (Integer) Instruction to use LEA instead. (#1971)
* Optimization | Modify Add Instruction to use LEA instead.

Currently, the add instruction requires 4 registers to take place. By using LEA, we can effectively perform the same working using 3 registers, reducing memory spills and improving translation efficiency.

* Fix IsSameOperandDestSrc1 Check for Add

* Use LEA if Dest != SRC1

* Update IsSameOperandDestSrc1 to account for Cases where Dest and Src1 can be same for add

* Fix error in logic

* Typo

* Add paranthesis for clarity

* Compare registers as requested.

* Cleanup if statement, use same comparison method as generateCopy

* Make change as recommended by gdk

* Perform check only when Add calls are made

* use ensure sametype for lea, fix else

* Update comment

* Update version #
2021-02-08 10:49:46 +11:00
LDj3SNuD
c3e0c41da3
CPU (A64): Add Fmaxnmp & Fminnmp Scalar Inst.s, Fast & Slow Paths; with Tests. (#1894) 2021-01-20 09:12:33 +11:00
LDj3SNuD
8a33e884f8
Fix Vnmls_S fast path (F64: losing input d value). Fix Vnmla_S & Vnmls_S slow paths (using fused inst.s). Fix Vfma_V slow path not using StandardFPSCRValue(). (#1775)
* Fix Vnmls_S fast path (F64: losing input d value). Fix Vnmla_S & Vnmls_S slow paths (using fused inst.s).

Add Vfma_S & Vfms_S Fma fast paths.
Add Vfnma_S inst. with Fma/Sse fast paths and slow path.
Add Vfnms_S Sse fast path.

Add Tests for affected inst.s.

Nits.

* InternalVersion = 1775

* Nits.

* Fix Vfma_V slow path not using StandardFPSCRValue().

* Nit: Fix Vfma_V order.

* Add Vfms_V Sse fast path and slow path.

* Add Vfma_V and Vfms_V Test.
2020-12-17 20:43:41 +01:00
LDj3SNuD
b5c215111d
PPTC Follow-up. (#1712)
* Added support for offline invalidation, via PPTC, of low cq translations replaced by high cq translations; both on a single run and between runs.

Added invalidation of .cache files in the event of reuse on a different user operating system.

Added .info and .cache files invalidation in case of a failed stream decompression.

Nits.

* InternalVersion = 1712;

* Nits.

* Address comment.

* Get rid of BinaryFormatter.

Nits.

* Move Ptc.LoadTranslations().

Nits.

* Nits.

* Fixed corner cases (in case backup copies have to be used). Added save logs.

* Not core fixes.

* Complement to the previous commit. Added load logs. Removed BinaryFormatter leftovers.

* Add LoadTranslations log.

* Nits.

* Removed the search and management of LowCq overlapping functions.

* Final increment of .info and .cache flags.

* Nit.

* GetIndirectFunctionAddress(): Validate that writing actually takes place in dynamic table memory range (and not elsewhere).

* Fix Ptc.UpdateInfo() due to rebase.

* Nit for retrigger Checks.

* Nit for retrigger Checks.
2020-12-17 20:32:09 +01:00
sharmander
3332b29f01
CPU: Implement VFMA (Vector) (#1762)
* Implement VFMA.F64

* Simplify switch

* Simplify FMA Instructions into their own IntrinsicType.

* Remove whitespace

* Fix indentation

* Change tests for Vfnms -- disable inf / nan

* Move args up, not description ;)

* Implementation Complete.

All Tests Pass (Slow / Fast Path)

* Move location of function in assembler + test updates.

* Shift params upwards

* Remove unused function

* Update PTC version.

* Add comments / re-oreder opcode table.

* Remove whitespace

* Fix nit

* Fix nit.

* Fix whitespace

* Wrong opcode was used by a bad merge.

* Addressed rip's comments.
2020-12-15 00:01:52 -03:00
gdkchan
47ba81c661
Fix pre-allocator shift instruction copy on a specific case (#1752)
* Fix pre-allocator shift instruction copy on a specific case

* Fix to make shift use int rather than long for constants
2020-12-14 17:56:07 -03:00
sharmander
36f6bbf5b9
CPU: Implement VFNMA.F32 | F.64 (#1783)
* Implement VFNMA.F<32/64>

* Update PTC Version

* Update Implementation & Renames & Correct Order

* Fix alignment

* Update implementation to not trigger assert

* Actually use the intrinsic that makes sense :)
2020-12-07 21:04:01 -03:00
LDj3SNuD
567ea726e1
Add support for guest Fz (Fpcr) mode through host Ftz and Daz (Mxcsr) modes (fast paths). (#1630)
* Add support for guest Fz (Fpcr) mode through host Ftz and Daz (Mxcsr) modes (fast paths).

* Ptc.InternalVersion = 1630

* Nits.

* Address comments.

* Update Ptc.cs

* Address comment.
2020-12-07 10:37:07 +01:00
sharmander
b479a43939
CPU: Implement VFNMS.F32/64 (#1758)
* Add necessary methods / op-code

* Enable Support for FMA Instruction Set

* Add Intrinsics / Assembly Opcodes for VFMSUB231XX.

* Add X86 Instructions for VFMSUB231XX

* Implement VFNMS

* Implement VFNMS Tests

* Add special cases for FMA instructions.

* Update PPTC Version

* Remove unused Op

* Move Check into Assert / Cleanup

* Rename and cleanup

* Whitespace

* Whitespace / Rename

* Re-sort

* Address final requests

* Implement VFMA.F64

* Simplify switch

* Simplify FMA Instructions into their own IntrinsicType.

* Remove whitespace

* Fix indentation

* Change tests for Vfnms -- disable inf / nan

* Move args up, not description ;)

* Undo vfma

* Completely remove vfms code.,

* Fix order of instruction in assembler
2020-12-03 20:20:02 +01:00
LDj3SNuD
0679084f11
CPU (A64): Add FP16/FP32 fast paths (F16C Intrinsics) for Fcvt_S, Fcvtl_V & Fcvtn_V Instructions. Now HardwareCapabilities uses CpuId. (#1650)
* net5.0

* CPU (A64): Add FP16/FP32 fast paths (F16C Intrinsics) for Fcvt_S, Fcvtl_V & Fcvtn_V Instructions. Switch to .NET 5.0.

Nits.

Tests performed successfully in both debug and release mode (for all instructions involved).

* Address comment.

* Update appveyor.yml

* Revert "Update appveyor.yml"

This reverts commit 27cdd59e8b.

* Remove Assembler CpuId.

* Update appveyor.yml

* Address comment.
2020-11-18 19:35:54 +01:00
FICTURE7
64088f04e3
Fix LiveInterval.Split (#1660)
Before when splitting intervals, the end of the range would be included
in the split check, this can produce empty ranges in the child split.

This in turn can affect spilling decisions since the child split will
have a different start position and this empty range will get a register
and move to the active set for a brief moment.

For example:

  A = [153, 172[; [1899, 1916[; [1991, 2010[; [2397, 2414[; ...

  Split(A, 1916)

  A0 = [153, 172[; [1899, 1916[
  A1 = [1916, 1916[; [1991, 2010[; [2397, 2414[; ...
2020-11-04 23:09:45 -03:00
FICTURE7
f60033e0aa
Implement block placement (#1549)
* Implement block placement

Implement a simple pass which re-orders cold blocks at the end of the
list of blocks in the CFG.

* Set PPTC version

* Use Array.Resize

Address gdkchan's feedback
2020-09-19 20:00:24 -03:00