Ryujinx/Ryujinx.Graphics.Shader/Decoders/Decoder.cs
riperiperi 7ff1f9aa12
End shader decoding when reaching a block that starts with an infinite loop (after BRX) (#2367)
* End shader decoding when reaching an infinite loop

The NV shader compiler puts these at the end of shaders.

* Update shader cache version
2021-06-15 02:09:59 +02:00

516 lines
18 KiB
C#

using Ryujinx.Graphics.Shader.Instructions;
using System;
using System.Collections.Generic;
using System.Linq;
using static Ryujinx.Graphics.Shader.IntermediateRepresentation.OperandHelper;
namespace Ryujinx.Graphics.Shader.Decoders
{
static class Decoder
{
public const ulong ShaderEndDelimiter = 0xe2400fffff87000f;
public static Block[][] Decode(IGpuAccessor gpuAccessor, ulong startAddress, out bool hasBindless)
{
hasBindless = false;
List<Block[]> funcs = new List<Block[]>();
Queue<ulong> funcQueue = new Queue<ulong>();
HashSet<ulong> funcVisited = new HashSet<ulong>();
void EnqueueFunction(ulong funcAddress)
{
if (funcVisited.Add(funcAddress))
{
funcQueue.Enqueue(funcAddress);
}
}
funcQueue.Enqueue(0);
while (funcQueue.TryDequeue(out ulong funcAddress))
{
List<Block> blocks = new List<Block>();
Queue<Block> workQueue = new Queue<Block>();
Dictionary<ulong, Block> visited = new Dictionary<ulong, Block>();
Block GetBlock(ulong blkAddress)
{
if (!visited.TryGetValue(blkAddress, out Block block))
{
block = new Block(blkAddress);
workQueue.Enqueue(block);
visited.Add(blkAddress, block);
}
return block;
}
GetBlock(funcAddress);
while (workQueue.TryDequeue(out Block currBlock))
{
// Check if the current block is inside another block.
if (BinarySearch(blocks, currBlock.Address, out int nBlkIndex))
{
Block nBlock = blocks[nBlkIndex];
if (nBlock.Address == currBlock.Address)
{
throw new InvalidOperationException("Found duplicate block address on the list.");
}
nBlock.Split(currBlock);
blocks.Insert(nBlkIndex + 1, currBlock);
continue;
}
// If we have a block after the current one, set the limit address.
ulong limitAddress = ulong.MaxValue;
if (nBlkIndex != blocks.Count)
{
Block nBlock = blocks[nBlkIndex];
int nextIndex = nBlkIndex + 1;
if (nBlock.Address < currBlock.Address && nextIndex < blocks.Count)
{
limitAddress = blocks[nextIndex].Address;
}
else if (nBlock.Address > currBlock.Address)
{
limitAddress = blocks[nBlkIndex].Address;
}
}
FillBlock(gpuAccessor, currBlock, limitAddress, startAddress, out bool blockHasBindless);
hasBindless |= blockHasBindless;
if (currBlock.OpCodes.Count != 0)
{
// We should have blocks for all possible branch targets,
// including those from SSY/PBK instructions.
foreach (OpCodePush pushOp in currBlock.PushOpCodes)
{
GetBlock(pushOp.GetAbsoluteAddress());
}
// Set child blocks. "Branch" is the block the branch instruction
// points to (when taken), "Next" is the block at the next address,
// executed when the branch is not taken. For Unconditional Branches
// or end of program, Next is null.
OpCode lastOp = currBlock.GetLastOp();
if (lastOp is OpCodeBranch opBr)
{
if (lastOp.Emitter == InstEmit.Cal)
{
EnqueueFunction(opBr.GetAbsoluteAddress());
}
else
{
currBlock.Branch = GetBlock(opBr.GetAbsoluteAddress());
}
}
else if (lastOp is OpCodeBranchIndir opBrIndir)
{
// An indirect branch could go anywhere, we don't know the target.
// Those instructions are usually used on a switch to jump table
// compiler optimization, and in those cases the possible targets
// seems to be always right after the BRX itself. We can assume
// that the possible targets are all the blocks in-between the
// instruction right after the BRX, and the common target that
// all the "cases" should eventually jump to, acting as the
// switch break.
Block firstTarget = GetBlock(currBlock.EndAddress);
firstTarget.BrIndir = opBrIndir;
opBrIndir.PossibleTargets.Add(firstTarget);
}
if (!IsUnconditionalBranch(lastOp))
{
currBlock.Next = GetBlock(currBlock.EndAddress);
}
}
// Insert the new block on the list (sorted by address).
if (blocks.Count != 0)
{
Block nBlock = blocks[nBlkIndex];
blocks.Insert(nBlkIndex + (nBlock.Address < currBlock.Address ? 1 : 0), currBlock);
}
else
{
blocks.Add(currBlock);
}
// Do we have a block after the current one?
if (currBlock.BrIndir != null && HasBlockAfter(gpuAccessor, currBlock, startAddress))
{
bool targetVisited = visited.ContainsKey(currBlock.EndAddress);
Block possibleTarget = GetBlock(currBlock.EndAddress);
currBlock.BrIndir.PossibleTargets.Add(possibleTarget);
if (!targetVisited)
{
possibleTarget.BrIndir = currBlock.BrIndir;
}
}
}
foreach (Block block in blocks.Where(x => x.PushOpCodes.Count != 0))
{
for (int pushOpIndex = 0; pushOpIndex < block.PushOpCodes.Count; pushOpIndex++)
{
PropagatePushOp(visited, block, pushOpIndex);
}
}
funcs.Add(blocks.ToArray());
}
return funcs.ToArray();
}
private static bool HasBlockAfter(IGpuAccessor gpuAccessor, Block currBlock, ulong startAdddress)
{
if (!gpuAccessor.MemoryMapped(startAdddress + currBlock.EndAddress) ||
!gpuAccessor.MemoryMapped(startAdddress + currBlock.EndAddress + 7))
{
return false;
}
ulong inst = gpuAccessor.MemoryRead<ulong>(startAdddress + currBlock.EndAddress);
return inst != 0UL && inst != ShaderEndDelimiter;
}
private static bool BinarySearch(List<Block> blocks, ulong address, out int index)
{
index = 0;
int left = 0;
int right = blocks.Count - 1;
while (left <= right)
{
int size = right - left;
int middle = left + (size >> 1);
Block block = blocks[middle];
index = middle;
if (address >= block.Address && address < block.EndAddress)
{
return true;
}
if (address < block.Address)
{
right = middle - 1;
}
else
{
left = middle + 1;
}
}
return false;
}
private static void FillBlock(
IGpuAccessor gpuAccessor,
Block block,
ulong limitAddress,
ulong startAddress,
out bool hasBindless)
{
ulong address = block.Address;
hasBindless = false;
do
{
if (address + 7 >= limitAddress)
{
break;
}
// Ignore scheduling instructions, which are written every 32 bytes.
if ((address & 0x1f) == 0)
{
address += 8;
continue;
}
ulong opAddress = address;
address += 8;
long opCode = gpuAccessor.MemoryRead<long>(startAddress + opAddress);
(InstEmitter emitter, OpCodeTable.MakeOp makeOp) = OpCodeTable.GetEmitter(opCode);
if (emitter == null)
{
// TODO: Warning, illegal encoding.
block.OpCodes.Add(new OpCode(null, opAddress, opCode));
continue;
}
if (makeOp == null)
{
throw new ArgumentNullException(nameof(makeOp));
}
OpCode op = makeOp(emitter, opAddress, opCode);
// We check these patterns to figure out the presence of bindless access
hasBindless |= (op is OpCodeImage image && image.IsBindless) ||
(op is OpCodeTxd txd && txd.IsBindless) ||
(op is OpCodeTld4B) ||
(emitter == InstEmit.TexB) ||
(emitter == InstEmit.TldB) ||
(emitter == InstEmit.TmmlB) ||
(emitter == InstEmit.TxqB);
block.OpCodes.Add(op);
}
while (!IsControlFlowChange(block.GetLastOp()));
block.EndAddress = address;
block.UpdatePushOps();
}
private static bool IsUnconditionalBranch(OpCode opCode)
{
return IsUnconditional(opCode) && IsControlFlowChange(opCode);
}
private static bool IsUnconditional(OpCode opCode)
{
if (opCode is OpCodeExit op && op.Condition != Condition.Always)
{
return false;
}
return opCode.Predicate.Index == RegisterConsts.PredicateTrueIndex && !opCode.InvertPredicate;
}
private static bool IsControlFlowChange(OpCode opCode)
{
return (opCode is OpCodeBranch opBranch && !opBranch.PushTarget) ||
opCode is OpCodeBranchIndir ||
opCode is OpCodeBranchPop ||
opCode is OpCodeExit;
}
private enum MergeType
{
Brk = 0,
Sync = 1
}
private struct PathBlockState
{
public Block Block { get; }
private enum RestoreType
{
None,
PopPushOp,
PushBranchOp
}
private RestoreType _restoreType;
private ulong _restoreValue;
private MergeType _restoreMergeType;
public bool ReturningFromVisit => _restoreType != RestoreType.None;
public PathBlockState(Block block)
{
Block = block;
_restoreType = RestoreType.None;
_restoreValue = 0;
_restoreMergeType = default;
}
public PathBlockState(int oldStackSize)
{
Block = null;
_restoreType = RestoreType.PopPushOp;
_restoreValue = (ulong)oldStackSize;
_restoreMergeType = default;
}
public PathBlockState(ulong syncAddress, MergeType mergeType)
{
Block = null;
_restoreType = RestoreType.PushBranchOp;
_restoreValue = syncAddress;
_restoreMergeType = mergeType;
}
public void RestoreStackState(Stack<(ulong, MergeType)> branchStack)
{
if (_restoreType == RestoreType.PushBranchOp)
{
branchStack.Push((_restoreValue, _restoreMergeType));
}
else if (_restoreType == RestoreType.PopPushOp)
{
while (branchStack.Count > (uint)_restoreValue)
{
branchStack.Pop();
}
}
}
}
private static void PropagatePushOp(Dictionary<ulong, Block> blocks, Block currBlock, int pushOpIndex)
{
OpCodePush pushOp = currBlock.PushOpCodes[pushOpIndex];
Stack<PathBlockState> workQueue = new Stack<PathBlockState>();
HashSet<Block> visited = new HashSet<Block>();
Stack<(ulong, MergeType)> branchStack = new Stack<(ulong, MergeType)>();
void Push(PathBlockState pbs)
{
// When block is null, this means we are pushing a restore operation.
// Restore operations are used to undo the work done inside a block
// when we return from it, for example it pops addresses pushed by
// SSY/PBK instructions inside the block, and pushes addresses poped
// by SYNC/BRK.
// For blocks, if it's already visited, we just ignore to avoid going
// around in circles and getting stuck here.
if (pbs.Block == null || !visited.Contains(pbs.Block))
{
workQueue.Push(pbs);
}
}
Push(new PathBlockState(currBlock));
while (workQueue.TryPop(out PathBlockState pbs))
{
if (pbs.ReturningFromVisit)
{
pbs.RestoreStackState(branchStack);
continue;
}
Block current = pbs.Block;
// If the block was already processed, we just ignore it, otherwise
// we would push the same child blocks of an already processed block,
// and go around in circles until memory is exhausted.
if (!visited.Add(current))
{
continue;
}
int pushOpsCount = current.PushOpCodes.Count;
if (pushOpsCount != 0)
{
Push(new PathBlockState(branchStack.Count));
for (int index = pushOpIndex; index < pushOpsCount; index++)
{
OpCodePush currentPushOp = current.PushOpCodes[index];
MergeType pushMergeType = currentPushOp.Emitter == InstEmit.Ssy ? MergeType.Sync : MergeType.Brk;
branchStack.Push((currentPushOp.GetAbsoluteAddress(), pushMergeType));
}
}
pushOpIndex = 0;
if (current.Next != null)
{
Push(new PathBlockState(current.Next));
}
if (current.Branch != null)
{
Push(new PathBlockState(current.Branch));
}
else if (current.GetLastOp() is OpCodeBranchIndir brIndir)
{
// By adding them in descending order (sorted by address), we process the blocks
// in order (of ascending address), since we work with a LIFO.
foreach (Block possibleTarget in brIndir.PossibleTargets.OrderByDescending(x => x.Address))
{
Push(new PathBlockState(possibleTarget));
}
}
else if (current.GetLastOp() is OpCodeBranchPop op)
{
MergeType popMergeType = op.Emitter == InstEmit.Sync ? MergeType.Sync : MergeType.Brk;
bool found = true;
ulong targetAddress = 0UL;
MergeType mergeType;
do
{
if (branchStack.Count == 0)
{
found = false;
break;
}
(targetAddress, mergeType) = branchStack.Pop();
// Push the target address (this will be used to push the address
// back into the SSY/PBK stack when we return from that block),
Push(new PathBlockState(targetAddress, mergeType));
}
while (mergeType != popMergeType);
// Make sure we found the correct address,
// the push and pop instruction types must match, so:
// - BRK can only consume addresses pushed by PBK.
// - SYNC can only consume addresses pushed by SSY.
if (found)
{
if (branchStack.Count == 0)
{
// If the entire stack was consumed, then the current pop instruction
// just consumed the address from out push instruction.
op.Targets.Add(pushOp, op.Targets.Count);
pushOp.PopOps.TryAdd(op, Local());
}
else
{
// Push the block itself into the work "queue" (well, it's a stack)
// for processing.
Push(new PathBlockState(blocks[targetAddress]));
}
}
}
}
}
}
}