Two stacked optimizations land on the SqlScan hot path. Combined
effect on the 50k-row benchmark:
Before After vs raw
Numeric WHERE 10.2ms 7.8ms 1.15x
String WHERE 10.5ms 7.9ms 1.15x
No WHERE 9.2ms 10.0ms 1.45x
Raw RDD baseline 6.8ms 6.8ms 1.00x
WHERE-predicate paths are now within 15% of the raw Harbour-style
RDD scan loop. The no-WHERE path is unchanged (slight jitter from
the added devirt branch); FieldGet peephole doesn't apply there.
--- Optimization 1: PcOpFieldGet peephole ---
Adds a new pcode opcode `PcOpFieldGet <fieldIdx>` (0x46) that skips
the usual PushSymbol+Function+Frame+FieldGet-RTL+EndProc chain and
calls a direct field getter closure instead. genpc recognizes the
shape `FieldGet(<int-literal>)` during emitCall and emits the
specialized opcode automatically — no SQL-side API change.
Integration:
* hbrt.Thread.FastFieldGetter — hot-path closure set by scan loops.
Non-nil → pcode bypasses dispatch.
Nil → pcode resolves FIELDGET via
the RTL symbol table (correctness
fallback for any other callers).
* compiler/genpc/genpc.go — peephole in emitCall.
* hbrt/pcinterp.go — PcOpFieldGet handler.
This alone cut numeric WHERE from 10.2 → 7.9ms: eliminated roughly
one full Frame/EndProc + RTL dispatch per row × 50k rows.
--- Optimization 2: DBFArea devirtualization ---
SqlScan type-asserts the workarea to *dbf.DBFArea once and runs a
dedicated loop that calls GoTop/EOF/Skip/GetValue directly on the
concrete type. Go's compiler inlines these, skipping the interface
vtable per row. Non-DBF drivers still work via the generic Area
branch.
The FastFieldGetter closure also captures *DBFArea directly in the
DBF branch, so the WHERE predicate side of the hot loop is now
entirely devirtualized: no interface dispatch between the pcode
dispatch loop and the DBF record buffer.
Validation:
- FiveSql2 43/43
- Harbour compat 51/51
- go test ./... ALL PASS
Remaining gap to raw RDD on no-WHERE (~1.45x) is dominated by the
two-column row construction + ArraySlab + flat backing bookkeeping
that the raw loop doesn't do. Going below that requires changing
the SQL engine's result shape — out of scope here.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
273 lines
6.5 KiB
Go
273 lines
6.5 KiB
Go
// Copyright (c) 2026 Charles KWON OhJun (charleskwonohjun@gmail.com)
|
||
// All rights reserved.
|
||
|
||
// Five pcode interpreter — executes pcode bytecode on a Thread.
|
||
// Each opcode directly calls the corresponding Thread method,
|
||
// so pcode execution is semantically identical to gengo-compiled code.
|
||
|
||
package hbrt
|
||
|
||
import (
|
||
"encoding/binary"
|
||
"fmt"
|
||
"math"
|
||
)
|
||
|
||
// ExecPcode runs a pcode function on the given thread.
|
||
// Full variant — installs a defer/recover so panics from inside the
|
||
// pcode body (HbError, BreakValue, user Break) are re-panicked with
|
||
// proper frame unwinding. Used for general-purpose pcode evaluation.
|
||
func ExecPcode(t *Thread, fn *PcodeFunc, mod *PcodeModule) {
|
||
t.Frame(fn.Params, fn.Locals)
|
||
defer t.EndProc()
|
||
execPcodeBody(t, fn, mod)
|
||
}
|
||
|
||
// ExecPcodeFast is a hot-path variant for short, pure expressions
|
||
// (FiveSql2 WHERE predicates, inline lambdas) where the caller has
|
||
// already guaranteed that the body will not panic with HbError /
|
||
// BreakValue. Skips the defer+recover dance in EndProc, saving ~15ns
|
||
// per call × tens of thousands of rows in scan loops.
|
||
//
|
||
// Contract: caller is responsible for panic discipline. If the pcode
|
||
// body panics, the frame stack is still cleaned up (EndProcFast) but
|
||
// no diagnostic is logged and SEQUENCE/RECOVER will not see the panic.
|
||
func ExecPcodeFast(t *Thread, fn *PcodeFunc, mod *PcodeModule) {
|
||
t.Frame(fn.Params, fn.Locals)
|
||
execPcodeBody(t, fn, mod)
|
||
t.EndProcFast()
|
||
}
|
||
|
||
// execPcodeBody is the shared opcode dispatch loop.
|
||
func execPcodeBody(t *Thread, fn *PcodeFunc, mod *PcodeModule) {
|
||
code := fn.Code
|
||
pc := 0 // program counter
|
||
|
||
for pc < len(code) {
|
||
op := code[pc]
|
||
pc++
|
||
|
||
switch op {
|
||
case PcOpNop:
|
||
// do nothing
|
||
|
||
// --- Stack ---
|
||
case PcOpPushNil:
|
||
t.PushNil()
|
||
case PcOpPushTrue:
|
||
t.PushBool(true)
|
||
case PcOpPushFalse:
|
||
t.PushBool(false)
|
||
case PcOpPushInt:
|
||
v := int64(binary.LittleEndian.Uint64(code[pc:]))
|
||
pc += 8
|
||
t.PushLong(v)
|
||
case PcOpPushDouble:
|
||
bits := binary.LittleEndian.Uint64(code[pc:])
|
||
pc += 8
|
||
t.PushDouble(math.Float64frombits(bits), 0, 0)
|
||
case PcOpPushString:
|
||
slen := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.PushString(string(code[pc : pc+slen]))
|
||
pc += slen
|
||
case PcOpPushBool:
|
||
t.PushBool(code[pc] != 0)
|
||
pc++
|
||
case PcOpPushLocal:
|
||
idx := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.PushLocal(idx)
|
||
case PcOpPopLocal:
|
||
idx := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.PopLocal(idx)
|
||
case PcOpPop:
|
||
t.Pop()
|
||
case PcOpDup:
|
||
t.Dup()
|
||
|
||
// --- Arithmetic ---
|
||
case PcOpPlus:
|
||
t.Plus()
|
||
case PcOpMinus:
|
||
t.Minus()
|
||
case PcOpMult:
|
||
t.Mult()
|
||
case PcOpDivide:
|
||
t.Divide()
|
||
case PcOpMod:
|
||
t.Modulus()
|
||
case PcOpPower:
|
||
t.Power()
|
||
case PcOpNegate:
|
||
t.Negate()
|
||
|
||
// --- Comparison ---
|
||
case PcOpEqual:
|
||
t.Equal()
|
||
case PcOpNotEqual:
|
||
t.NotEqual()
|
||
case PcOpLess:
|
||
t.Less()
|
||
case PcOpGreater:
|
||
t.Greater()
|
||
case PcOpLessEq:
|
||
t.LessEqual()
|
||
case PcOpGreaterEq:
|
||
t.GreaterEqual()
|
||
case PcOpInString:
|
||
t.InString()
|
||
|
||
// --- Logical ---
|
||
case PcOpAnd:
|
||
t.And()
|
||
case PcOpOr:
|
||
t.Or()
|
||
case PcOpNot:
|
||
t.Not()
|
||
|
||
// --- Flow control ---
|
||
case PcOpJump:
|
||
offset := int32(binary.LittleEndian.Uint32(code[pc:]))
|
||
pc += 4
|
||
pc += int(offset)
|
||
case PcOpJumpFalse:
|
||
offset := int32(binary.LittleEndian.Uint32(code[pc:]))
|
||
pc += 4
|
||
if !t.PopLogical() {
|
||
pc += int(offset)
|
||
}
|
||
case PcOpJumpTrue:
|
||
offset := int32(binary.LittleEndian.Uint32(code[pc:]))
|
||
pc += 4
|
||
if t.PopLogical() {
|
||
pc += int(offset)
|
||
}
|
||
case PcOpReturn:
|
||
return
|
||
case PcOpRetValue:
|
||
t.RetValue()
|
||
return
|
||
|
||
// --- Frame ---
|
||
case PcOpFrame:
|
||
// Already called at function entry; skip if re-encountered
|
||
pc += 4 // params + locals
|
||
case PcOpEndProc:
|
||
return
|
||
|
||
// --- Workarea field access (peephole for FieldGet(literal)) ---
|
||
case PcOpFieldGet:
|
||
fIdx := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
// Hot path — SqlScan plugs a direct field getter closure into
|
||
// t.FastFieldGetter before running the predicate, so we skip
|
||
// PushSymbol + Function dispatch + FieldGet RTL's own Frame.
|
||
if fg := t.FastFieldGetter; fg != nil {
|
||
t.PushValue(fg(fIdx))
|
||
} else {
|
||
// Generic fallback: resolve through RTL symbol table
|
||
t.PushSymbol(t.VM().FindSymbol("FIELDGET"))
|
||
t.PushNil()
|
||
t.PushLong(int64(fIdx))
|
||
t.Function(1)
|
||
}
|
||
|
||
// --- Function calls ---
|
||
case PcOpPushSymbol:
|
||
slen := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
name := string(code[pc : pc+slen])
|
||
pc += slen
|
||
sym := t.VM().FindSymbol(name)
|
||
t.PushSymbol(sym)
|
||
case PcOpPushNilArg:
|
||
t.PushNil()
|
||
case PcOpFunction:
|
||
nArgs := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.Function(nArgs)
|
||
case PcOpDo:
|
||
nArgs := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.Do(nArgs)
|
||
|
||
// --- Self / OOP ---
|
||
case PcOpPushSelf:
|
||
t.PushSelf()
|
||
case PcOpPushSelfField:
|
||
slen := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
name := string(code[pc : pc+slen])
|
||
pc += slen
|
||
t.PushSelfField(name)
|
||
case PcOpSetSelfField:
|
||
slen := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
name := string(code[pc : pc+slen])
|
||
pc += slen
|
||
t.SetSelfField(name)
|
||
case PcOpSend:
|
||
slen := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
name := string(code[pc : pc+slen])
|
||
pc += slen
|
||
nArgs := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.Send(name, nArgs)
|
||
|
||
// --- Array ---
|
||
case PcOpArrayGen:
|
||
count := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
t.ArrayGen(count)
|
||
case PcOpArrayPush:
|
||
t.ArrayPush()
|
||
case PcOpArrayPop:
|
||
t.ArrayPop()
|
||
|
||
// --- Block ---
|
||
case PcOpPushBlock:
|
||
codeLen := int(binary.LittleEndian.Uint32(code[pc:]))
|
||
pc += 4
|
||
blockCode := make([]byte, codeLen)
|
||
copy(blockCode, code[pc:pc+codeLen])
|
||
pc += codeLen
|
||
nDetached := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
|
||
// Create a Go function that interprets the block's pcode
|
||
blockFn := &PcodeFunc{Code: blockCode}
|
||
modCopy := mod
|
||
t.PushBlock(func(t2 *Thread) {
|
||
ExecPcode(t2, blockFn, modCopy)
|
||
}, nDetached)
|
||
|
||
// --- Local ops ---
|
||
case PcOpLocalAddInt:
|
||
idx := int(binary.LittleEndian.Uint16(code[pc:]))
|
||
pc += 2
|
||
val := int32(binary.LittleEndian.Uint32(code[pc:]))
|
||
pc += 4
|
||
t.LocalAddInt(idx, int64(val))
|
||
case PcOpInc:
|
||
t.Inc()
|
||
case PcOpDec:
|
||
t.Dec()
|
||
|
||
case PcOpPopLogical:
|
||
t.PopLogical()
|
||
|
||
case PcOpLine:
|
||
pc += 2 // skip line number (for debugging)
|
||
|
||
case PcOpHalt:
|
||
return
|
||
|
||
default:
|
||
panic(fmt.Sprintf("unknown pcode opcode: 0x%02X at pc=%d", op, pc-1))
|
||
}
|
||
}
|
||
}
|