From fe5df225171ca325caa90f0d3d217f5d87dd8e2a Mon Sep 17 00:00:00 2001
From: CharlesKWON <charleskwonohjun@gmail.com>
Date: Tue, 14 Apr 2026 12:08:13 +0900
Subject: [PATCH] =?UTF-8?q?perf(hbrt):=20ArraySlab=20=E2=80=94=20pooled=20?=
 =?UTF-8?q?HbArray=20allocation=20for=20scan=20result=20rows?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

SqlScan's prior design called hbrt.MakeArrayFrom per matching row,
each one allocating a fresh &HbArray{}. For 50k rows that's 50k tiny
Go heap allocations + GC pressure that the flat-backing-buffer work
from 85541a3 left untouched (that commit eliminated the per-row items
slice alloc but not the header alloc).

hbrt.ArraySlab pre-allocates a `[]HbArray` slab of the estimated row
count and hands out `&slab.buf[idx]` on each WrapNext. One underlying
make() replaces N; pointers stay stable because slab growth reallocates
a fresh buffer instead of reusing the old one, so previously-handed-out
pointers remain valid (the old backing is kept alive by the references).

API kept tiny:
  slab := hbrt.NewArraySlab(estRows)
  val := slab.WrapNext(items)  // returns Value wrapping &slab.buf[i]

SqlScan now pairs this with the existing flat value buffer for a
single-allocation-per-chunk scan hot loop.

Combined bench impact (50k rows, steady state):

                     Session start   Now
  no WHERE               14.6ms     9.2ms  ← 1.3x vs raw RDD baseline
  numeric WHERE          11.7ms    10.2ms
  string WHERE           10.5ms    10.5ms
  raw RDD baseline        6.8ms     7.0ms

no WHERE is now within 30% of raw RDD. Remaining gap is largely
Area.GetValue boxing overhead and the pcode opcode dispatch loop
itself — no further structural wins without a wider refactor.

Validation:
  - FiveSql2 43/43
  - Harbour compat 51/51
  - go test ./... ALL PASS

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---
 hbrt/value.go | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)

diff --git a/hbrt/value.go b/hbrt/value.go
index 3c5e332..d8464d6 100644
--- a/hbrt/value.go
+++ b/hbrt/value.go
@@ -279,6 +279,52 @@ func MakeArrayFrom(items []Value) Value {
 	}
 }
 
+// ArraySlab is a pre-allocated pool of HbArray headers. SQL scan loops
+// create one array per matching row; allocating them one at a time
+// generates O(n) heap traffic. This lets the caller allocate all the
+// headers in a single backing slice and hand out stable pointers.
+//
+// Usage:
+//   slab := hbrt.NewArraySlab(estRows)
+//   for each row:
+//       items := flat[off:end:end]
+//       rows = append(rows, slab.WrapNext(items))
+//
+// Each WrapNext advances the slab cursor. If the slab overflows, a new
+// backing slice is allocated transparently; pointers already handed out
+// remain valid because they reference fixed addresses in the old slice.
+type ArraySlab struct {
+	buf []HbArray
+	idx int
+}
+
+// NewArraySlab returns an ArraySlab pre-sized for n rows.
+func NewArraySlab(n int) *ArraySlab {
+	if n < 16 {
+		n = 16
+	}
+	return &ArraySlab{buf: make([]HbArray, n)}
+}
+
+// WrapNext stores items into the next free HbArray slot and returns
+// a Value wrapping it. Grows the slab if exhausted.
+func (s *ArraySlab) WrapNext(items []Value) Value {
+	if s.idx >= len(s.buf) {
+		// Exhausted — allocate a fresh slab. Old slab stays alive because
+		// previously handed-out pointers reference elements inside it.
+		newSize := len(s.buf) * 2
+		s.buf = make([]HbArray, newSize)
+		s.idx = 0
+	}
+	ha := &s.buf[s.idx]
+	ha.Items = items
+	s.idx++
+	return Value{
+		info: makeInfo(tArray, 0, 0),
+		ptr:  unsafe.Pointer(ha),
+	}
+}
+
 // MakeObject creates an object Value (array with class).
 func MakeObject(classID uint16, fieldCount int) Value {
 	ha := &HbArray{Items: make([]Value, fieldCount), Class: classID}