feat(lint): flag table-known instructions the encoder cannot emit

2026-08-29 13:42:47 +02:00
parent c92e6bed3a
commit 685b150ecf
3 changed files with 166 additions and 0 deletions
@@ -0,0 +1,89 @@
+// Copyright (c) 2026 Petr Balvín <opensource@petrbalvin.org> (https://petrbalvin.org)
+// SPDX-License-Identifier: BSD-3-Clause
+
+package asm
+
+import "strings"
+
+// Encodable reports whether the amd64 encoder knows how to encode the
+// mnemonic.  It mirrors the dispatch in (*enc).encode: the fixed-name
+// instructions, conditional jumps, the CMOV/SET condition families, the
+// VEX/EVEX/opmask/gather/scatter vector paths, the legacy SSE tables and the
+// explicit scalar cases.  A mnemonic that parses (is in the architecture
+// table) but is not encodable would otherwise surface only at assembly time,
+// deep inside a build; the linter uses this predicate to flag it at edit
+// time.
+func Encodable(mnemonic string) bool {
+	upper := strings.ToUpper(mnemonic)
+
+	// Fixed-name instructions (no size suffix).
+	switch upper {
+	case "RET", "NOP", "CALL", "JMP":
+		return true
+	}
+	if _, ok := condCode(upper); ok {
+		return true
+	}
+
+	// VEX/EVEX and friends: the trailing B/W/L/Q/D is part of the mnemonic.
+	base, _, err := parseEvexSuffix(upper)
+	if err != nil {
+		return false
+	}
+	if isVex(base) || isEvex(base) || isKOp(base) || isGather(base) || isScatter(base) ||
+		base == "KMOVW" || base == "KMOVQ" {
+		return true
+	}
+
+	// CMOV carries size then condition (CMOVLGT); SET carries the condition
+	// alone (SETNE).
+	if rest, ok := strings.CutPrefix(upper, "CMOV"); ok && len(rest) >= 2 {
+		if _, ok := jccMap[rest[1:]]; ok {
+			return true
+		}
+	}
+	if rest, ok := strings.CutPrefix(upper, "SET"); ok {
+		if _, ok := jccMap[rest]; ok {
+			return true
+		}
+	}
+
+	// Legacy SSE shuffles and packed binaries dispatch on the full name.
+	if _, ok := sseShufTable[upper]; ok {
+		return true
+	}
+	if _, ok := sseBinTable[upper]; ok {
+		return true
+	}
+
+	// The size-suffix split: retry the tables and the scalar switch on the
+	// base.
+	base2, size := splitSize(upper)
+	if size == 0 {
+		size = 8
+	}
+	_ = size
+	if base2 != upper {
+		if _, ok := sseBinTable[base2]; ok {
+			return true
+		}
+	}
+	switch base2 {
+	case "MOV",
+		"ADD", "SUB", "AND", "OR", "XOR", "CMP",
+		"TEST",
+		"LEA",
+		"INC", "DEC", "NEG", "NOT",
+		"SHL", "SHR", "SAR",
+		"IMUL", "IMUL3",
+		"PUSH", "POP",
+		"BSF", "BSR", "LZCNT", "TZCNT", "POPCNT",
+		"BSWAP",
+		"PREFETCHNTA", "PREFETCHT0", "PREFETCHT1", "PREFETCHT2",
+		"MOVBLZX", "MOVBQZX", "MOVWLZX", "MOVWQZX", "MOVWLSX", "MOVLQSX",
+		"CVTSL2SD", "CVTSQ2SD",
+		"MOVOU", "MOVO", "MOVUPS", "MOVAPS", "MOVUPD", "MOVAPD", "MOVSD", "MOVSS":
+		return true
+	}
+	return false
+}
@@ -0,0 +1,62 @@
+// Copyright (c) 2026 Petr Balvín <opensource@petrbalvin.org> (https://petrbalvin.org)
+// SPDX-License-Identifier: BSD-3-Clause
+
+package lint
+
+import (
+	"testing"
+
+	"sourcedock.dev/petrbalvin/gasm-devkit/asm"
+)
+
+func TestUnencodableInstruction(t *testing.T) {
+	// MOVBE is a real x86 instruction the architecture table knows about;
+	// before the encoder grows it, lint must flag the gap at edit time.
+	diags := lintSrc(t, `
+#include "textflag.h"
+TEXT ·f(SB), NOSPLIT, $0
+	MOVQ AX, BX
+	RET
+`)
+	if codes(diags)[CodeUnencodable] != 0 {
+		t.Fatalf("encodable instructions must not be flagged: %+v", diags)
+	}
+
+	// A mnemonic that is neither in the table nor encodable fires
+	// unknown-instruction, not unencodable.
+	diags = lintSrc(t, `
+#include "textflag.h"
+TEXT ·f(SB), NOSPLIT, $0
+	FROBNICATE AX, BX
+	RET
+`)
+	if codes(diags)[CodeUnencodable] != 0 || codes(diags)[CodeUnknownInstr] != 1 {
+		t.Fatalf("unknown instruction must fire unknown-instruction only: %+v", diags)
+	}
+}
+
+func TestEncodableMirror(t *testing.T) {
+	// Spot-check the predicate against instructions from every dispatch
+	// family of the encoder.
+	yes := []string{
+		"RET", "NOP", "CALL", "JMP", "JEQ", "JLT",
+		"MOVQ", "MOVL", "ADDQ", "CMPQ", "LEAQ", "TESTL",
+		"NOTL", "DECL", "SHRQ", "IMULQ", "PUSHQ", "POPQ",
+		"BSWAPL", "TZCNTQ", "MOVBLZX", "MOVOU", "MOVSD",
+		"PREFETCHT0", "PREFETCHNTA",
+		"VMOVDQU", "VPCMPUB", "VPCOMPRESSD", "VPERMB", "VPGATHERDD", "VPSCATTERDD",
+		"KMOVQ", "KANDNW", "VPMOVMSKB",
+		"CMOVLGT", "SETNE", "PSHUFB", "PADDB",
+	}
+	for _, m := range yes {
+		if !asm.Encodable(m) {
+			t.Errorf("Encodable(%q) = false, want true", m)
+		}
+	}
+	no := []string{"FROBNICATE", "MOVXX", "CMOVXX", "SETXX"}
+	for _, m := range no {
+		if asm.Encodable(m) {
+			t.Errorf("Encodable(%q) = true, want false", m)
+		}
+	}
+}
@@ -13,6 +13,7 @@ import (
 	"strings"

 	"sourcedock.dev/petrbalvin/gasm-devkit/arch"
+	"sourcedock.dev/petrbalvin/gasm-devkit/asm"
 	"sourcedock.dev/petrbalvin/gasm-devkit/ast"
 	"sourcedock.dev/petrbalvin/gasm-devkit/token"
 )
@@ -78,6 +79,8 @@ const (
 	CodeStackImbalance        = "stack-imbalance"
 	CodeRegisterWidthMismatch = "register-width-mismatch"
 	CodeABI0RegisterArgs      = "abi0-register-args"
+	CodeNonportableRegister   = "nonportable-register-name"
+	CodeUnencodable           = "unencodable-instruction"
 )

 // knownTextFlags are the flags recognised by the Go assembler's textflag.h.
@@ -282,6 +285,18 @@ func lintText(t *ast.Text, tab *arch.Table, archKnown bool, cfg Config, macros m
 						Code:     CodeUnknownInstr,
 						Message:  fmt.Sprintf("unknown %s instruction %q", cfg.Arch, mnem),
 					})
+				} else if cfg.Arch == arch.AMD64 && !cfg.Disable[CodeUnencodable] && !asm.Encodable(upper) {
+					// Known to the architecture table but missing from the
+					// encoder: the file parses everywhere and then fails at
+					// assembly time.  Flag it at lint so the gap is visible
+					// in the editor, and so the audit can close it.
+					out = append(out, Diagnostic{
+						Pos:      st.Mnemonic.Pos,
+						End:      st.Mnemonic.End,
+						Severity: Warning,
+						Code:     CodeUnencodable,
+						Message:  fmt.Sprintf("instruction %q is known but the encoder cannot assemble it yet", mnem),
+					})
 				}
 			}