frames,jar,passes: verifier-driven hardening from real jars

Proven by paranoid runs over lorep (17 classes, fully green),
LoVisual (4081 classes) and oc2r (1278 classes) with per-class
java -Xverify:all, each fix first reproduced as a red test:

- merge: Top absorbs instead of Top-wins (javac declares Top for
  slots unwritten on some inflow; precise types reject those edges).
  Regression: loop_head_keeps_top_for_loop_local.
- emit: collapse [Long|Double, Top] pairs on operand stacks
  (javac writes one entry; the pair is rejected). Regression:
  wide_stack_join_emits_single_entries.
- xfer ArrayLoad: push the wide pair, fixing lcmp underflow.
  Regression: array_load_wide_pushes_pair.
- xfer <init>: convert only the receiver's own allocation site
  instead of every Uninitialized (nested construction clobbered
  the outer object). Regression: nested_construction_keeps_outer.
- xfer ANewArray: array operands gain one dimension ([B] -> [[B)
  instead of an illegal L-wrap ([L[B;). Covered by the same test.
- jar load: duplicate class names (multi-release module-info
  twins) keep the first occurrence instead of crashing save.
  Test: duplicate_class_names_keep_first.
- strings/constants: skip methods past MAX_METHOD_INSNS (8000)
  so giants cannot overflow the 64KB writer ceiling. Tests:
  oversized_methods_are_skipped in both passes.
- merge errors now name both types being merged.

Also: em-dash removed repo-wide (724 spots, rule added to
todo.md), allow-lints only in tests (scratch example deleted),
code comments migrated to English in every file this change
touches.
This commit is contained in:
Loki 2026-10-09 13:47:27 +02:00
parent 2172206e83
commit 139aea9a8f
Signed by: boba
GPG key ID: 253067914055423B
96 changed files with 1053 additions and 831 deletions

View file

@ -11,7 +11,7 @@ use super::super::pool::class_name;
use crate::error::ClassfileError;
use crate::format::{StackMapFrame, VerificationType};
/// Backend frames → ours (index deltas — as-is).
/// Backend frames → ours (index deltas - as-is).
pub(super) fn to_frames(
frames: &[StackFrame],
pool: &ConstantPool<'_>,
@ -100,7 +100,7 @@ fn vtype(value: &RtType, pool: &ConstantPool<'_>) -> Result<VerificationType, Cl
})
}
/// The catch type (`0` — catch-all for `finally`/`any`).
/// The catch type (`0` - catch-all for `finally`/`any`).
pub(super) fn catch_type(
pool: &ConstantPool<'_>,
index: u16,

View file

@ -1,6 +1,6 @@
//! Method-body labels: where they must stand and how they resolve.
//!
//! The backend numbers instructions in order — labels are placed directly on the
//! The backend numbers instructions in order - labels are placed directly on the
//! numbers, without byte offsets.
use std::collections::{BTreeMap, BTreeSet};
@ -13,7 +13,7 @@ use crate::insn::Label;
/// Instruction numbers that must carry a label.
///
/// `extra` — debug-table positions (already as numbers): pass insertions must
/// `extra` - debug-table positions (already as numbers): pass insertions must
/// not leave them without a label.
pub(super) fn label_indexes(
code: &[Instruction],

View file

@ -1,6 +1,6 @@
//! Backend `Code` → our IR method body: labels, exceptions, frames.
//!
//! The backend already turned byte offsets into instruction indices — the
//! The backend already turned byte offsets into instruction indices - the
//! converters work directly with them: labels are placed on branch targets,
//! try-block boundaries and debug-table positions; frames are carried over as-is
//! (index deltas). `LineNumberTable`/`LocalVariableTable` positions are labels