Drew T
5.3 KiB
§264 — FOUR TELLS-LANE C RECIPES, EACH DRIVEN TO MATCH (P31 S59)
The extend-tell / swaprepeat-tell / s16-div-tell population banks well below the default lane,
and the measured cause is not the models: the card names a lever that appears nowhere in this
cookbook (see the LANE ALIASES block at §172b) and the drafts stack 3-5 idioms. These four were
proven this session by taking real failed wave drafts to MATCH; full derivations, transcripts and
probe files in docs/sunset/tool-designs/tells-lane-s59.md and .run/s59_tells/.
1. The (s16) promotion belongs INLINE in the call argument (func_8017D8E8, 33/33). Spelled
inline, its sra lands in the jal delay slot; hoisting it into a temp kills that fill. Three
companions in the same function: hold the table base in a source pointer local so it owns its
$sN across calls (gcc-2.7.2 has no cross-bb CSE, L2425); index with p[i] so (field<<3)+base
stays ONE association unit (writing base + f8 + i2 lets fold re-associate the constant out); read
an s16 field directly in the expression for a plain lh. The failing draft had pinned EIGHT
registers to transcribe the asm 1:1 and plateaued at near-2 over 23 oracle calls — pins are the
last resort (§17), not the opening move; they block the parm-copy and arg-copy shapes.
2. NEW LAW — a save-order-only residual is a bb0 anti-dependence (func_801808AC, 67/67).
Symptom: your prologue emits sw $s1 before sw $ra where the target has $ra first, and nothing
else differs. Saves are emitted descending-regno and woven at sched2, so an anti-dependence from a
basic-block-zero write of that $sN promotes its store. Cure: move the first write of that
variable out of bb0 and into a branch arm — i = 0; if (…) goto found; i = 1; becomes
if (…) i = 0; else if (…) i = 1; … else return;. reorg then re-creates the target's
addu $s1,$zero in the beq delay slot by thread fill.
3. NEW LAW — an opaque bound must live in a LOCAL, assigned LATE (func_80180E58, 34/34).
if (v < 0x10000) and if (v <= 0xFFFF) both canonicalize to li 0xFFFF; slt(0xFFFF,v); bnez.
The target's lui 0x10000; slt(v,K); beqz — plus the arm's store REUSING that register — is only
reachable with the bound in a local, and assigned late, so its live range starts after the block
temps die (assigned at the top it takes $a1, not $v1). Two companions: use the PARAMETER as the
multi-use variable (its expand-time parm copy survives exactly when $aN is re-used; a
self = arg0 local emits double moves), and name the inner term of a nested a - (b - K) as a temp
or fold distributes it away.
4. §172b-1 made concrete — a naked sll/sra pair on a REGISTER is a branch-merged multi-def
variable (func_8018A6A0). Not a cast of a fresh load: mirror the memory slot with a variable
written on BOTH paths (vec0[1] = w = *s1; at entry, w = vec0[1] + …; vec0[1] = w; in the loop)
and compare (s16)w after the join. (s16)vec0[1] instead emits a fresh lh — store-forwarding
cannot cross a two-predecessor label. This flipped the residual from structural to a 3-instruction
commutative register tie, i.e. from an idiom problem to permuter fuel.
Two honest walls recorded so nobody re-spends them: func_801AE3A4's missing same-value copy is
reachable by an "0"(b) re-tie but sits behind §172's +8-frame canonicalization wall; and the
swaprepeat siblings func_80187E08/func_80182FD0 already carry the right doubled-compare
structure — their 2-10-instruction scratch permutation is the S11 knife-edge, permuter fuel, and one
fix banks both by remap.
Addendum (P31 S63 t5e-t5i, func_8017EDD4): §264-2's save-order-only residual has a second, cheaper cure that needs no control-flow surgery: inside bb0, move the def statement to the far side of the adjacent BLKmode aggregate-copy statement. func_8017EDD4 (ov_SC03_030, 67 ins) already carries the §30#3 boost-kill re-tie __asm__ volatile("" : "=r"(p)) at end-of-function — bought for the §201-C addiu-sinks-past-jal shape — and still emits the inverted pair: idx3/4 sw $s1,28($sp) / sw $ra,32($sp) against the target's sw $ra,0x20($sp) / sw $s1,0x1C($sp), SCHEDULE-REORDER/2, span [3,4], every other word already exact. Writing the copy first — buf = D_801E0764; (the §48-C2 lwl/lwr+swl/swr align-2 struct assign) before p = (u8 *)&buf; instead of after, re-tie unchanged and nothing else touched — is the whole delta from closeness 2 to MATCH 67/67 (.run/t5i/stage_union/ov_SC03_030/func_8017EDD4.c:27-28,54). TRIAGE — do not read this residual as the re-tie failing. With the re-tie present, a clean 2-instruction prologue swap is no longer the call/branch delay-slot contention the re-tie was spent on, and re-spending §30#3 / §257-1 placement levers on it costs a wave: the same re-tie moved into the THEN arm regressed to closeness 33 / LENGTH-DRIFT, and replacing it with a plain second p = (u8 *)&buf; also gave 33 (both re-confirming that only the volatile re-tie survives cse). Move the def, not the re-tie. Honest scope: n=1, one single-axis A/B pair; the agent's sw $sN→def anti-dependency account agrees with §264-2's mechanism but no RTL dump was taken here — the position-sensitivity is byte-proven, the attribution is not.